Front Cover
 Title Page
 Table of Contents
 A pilot study of the mountain gorilla...
 The death of Bishop Hannington....
 The eclipse at Biharwe
 Ancient capital sites of Ankol...
 The composition of the Buganda...
 The eastern boundary of Uganda...
 The Kagera Triangle and the Kagera...
 Bark-cloth hammers
 Notes on contributors
 Back Cover

xml version 1.0 standalone yes
PageID P46
ErrorID 4

The Uganda journal
External Link ( Journal issues, 1999-2002 )
Full Citation
Permanent Link: http://ufdc.ufl.edu/UF00080855/00066
 Material Information
Title: The Uganda journal
Abbreviated Title: Uganda j.
Physical Description: v. : ill. ; 22-24 cm.
Language: English
Creator: Uganda Society
Publisher: Uganda Society etc.
Uganda Society etc.
Place of Publication: Kampala etc
Publication Date: 1959
Frequency: irregular[1976-<1995>]
semiannual[ former 1934-]
annual[ former <1948>-1973]
completely irregular
Subjects / Keywords: Periodicals -- Uganda   ( lcsh )
Genre: serial   ( sobekcm )
periodical   ( marcgt )
Spatial Coverage: Uganda
Dates or Sequential Designation: v. 1- 1934-
Numbering Peculiarities: Suspended May 1942-Mar. 1946; replaced by the Society's Bulletin.
Numbering Peculiarities: One issue published every few years, v. 38 (1976)-<v. 42 (1995)>
Issuing Body: Vols. 11-13 published in London.
General Note: Journal of The Uganda Society (formerly The Uganda Literary and Scientific Society).
 Record Information
Source Institution: University of Florida
Holding Location: University of Florida
Rights Management: All rights reserved by the source institution and holding location.
Resource Identifier: ltuf - ABS8002
oclc - 01644239
alephbibnum - 000301510
issn - 0041-574X
lccn - 52026895
System ID: UF00080855:00066
 Related Items
Related Items: Uganda journal

Table of Contents
    Front Cover
        Front Cover 1
        Front Cover 2
    Title Page
        Title Page 1
        Title Page 2
    Table of Contents
        Table of Contents 1
        Table of Contents 2
    A pilot study of the mountain gorilla in south-west Uganda
        Page 1
        Page 2
        Page 3
        Page 4
        Page 5
        Page 6
        Page 7
        Page 8
        Page 9
        Page 10
        Page 11
        Page 12
        Page 12a
        Page 12b
        Page 13
        Page 14
        Page 15
        Page 16
        Page 17
        Page 18
        Page 19
        Page 20
        Page 21
        Page 22
        Page 23
        Page 24
        Page 25
        Page 26
        Page 27
        Page 28
    The death of Bishop Hannington. Supplementary evidence
        Page 29
        Page 30
        Page 31
        Page 32
        Page 33
        Page 34
        Page 35
        Page 36
        Page 37
        Page 38
        Page 39
        Page 40
        Page 41
        Page 42
        Page 43
    The eclipse at Biharwe
        Page 44
        Page 45
        Page 46
        Page 47
        Page 48
        Page 49
        Page 50
    Ancient capital sites of Ankole
        Page 51
        Page 52
        Page 53
        Page 54
        Page 55
        Page 56
        Page 57
        Page 58
        Page 59
        Page 60
        Page 61
        Page 62
        Page 63
    The composition of the Buganda Lukiko in 1902
        Page 64
        Page 65
        Page 66
        Page 67
        Page 68
    The eastern boundary of Uganda in 1902
        Page 69
        Page 70
        Page 71
        Page 72
    The Kagera Triangle and the Kagera Salient
        Page 73
        Page 74
        Page 75
        Page 76
        Page 77
        Page 78
    Bark-cloth hammers
        Page 79
        Page 80
        Page 80a
        Page 81
        Page 82
        Page 83
        Page 84
        Page 84a
        Page 85
        Page 86
        Page 87
        Page 88
        Page 89
        Page 90
        Page 91
        Page 92
        Page 93
        Page 94
        Page 95
        Page 96
        Page 97
        Page 98
        Page 99
    Notes on contributors
        Page 100
    Back Cover
        Back Cover 1
        Back Cover 2
Full Text

The Uganda Journal


VOLUME 23, No. I

MARCH 1959

A Pilot Study of the Mountain Gorilla in South-west Uganda
The Death of Bishop Hannington, Supplementary Evidence
Kimera C. C. WRIGLEY
The Eclipse at Biharwe - J. SYKES
Ancient Capital Sites of Ankole - ROLAND OLIVER
The Composition of the Buganda Lukiko in 1902 A. Low
The Eastern Boundary of Uganda in 1902 G. BENNETr
The Kagera Triangle and the Kagera Salient H. B. THoMAs
Bark-cloth Hammers E. C. LANNINO

A Rain Cape common to South Asia and Africa F.J. SuIooNs
On Avoiding Detection H. S. S. FEW

Lwoo Traditions -
Rock Paintings in Teso
Historic Sites in Ankole
Runyankore Vowels

J. E. CosMrroN and C. J. HELLBERG

The Making of Modern Uganda (by K. Ingham) G. BENNErr
A Bird Watcher in Kenya (by V. D. van Someren) A. C. BROOKS
An Atlas of African History (by J. D. Fage) B. W. LANGLANDS
A Runyankore Grammar (by H. F. Morris and B. E. R. Kirwan)
A Linguistic Bibliography of East Africa (by W. H. Whiteley
and A. E. Gutkind) H. F. MoRRI
Makerere Journal W. T. NEWLYN
The Growth of the Church in Buganda (by John V. Taylor)

Published by
Price Shs. 15 (15s.)










His Excellency Sir Frederick Crawford, g.C.M.a., O.a.E.

Dr. H. F. Morris

The President
The Vice-President
The Hon. Secretary
The Hon. Treasurer
The Hon. Editors
The Hon. Librarian
Dr. W. Elkan
Mr. H. S. S. Few
Mr. J. S. Kasirye
Mr. E. Kironde
Hon. Secretary:
Hon. Treasurer:
Hon. Editors:

Dr. A. W. Southall


Hon. Librarian:
Hon. Auditor
Mr. G. Wright
Corresponding Secretary at Mbale:
Corresponding Secretary at Masaka:
Corresponding Secretary at Tororo:

Mr. B. Kirwan
Mr. S. W. Kulubya, ..B.E.
Mr. D. K. Marphatia
Mr. R. J. Mebta
Mr. C. N. Mukuye
Mr. P. Tamukedde
Dr. H. C. Trowell, O.B.E.
Mrs. K. M. Trowel], M.B.E.
Mr. M. Barrington Ward
Dr. C. Ehrlich
Mrs. M. M. Wallis
Mr. J. C. D. Lawrance
Dr. E. M. Chenery
Mr. M. D. Ross
Hon. Legal Adviser
Mr. C. L. Holcom
Mr. F. Lukyn Williams
Mr. E. C. Lanning
Dr. W. H. R. Lumsden

Hon. Vice-Presidents:
H.H. Mutesa II, 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.B.E.
Lord Twining of Tanganyika and Professor A. W. Williams
Godalming, G.C.M.G., M.B.E.

Past Presidents:




Sir A. R. Cook, C.M.G., O.B.E.
Mr. E. J. Wayland, C.B.E.
Dr. H. H. Hunter, c.B.E., LL.D.
Mr. H. Jowitt, C.M.a.
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.B.E.
Mrs. K. M. Trowell, M.B.E.
Dr. W. J. Eggeling
Dr. G. ap Griffith
Rev. Fr. F. B. Gaffney, w.P.
Professor A. W. Williams
Sir J. B. Hutchinson,
C.M.G., P.R.S.
Mr. J. D. Jameson, O.B.E.
Dr. Audrey I. Richards, c.B.E
Dr. H. C. Trowell, o.B.E.
Mr. D. K. Marphatia
Mr. M. Barrington Ward

Mrs. K. M. Trowell, M.n.e.

Mr. B. K. Mulyanti

Mr. G. P. Saben

Secretary: Mrs. M. M. Wallis


Uganda Journal



No. 1

MARCH 1959

(assisted in England by H. B. THOMAS, O.B.E.)

Published by



A Pilot Study of the Mountain Gorilla in South-west Uganda
The Death of Bishop Hannington. Supplementary Evidence

Kimera -
The Eclipse at Biharwe
Ancient Capital Sites of Ankole - -
The Composition of the Buganda Lukiko in 1902
The Eastern Boundary of Uganda in 1902
The Kagera Triangle and the Kagera Salient -
Bark-cloth Hammers -

A Rain Cape common to South Asia and Africa
On Avoiding Detection -

Lwoo Traditions -
Rock Paintings in Teso
Historic Sites in Ankole
Runyankore Vowels


A. Low

H. S. S. FEW


The Making of Modern Uganda (by K. Ingham) G. BENNETT
A Bird Watcher in Kenya (by V. D. van Someren) A. C. BROOKS
An Atlas of African History (by J. D. Fage) B. W. LANGLANDS
A Runyankore Grammar (by H. F. Morris and B. E. R. Kirwan)
A Linguistic Bibliography of East Africa (by W. H. Whiteley
and A. E. Gutkind) H. F. MORRIS
Makerere Journal W. T. NEWLYN
The Growth of the Church in Buganda (by John V. Taylor)



[Reproduced with the permission of the South African Association for the Advancement
of Science from South African Journal of Science, Vol. 54, No. 8, August 1958, pp. 195-217.
In an introductory article ibidd., pp. 192-4), Professor Raymond A. Dart of the University
of the Witwatersrand remarks:
"Towards the end of 1956 the University of the Witwatersrand received a letter from
W. M. Baumgartel, the proprietor of Travellers' Rest at Kisoro in Uganda stating that his
inn lay at the foot of the volcanic range of Muhavura, Mgahinga and Sabinio separating
Uganda from the Belgian Congo; and that gorillas live on these mountains in the Forest
Reserve which is continuous with the Parc Nationale Albert. The letter also carried the
practical news that Travellers' Rest would accommodate at the proprietor's expense one
or even two serious students who were competent and prepared to study the habits of the
This prodigy of a hostelry snugly situated alongside a gorilla reserve and under the
tutelary direction of a scientifically-minded, and anthropologically-conditioned spirit
augured so propitiously for future studies in higher primate behaviour that Professor
Dart, to whom the letter had been referred, was elated to enter into direct negotiations
with the presiding genius of the Kisoro sanctuary. Gradually it emerged that Baumgartel's
interest in South Africa arose from having lived aforetime first in Cape Town and then in
Johannesburg after having fled betimes from Hitlerian Germany. When the second World
War began he left Johannesburg to serve throughout the North African campaign. After
the victory he settled in England but before long found that Africa held him permanently
in her toils.
In May 1955 in response to an advertisement he had taken a partnership in Travellers'
Rest, which became his sole responsibility in 1956. At Kisoro he had found to his delight
that the mountain gorillas were his next door neighbours. He became seized with the
idea that, if he could establish sufficiently friendly relations with them, they could become
both an attraction to tourists and a source of information to scientists.
So he employed local native inhabitants familiar with the ways of the gorillas and of
whom Reuben is chief; strewed bananas, sugar-cane and other hopeful dainties in the
vicinity of the gorillas' movements; and prepared salt-licks on the saddle-back between
Muhavura and Mgahinga. Gorillas' knuckleprints were found near the hoof-marks of
buffaloes that visited a salt-lick, but the vegetable sweetmeats found no welcome from
gorillas. The expense involved did not warrant the use of fruit, but the salt-lick success
is being pursued.
Meanwhile, through the personal interest of Dr. L. S. B. Leakey, Baumgartel was joined
in October 1956 by Miss Rosalie Osborn. With her help systematic trips were made up the
mountains with the native guides and visitors to the hotel, staying overnight for a few
days at a rest hut erected on the saddle-back, and making documentary ecological records
of the many animals seen and heard, the plants eaten and the nests and movements of the
gorillas. In the following February Miss Jill Donisthorpe, B.Sc. (Bristol), replaced Miss
Osborn and continued her making of records and collections until October 1957. Thus
through the persistence and enterprise of these two women there has been gathered
systematically for the first time information gleaned over the course of a year about the
movements and associations of the gorillas in south-western Uganda. The zeal of Miss
Donisthorpe especially led to the collection of plants used by the gorillas as food and a
better understanding of their groupings. Her work was rendered the more onerous through
Baumgartel's ill-fortune in suffering a fracture of his right leg on the very day she assumed
her duties. She, like Miss Osborn, had originally intended to stay merely for a few holiday
months; but the happily temporary plight of her good host and the interest of her new
surroundings tempted her to stay until the full year's cycle of the events observed had
been recorded.
In August favourable news about Kisoro and its value to students of botany and
geology as well as geography, anthropology and zoology persuaded the Research Commit-
tee of the University to support the Uganda enterprise to the extent of 50 per month in

collaboration with our numerous colleagues at Makerere Medical School; and to ask
Professor Alexander Galloway-previously Senior Lecturer in Anatomy at the University
of the Witwatersrand and presently Dean of the Medical School and Professor of
Anatomy at Makerere University College-to establish a local committee in Uganda to
advise upon the allocation and expenditure of this fund and any further monies that might
accrue to it in the future.
Kisoro is particularly well situated for such an ambitious enterprise as Mr. Baumgartel
has launched and it appears tolerably certain that these studies will evoke considerable
international collaboration. Placed approximately midway a day's journey by motor car
between Kampala (where Makerere College is situated in Uganda) and Astrida (where the
Anthropological Research Sub-station of I.R.S.A.C., i.e. Institute de Recherche Scientifique
en Afrique Centrale, is situated in Ruanda-Urundi) such collaboration should easily be
developed between scientists whether in Africa south of the Sahara, South Africa, Rhodesia,
Uganda or Belgian Congo. EDS.]
In 1956, Mr. M. W. Baumgartel, the proprietor of a small hotel in Kisoro,
S.W. Uganda, conceived the idea of trying to tempt the elusive mountain gorillas
of the nearby sanctuary into clearings by means of special foods, so that they
could be seen by visitors. Permission was obtained from the Game Department,
but the experiment was not a success and attempts to feed the animals were
abandoned after a few months. However, much interest had meanwhile been
aroused in the gorillas. Mr. Baumgartel had been in touch with various scientific
bodies in Great Britain and South Africa and the suggestion that a scientific
expedition should come to the sanctuary had been favourably received. While
the financial possibilities for an expedition were being examined, Mr. Baum-
gartel was asked to continue his activities as far as possible in the form of a
pilot study, so that some information might be available if an expedition were
organized. The development of the pilot study took place gradually between
October 1956 and the end of the year, during which time Miss R. Osborn
from Nairobi was assisting with the feeding scheme. When I replaced her at
the end of January 1957, feeding as a means to a commercial attraction had
been abandoned, though as an aid to the observation of gorillas it was still
regarded as a possibility. When I left at the end of September 1957, a year's
preliminary observation had been completed.

The Uganda Gorilla Sanctuary lies on the northern slopes of three volcanoes
-Muhavura, Mgahinga and Sabinio, which lie in an east-west line and form
part of the Birunga or Bufumbira group on the borders of Uganda, the Belgian
Congo, and Ruanda-Urundi. Of the eight volcanoes, six are extinct but two in
Belgian territory are still active, and a small new crater erupted in August 1957
at the base of Sabinio on the Belgian side. The slopes of the volcanoes which
lie in Belgian territory form part of the Pare National Albert (see Map 1); this
part of the park is closed to the public.
Muhavura, 13,547 feet, and Mgahinga, 11,400 feet, are separated by a saddle
about a mile wide at an altitude of approximately 10,000 feet. On this saddle
a small camp was built in October 1956 to facilitate observations of the gorilla.
Sabinio lies some four miles west of Mgahinga, the lowest point of the inter-
vening slopes being at about 8,000 feet (see Map 2).

From the saddle, an artificial watercourse runs down the hillside to the so-
called Schandl's camp and a nearby dam from which water is piped to cultiva-
tions on the plain. Alongside the watercourse is a path leading to the saddle
camp and beyond it to an open swamp (see Map 2).
The crater of Muhavura is a shallow depression containing a small lake. The
crater of Mgahinga is about a quarter of a mile across and has a grassy floor
with a swamp at one end. Sabinio, which consists of a number of tooth-like
peaks, has no definite crater. The slopes of all three mountains approximate in
places to 40 degrees.
The sanctuary is accessible from the small township of Kisoro (or Kisolo)
(altitude 6,300 ft., lat. 10 15' S., long. 29 40' E.) which lies seven miles from
the Congo border, five miles from the Ruanda-Urundi border, and six miles
from the base of Muhavura and Mgahinga. The approach lies along three miles
of dirt road followed by three miles of grass track to just over 7,000 feet, the
highest point which can be reached by car. Here, at the base of the path to
the saddle, is the disused camp built by the watercourse engineer and so named
after him Schandl's Camp. Sabinio is less easy to reach; excursions up its slopes
are best made from a camp at its base which is two hours' walk from the
nearest point accessible by car.
The boundaries of the sanctuary are officially defined as follows:
"The area bounded by a line from the highest point of Sabinio Mountain
following the Anglo-Belgian International boundary in a north-easterly
direction to the intersection of that boundary with the parallel 1 20' 15"
south latitude; thence in an easterly direction in a straight line to a pillar
erected on the international boundary between the Uganda Protectorate and
Belgian Ruanda situated between the hills, Nyarnbebsa and Mussongo;
thence following the international boundary between the Uganda Protector-
ate and Belgian Ruanda in a south-west and westerly direction to the point of
The southerly boundary of the sanctuary is therefore along the frontier line
which connects the three peaks, while the northerly (lower) boundary very
roughly follows the 7,500 foot contour line. The area is approximately 18 square
The side of Muhavura is scored by ravines, some comparatively shallow,
some more than 100 feet deep, down which torrents flow after rain. In the
centre of the north face is one exceptionally large ravine with steep, unclimbable
sides and a barren, stony floor, to the east of which gorillas are not found. So,
though officially the sanctuary extends to the Ruanda border on the east side,
for all practical purposes it ends at the ravine.

Forest Reserve
The upper part of the slopes constitutes a Forest Reserve. The boundary
of the Reserve is demarcated by a cutting some 15 feet wide, which encircles
Muhavura and Mgahinga at about 9,000 feet, dropping to around 8,000 feet
towards Sabinio. This was not always so. The original boundary, as gazetted in
1939 and marked in 1941, ran along the 8,000 foot contour line. In 1950, owing

to the rapidly expanding population of the district, the boundary of the Forest
Reserve was moved a mile uphill and the area reduced from 13 square miles to
nine. Cultivation is allowed on the slopes outside the Forest Reserve, and on
the Muhavura side extends practically up to the demarcation line. On the
Mgahinga side the cultivation limit is considerably lower but it still extends
some way inside the boundary of the sanctuary.

Since only two visits were made to Sabinio, the description that follows
applies only to that part of the sanctuary which lies on Muhavura and
The vegetation of the slopes varies with altitude. Although several belts can
be distinguished, their limits are not clearly defined or even horizontal. Very
often a tongue of vegetation from one belt extends into another, and many
plants appear in several belts. The main zones are:
(a) Cultivations. These extend from the base to about 8,000 feet on Mgahinga
and to almost 9,000 feet on Muhavura. They are not continuous, but owing to
tribal fragmentation customs, odd patches of cultivation may be found sur-
rounded by a tangle of bush or forest. Neglected cultivations become quickly
covered with secondary growth. More tree clearing has been done on Muhavura,
so that here the overgrown areas between cultivations are less wooded than on
(b) Lower Forest. The slopes immediately above cultivation level are covered
by forest. Though the trees are only medium-sized, they are festooned with
lianas which, added to the thick bush and undergrowth, give the forest an
appearance of great density. A tangle of creepers and ramblers covers fallen
trees, bushes and old stumps, and in places there is a solid matting of vegeta-
tion several feet above the ground. Bamboo clumps are found scattered through-
out the region.
(c) Bamboo belt (native name Rugano). The clumps of the lower forest
merge into a pure bamboo belt just below the Forest Reserve line on Mgahinga,
but some way above it on Muhavura. The belt extends to the level of the
saddle, though not much bamboo exists on the Uganda side of the saddle
itself. Throughout the zone, particularly on the Muhavura side, there are clear-
ings containing heather, Red Hot Poker and a sprinkling of small Hypericum
(d) The saddle is an open swampy area surrounded by thickets of woody
shrubs and lichen-covered Hypericum trees, with an undergrowth of nettles.
On the north-west side there is a bamboo forest interspersed with swampy
clearings. Rushes are found throughout the area.
(e) Upper forest or Hypericum woodland (native name Rugeshi). From the
saddle to approximately 11,000 feet there is another wooded zone. The vegeta-
tion here is less varied than in the lower forest, consisting mainly of the tree
Hypericum lanceolatum and a mass of woody shrubs. Owing to the thickness
of the shrub layer, there is less undergrowth and ground cover than in the
lower forest.
(f) Tree-Heather or Sub-Alpine zone (native name Rutiti). On Mgahinga

this zone more or less coincides with the crater rim, but on Muhavura it
extends beyond the upper forest for almost 1,000 feet. Tree heather grows
10-20 feet high, its branches often covered with beard lichen and moss.
(g) Alpine zone. Here one finds mainly Giant Lobelia, Giant Groundsel and
everlasting flowers, with a ground covering of succulents and short grass. The
alpine plants are intermixed with the heather, but on Muhavura extend above
it to the summit.
Amongst big game, elephant, buffalo and leopard inhabit the slopes. Red
Forest Duiker and Yellow-backed Duiker may be seen, the former frequently,
the latter only rarely. Tree Hyrax and a large rat are amongst the smaller
animals recorded, and tracks of pig and hyena are often found. The Golden
Monkey, rare in other parts of Africa, is often seen in the bamboo. Snakes were
thought to be non-existent, but in February two visitors saw one and in July
other visitors caught an adder in the lower forest.
Elephant were thought to live mostly in Belgian territory and to be only
occasional visitors to the Uganda side of the slopes, but this was not found
to be the case. Two large invasions took place, one in April and one in August,
while smaller numbers were recorded in February, July and September. Miss
Osborn also records the presence of elephant in October. Owing to the ravines,
the elephant do not penetrate the lower slopes of Muhavura to the east, but
they trample all over the saddle, go to the top of the heather zone on Muhavura
and to the crater of Mgahinga, and frequent the lower forest on Mgahinga
down to cultivation level. Their presence hampers gorilla observations in that
they prevent the observer from moving freely, they obliterate previous gorilla
tracks and they make even subsequent tracks difficult to see.
Buffalo occur in large numbers on the saddle where they trample the muddy
areas into a morass. They are also found in those parts of the upper forest
adjoining the saddle and in the bamboo, but only occasionally in the lower
Leopard live in the numerous caves which exist on the slopes. Their tracks
are often found but the animals themselves were not seen.
Hyena tracks were found several times but the animals were not encountered.
According to the local natives, they find so few opportunities for scavenging
that they take bites out of living cattle and even buffalo.
The remaining animals play no part in gorilla observations.
Records show Kisoro as having an average rainfall over 15 years of 66-06
inches per year; that in the mountains is probably very much higher. The rainy
season lasts from September to May with a short dry spell just after Christmas.
The mornings are usually clear and fine, clouding over towards mid-day, with
rain and thunder during the afternoon and evening. Rain normally starts be-
tween 2 and 3 p.m., sometimes as early as 12 noon. Storms occasionally occur
during the mornings, but on only one day, 9 April, was it too wet to go out at all.
Hail-storms are frequent on the mountains, and the saddle camp is usually cut
off from the plains by cloud and mist. During fine spells, visibility is excellent.

Distant ranges stand out clearly and Lake Edward, and occasionally the
Ruwenzori Mountains, can be seen from the slopes.
The dry season usually lasts from June to August, though in 1957, September
also was fairly dry. During this season the atmosphere becomes hazy and for a
short time during August the mountains are invisible from Kisoro. There were
short wet spells during the dry season of 1957 as follows:
19-22 June.
26-29 July.
1-5 and 26-27 August.
During June and early July the vegetation was extremely dry, but an increase
of cloud and mist towards the middle of the month freshened it up again.
After a storm many waterfalls appear on Muhavura and water rushes down
the ravines. On 6 May a landslide brought rocks and boulders down one of the
ravines to the plains, blocking the track for two days. In September landslides,
probably caused by earth tremors, occurred in two ravines and were mistaken
by natives for eruptions. During bad weather the track is often too slippery
for an ordinary car and the last mile or so has to be walked.
The prevailing wind is from the east, veering to south-east during the dry
No temperature records are available, but the temperature at Kisoro varies
from pleasantly cool to warm. Evenings are sometimes raw when it is wet,
though the actual temperature is not low. Records for Kabale, a township 50
miles to the east, at an altitude of 6,139 feet are:
Mean air 61-3F.
Mean max. 75-4F.
Mean min. 50-4F.
On the lower slopes of the mountains temperatures are similar to those of
the plains, but on the saddle and above it is uncomfortably cold in mist and
rain and also cold at night when a strong wind is blowing. Akeley's 1926 expedi-
tion recorded temperatures of 460 by day and 320 by night at their camp at
11,000 feet. In fine weather the summits are not cold by day but during the
night of 19 July, when I camped in the crater of Muhavura, there was a hard
frost with ice round the edges of the lake.
The Mountain Gorilla was first discovered at the beginning of the century
when Captain Oscar von Beringe, a German East Africa Railways' official on
a hunting holiday, shot one on Sabinio. It was examined scientifically by
Matschie who found it differed from the Forest Gorilla, and in 1903 it was named
Gorilla gorilla beringei after its discoverer.
The first serious attempt to study the Mountain Gorilla was made by the
American naturalist, Carl Akeley, who organized two expeditions to the Congo
volcanoes. During his first, in 1921, he collected some specimens for museums
in the United States and made the first, and probably the only, cinema film of
them. He died tragically during his second expedition in 1926 and was buried
on the saddle between Mikeno and Karisimbi, the two highest peaks of the
Birunga group. On this expedition he was accompanied by Dr. J. M. Derscheid,

Belgian zoologist and cartographer, who stayed on in the area after his death,
first with other members of the expedition, and later with the Harvard Africa
expedition of 1927. Altogether Derscheid spent seven months in the area and
saw gorilla 33 times. A section of the Prince William of Sweden expedition of
1921, which spent a short time in the volcanoes, encountered gorillas; otherwise
the only accounts of the animals come from hunters among whom T. Alexander
Barns,(l) Benjamin Burbridge(2) and Edmund Heller have recorded their
Few visits seem to have been made to the Uganda part of the volcanoes. It
was not until 1919 that gorillas were suspected in British territory and they
were not officially acknowledged until 1929.
Outside the volcano area, the mountain gorilla is now known to exist in
three pockets in the eastern Congo and in the Kayonza Forest of Uganda.
Since Sabinio is not easily accessible from Kisoro, all observations were made
on Muhavura and Mgahinga.
The start of the investigation was influenced by the opinion, widely held
amongst previous observers, that the gorillas lived for the most part on the
saddle or above, making temporary excursions into the bamboo. If, as was
thought, they spent most of their time between 9,000 and 11,000 feet, they
would be most conveniently observed from the saddle. I therefore started off by
living in the saddle camp for three or four days at a time, but it soon became
apparent that the animals, wherever they may have been previously, were now
very definitely in the lower forest. When they had been encountered four times
out of five below the Forest Reserve, it appeared more convenient to visit them
daily from Kisoro. This was done tentatively from 27 February until the end
of March; then as there were no signs of a change of habitat, the camp was
closed, partly for convenience and partly for economy. The camp was not
occupied again until 12 August, when it was reopened for a month, though
visits to the higher altitudes were made regularly.
At first, knowing nothing of the conditions, my sole aim was to find gorillas
and observe them. I was accompanied at all times by an African guide and a
tracker, who were expert at finding gorilla spoor. The system followed, which
proved effective throughout, was simply to walk through the forest at random
until a track was found, then follow it. The important thing is to start walking
in roughly the right area, and for this information I relied on the advice of the
guide, who was also helped on occasions by information from the local people.
To be noticed by the local people, of course, the gorillas must be below the
Forest Reserve, which was the case most of the time.
But I soon discovered that, though the gorillas were not difficult to find, they
were situated in such thick vegetation that watching them for any length of
time was impossible. I therefore decided to follow lines of investigation that
did not depend on prolonged views of the animals. Possible subjects were:
1. Migration, seasonal patterns, daily movement, distances covered, group
2. Food: plants eaten, in what quantities and in what seasons.

3. Nests: construction, siting, vantage points, daytime resting places.
shelters. Nests as an indication of size and relationships in a group.
The remaining subject for investigation, which required at least some
sight of the animals was:
4. Behaviour, including posture, tree climbing, reactions to human beings,
escaping, aggressiveness, relationships within the group, between groups
and with other animals.
Both from our own point of view and that of the Game Department, we had
to be careful not to frighten or harass the animals, and for this reason, once we
had made contact, we left them alone for the day. Further following might
not only have annoyed them, but might also have driven them to unaccustomed
places so that we should have been observing them under artificial rather than
true conditions.
A diary of observations was kept daily and the locality of all groups en-
countered was plotted on a sketch map (see Map 2). Gorillas were entered as
located when:
i. Seen.
ii. Their cry was heard.
iii. Smelt.
iv. Warm droppings were found.
Positions were estimated by timing from known landmarks but, needless to
say, are only approximate.
Gorilla foods were collected and pressed, partly eaten specimens being chosen
to avoid confusion. They were sent for identification to the Agricultural Depart-
ment's Research Station at Kawanda, near Kampala, of which Dr. E. W.
Chenery is the present Director.
No feeding was attempted at this stage because in the lower forest it was
impracticable, and once the camp was closed a frequent check of food dumps
on the more open upper slopes would not have been feasible.

General Observations
(a) Track. The gorilla track is indistinguishable from a human track except
for the presence at intervals of partly-eaten food plants, pulled-down creepers
and occasional areas of beaten-down vegetation. It can be confused for a short
time with a buffalo track, though in the latter case signs of grazing will soon be
noticed. A small elephant track somewhat resembles a large gorilla track until
footprints, droppings and general destruction make its identity obvious. Some-
times a gorilla group may be followed along a single track, since the animals
often walk in single file, but when feeding their tracks diverge and rejoin,
sometimes forming a complex network. The droppings have been described as
triangular in section and something like those of a pig. The footprints on vege-
tation can only be recognized by an expert tracker, but on soft earth they stand
out clearly. The footmark is not noticeably larger than a human being's but
is very much broader. Sometimes the mark of the separate big toe can be seen.
The knuckle prints are unmistakable (Fig. 1). Along the track groups of nests
will be found. These will be described later.

(b) Physical characteristics. First glimpses of gorillas confirmed descriptions
by previous observers and may be summed up as follows: The coat is jet black
except for white markings on the back of the male, sometimes in the form of a
spinal stripe, sometimes a completely silver back. The amount of white is said
to increase with age. The animals walk in a stooping position with the main
weight on the legs and the knuckles of the hands resting lightly on the ground.
When erect, the males average six feet in height, the females much less, though,
since females do not stand still to be looked at, it is difficult to judge their size.
The males are tremendously broad and powerful-looking and one can well
believe that their estimated weight of 400 lb. or more may be correct. The male
has a crest of hair growing from the forehead which gives added height. Gorillas
have a very strong smell which is difficult to describe but is unmistakable. It
is rather like a combination of very pungent human sweat, manure and charred
(c) Groups. Groups of animals from three to twelve in number were en-
countered. Groups are said to have an acknowledged leader, though this was
not always apparent to the observer. Out of 30 visual meetings with gorillas, on
only 8 occasions was there any kind of demonstration, presumably by the leader
of the group. On another occasion, while four animals walked away in single
file, a fifth and larger one remained behind watching us, though no noise was
(d) Senses. Gorillas are said to rely mainly on sight, their smell and hearing
being poor. Their hearing did not seem good though it is difficult to judge. We
were certainly able to approach them under conditions where noise was
unavoidable, but this may have been because they mistook our noise for that
of their companions. We once managed to get within 12 feet of a gorilla, at
which distance one would expect our scent to have been obvious, but it was
unaware of our presence until it saw us. Normally, one is detected at a distance
of within 15 to 30 yards.
(e) Sounds. The sounds made by gorillas may be divided into those made
between themselves and those directed at, or evoked by the presence of an
It has been said that gorillas talk to one another, but it is difficult to pick on
any sound as indicative of particular communication. I have heard several-
once a kind of hoarse croak, when the animals had no idea we were near, and
on two occasions a burbling sound, like indistinct human voices. In the first
case the sound was like women's voices, so much so that the guide, who had
lost the trail, said he would go and ask the women whether they had seen any
gorillas. He went in the direction of the sound and saw, as he thought, five
people sitting round a tree; on closer inspection, they proved to be five gorillas!
A few minutes later we all saw them. On the second occasion the sound was
deeper, more like male voices. As we followed the track, this sound was
replaced by chewing and crunching as the animals fed beyond a screen of
creepers. Both times the burbling sound was connected with a group which was
sitting eating. It may have no connection with gorillas at all.
There is another sound, which the guide states is made by gorilla but about
which I have always been sceptical. It is a regular high-pitched "toot", much

more like the sound of a bird. However, on three occasions fresh tracks were
found leading towards the sound, though I was never able to follow them far
enough to establish anything. Ionides(3) mentions a sound "like two empty
bottles being struck together, made by the cupped hands striking the cheeks",
a description which more or less fits this sound.
A squeaking noise, of a different sort, is made by a baby gorilla. It is not
regular but intermittent. The squeaks, when I heard them, were punctuated by
a slapping sound which the guide said was made by the mother either playfully
slapping the baby or slapping her own cheeks. Other observers have both seen
and heard this cheek-slapping, though its object is not clear.
A peculiar sound, also heard by Miss Osborn, might be described as lip-
shaking. It is something between the flapping of a bird's wings and the noise
made by a horse blowing through its mouth. Its purpose is also obscure, though
on the occasion I heard it, it was made by the leader of the group who had seen
us from a distance, possibly as a warning to the rest or to tell them where he
was, in case of trouble. This was verified later by the finding of a large resting
place on the edge of a ravine in the direction from which the noise had come.
When. confronted by an intruder, the sounds made by gorillas vary from a
guttural bark to a roar or scream, depending on the extent of their fright or
anger. A bark is usually made by the first animal to see one, as a warning to the
others. A deep roar is made by an aggressive male; in between the two is a
sound something like the grunt of a lion with two notes in it, as though it were
made by the indrawn and outgoing breath.
Females, when vocal, give a piercing scream, and even some males when
alarmed, get a high note into their roar, which is terrifying until one realizes
it denotes fright rather than rage. On one occasion I heard from a distance a
great outcry from a female, lasting several minutes. Whether this was due to
human or animal interference or a family squabble one could not guess.
I was never fortunate enough to hear chest drumming.
(f) Numbers. It has been estimated by various authorities that there are
some 50 gorillas in the Uganda sanctuary. How this figure is arrived at is not
clear, since as far as is known, no observers other than ourselves have spent
more than a few days in the sanctuary. To mean anything, an estimate of this
sort would have to include the Belgian side of the slopes as well, since the
populations probably intermix. An assessment would be complicated, but the
following figures may be of interest:
Days spent in sanctuary 122
Days on which gorillas were located 41
Days on which fresh tracks were found (frequently several, not
necessarily connected with one another) 99
Days on which tracks were found by random walking, not following
those of previous day 44
Times gorillas were encountered by accident without tracking 7
The area regularly searched covers about 10 square miles of thick forest
where it is easy to miss a track by a few yards. On the upper slopes, owing to the
impenetrable bush, one tends to follow the same few paths all the time, and it
is surprising how often signs are found on these paths. Taking all things into

consideration-the area, the terrain and the conditions, I would say that the
high proportion of encounters indicates a larger number of animals than cur-
rently estimated.
(a) Preconceived ideas. It has already been mentioned that the true gorilla
habitat was thought to be the Hypericum woodland on the saddle and above.
It was also widely believed that during the dry season gorillas either disap-
peared across the border into Belgian territory, or migrated high up the slopes
into the alpine zone. The first idea was not held by the local people, who accord-
ing to the one writer,(4) name the lower forest 'Rwengaji'-the place of the
gorillas. The guide was inclined to support the second theory, though entries
in his visitors' book showed that gorillas had been seen in July and August 1956.
Questioned on the subject he suggested that seasonal movements of gorillas
were as follows:
June to August High up or over the border
September to December Moving down
January to April Low down
May Moving up
(b) Seasonal movements. The first three months' observations coincided with
his forecast, and towards the end of April I began to look for signs of an up-
ward migration. Encounters in late April and early May just below the saddle
camp gave support to the theory, and during the first week in May nests and
tracks were found on the saddle. But on 9 May tracks found high on Muhavura
led downwards and on 10 May a group was located in the lower forest on
Mgahinga. Later searches on the saddle revealed nothing, but tracks were
found low on Muhavura and natives reported having seen a group just above
the natives' cultivations (see Map 2).
During June, three out of four encounters were below the Forest Reserve and
the fourth was below the saddle camp. A porter related seeing three gorillas
towards the end of June on the saddle, but other searches of the upper regions
revealed no tracks. Early in July, recent signs were seen on the upper slopes
and around the crater of Mgahinga but at the same time news was heard of a
group low down on Muhavura, and fresh tracks and nests were found. During
the next two weeks tracks and nests were found in the lower forest of both
Muhavura and Mgahinga, and though signs were seen above the saddle towards
the middle of the month, the tracks led downwards and a large group was
eventually contacted in the lower forest on 21 July. From this date onwards
signs increased and it became apparent that the population of the lower forest
was increasing. Regular visits were still made to the upper levels and on
12 August the camp was re-opened to enable a thorough check to be made, but
results were negative. During August and September there were many en-
counters in the lower forest and on 5 September the lowest tracks of the entire
period were seen. On 18 September tracks were seen at intervals from the
cultivation level to just below the saddle camp, but no signs were found above it.
The drawback with a single observer is that it always seems more important
to investigate existing signs than to wander about at random. Thus, although
I was constantly looking for tracks leading upwards, when none were found I


concentrated on following up what signs there were. As a result, signs of
gorillas on the upper slopes were not found until after they had left, but it can
be fairly clearly established that they were on the saddle in May, at the top of
Mgahinga in June and at the top of Muhavura in July. Month by month results
can be summarized as in Table I.


Upper levels Lower levels
Month (saddle&above) (below saddle) Remarks
Searches Signs Searches Signs

February 2 nil 12 12 No signs on upper levels. Groups found
7 times in lower forest.
March 4 1 20 20 Ten encounters in lower forest, one on
April 3 1 10 9 Signs in lower forest for most of month
but in bamboo towards end. One track on
17th and signs found in May indicate
population on saddle mid-April.
May 5 3 12 8 Numerous signs on saddle and in bamboo,
also a track leading down from high up
on Muhavura. Tracks in lower forest.
June 2 nil 12 9 In bamboo, though below Forest Reserve
on 3 occasions. Signs found on crater of
Mgahinga early July indicate gorilla
visitation in June.
July 5 3 12 11 Although groups were present in the lower
forest, a climb to the top of Muhavura on
the 19th and 20th revealed very recent
gorilla signs as far as the crater rim.
August 6 1 7 7 A track leading down from the upper
slopes of Muhavura on 13 August was the
last sign of gorillas in the higher regions.
After this date signs increased in the lower
forest and the general set-up was similar
to that of February.
September 4 nil 10 10 Lowest recorded track found 5 Septem-
ber, also highest number of contacts-8
out of 10 on lower slopes. Very similar to

It would be tempting to conclude from these findings that there is an upward
migration from the end of April, reaching a peak in June and July with a
downward movement again in August. But the evidence is not really conclusive.
Arguments for and against are:

Signs from mid-April to mid-August:
eight positive and two possible out of a
total of 16 visits to upper regions. During
the rest of the period there was only one
positive sign out of 15 visits.
The fact that signs in the lower forest be-
came fewer during May and June. In May
fresh tracks were found only 8 times out
of 12 and in June 9 out of 12. In April
and July the figures were 9 out of 10 and
11 out of 12, while in all other months
fresh tracks were seen every day.

There was no month without some signs
in the lower forest.
16 excursions to the upper levels from
mid-April to mid-August resulted in no
actual contacts with gorillas, indicating
that perhaps the visitations were of short
duration only.
The fact that other observers(5) have seen
gorillas on the saddle in February, March
and September in other years.

FIG. 1
Knuckle print.

FIG. 3
Ground nest of bamboo.

FIG. 5
Baby's nest in bush.

FIG. 6
Shelter resting place.

FIG. 2
Hole made by scooping out root. The plant (Cyno-
glossum amplifolium) with root broken off is on
the left.

FIG. 4
Bamboo platform.

For conclusive results more observers, covering all altitudes simultaneously,
would be needed.
(c) Daily Range. Groups which were followed for several days in succession
were found to move very short distances-less than half a mile a day-for as
long as a month, though much greater distances were covered on occasions.
Longer journeys did not necessarily indicate a change of habitat, one such
comprising a circular tour of about seven miles, returning to within a short
distance of the starting point. Examples:
1. Between 20 and 28 February, a group of five was followed in the lower
forest on Mgahinga. During this time the group covered not more than two
miles in a straight line (though considerably more in meanderings), mov-
ing roughly west to east.
2. From 13 to 16 March a group was followed from east to west in the same
area, covering during this time perhaps a mile and a half.
3. Some time between 16 and 18 March this same group took off on an
excursion of about seven miles up through the bamboo to camp level,
across to the Muhavura side and back again, returning down to within
about 100 yards of the starting point. No nests were found along the route,
indicating that the journey was made in a day.
4. A similar, though shorter trek was done on 27 March by a group, possibly
the same as above, in the lower forest.
5. A small group was followed on 31 March. Following tracks known to be
five days old, we overtook the group after three hours. The distance in a
straight line was less than a mile.
6. On 9 May tracks found leading down from the upper slopes of Muhavura
were followed back and up, but in 4 hours of climbing no change of
direction was found. Since a deep ravine separated the tracks from the
saddle it is possible that they came from over the top of the mountain.
(d) Group territories. It has been suggested that gorilla groups keep to their
own territories and resist intrusion by other groups. I saw no signs to support
this suggestion. Towards the end of March a group of three and a large group
were known to be in the same area, moving in opposite directions. By 31 March
they had passed each other within 100 yards. Had there been any antagonism
one would have expected the smaller group to retreat.
(e) Reasons for migrations. It has always been assumed locally that migra-
tions are caused by food requirements. But assessment of the vegetation makes
this rather doubtful, for most of the gorilla foods are widely distributed over the
slopes. Much the same foods are found in the upper forest as in the lower,
though in different proportions; the woody shrubs which are much eaten on the
saddle grow both above and below it; bamboo occurs at all levels from 8,500 to
10,500 feet. The ripening of fruits might have some bearing on the matter,
though with one doubtful exception, no fruit foods were found. With constant
rain for nine months, there seems to be no definite season for the appearance
of fresh vegetation or new shoots. There seems no obvious reason why gorillas
should leave the lower forest during the dry season (if they do) "in search of
food" for there is abundant food at all times, though possibly clouds might
make the vegetation of the upper altitudes fresher. Furthermore, on a long

journey, fewer droppings and signs of eating are found, even though food plants
may be available along the trail. This would seem to indicate some reason other
than food shortage for the trek. Other possibilities might be the need for fresher
vegetation, exercise, or a change of climate, or in dry seasons the greater
succulence of foods higher up the slopes.
(a) Feeding habits. For the most part gorillas do not eat while on the move,
but collect a handful of food which they eat sitting down in a resting place
nearby. The whole of the picked portion is not eaten; the hard, unpalatable
parts are discarded, a habit which greatly facilitates the recognition of their
foods by an observer. The pieces are not thrown or scattered about but are
placed in a fairly tidy pile. As far as one can judge, mixtures of food are not
collected, but one type of plant only is eaten at each sitting place. From the
evidence, it looks as though the more sparsely distributed plants may be eaten
on the move, since making a collection would consume the time and patience
of the gorillas.
(b) Food plants. The food plants identified are described in Table II.
Foods according to zones may be summarized as follows:
i. Upper slopes: Anthriscus sylvestris, giant lobelia and giant groundsel in
about equal quantities (June-July).
ii. Upper forest: Woody shrubs supplemented by roots, creepers, bramble
iii. Saddle: Woody shrubs with tit-bits of giant lobelia or rushes (April-May).
iv. Bamboo forest: Bamboo as main diet with tit-bits of red hot poker in
clearings during flowering season and possibly orange fungus (May-June).
v. Lower forest: Celery as main diet, closely followed by bamboo, supple-
mented by climbers, roots and other occasional tit-bits. A far greater
variety of food is available in the lower forest, which may account for the
preponderance of time spent there.
When in a certain zone the main food of that zone is eaten, but when the
gorillas are on trek and moving fast much less feeding takes place; there seems
to be no dominant food, but a variety of odd scraps is eaten. During the trek
of 16-18 March, the following foods were eaten in roughly equal quantities:
woody shrubs, nettle tops, roots, dock, bramble leaves, creepers, celery, red
hot poker and possibly fungus.
(a) Construction. Nests are built of branches either on the ground (Fig. 3),
on bamboo platforms (Fig. 4) or occasionally on trees (Fig. 5). Fresh nests are
said to be built each night and each animal, even a baby, has a separate nest.
A nest may be distinguished from a resting place (Fig. 6) by the droppings
it contains. Since so many observers(6) were of the opinion that the vast
majority of nests were built on the ground, it seemed of interest to make a list
of what was seen. The result was:
Total number of nests counted 225
Number on the ground 121
Number above ground 104


Peucedanum spp.
Giant celery
Native name-
Arundinaria alpina

Vernonia spp.
Native name-

Pychnostachys goetzenii
Native name--
Basella alba



Abundant in lower forest, often occurring in
large patches. Like a much branched garden
celery, average height 2-3 feet but sometimes
6-8 feet.

I tried to discover a season for the appearance
of young shoots but could not. There seem to
be some new shoots all the time, even during the
dry season.

Woody shrub found from 7,000 feet upwards
but dominant on the saddle. It was at first
thought to be Erlangea tomentosa, but the latter
has smaller inflorescences and has never been
recorded above 7,000 feet. Two species were
found, possibly V. lasiopus and V. syringifolia,
having mauvish-white heads, the former some-
what larger.
Woody shrub with purplish spikes. Leaves and
growth similar to above.

Climber with light green, heart-shaped, leaves
and small white flowers, found in lower forest.

Herb about 2 feet tall with flowers like a forget-
me-not but very much larger and broader leaves.
Found in lower forest.

How Eaten
Collected. Outer fibrous portions of stem dis-
carded and soft inner parts eaten.

Collected and eaten in two ways: i. The whole
of the young shoot is eaten, the papery sheath
being discarded.
ii. Larger stems about an inch in diameter are
broken across, then split and the white pith,
which has the consistency of coconut, is eaten.
Collected, bark peeled off and inner portion of
stems eaten. Mostly eaten on saddle and just
above, where there is little choice.

As above.

Leaves are eaten, possibly on the spot, for the
plant is often found with a number of leaf stalks
about an inch long remaining. Sometimes the
plant is hauled down from a tree and in this
case a resting place is often found at the base
of the tree, showing that a meal has taken place.
The root is eaten and the rest of the plant dis-
carded. It is scooped out of the ground leaving
a characteristic round hole which helps to
identify a gorilla trail. Discarded plants are
found by each hole, indicating that the roots are
eaten on the move (Fig. 2).

Name Description
Anthriscus sylvestris Plant 12-18 inches tall with feathery leaves, like
(Umbelliferae) a cross between a fern and garden parsley. The
stalk bearing the inflorescence is much taller
(2-21 feet). The bases of the leaves form a sheath
like that of celery and the flavour is more like
celery than is Peucedanum. Found on upper
slopes and in sub-alpine zone.

How Eaten
The sheath-like leaf bases are eaten. They are
not pulled off separately, but the whole plant is
pulled up and the "heart" bitten. Probably eaten
on the move.

Bramble Similar to British type though restricted to a Young leaves, and very likely fruits also, are
(Rosaceae) few branches, not a large bush. eaten, leaving the whole branch bare. A favour-
ite tit-bit.
Droquetia iners Creeper with toothed, nettle-like leaves, but no Stems eaten.
(Urticaceae) sting.

Stephania abyssinica Climber with roundish leaves. No flowers were Leaves eaten.
var tomentella found.

Piper capense Climber with large dark-green roundish leaves Young stems are eaten, also the bark of older
with palmate venation and a greenish-white stems.

Kniphofia The same plant which is cultivated in gardens The succulent, almost fibreless stems are eaten
Red hot poker grows wild on the mountains, in clearings in during the flowering season from February to
bamboo forest, or Hypericum woodland. April. The heads are found scattered about,
indicating that the stems are eaten on the spot.

Giant lobelia Grows mainly in the alpine and sub-alpine zones, The bases of the leaves, which have a bitter,
but also on the saddle swamp with a few isolated peppery flavour like a chilli, are eaten.
plants on the lower slopes. The narrow leaves
form a rosette with the flowering spike growing
6-9 feet tall from the centre. The old, hollow
spikes contain water which gorilla are said to
drink, though no evidence of this was seen.
Senecio Found in alpine and sub-alpine zones only. Has
Giant groundsel a soft pulpy stem with clusters of large floppy Leaf bases on low plants are eaten.
(Compositae) leaves and yellow inflorescences which bloom
during the dry season.

Name Description How Eaten
Carex petitiana A rush which grows on the saddle swamp and is The soft white bases of the leaves are eaten and
much eaten by buffalo, the tough upper parts are laid in piles indicating
a sit-down meal.

Laportea alatapes Large and virulent type of stinging nettle found Tops eaten. Lower parts of stem and root seen
Nettle at all levels up to the tree line. eaten once.
Rumex nepalensis Perennial with large leaves growing 4-5 feet tall, Stem and leaf base eaten.
Giant dock mainly in lower forest.
Rumex usambarensis A type of bushy dock with thin pointed leaves Thick part of stem eaten.
and fruits a small red siliqua. Occurs as second-
ary growth in neglected cultivations.
Thistle A very tall variety growing 7-8 feet high often Stem and leaf eaten.
in large patches in clearings in lower forest.
Bracken Found in clearings in lower forest. On two occasions the stem was seen bitten but
not broken off, as if the animal had sucked the
juice and fibres.
Lactuca paradoxa Scrambler with narrow indented leaves and yel- Where these plants occur, numerous resting
(Compositae) low inflorescences, found 7-8,000 feet. places are usually found and wide areas are
beaten down, but no piles of remains are to be

Crassocephalum mannii Herb with yellow head found in lower forest.

??? Climber with deeply indented vine-like leaves Fruit was once found partly eaten on a gorilla
and a red cucumber-like fruit. track, but it cannot be definitely established that
it was eaten by gorillas.
??? Orange-coloured fungus about the size of a Often found broken up on gorilla trails, but no
grapefruit which grows on bamboo, proof that it is actually eaten.

Breaking down the results into regions we get
Place On ground Above ground Total
Saddle 47 14 61
Bamboo belt 10 23 33
Lower forest 64 67 131
The reason that most nests on the saddle are built on the ground is that
probably no bush is strong enough to support a fully grown gorilla's weight.
The few nests found on bushes on the saddle were babies' nests. Conversely,
in the bamboo forest, there is little space on the ground, so it is more convenient
to bend bamboo stems over in a platform and to make a nest on that. In the
lower forest there is a good deal of bamboo scattered throughout the other
vegetation and here the ratio of ground to platform nests is about 50:50. In
each group there will usually be some nests on the ground and some above,
though very often even those on the ground are built of bamboo.
It has been suggested that the leader of the group invariably sleeps on the
ground. This could be so, for no group of nests was seen without at least one on
the ground, and six groups of over three were seen with one ground nest only.
(b) Materials. Nests are built of any handy material. Bamboo, when available,
is a favourite. On the saddle, woody shrubs are generally chosen though some-
times nests are made amongst the rushes of the swamp. In the upper forest
woody shrubs again predominate. I only once saw nests built above the tree
line and they were very old ones near the top of Muhavura, made of giant
groundsel leaves.
The bamboo platforms are made by bending bamboo poles over at any
height from three to eight feet above the ground-within the reach of the
extended arm. The platform is then covered with leaves and branches, usually
of bamboo but occasionally of some shrub. Even in the lower forest, where a
wide variety of materials is available, most nests are made either from bamboo
or woody shrubs (Pychnostachys goetzenii or Vernonia spp.) though occasion-
ally the branches of some other type of tree may be used, or where the ground
is very soft and mossy, a handful of creepers may be pulled down to form a
Tree nests are rare. I was shown two old ones on the saddle (on Hypericum)
when I first arrived, but except for one on 23 March, I saw no more until Sep-
tember. Then, strangely enough, there was a spate of them--13 within a month.
As all were found in the same area, it is possible that they were built by one
(c) Siting. It has often been suggested that nests are built with a good view of
all approaches, or that when one of a group is built high up it serves as a
lookout. I did not see much evidence of this. Very often nests have no view at
all, and it seems more likely, judging from their size, that the higher nests are
not lookouts at all but are babies' nests, put out of harm's way. On a few occa-
sions nests were seen overlooking a ravine, where they would certainly have a
good view downwards, but since any attack would come from above, there
seems to be little object in this. In any case, the time of greatest danger would
be during the hours of darkness when a vantage point would be virtually useless.
(d) Frequency of building. It is said that a fresh nest is built each night, and

since the animals are constantly on the move and the nest is probably built
very quickly, this seems a plausible theory. When following a known group I
rarely found sufficient nests to account for the number of nights, e.g. on
31 March, following a group seen on 27 March, we found two lots of nests
instead of four. But this may be for a variety of reasons:
i. Nests are difficult to spot in thick forest and may be easily overlooked
when there are many side tracks, some of which may lead to nests while
others by-pass them.
ii. Natives admit that they sometimes cut nests down or cover ground nests
with branches to hide them, for they fear that the land may be taken away
and returned to Forest Reserve if it is realized how low the gorilla come
down the mountains.
iii. Recognizable nests may not always be built. Where the ground is soft
and covered with undergrowth, they may simply rest on the ground and
their droppings may be trampled by elephant or buffalo. I once saw
depressions in soft grass and moss, which would have been taken for
resting places except that they had dung in them. Such nests could be
very easily missed.
iv. Groups may scatter for the night more often than realized, so that the
group of nests one is looking for is missing, and the isolated nests are
overlooked. Scattering was observed on two occasions, and may perhaps
take place frequently.
There is no evidence that nests are used more than once. On the contrary:
i. If they were, one would expect to find two lots of droppings of different
ages in a nest. This was never seen.
ii. Nests of obviously different dates are sometimes found near together,
showing that although they have seen the old nests, the animals prefer to
build new ones.
iii. Groups usually move in a set direction, not a circle.
(e) Babies' nests may be recognized by the small-sized droppings in them.
The depression of the nest may be little more than 12 inches in diameter but it is
often very deep, so that the nest is like an inverted thimble. Very small nests are
often seen over six feet up on bamboo, indicating that they are built by the
parents. Four babies' nests seen on the saddle were six feet up in shrubs which
could not have supported an adult. Of two tree nests seen on the saddle, one was
very small and the other was about the size of a stork's nest. But a more
puzzling thing was seen in the lower forest. Here, six nests were seen 20-30 feet
up in trees, but only one was found below although a careful search was made.
This means that some adults must have slept in the tree nests. Possibly females
are sufficiently light to do so. The trees in this case were not large and the
branches, on which the nests were built, were not more than four inches in
diameter. Babies are said never to foul their nests. This would surely depend on
the age of the baby and on its diet. I saw only one clean baby's nest, though
there were many above eye level that I did not look into.
(f) Resting places are of various kinds. Some are as carefully built as nests
and only distinguishable from them by the absence of droppings. These are
thought to be used for resting in during the heat of the day, and are usually

found singly, as if used only by the leader. They are different from the feeding
type, which are depressions in the ground cover, perhaps at the foot of a tree,
with evidence of the meal nearby. Some resting places are found beneath a
curtain of creepers and may be used as shelters during a storm. On one occasion
I saw five distinct resting places made by placing bamboo leaves over a
carpet of nettles.
Though there is no evidence that nests are built as look-out places, it seems
probable that resting places are. Five such places were seen and in one case
the occupant saw us coming and gave a warning. Three were more puzzling,
since they were on bamboo platforms near groups of nests. The question then
arises, if they were occupied by a guard while the nests were being built, why
did he not remain for the night? A possible explanation may be that in spite of
the absence of droppings, they were actually nests. An authority(7) on the
plains gorilla states that the animals get up earlier during the dry season and so
do not foul their nests. The resting places in question were seen in July.

(a) Climbing. Gorillas were seen in trees four times, and evidence of fallen
trees and pulled down vegetation indicates that they climb fairly frequently
in search of creepers. As has been stated by various authorities, the reaction of
gorillas when startled is to escape not upwards but downwards.
The first gorilla I saw in a tree was a large male, about 30 feet up. As soon
as he spotted us he leaped down in three or four huge strides, with his back
to the trunk. It happened very quickly, but I have the impression that he did
not swing or use his arms to any extent, but leapt down in a more or less
upright position, like someone going downstairs three at a time. As soon as he
was on the ground he assumed the normal crouched position.
The second tree-climbing gorilla was also a male but this time he was on
quite a low branch, holding on to a creeper which he was obviously about to
pull down, when he saw us and jumped off. On two other occasions several
animals were climbing; one group was clambering around some fallen trees,
only a few feet above the ground and when seen they came slowly down. The
other group was about 20 feet up and unaware of our presence. I saw only one
clearly, a female who appeared to be sliding down with her arms and legs
wrapped round the trunk.
On one occasion I saw a gorilla standing upright with his arms on a branch,
watching us.
Altogether it would seem that gorillas spend perhaps 10 per cent of their
times in trees, either collecting creepers or building nests.
(b) Reaction to human beings. The normal reaction of gorillas is to retreat
from human beings, but the speed and manner of their retreat depend on a
number of factors including the character of the individual animal. These rules
seem to be fairly general:
i. If they recognize an intruder at a distance of 50 yards or more, gorillas
will slip silently away. They do not seem unduly disturbed and will
resume feeding as soon as sufficient distance has been put behind them.
On three occasions after making contact with a group, I heard them

within a few yards for some time afterwards, and once when I sat down to
have lunch, a gorilla from a group I had just seen came walking down the
path towards me!
ii. Between 20 and 50 yards the retreat will usually be accompanied by
noise and possibly a demonstration from the leader. There may be barks
or roars and perhaps the scream of a female, though on the whole I find
that females rarely make a noise unless disturbed away from the males.
As they retreat the males will stop every few yards to bark and eye the
intruder, making off when they are sure they are not being followed. The
demonstration by the leader consists of a stand of several minutes, during
which he roars aggressively, perhaps dancing about as if contemplating a
charge. Having allowed time for the family to escape, he will eventually
follow them. At any attempt to approach during this time, he will either
bolt or make a mock charge, depending on his individual courage and the
needs of his family, those with young ones in their care no doubt being
more aggressive than those without.
iii. To attempt to approach closer than 20 yards, once sighted, is to provoke
a mock charge. This consists of a rush forward followed by a pause to
watch the result. If the rush has not frightened the intruder away, the
animal will be puzzled and indignant and may charge again, this time
much closer. But the closer he gets, the less he likes it, and the rush will
be followed by a rapid retreat to a more comfortable spot.
Although these rules hold good in general, there is a variation in the reaction
of different groups and individuals, and the amount of noise in particular varies
a good deal. Some visitors I took into the sanctuary admit quite candidly that
they were shaking too much to attempt a photograph, whereas others, who have
heard only a short bark and glimpsed a fleeing form, assume more alarming
descriptions to be exaggerated. I found that large groups were less aggressive
than small, though possibly this was because the small group of three that I
saw contained a young baby. The male of this group was the most aggressive
individual that I met. Out of 30 occasions when gorillas were aware of my
presence, there were 8 demonstrations including two charges. Both charges
were made by this male as were four, and probably five, of the demonstrations.
Only twice did I contact this male without his making a stand. On both
occasions he was at the bottom of a steep slope with us at the top; and therefore
at a disadvantage. The more peaceful reaction of a large group, on the other
hand, may simply be due to the fact that the group is scattered and one does
not always make contact with the leader.
I had the impression that surprise was a strong factor in causing precipitate
flight, a group which had been comparatively calm one day, rushing off at speed
the next, if taken unawares. There are probably many occasions when gorillas
sense the presence of something strange long before they bark, and during
this time they prepare to escape, doing so later with less fuss than if surprised.
If they can see the intruder clearly they are much less worried; on the two
occasions when I saw gorillas in clearings, so that we had a good view of each
other, they walked slowly away, looking back, but did not run.
(c) Relationships with other animals. According to the local people gorillas

are afraid of elephant, buffalo, leopard and hyena. As far as the first two are
concerned, I saw no evidence that the vicinity of these animals was avoided.
On the contrary, both elephant and gorilla were present in the same part of the
lower forest in early April and again in September, both were present on the
saddle over Easter, buffalo and gorilla occupied the same area of the saddle
during May. Leopards are said to prey upon young gorillas and the frequent
building of babies' nests above the rest lends support to the statement. Hyenas
are also said to attack gorillas, not by killing but simply by biting live animals.
On two or three occasions we saw hyena tracks following gorilla tracks, though
whether this was deliberate or not one cannot say.
It seems certain that gorillas do not like dogs. A number of stories involving
attacks on men by gorillas show the attack to have been provoked by the
native's dog.
(d) Group relationships. It is not known exactly what constitutes a gorilla
group or what holds it together. Only a few sketchy surmises can be made:
Q. Were any particular groups recognized?
A. One family of three was followed for some days in February, again in
March and in June. It was distinguished by including an unusually large
male with a white stripe down his back. If it should be established that
whiteness increases with age this would indicate that he was young and
would account for the family being small. It was also characteristic of this
small group that the male's nest was often some distance from the others and
that he often made a large nest-type resting place. Other groups seen con-
sisted of 5, 6, 11 and 12 animals. Possibly the first two and the second two
were sometimes confused, but what does seem to be established is that a
large group has on several occasions split up into two or more smaller ones.
Q. What is the average proportion of males, females and young in a group?
A. One can estimate the young in a group from the number of nests but the
proportion of males to females is more difficult. From what I saw and heard
I judge that an average group consists of one old male and four adults of
which at least two are females, and one or two young ones. This was
estimated from the sizes of the nests and droppings and the animals actually
seen at one time for example:
i. Four females together in a group thought to number six and to contain
one young one.
ii. One male with a white stripe, presumably not the leader, in a group of
iii. Of five animals, all seen at once, four were together in single file, and some
way behind, facing us, was the leader. Of the four, the front one appeared
larger than the others and the rear one smaller. The two middle ones were
probably females and the front one might have been a larger female or a
young male, for I did not see any white on it.
Q. Is there usually more than one fully grown male in a group?
A. It would seem so,
i. see ii above;

ii. from the sounds made by a group-low pitched barks coming from
several directions-there would seem to be several males in some groups;
iii. males are encountered more often than females which one would not
expect if there was only one per group.
Q. If so, how long do the males remain in the group. Do they leave after mating?
A. Difficult to say, but the separate family of three supports the theory that
they break away from larger groups and form new groups of their own. On
the other hand, the very large groups met by early observers suggest the
Q. Were solitary males ever met with?
A. No.
Q. Do large groups divide and rejoin again?
A. This is suggested from the evidence:
i. A group of 12 nests was found on the saddle on 2 May, the guide giving
his opinion that it was made up of two groups of six. Two days before a
group of six had been found in the bamboo with tracks leading towards
the saddle, and some time later another group of six was found on the
ii. On 23 March a group of 10 or 11 nests were found. On 27 March a large
group was seen, and on 29 March the tracks were followed. Twelve nests,
all said by the guide to be of the same night, were found scattered up the
slope within about 1 mile of one another. They were in groups of 5, 3
and 4. On 10 April another group of 12 was found in the same area.
iii. On 17 July two groups of nests were found within a short distance of each
other, one of six and one of five, both of the same night. On 21 July,
having followed the same tracks, visitors saw 11 gorilla, these two groups
presumably having linked up. Following this group on subsequent days,
I found a group of four nests (night of 21/7) groups of 8 and 3 (night of
23 or 24/7) and a group of five (night of 28/7). This indicates that a
group does not always split up into the same component parts. If the
sub-group of a large group were family units, one would expect them
always to divide up in the same way. Perhaps for longer separations they
do. In the case noted the tracks joined up again each day, so that the
group as a whole kept in touch.
It is difficult at this stage to piece the picture together without involving
much guesswork.

1. Days spent in the Kisoro Gorilla Sanctuary during the eight months Febru-
ary to September 1957 numbered 122. Fresh tracks (not more than two days
old) were found on 99 days. Gorillas were encountered 41 times, of which
seven were accidental.
2. General observations of the gorillas confirmed statements of previous
observers as to their physical characteristics.

3. Contrary to expectation, gorillas were found to live in the lower forest,
between 7,500 and 9,500 feet for most of the period under observation.
There were signs that they visited altitudes of 10,000 feet and over during
May, June, July and early August but there is insufficient evidence to des-
cribe this as a definite seasonal migration. Changes of habitat do not appear
to be caused by food shortages, but may be due to the need for different
food, moister food, exercise or a change of climate.
4. The daily range while feeding in the lower forest is very small, probably
not more than half a mile a day. Some longer journeys, for 3-7 miles,
involving a slight change of altitude, were tracked.
5. No evidence was seen of gorilla groups guarding their own territories.

6. Gorillas feed largely by collecting foods and taking them to a resting place
to eat. Twenty-two food plants were collected and identified. The main
foods appear to be giant celery, bamboo shoots and woody shrubs, supple-
mented by a variety of herbs, creepers and a root.

7. Nests were found built in roughly equal numbers on the ground and above:
Total seen 225
On the ground 121
Above the ground 104
The majority of those on the saddle were on the ground, the majority in
bamboo were on platforms, while in the lower forest the number on and
above ground were about equal. Tree nests numbered 16, or 7 per cent of
the total. Higher nests are usually found to be those of babies, though
females may occasionally sleep in trees. Males are thought to prefer ground
8. The majority of nests are built from bamboo or woody shrubs; a few are
made from branches of other trees and a few, on soft ground, from a tangle
of creepers, ferns and moss. With the possible exception of small babies,
gorillas foul their nests.
9. There was no evidence that nests are used as lookouts or are built with a
good view of all approaches.
10. Insufficient nests were found to prove the theory that fresh ones are built
each night, but this lack of proof may be due to a variety of reasons.
11. Nest-type resting places, which are not fouled, are sometimes built for day-
time use. Less elaborate seats are used for meals. Some resting places have
the appearance of shelters.

12. Gorillas were observed to spend about 10 per cent of their time in trees,
either collecting creepers or building nests.
13. The normal reaction to human beings is retreat, but on eight occasions the
retreat of the group was accompanied by a demonstration on the part of the

leader. Four of these demonstrations, including two mock charges, were
made by one individual.
14. Surprise seems to increase their fear, as does lack of space in which to see
the intruder clearly.
15. There was no evidence that gorillas fear any animals other than man.
16. The proportion of young in a group seems to be between a fifth and a
quarter. It was not possible to find out the relationships within the group,
or at what stage new groups are formed.
17. Large groups sometimes divide and rejoin, but they do not always split into
the same component parts.

Results of human interference
18. Chance encounters with human beings do not seem to worry gorillas and
they seldom move far as a result. Persistent following on the other hand,
might both annoy and drive them away.
19. Cultivation inside the original sanctuary area has not had the effect of
driving the animals off; though, according to the guide, many more gorillas
were seen low down the mountain before the Forest Reserve boundary was
moved higher up.
20. It would be valuable to study the gorillas on Sabinio, where there is as yet
no human interference either through cultivation or wood cutting.

In conclusion I desire to express my indebtedness to Mr. Walter M.
Baumgartel, the host of Travellers' Rest, Kisoro, for accommodation and the
numerous other assistance that made this research possible, to Professor R. A.
Dart and Dr. P. V. Tobias for their interest throughout the project and for
seeing this paper through the press, to Mrs. M. E. de Beer for her preparation of
the maps, to the Wilkie Foundation for the camera used for the photographs, to
Miss V. de Wet for redrawing the graphs for publication and to any others whose
help I have failed to recall here.

1. Barns, T. A. The Wonderland of the Eastern Congo (Putnam, 1922).
2. -. Across the Great Craterland to the Congo (Ernest Benn, 1923).
3. -. An African Eldorado (Methuen, 1926).
4. Burbridge, B. Gorilla: Tracking and Capturing the Ape-man of Africa (Harrap,
5. Ionides, C. P., Senior Game Ranger, Tanganyika, who collected the gorilla for the
Coryndon Museum, Nairobi, in 1946.
6. Milton, O., former Game Ranger, Tanganyika, in Animal Kingdom, the magazine
of the New York Zoological Society, Vol. LX, No. 2, March-April 1957.
(The Last Stronghold of the Mountain Gorilla in East Africa.)
7. Blower, J., Game Ranger, Uganda, Oliver Milton and entries in the guide's visitors
8. Akeley, Derscheid, Blower, Baumgartel, etc.
9. Merfield, F. Gorillas were my Neighbours.


0 1-

.. Y
< z0 z
D z j

Av ;

o ^- \ -.

a ni .'

Map 1.-Sketch map to illustrate approximate location of the Uganda KISORo GORILLA
SANCTUARY and the roads thereto in relation to Lake Kivu and the international boundaries
of Uganda, Ruanda-Urundi and Belgian Congo.


Map 2.-Sketch map of the KISORO GORILLA SANCTUARY to illustrate sightings of the
gorillas, their nests and tracks on the different days of the eight months February to
September 1957. The figures in the circles give the date and number of the months, e.g.
6/9 represents 6 September 1957.


Diagram I.-Month by month results shown as a percentage of searches.
Upper levels: Tracks, interrupted line with dots; Contacts, continued line with dots.
Lower levels: Tracks, plain interrupted line; Contacts, plain continuous line.


Diagram II.-Approximate ratios of bamboo to celery eaten on lower slopes (bamboo
forest and lower forest) month by month as percentage of total food.

IN volume 8 of The Uganda Journal1 it was sought to reconstruct the circum-
stances of the death of Bishop Hannington on 29 October 1885. The Rev.
E. C. Dawson's standard Life of James Hannington appeared in November 1886,
only about nine months after the news of his death was confirmed in England.
In later editions there is reference to further evidence regarding the Bishop's
death which had been gleaned from some of his African followers. It gradually
emerged that apart from the four who brought the news to the Bishop's
African chaplain, the Rev. W. H. Jones, waiting at Kwa Sundu and who were
at first thought to be the only survivors, a number of others were spared or
escaped. When, in February 1886, the Rev. A. Downes Shaw, the C.M.S.
Secretary at Frere Town, examined these four (the names of Serenge of Rabai
and Phillip of Frere Town survive) he reached the conclusion that none of them
had witnessed the Bishop's death, one admitting that he had made his escape
before there was any attempt to kill the Bishop. Later two of them were
questioned in Zanzibar and thought that "about ten men were spared as being
useful slaves . some because they could read English, others masons, one
gunsmith, and the Bishop's boy, Almass, because the Wasoga could not unlock
the packages without him". (C.M.I. April 1886, p. 206). Indeed it may be that
of the Bishop's caravan of fifty men nearly twenty in some manner or other
Sir John Gray has reprinted from the archives of the Zanzibar Consulate
(U.J. 13 (1949), 1-22) certain letters, mostly from missionaries in Uganda, which
throw further light on the background and current understanding of events
before and following the Bishop's death. In a letter to Sir John Kirk, dated
Buganda, 26 December 1886 (loc. cit. p. 20) A. M. Mackay reports that he had
been in contact with some of those who escaped. Their recorded evidence and
other later information are to be found scattered through the files of the
Church Missionary Intelligencer. They give an impression which differs in some
details from the previously accepted story. It seems that while the Bishop was
confined and being closely guarded his followers were allowed to move about
freely in Luba's village; only when the messengers with their fatal instructions
returned from Mwanga were they secured. Most of the survivors claim that they
witnessed the fracas-one can picture the scuffling and shouting-in which the
Bishop and Pinto, stripped of their clothing, met their end. They do not refer
to any separate place of execution for the bulk of their companions; but this
may well be for the reason that those who escaped or were spared did not
reach the scene of the ultimate massacre. In all this it is right to recollect that
for events subsequent to the last brief entry in his journal, made by the Bishop
presumably on the morning of 29 October 1885 before being led out for
1 The Last Days of Bishop Hannington by H. B. Thomas, U.J., 8 (1940-1), 19-27.


execution, we are dependent entirely on African testimony, much of it at second-

In C.M.I., August 1887, is a long letter from Mackay to the Rev. R. P. Ashe
(then in England) despatched from Buganda early in January 1887. The follow-
ing extract (pp. 490-1) somewhat amplifies Mackay's report of 26 December
1886 to Sir John Kirk which is referred to above.

The Arab's dhow brought me a note in half-English from one Christopher Boston, a
mission boy, or as he calls himself 'missionary'. He was one of the Bishop's lads, and had,
with two others (Alimas and Kutu), made his way round the east side of the lake to Magu
in Usukuma. He begs cloth, etc., and medicine for a wound. These I have sent with
instructions to them to proceed to Msalala [sc. C.M.S. station, south of Mwanza, then
manned by Rev. E. C. Gordon and Charles Wise].
More recently a Mgwana [sc. coast man] who had also been of the Bishop's party,
arrived here [sc. Natete]. He was one of thirteen (or fourteen) who, he says, were spared
as being juniors. The King's order, he says was to kill them all, but the Basoga saved them
as slaves. His name is Hamis Turki. He gives the names of seven of his companions who
are still at Luba's in custody. Some five or six more, he says, have escaped since the
massacre, running to Wakoli's. Of these, Ibrahim, the Bishop's head-man was one. I am
sending to the Consul the names of all whom Hamis knows to have been killed, and of
those saved alive. I can get almost no fresh information out of him. Only he avows that
the Bishop and Pinto were the first to be killed. They were stripped first of all clothing and
speared. The others were some speared, some shot. He says that the Bishop kneeled down
and clasped his hands as if in prayer, and was speared in both sides at once when in that
attitude. I shall (D.V.) send this man to Msalala in the Eleanor. Poor fellow, he had no
easy task in escaping from Luba's and getting over the Nile in disguise. I shall be glad
when he is out of the country as he may be discovered here and murdered.

In due course Christopher Boston made his way to the coast. At Frere Town
he was examined about May 1887 by the Rev. A. D. Shaw. His statement is
printed at C.M.I., August 1887, pp. 493-6. It is the most detailed account we
Statement by Christopher Boston.
October 12th (1885). After leaving Mr. Jones and the bulk of the caravan at Kwa
Sundu we came to a place called Kwa Mtindi. Mtindi, the chief man of the place treated
us well and gave us a camping ground.
13th.-The next place we halted at was Kwa Mtunga: this man is a Sultani Mkubwa
(great king) and possesses a large dominion.
14th.-Mtunga said he would take us to the lake, so we started, he acting as guide.
We saw the lake like a big sea. We turned back and slept near Mtunga's village.
15th.-The next day we left early and walked a long way. We slept porini (in the desert).
16th.-On getting up we found we were near a banana-grove, which owing to the
darkness, we had not noticed last night. We went to it and found many people, the
Wakoli, but did not see the king.
17th.-We reached the lake and slept in a large town.
18th.-This was Sunday, so we remained here at the large town.
19th.-We left early with a guide. About noon we found the Sultan. He was with his
soldiers on a fighting expedition. He offered the Bishop pombe [beer] and wanted him to
encamp with the soldiers, but they were very drunk, and pushed us about, so the Bishop
chose a place in a banana grove, and there we stayed the night.
20th.-We marched till 4 p.m. when we came to a village, where Kisoma, the son of
Luba, ruled.
21st.-In the morning we left before the people were up and went along the main road.
We were soon overtaken by soldiers from Kisoma's. At 7 a.m. we arrived at Luba's place.
Here was a great baraza (council) on. The Bishop took his chair and went and sat down
near the baraza. After a time a man came forward and said he was the Sultan. The Bishop
would not believe him; then another man did the same, but with the same result. Then a


young man came and said "I am the king's son" but the Bishop said "No you are not."
Then the people were astonished and said "Why! what kind of man is this; he is a
stranger; he has never been here and we don't know him, yet he knows all about us?"
After a time a messenger came and demanded ten barrels of powder and ten guns. The
Bishop said "I cannot give you these: I have left my goods at Kavirondo." He sent one
barrel and a cloth, but these were returned. The Bishop then ordered the caravan to
proceed. We went on a short way, but the people surrounded us and clamoured for our
return. The Bishop at last consented to go back. On reaching the town the Sultan pre-
sented himself for an interview. He said "You must stay here and I will send a messenger
to Mwanga; if he says you are to proceed you shall go on; but if he says you are to go
back, then I shall send you back." To this the Bishop agreed. Then the Sultan showed us
a place to camp and said "You must stay here". The Bishop said "No; I want to stay
outside the town". He chose a place outside on the main road to Uganda. Here we camped.
The Bishop went off exploring, hoping he could find the road to Uganda and thus be
independent of the King's guides. On his return we told him that the people had stolen
seven guns. He said "Oh! I will soon put that to rights, as the king will make them restore
them." He took Ibramu, his head-man, and went off to see the king. After a time a
runaway Swahili, who was living here, came and said "Your master is taken prisoner, you
had better take your goods up to the town." Then there was a shauri (consultation) among
ourselves. "Shall we fight or shall we not fight?" Some said "Let us fight: we can escape";
others "if we fight they will kill our master. Even if we get back to the coast, people will
say 'Where is your master; what have you done with him?' So we determined not to
fight, but each took his load and we went back to the town. Here we were shown a large
house and we put all our loads into it. We all had our guns and our liberty. We walked
about the town and bought food in the market. We were allowed to go and see the Bishop
one by one, but could not go in a body. We took his food to him and if anyone was called
by the Bishop he could go and see him. We asked "What was our fault that we were thus
treated?" They said you have done no wrong, but it is such a strange thing for a European
to come to see Mwanga by this way; all the other Europeans go across the lake. The king
may not like your going this way, so you will be kept here till we hear from Mwanga what
you are to do."
28th.-The Bishop was kept a prisoner for seven days. We were all quite free to walk
about. We had our guns, and all the loads were left in our house: nothing was taken
away; only there was a soldier there always to see that we did not take anything away.
On the seventh day the messengers returned from Mwanga, and there was much firing of
guns. We asked what the news was and were told that Mwanga had refused to give us
permission to go on to Uganda, that we were to go back the way we came, and that on
the morrow we should start. We all slept well that night.
29th.-About 7 a.m. some soldiers came and began to bind us. Some of us struggled a
good deal, and then those who did, had their hands tied behind, and were put in wooden
slave collars, but those who submitted were only tied with their hands in front. Some
Waganda, whom we had not previously seen (they had come back with the messengers)
came and talked to us. They asked "Who gave you permission to come this way? You
have come without leave, and must return at once." About 2 p.m. the Sultan came to see
us: he had the Bishop's umbrella in his hand, and when it rained he put it up. He divided
us among his soldiers putting one of us to two soldiers, and then we were taken away,
each one to the soldier's house who had charge of us. At 3 p.m. we were brought out
and put together in a line and marched off, taking a road leading in the way by which we
had come. Before leaving the houses our guards had taken away our clothes, and gave us
pieces of bark-cloth to wrap round our loins. We were marched a long way-it took us
more than two hours to reach the spot where we halted. Shortly before reaching that place
we saw in front of us the Bishop and his boy Ikutu, who carried his chair; they were
surrounded by a great many soldiers. Pinto, the Bishop's cook, was with us, with his
hands tied behind him. We came to a place where there were many trees on one side and
a valley on the other. Here the Bishop was with the soldiers. We stopped within a few
yards of where he stood and could see him quite plainly. He tried to sit down but the
soldiers would not let him. They began to pull his clothes off him. They took away all his
clothes and left him naked, with only his boots on. This they did for they wanted hiN
clothes. Then most of the soldiers left the Bishop and came and stood near us. Suddenly
a gun was fired off as a signal: then two soldiers who were standing on either side of the
Bishop stabbed him in his sides with their spears, and he fell down on his back.
Mr. Shaw proceeds:
Here all was lost to the poor fellow. He said "When I saw the Bishop stabbed I
trembled because I knew they would kill me." The soldiers set on the forty-odd helpless

men with great ferocity, and he was stabbed in the side and fell down as dead. Seeing him
fall, the man who struck must have imagined him dead, and left him to attack another.
Christopher said: "All was dark to me until I woke up about 3 a.m. the following
morning. It was very cold and the cold had revived me. I sat up and found that my bowels
were coming out of a wound in my side. I picked a broad plantain-leaf, and having put
my bowels in, I bound the leaf over the wound to keep them inside. I saw my dead
comrades round, but did not look for the Bishop's body. I was too ill to do that. I could
not stand up, but crawled away in the direction of the country where we left Mr. Jones.
I knew that three of our people had not been brought to the slaughter-Ibramu,
Mukanyaga and Hamisi Turki-but I thought that all the others were dead. I struggled
along for some days. I saw many people: the women pitied me, and gave me food. Of
the men, some said "Kill him;" others "No, we don't want his blood in our land: let him
go on, and he will die somewhere else." I found a man who was kind to me and let me
live with him for a time; he knew all about the murder of our people, and used to tell
me lots of things. I asked why they killed the Bishop with spears, and did not shoot him.
He said "Ah! the gun is the weapon of the white man, they make it, and they know
what charm to use so that it will not kill them, so that it would be no use trying to kill
him with a gun; but the spear is our weapon and the European has no charm against
it, therefore, the white man was killed with the spear."
After many days-I do not know how many-I reached Kaunyi, where the Bishop
slept on the 19th: here I was treated very kindly and stayed some time. After several
days two of my comrades came-their names were Ikutu and Almasi; they said that
just as the soldiers were attacking the party a messenger came from the king saying that
some young men were to be saved, therefore ten were snatched away from the party and
the rest were killed: of these ten were Ikutu and Almasi. They said they made their escape
and came here, hoping to get to Mr. Jones. I heard that Ibramu was given the young men
who were saved, and he was told to make guns. He said he would, and they gave him
ground and a house. After a short time Ibramu ran away and they did the same a few
days after.
We stayed at this place many days, and then a Swahili man who was hunting came
across the lake in a canoe. He was a friend of Mackay and I asked him to take me across
the lake in a canoe ... we landed on the south side of the lake, and then I sent a letter to
Mr. Mackay and told him where I was. He sent me some cloth and told me to go to
Mr. Gordon. [sc. The Rev. E. C. Gordon then at Msalala.] I went to Mr. Gordon and
after eight days went on to Uyui [C.M.S. station near Tabora]. There I found Mr. Taylor
[the Rev. W. E. Taylor of C.M.S.] and came with him down to the coast, and so on here.
Mr. Shaw adds:
May 24th 1887.-The narrative was taken down by me from the man himself,
Christopher Boston. He knows a very little English but spoke Kiswahili. He is one of our
Frere Town boys, having years ago been released from a slave dhow, trained in our
schools, and when of age, sent out to earn his living. When we were commencing our
work at Taita, there was a need for promising lads to go up to assist Mr. Wray.
Christopher was one of them. When I went to Taita to visit Mr. Wray I found that
Christopher was making himself very useful in many ways. When Bishop Hannington
was making up his caravan to go to Uganda, Christopher volunteered to go. Of his
thrilling adventures and wonderful escape, the above is a very brief outline, but it is
sufficient to show that God's loving care is extended to black as well as white and that ...
He can bring us safely through all, as He did His young servant, Christopher.
Our dear friend the Bishop, therefore, was not shot with his own rifle, as we had sup-
posed, but was speared like his followers. The touching incident of his being on his knees
in prayer at the time is not mentioned by Christopher, but only by the coastman examined
by Mr. Mackay.

It is not clear whether Ukutu, the Bishop's chair-bearer, went down to the
coast after he turned up in Usukuma. He seems to have attached himself to the
Rev. J. Blackburn (who died at Usambiro, 12 March 1888) for his 'boy Ikutu'
is noted as assisting at his burial by Bishop Parker (C.M.I., July 1888, p. 440).
A letter from Bishop Parker (who also died at Usambiro a fortnight later) is
quoted in The Last Journals of Bishop Hannington, p. 238-
Ukutu, who was with Bishop Hannington constantly during his imprisonment, and
undid his hands when they bound him to lead him off to the spot where he was murdered,

told us that as the Bishop walked to the spot he was singing hymns nearly all the way.
As they were in English, he did not know their meaning; but he noticed that in them the
word JESUS came very frequently.

One of the messengers who passed between Mwanga and F. J. Jackson at
Mumias in 1889-90 was Marko. At p. 717 of C.M.I., October 1890 is an account
of the recovery of Bishop Hannington's remains written by the Rev. R. H.
Walker from Buganda about March 1890.
When Bishop Hannington was murdered, the same day his body was carried to another
place, because the people feared that the dead body of a white man might bring evil to
them. But the people of the next place refused to have it; so it was carried from place to
place, each refusing to allow it to remain in their country. A coastman, who we understand
was one of Bishop Hannington's porters, accompanied the corpse. At last it reached a
place on the boundary of Busoga, or in the country of the Bakeddi. Here they agreed
to build a house for it, and on a framework, or bedstead, such as they make for smoking
meat and fish on, the body was laid, and left to decay. An agreement was made with the
coastman to live at this house and to take care of it, and in return the people would
give him food.
To this place Marko came on his way with letters to Mr. Jackson of the Imperial
British East Africa Company. He seems to have heard that the people there had
experienced bad harvests and drought of late years, and that they attributed this to the
fact of their having the white man's bones; and he suggested that he would get rid of the
bones by taking them to the white man. He passed the place twice, and I fancy it was on
the second journey that he took the bones to Mr. Jackson. The above is the substance of
what Marko says. Some of the details may be incorrect.2
Sir Frederick Jackson has related in detail the circumstances of the arrival
and temporary burial of Bishop Hannington's remains at Mumias. (Early Days
in East Africa, pp. 76-7.) The manuscript diary of his companion, Ernest Gedge,
which it has recently been possible to consult, fills in the story.
1st December 1889. Messengers came in from Tungu. They say the Bishop's bones will
arrive here tomorrow, also one of Kwa Sundu's men who was taken prisoner at the time
of the disturbance.
5th March 1890. In the afternoon some relics of Bishop Hannington turned up. An india-
rubber water bottle, etc. and the skull-poor fellow. Mumia does not want to have them
in his village fearing it will bring bad luck to the crops.
6th March 1890. The boy who was with Bishop Hannington at the time of his death
and who brought in the remains yesterday told us further particulars about the murder.
They are almost identical with those recorded in the 'Memoirs' [sic. this may refer to
Dawson's Life]. His skull was easily identified by some stopping in one of the upper teeth.
Curiously enough though nearly all the teeth were missing this one had remained.
From the above it seems that the existence of the remains in the custody of
Tunga became known to Jackson and Gedge before they set out for Elgon and
the Turkwell River: but that Tunga would not then hand them over. But,
following Jackson's return from Elgon, they were brought into Mumias on
5 March (actually six days before, not, as Jackson says, "the day before", they
left for Buganda).

2 Marco comes to life in Carl Peters' New Light on Dark Africa (1891). On 9 February
1890 he made himself known to Peters near Wakoli's, Busoga, where he was perhaps
awaiting further letters from Buganda. For, though Peters was hurrying towards Mengo,
Marco was still on hand four days later-"On February 13th .. suddenly some Waganda
handed me four letters addressed to gentlemen of the English expedition. I saw H. M.
Stanley's name as the sender of one. Marco tore it open. I then read Stanley's letter of
4 September 1889 from C.M. Station, Makolo, announcing his arrival there with Emin
on 28th August, 1889." Marco spoke Swahili.


Extracts from the Rev. John Roscoe's journal from Mengo are printed in
C.M.I., June 1892. A large caravan of the I.B.E.A. Company had arrived at
Kampala on 9 December 1891, and Roscoe records (p. 422):

11th December. Mr. Martin, the Company's caravan leader told Pilkington last night
that Bishop Hannington's bones are at a place fourteen days' journey from here. He wishes
to know what we wish done with them. The boy who was found guarding them is with
him, and is to come over here to see us. I mean to try and get him for my servant.
12th December. At noon Pilkington brought the boy who was with Bishop Hannington
when he was murdered, and who subsequently tried to carry some of his bones to the
coast. He said he belonged to Mumia of Kavirondo; his name was Mcharo, but now he
goes by the name of Bishop. His account of the murder coincides with the facts already
published. After the death of the Bishop, Luba, Chief of Usoga, kept him as a slave.
For two years he had to work in the fields, etc., but in his third year he was made a
soldier. At the end of his third year there was a tremendous famine. Luba asked the
medicine-man the cause. He attributed it to the fact of the Bishop's bones having never
been buried. The custom of the Basoga is, if a chief dies when on a journey or away from
home to cut off his head and bring it back to his country. I found this out when at
Frere Town. A chief in charge of a Busoga caravan died, his followers cut off his head
to bring back to his home. The chief, Luba, thereupon told Mcharo to take the Bishop's
head and an arm-bone, also the head of his Goanese cook, and carry them to the coast,
that his country might be saved from famine. The boy got the skulls, etc. Luba also gave
him Hannington's water-bottle and a pair of boots to take down to the coast. When
Mcharo reached his own town in Kavirondo the chief refused to allow him inside the
town. He was permitted to build a house outside the town and there he lived for two years.
The bones he wrapped in fibre from the banana and buried in his house, awaiting the time
when he could go to the coast. There it was he was found by Mr. Jackson. Mcharo
handed the bones over to him, and says they were to be taken to the coast. Mr. Martin
affirms they are still at Mumia buried in an iron box. The rest of the bones are still lying
where they fell. Captain Williams has promised to have them collected and kept in a
place of safety until we receive instructions where they are to be buried.3

The account of the recovery of the remains on 9 December 1892 as given by
Bishop Tucker in volume I of Eighteen Years in Uganda and East Africa,
(pp. 216-9) has been 'written up' from his more detailed contemporary journal
which is printed in C.M.I., April 1893, pp. 272-3. He relates how Mumia
affected ignorance of the whereabouts of the box: but that a young man "who
had been with Bishop Hannington on his journey to Busoga-in fact he had
acted as his guide", furtively provided the clue which led to a ruined hut under
the floor of which the box was found buried. Before leaving Mumia's, Bishop
Tucker took down a statement from this young man or lad (C.M.I., April 1893,
p. 276).

Statement by Bishopo, the lad who guided Bishop Hannington into Busoga.
Taken down in December 1892, by Bishop Tucker at Mumia's village, Kavirondo.
They first of all went to Tindi's village, then to Tunga's, thirdly to Samia, then they
entered Busoga at Hanaahola, then on to Bugalo. There was fighting there, so turned off
the road; then to Hjayah, then to Luba's. Remained there five days. All were then seized.
When the Bishop was taken, four lads went and took down the tent and put it up where
he was. On the sixth day, the Waganda came with the message. Early in the morning they
were called. They went, but the Bishop did not come. They were told to bring the guns.
They refused, and said they were only children, and could not without permission of the
Bwana (master). They went to the Bishop and told him. He said "Give the guns; we can
do nothing." The Wasoga then asked for knives: they took them all. They were then all
called together. When they were brought together he (the boy) was told to remain with the
3 James Martin was with Jackson and Gedge when the Bishop's remains were secretly
buried at Mumias in March 1890. Regarding the 'Busoga caravan', see note 5 post.

goods. They were taken to the Sultan. He could hear the noise of them all being seized.
A woman said he was to go and look after her goats. He went for five or six days, and
after that he was hidden for very many days.
The Bishop was killed about thirty or forty yards from where he was standing. He saw
it all. The Bishop was first of all struck in the lower part of the body with the point of
the spear, and then the iron end of it was driven into his throat. The body was left
lying on the ground until it got quite dry. Then there came great hunger on the land. The
people went to the medicine-men who said, "The remains of the white men being in the
country is the cause of the famine." The people then gave him the remains, and he was
sent away. He remained two years in Tunga's country, and had the bones all the while.
How could he leave them and throw them away, when they had been given into his hand?
His brothers would come some day and ask him for them, and how could he answer them
if he had lost them?

What can be pieced together regarding Mcharo alias Bishopo?
Even before Hannington's forerunner Joseph Thomson in 1883 reached Kwa
Sundu (Mumia's) which was Mcharo's home, Swahili caravans were a not
unfamiliar feature in Kavirondo. Like many another young African, Mcharo
had perhaps been seized by wanderlust: his name (coast Swahili, charo 'the
caravanner') suggests that he may already have accompanied one of these
Swahili caravans. He had probably picked up some Swahili, and would thus
have a useful qualification to offer to Bishop Hannington. How old was he in
1885? He was "a small Kavirondo boy" when he appeared before Jackson in
March, 1890 (Early Days, p. 76) while Gedge refers to him as "the boy who was
with Bishop Hannington". He was still "a boy" when Roscoe met him on
12 December 1891. Bishop Tucker, on 9 December 1892, describes him both as
"the lad" and as "the young man". It can be suggested that he may have been
ten years old in 1885. At that age many a toto is highly intelligent, and he would
be serviceable to Hannington as much as an interpreter as a guide.

Of Mcharo Bishopo there seems to be no later record. But explorers of the
by-ways of history at this period may be confused by the appearance of what is,
I am now satisfied, another contemporary 'Bishop'.
Jackson and Ernest Gedge had in their caravan a number of men recruited
from the C.M.S. freed slave colony at Rabai. It is probable that one of them
would be Cuthbert, whose name first appears in Gedge's journal in July 1890.
Gedge had been alone at Mengo since Jackson started his return to the coast
on 14 May 1890. When in August Gedge set out to visit the south of Lake
Victoria, Cuthbert was left in charge of his stores. A holograph letter from
Cuthbert dated 4 October 1890, which is preserved among Gedge's papers,
shows that he had a smattering of English such as might have been acquired
from the missionaries at the coast and sufficient to convey his news to his master
in Usukuma.
Cuthbert "who spoke Kiganda well" (Lugard. Rise of Our East African
Empire, ii, 82) remained at Mengo after Gedge left for the coast in January 1891.
When recently, on an examination of the proceedings of Macdonald's enquiry
into the Uganda disturbances convened in December 1892, it was noted that one
Cuthbert Bishop was sworn as 'Kiganda-English' interpreter it was natural to
wonder if this was Mcharo in a new role. But such a surmise can be dismissed.
For "the Kavirondo lad" who was at much the same time in contact with

Bishop Tucker at Mumias can have had no opportunity of learning English.
Cuthbert Bishop, the interpreter, is clearly Gedge's Cuthbert. How he came to
assume the name of Bishop is likely to remain unresolved; though he may well
have had some earlier contact with Bishop Tucker, who spent most of January
1891 at Mengo.

What happened to Hannington's and Pinto's bones during the earlier part of
the nearly four and a half years between October 1885 and March 1890? This
question has already been posed in U.J., 8 (1940-1), 26.
It is unrealistic to accept, as do some accounts that they were left untended at
the place of execution and were retrieved only after an interval of some two
years; for so left, they would surely have been scattered and crunched to
fragments by hyenas within a few days.
Clearly they must have been collected without delay and of the various
accounts that related to Roscoe (already an experienced Africanist) by Mcharo,
coupled with that of Marko, an intelligent and independent witness, may well
comprise much of the truth. There is not a little evidence that we have here an
example of the common burial customs among a primitive people.
Lionel Decle, the French traveller, passing through Busoga in February 1893
records (Three Years in Savage Africa, 1898, pp. 461-2):
Lubwa gave me some curious information about the Government of Usoga . .After
the dead chief has been buried for some time his bones are dug up again. Anyone who has
the misfortune to pass near the tomb while the operation is going on-man, woman or
child-must die. The skull and the larger bones are put aside while the smaller ones are
taken away to be used as drumsticks. When the chief's bones have been disinterred they
are laid in a hut built especially for the purpose, upon a bed, and covered with fine bark-
cloth; then the drums are beaten with the small bones, and all chant that their chief has
come back again. The reigning chief appoints a certain number of people to watch the
hut and the bones.
This corroborates Roscoe's account (p. 34 ante) of the preservation of the
remains of a chief. Criticism has on occasions been levelled against some of
Roscoe's anthropological statements regarding tribes other than the Baganda;
but his account of the burial customs of the Basoga in The Northern Bantu
(p. 227-8) certainly merits credence, for he was stationed in Busoga in 1892 and
for a while resided at Luba's village.
Very much to the point also is Sir James Frazer Fear of the Dead in Primitive
Religion, 1933 (vol. i, pp. 59-60, 127-9). More recently Yekonyia K. Lubogo in
U.J., 2 (1934-5), 120-32 and 255 describes Basoga Death and Burial Rites, and
though he says nothing about removing the skull or the jaw-bone he does
emphasize the importance of propitiating a dead man's spirit lest it wreak
vengeance on the living.

By way of postscript one other of Bishop Hannington's servants, who did not
however accompany him on his last journey, may be salved from oblivion.
Robert Livingstone was a rescued slave cared for by the C.M.S. at Frere
Town where he received some education. He comes to notice as Hannington's
servant on his expedition to Taita and Mochi in March-April 1885 where

Hannington jotted in his note book "Made my boys, Robert Livingstone and
Legh Richmond, wash, giving them a lesson in the art"; but there is no evidence
that they accompanied the Bishop when he set off for Kavirondo and Uganda in
He was with Bishop Parker's caravan which left the coast for Lake Victoria
by the southern route in June 1887 and was at the Bishop's bedside when he
died at Usambiro on 26 March 1888. In August he was back at the Coast as a
teacher at Frere Town (C.M.I., October 1888, p. 672), and in November he was
helping the Rev. A. G. Smith to found a new C.M.S. station at Mbungu some
23 miles north-west of Rabai. But he gave up the idea of becoming a minister
and was engaged as interpreter by J. R. W. Pigott of the I.B.E.A. Company
(later the Company's last Administrator at Mombasa), and was with him on the
expedition to the River Tana and Ukamba in February to July 1889 and later
at Malindi. For a while, he learned blacksmithing in the Company's workshops.
He joined Bishop Tucker in September 1890 for his journey to Uganda. For
some alleged misdemeanor he was discharged from the caravan and returned
to Mombasa, but Pigott had a high opinion of him and recommended him as
interpreter to C. W. Hobley. He accompanied Hobley to Taita and Kiliman-
jaro, and towards the end of 1894 went on to Uganda with Hobley who was
stationed at Mumias. He remained in the service of the Uganda Government
ultimately becoming a clerk in the Kampala office. In January 1900 he assisted
in the translation of submissions made by the Regents to Sir Harry Johnston
during the earlier phases of the Uganda Agreement negotiations. He left
Kampala for leave in Mombasa in May 1900 and was travelling through the
Nandi country when his caravan was attacked and he was killed. This was one
of the outrages which provoked the Third Nandi Expedition launched by
Colonel Evatt in July 1900.
Robert Livingstone's story4 is typical of the adventures, hardships and often
dangers which were the common lot of so many of the humble African followers
who made possible the achievements of the explorers of eastern Africa 50 to 100
years ago.
4 A conflation of (i) a note in Mengo Notes for August 1900 reprinted in C.M.I.,
January 1901, p. 39; and (ii) a letter from J. R. W. Pigott, dated H.B.M. Consulate,
Paramaribo, 27th January, 1901, printed in C.M.I., April 1901, p. 327. With Captain
Claude Sitwell's caravan up-country in 1895 was a certain 'Robert' (Sitwell's Diary:
18 July, "Gave Robert some tobacco"; 25 July, "Robert had pains"). This could well be
5 In April 1891 Martin, who had lately brought up a caravan from the Coast, left Mengo
with Lugard's orders to recruit porters in Busoga and hasten to Mombasa to bring up
ammunition. He joined up with Captain Eric Smith and arrived in Mombasa with a
considerable party of Basoga on 16 July 1891.
Roscoe was hurrying out from England to overtake Ashe's party of C.M.S. missionaries
which had already started for Uganda by the southern (German) route. It was doubtless
this caravan of Basoga which Roscoe encountered at Frere Town. As has been indicated
Martin, travelling by the northern route reached Mengo again in the following December,
Roscoe having arrived from the south of the Lake just a fortnight before (28 November

ONE of the best-known episodes in the traditional history of Buganda is the
story of Kimera, the third reigning member of the dynasty. It may be
recalled that after the disappearance of Kintu his successor Cwa quarrelled
with his young son, Kalemera, and banished him to the court of king Winyi of
Bunyoro. While there he seduced his host's wife, Wanyana, who in due course
gave birth to a son. This child, who was called Kimera, was exposed in a clay-
pit, but was rescued by a potter and brought up as the son of Wanyana's servant,
Mulegeya. After he had reached manhood, however, he was summoned by the
Baganda chiefs to take possession of the kingdom, Cwa having in the meantime
disappeared and Kalemera having died without other issue. He accordingly set
out with his mother and a number of companions, and when he presented
himself in Busiro the regent Sebwana fled before him.
It was suggested by Sir John Gray (1935) that this story conceals the conquest
of Buganda by a new ruling group of Bito origin, the coming of Kimera from
Bunyoro being a historical fact but his Ganda parentage a patriotic fiction.
This theory has great plausibility: the story undoubtedly is very suspect, and
there are good grounds for believing that the dynasty is of Bito origin. If I
venture to dispute it, the reason is that, to supplement the somewhat bald
narrative of Sir Apolo Kagwa (1901), we now have from M. B. Nsimbi (1956,
pp. 38ff.) a fuller account, which, without contradicting Kagwa on any point,
enables us to see the matter in a very different light. This is how he describes
the arrival of Kimera in Buganda. (My translation.)
"Kimera and his people camped a short distance from Ganda, where Naku
and Sebwana lived. And having camped they sent messengers to say to
Sebwana: 'A stranger has come from Bunyoro, he is yonder' ... So Sebwana,
learning that he had a distinguished visitor, bestirred himself and sent mes-
sengers in all directions to seek guest-gifts of every kind, that he might entertain
his guest. He was not aware that the stranger for whom he was collecting
presents had come to deprive him of the kingdom. It was the lady Naku who was
the leader of the conspiracy to wrest the kingdom from Sebwana. The plan
which she devised was this: to deceive the messengers, saying, 'When he sends
you out, do not return.' So, seeing that his men had not come back, Sebwana
left the place himself. Whence came the saying,
'They whom you send do not return.
Go yourself and follow after.'
So he went, and quickly found beer for his guest, and poured it into jars, and
he found men to carry them, and they went ahead of him. But as soon as he
had stepped off the royal Carpet, Naku had sent quickly to Katumba, telling
him to bring Kimera thither. And Kimera stepped on to the Carpet, wearing a
calfskin, for that was the clothing he had brought from Bunyoro. And as
Sebwana came back he heard the drums of the kingdom sounding, and the
people immediately told him:

'The stranger for whom you have brought beer is the grandson of Cwa
He has come to inherit the throne of his forefathers.'
And when Sebwana heard these words, he cast down the jar which he held,
and fled."
In my judgment this does not read like the narrative of a historical event. It
is the outline of a play. The stage is set before us, with the royal Carpet at its
centre, and the leading actors, prompted by the chant of the chorus, make their
ceremonial entrances and exits. But what was it that was enacted on this stage?
The first clue is to be found in the name of the protagonist. It is difficult not
to derive 'Kimera' from the verbal -mera, which denotes the springing up of
vegetation. Then there is the tradition that his mother brought with her from
Bunyoro a stock of finger millet and planted it in a field which was ever after-
wards religiously resown. To clinch the matter we may refer to an earlier
incident in his career. As a young boy, M. B. Nsimbi tells us, still unknown to
the mother from whom he had been separated at birth, Kimera allowed the
cattle he was herding to stray into her millet-field. He followed them into the
corn and was there discovered by his mother's maidens. The simple symbolism
of the boy found in the cornfield hardly calls for explanation. We are in the
presence of the oldest of agricultural rites, the dance of the women who greet, or
more probably summon, the young green shoots, fusing into a single emotion
their desire for the growth of the crop which they have planted and for the
growth of the sons whom they have brought to birth. 'Kimera' started life as a
cry of joy or hope.
When the girls reported to their mistress the straying of the cattle and the
finding of Kimera, her only comment was: "The millet is Wanyana's and the
cattle are Wanyana's also. Leave them alone, let them eat." Strange words for
the owner of a millet-field, but appropriate to her-for was she not herself
Wanyana, the Heifer, kinswoman of Hathor and of 'cow-faced holy Hera',
goddess of fertility in plant and beast and man?
When the springing corn has been conceived of as a growing boy, and still
more when it has come to be impersonated by an actual human youth, it is
natural for people to conclude that as it comes to maturity it, or he, must pass
through manhood-initiation rites.' Such rites are generally of much greater
importance than any which attend the birth of a child. For, among other
reasons, just as the risk that the seed will fail to germinate is much less than the
risk that the grain will fail to ripen, so, in primitive societies, in which the
majority of children die, it is only when a youth has come to manhood that a
new member can really be said to have been added to the group.
In Buganda, of course, boys were not ordinarily initiated. But the king was,
in the complex ceremony called okukula kwa Kabaka, the growing up of the
king.2 He spent a night alone in a hut, played childish games for the last time,
was instructed by Naku in the mysteries of marriage; and the proceedings
closed with a solemn enunciation of the ancient ban, which lies at the heart of
manhood-initiation and of human society, the condition of co-operation between
1 Cf. Harrison (1912).
2 Kagwa (1905).

father and adult son:3 "Kakano okuze, era tolabanga ku nyoko akuzala-Now
thou are grown, never again shalt thou look upon thy mother."
The origin of these royal ceremonies was not ascribed to Kimera but to his
successor Tembo, who probably stood for the young plantain grove, ntembo.
There are nevertheless indications that Kimera was subjected to a similar
process. On his way from Bunyoro he parted from his mother. The next night
he spent at 'Temangalo', or 'Chop-finger'4-that is to say, he suffered an
operation which was an ancient and widespread feature of initiation rites. The
culmination of the process is set before us with the greatest clarity in the scene
described above. Kimera has come to manhood, and the time has come for
Sebwana to yield his place. But he cannot, of course, simply depart. The drama
of supersession must first be enacted. Naku, who has transferred her affections
to the young prince, tricks Sebwana into leaving the Carpet unguarded. Kimera
steps forth in his regal calfskin, the drums roll, the chorus hails him king.
Sebwana comes back at the tail of a procession carrying jars of beer, then turns
in flight, dropping his jar. The grail is shattered, the new reign has begun.
Kimera has come to his people; the fear of famine is over for another season;
the days of feasting and dancing can begin.
I suggest, in short, that Kimera was not a person but a persona, the hero of
a perpetually repeated ritual drama which symbolized, and was designed to
ensure, the perpetual recurrence of life. This drama was far older than the
kingdom of Buganda in its present form; it was probably already old when the
first Pharaoh ruled over the two lands of Egypt.
One question remains to be answered: who was Sebwana? The answer is
probably again contained in the name-Chief of Infants. The Greeks told a
somewhat similar story, in which the infant Dionysus was first set upon the
throne and then tricked by Hera into leaving it.5 Thus Sebwana was, I think,
originally a child, who had to give way to the full-grown Kimera because he
represented the childhood which Kimera had left behind.
To this interpretation of the story of Kimera certain objections might be
raised. In the first place, it is hardly to be questioned that at some stage the
country that is now Buganda fell under the dominion of an alien group, the
Bito, who came proximately from Bunyoro and ultimately from beyond the
Nile. It is improbable that this event should not have figured, directly or
obliquely, in the traditions of the people, but to the incident which has
generally been taken to refer to this conquest I have assigned an altogether
different meaning. If Kimera was not the historical founder of the present
dynasty, who was? The answer, I suggest, is Mulondo.
According to the accepted tradition, Nakibinge, eighth in the list of kings, led
a heroic resistance to an invading army of Banyoro, by whom he was eventu-
ally defeated and slain. It is true that tradition goes on to claim that despite this
disaster the dynastic continuity of the kingdom remained unbroken, Nakibinge
being succeeded by his son Mulondo, who was still a child. Indeed there is no
further mention of the Banyoro in Kagwa's narrative until the reign of Kamanya,
3 Cf. Coon (1955), p. 66.
4 Kagwa (1901), p. 9.
5 Guthrie (1935), p. 108.

nine generations later. This fact is in itself somewhat suspicious. It seems as
though Baganda tradition-makers may have behaved rather like Roman his-
torians, who made much of the valour of Horatius but omitted to mention that
his resistance was unsuccessful and that Rome long remained a Tuscan vassal
state. There are certain indications that Mulondo was indeed the founder of a
new dynasty: he was the first of a long line of kings to be buried at Gombe, and
it was for him that the Namulondo throne was made. We may note also that
'Mulondo' was later the title of the chief of Bulondoganyi, in south Bugerere
and north-east Kyagwe. That it was from this direction that the invaders came
is suggested by the fact that they are said to have included Basoga as well as
Banyoro. Now Kagwa tells us that before the outbreak of war Nakibinge con-
cealed his wives and children at Kyajinja in the north of Kyagwe, right on the
line of march from Bulondoganyi to central Buganda. This is the kind of story
that would have been invented by those who could not deny that Mulondo
entered the country from the north-east but wished to assert that he was never-
theless a native prince. In reality he was probably the leader of one of the Bito
bands which had founded or were in the process of founding the principalities
of northern Busoga, and his particular band turned south-westward to invade
There is a second difficulty, the solution of which opens up wider perspec-
tives. Corn-rites and corn-kings are of course singularly inappropriate to
Buganda. It is possible that there was once a time when cereals in general and
finger-millet in particular played a larger part in the economy than they have
played in recent centuries. But even if this were so it would be hard to believe
that the emergence of the young green shoots could have here seemed
sufficiently dramatic to have served as the focus of ritual and myth. The natural
home of Kimera would have been in a region of dry-season desert such as the
southern Sudan. It seems therefore possible to conjecture that he was brought
southward from an earlier home on the middle reaches of the Nile, and that in
the unfavourable equatorial climate he languished and eventually died. For by
the nineteenth century, and probably far earlier, the drama of his accession had
ceased to be performed, and the associated myth was handed down as a quasi-
historical tale, the real meaning of which had been forgotten.
The myth of Kimera is in many ways closely paralleled by the Nyoro myth of
Ndahura.6 King Bukuku, it is said, had an only child, a daughter called
Nyinamuiru. Warned that her son would destroy him, he kept her in close
confinement. A hunter named Isimbwa nevertheless succeeded in making his
way into her hut, and in due course she gave birth to a son, who was later called
Ndahura. The child was cast into the river, but was caught up in the rushes and
rescued by a potter, who brought him up as his own son. When Ndahura had
come to manhood he encountered his grandfather, slew him and succeeded
him as king. This story is almost identical with the Greek myth of Acrisius,
Danae, Zeus and Perseus. The woman is the buried seed, her cruel father the
confining earth, her lover the fertilizing rain (Mbwa is good Bantu, though not
modern Nyoro, for 'rain', and Isi-mbwa stayed with Nyinamuiru for five months,
6 Fisher (1911); Bikunya (1927); Nyakatura (1947).

the duration of the rainy season.) Ndahura-Perseus is the young corn which
emerges from the seed and thrusts its blade through the body of the earth.
There are, however, certain differences between Kimera and Ndahura.
Kimera, instead of killing his grandfather, was killed by his own grandson,
Tembo. This is of no great importance; the same cycle is being looked at from a
different point of view. Ndahura was the maternal and Tembo the paternal
grandson of his victim, but this was a change which became necessary when the
story acquired a historical, dynastic colouring and had therefore to conform to
the rule of patrilineal succession. The river-god Lumansi, who figures as the
son of Kimera and the father of Tembo, probably started by being Kimera's
son-in-law-if he belonged to the story in its original form at all.
There is, however, a more important difference. It seems that primitive
agricultural ritual contains the germs of two distinct developments. Attention
may be focused either on the human actor at the centre of the ritual or on the
unseen force which animates both the ritual and the crop. In the one case the
ultimate product is sacral kingship, in the other it is true divinity. Now Ndahura
was, as Kimera was not, a god. In a previous article7 I suggested that he was a
sun-god, but this was clearly an error. It is not the sun but the moon which
spends two nights in the land of darkness, as Ndahura did. It is the moon, too,
which by its waxing and waning regulates the lives of women and has therefore
universally been taken as the primary symbol and cause of fertility. Ndahura is
the growing corn, but by a natural transition he is also that power which causes
the corn to grow. Hence his name 'I will give increase'. The power which
resides in the horned moon is also manifested in the bull, so that it comes as no
surprise to read that Ndahura seized an enemy "and butted him with the horns
which he wore on his head, so that he died".
As long ago as 1931 Dr. Marianne Schmidl of Vienna put forward the view
that the rituals of kingship described in the ethnographical literature on the
monarchies of central Africa were intimately linked with agricultural fertility,
and, therefore, that the monarchies themselves could hardly have been origin-
ally created, as the orthodox theory held, by pastoral invaders. The study of
myth appears to lend support to her argument. For this god who is corn and
moon and bull, and who stands at the centre of the mythology of the so-called
Hima states, is clearly not a concept that could have been arrived at by a
purely pastoral people.
There is one further point, which I make with hesitation because it is, on the
face of it, extremely odd. At several points in this article analogies have been
drawn with Greek myths. These are not the only parallels which could be
found. Thus there is the curious story of the Cwezi hero Mulindwa. His step-
mother, it is said, bore him a grudge because he excelled her own son, Mugenyi,
in popular acclaim. So, having induced him to visit her as a lover, she dug a pit
in her hut and filled it with boiling water, into which he duly fell. There is a
clear echo here of the strange statement of the Greek mythographer: "But after
his bath Minos was undone by the daughters of Cocalus; some say, however,
that he died through being drenched with boiling water."8 Minos again is said
7 Some Thoughts on the Bacwezi. Uganda J., 22 (1958), 11.
8 Frazer (1921), 2, 143.

to have reigned for nine years, and so is Rukidi, the first Bito king of Bunyoro.
According to another account, however, Minos did not die after his nine-year
reign, but retired to a mountain cave and, having communed with Zeus, returned
with new vigour to continue his rule. So also the second Bito king, Oyo, was
allowed to go on living and reigning after the conclusion of his term, but only
by virtue of the performance, on a hill-top, of elaborate and terrible rites.
The emotion generated by the dance of the women in the cornfield may
result in the projection of their desire for crops and for sons, in which case the
outcome is a male deity. But it may, alternatively, result in the projection of the
women's collective self. Then not only the seed but the springing corn itself is
seen as female, and Perseus becomes Persephone. There is no trace of Perse-
phone in Uganda, but she is very clearly present in the Zulu cult and myth of
Inkosazana, the royal Maiden.9
Although I cannot affirm this with any confidence, it seems to me that, taken
together, these correspondences go beyond mere analogy, and imply a definite
cultural link. There is perhaps nothing really improbable in the hypothesis that
Aegean and Bantu culture may have had a partial common ancestry in the early
neolithic of the Nile valley.


Bikunya, P. (1927). Ky'Abakama ba Bunyoro. London and Kampala.
Bryant, A. T. (1949). The Zulu People. Pietermaritzburg: Shuter and Shooter.
Coon, C. S. (1955). The History of Man. London: Cape.
Fisher, Mrs. A. B. (1911). Twilight Tales of the Black Baganda. London: Marshall.
Frazer, J. G. (ed) (1921). Apollodorus: The Library. London: Heinemann.
- (1924). The Golden Bough. London: Macmillan.
Gray, J. M. (1934-5). Early History of Buganda. Uganda J., 2, 266.
Guthrie, W. K. C. (1935). Orpheus and Greek Religion. London: Methuen.
Harrison, J. E. (1912). Themis. Cambridge: University Press.
Kagwa, A. (1901). Basekabaka Be Buganda. 4th ed. (1953). Kampala: Uganda
Bookshop. London: Macmillan.
- (1905). Mpisa za Baganda. 2nd ed. (1918). Kampala: Uganda Bookshop.
Nsimbi, M. B. (1956). Amannya Amaganda n'Ennono Zaago. Kampala: East
African Literature Bureau.
Nyakatura, J. W. (1947). Abakama ba Bunyoro-Kitara. St. Justin, Canada.
Raglan, Lord (1936). The Hero. London.
Schmidl, M. (1931). Die Mondk6nige in Ostafrika. Congrbs de rlnstitut International
des Langues et des Civilisations Africaines. Paris.
Tracy, H. and Masinga, K. E. (1944). Chief Above and Chief Below. Pietermaritz-
burg: Shuter and Shooter. (Summary in Africa, 14 (1944), 422-3.)
9 For the cult see Bryant (1949). For the myth see Tracy and Masinga (1944).

I AM grateful to Sir John Gray (1956) for advancing in his article on Kibuka,
a tentative solution of the problem of the date of the solar eclipse seen at
Biharwe,' Ankole, by Mukama Olimi I of Bunyoro, as recorded by K.W.
(1936-7, pp. 69, 79). It is a problem which has long intrigued me, and I now
venture on a few further comments, being encouraged to do so by Mr. E. B.
Haddon, to whom I am much indebted for advice and for information obtained
from Professor F. J. M. Stratton of the Solar Physics Observatory, Cambridge.
As we must of necessity calculate dates in African tribal history in terms of
generations, there is need certainly for consistency and exactitude of definition.
As things are, there seem to be differing views as to what constitutes a genera-
tion. Thus Sir John Gray (1956, p. 69) talks of the "customary allowance of
three generations to a century" and in this he has the support of no less an
authority than Herodotus (Book 2. 142).2 On the other hand, H. F. Morris
(1955, p. 205; 1957, p. 2) talks of "allowing four generations to a century".
Another point to determine is how we should reckon a series of generations.
My own view is that it should be from the birth of the first of a series to the
birth of the last, or alternatively from the death of the first to the death of the
last. In African history it must usually be the latter, as we are much more likely
to have information about the death dates of eminent people than about their
birth dates.
Sir John Gray (1956, p. 69) describes Mukama Kamurasi, who died in or
about 1869, as a "lineal descendant in the thirteenth generation from Olimi I"
(the italics are mine), and arrives at a hypothetical date, 1436, for the death of
Olimi by reckoning 13 generations between them, averaging approximately
33 years. Reference, however, to the Table at Uganda J., 5 (1937-8), p. 69 shows
that Kamurasi was the twelfth descendant of Olimi; hence, if we reckon, as
postulated above, from the death of the first to the death of the last in the
series, we should call the difference between Olimi and Kamurasi twelve
generations only; hence, if 1436 is to be taken as the date of the former's death,

I I have retained the spelling Biharwe as given in Sir John Gray's article. I am advised
by Mr. E. B. Haddon that the alternative of Bihalrwe mentioned by Sir John, owing to its
combination of I and r, is most unlikely, though Biharrwe would be possible. Bihalrwe,
however, is the spelling given on the East African Survey Group's 1:1 m. Map of 1943.
On the Uganda Survey Department's 1:1 m. Map of 1928, as also on its 1:500,000 Map
of 1923 the spelling is Baharwe.
I have also consulted Mr. Z. C. K. Mungonya, C.B.E., the Minister of Land Tenure, and
former Nganzi of Ankole, who informs me as follows:
"The correct spelling of the word is Biharwe. This correct spelling appears on the Land
Office map of 1954.
"Although the word, as it is, is never used except as a name of a place, it derives its
meaning from the word okuhara-to scratch, or rather scrape. Actually Biharwe is the
name of a hill in that area, and all the land round about has taken the name. The hill is
covered with numerous rocks, and green grass can hardly be seen, so that the whole hill
looks like ground over which a scraper has been working, i.e. of reddish colour."
2 yYeeal Y&p Tpe4i b&vtpVu cKaTrbv gred o"rt.

the average duration of a generation is increased from 33 years to approxim-
ately 36-1 years. Alternatively the date should be advanced to about 1469.
I am of the opinion that an allowance of even 33 years to a generation is too
much, and that the allowance should certainly not be more than 30 years. This
seems to be borne out by the record over a long period of time of certain
historic British families. For instance the Duke of Windsor, who was born in
1894, is 37th in descent through both the Lancastrian and Yorkist lines from
Egbert King of Wessex, who was born about 780. This gives an average per
generation from birth to birth of 30-0 years. Again, the Erskine family is one of
the oldest in the British Isles and can also trace its descent back to Egbert of
Wessex through the Kings of Scotland and England. Examination of three
different branches of it, in each instance also over 37 generations, gives much the
same answers as in the case of the Royal Family, namely 29-4, 29-7 and 29-7
years. The average of these four averages is 29-7 years, and it is this factor which
I now propose to apply to the pedigrees of the Bakama of Bunyoro and the
Basekabaka of Buganda in general and to the particular problems (a) of the
dates of Mukama Olimi I and his contemporary Kabaka Nakibinge and (b) of
the date of the eclipse that Olimi is recorded by K.W. as having seen.
Taking first the case of Bunyoro, if we work back at an average duration of
29-7 years per generation from the dates of the deaths of four recent Bakama,
1862 (Nyabongo II), 1869 (Kyebambe IV Kamurasi), 1923 (Chwa II Kabarega),
and 1924 (Andereya Duhaga II), we get the following possible dates for the
death of Olimi 1-1535, 1513, 1537, 1508. The average of these dates is 1523,
and if it is an approximately correct date for Olimi's death, it implies that he
could not have seen the eclipse that occurred in 1462, though he could have seen
that of 1492, the one favoured by Sir John Gray (1956, p. 70). Referring to the
list of eclipses, which Haddon (1957, p. 116) obtained from Professor Stratton,
the only other possible eclipses with paths crossing the 30th Meridian of east
longitude within a reasonable distance of the latitude of the Equator, which he
could have seen, seem to be the total eclipses of 17 April 1520, and 20 July 1504.
A check with Oppolzer's Canon der Finsternisse (1887) (see Appendix) shows
that the date 1504 is a misprint for 1506, and it is a better date for our purpose
than 1520. For one thing it is earlier in Olimi's life; K.W. (1936-7, p. 79) says
that "when he had explained the eclipse" he went to fight in countries far away,
probably against the Azande, and this seems to suggest that he lived for some
time after it. Secondly, the line of totality in 1506 was only 2 south of the
Equator, as against 40 south in 1520. 20 south would be about 102 miles south
of Biharwe, reckoning a degree of latitude at 69 miles.
I conclude therefore that so far as Nyoro chronology is concerned the most
likely dates for Olimi's eclipse are 1492 or 1506. The latter was a total eclipse
and Sir John Gray (1956, p. 69) quotes Oppolzer as saying that the former was
total also. He appears, however, to have misquoted him, as I find on checking
that Oppolzer actually lists the 1492 eclipse as annular and that of 1462 as
total, the opposite of Sir John's version (see Appendix). I do not think, however,
that the point of whether the eclipse which Olimi saw was annular or total is
material. An annular eclipse may cause a blackout quite sufficient to be
impressive, and to an untutored mind even alarming. The reactions of the local

inhabitants to the total eclipse that occurred in Uganda in May 1947 are not
without relevance (Crabbe (1948)).
I come now to an examination of Buganda chronology, which it is most
difficult to reconcile with that of Bunyoro, so raising grave doubts of the
validity of the conclusions we have so far reached.
The application of our factor of 29-7 years per generation backwards from
the death dates of recent Basekabaka, 1856 (Suna II), 1884 (Mutesa I), 1903
(Mwanga II) and 1939 (Sir Daudi Chwa II) would give the dates for the death
of Kabaka Juko of 1678, 1676, 1665 and 1672. As it is generally accepted that
Juko was alive in 1680 when the eclipse at Baka took place,3 all these dates
are too early and our factor, at any rate over two and a half centuries, is
apparently too high.
An alternative factor can be found by calculating the actual average years
per generation of Basekabaka. This involves accepting the tradition that
Kabaka Kimera was the twin brother of Mukama Isingoma Mpuga Rukidi, the
first of the Babito Dynasty in Bunyoro (K.W. (1935-6), pp. 154, 160), and
assuming that both of them died about the same date. Estimating an average
date for Isingoma Rukidi's death, and consequently that of Kimera also, in
exactly the same way as we have done above for that of Olimi I, gives us the
answer, 1434. As Sir Daudi Chwa II, who died in 1939, was 19th in descent
from Kimera, the average duration of a generation of Basekabaka is 26-6 years,
and if we work back to the date of death of Juko on this figure, again from the
death dates of Suna II, Mutesa I, Mwanga II and Sir Daudi Chwa II, we get
the years 1696, 1698, 1690 and 1700, all well after the date of the Baka eclipse.
On the average of 26-6 years per generation we get as the date for the death
of Nakibinge, who was six generations after Kimera, the year 1594 (i.e. 1434+
160), being a date 71 years later than the date we arrived at for the death of
his supposed contemporary Olimi I of Bunyoro, and it must be admitted at
once that there seems to be no really satisfactory explanation of this dis-
The fact is that the generations of the Babito Bakama and Basekabaka get
out of step almost from the beginning and K.W. actually confirms that this was
so. He frequently mentions the names of the Baganda contemporaries of par-
ticular Bakama and these are mostly three, and in two cases four, generations
later, though things level out somewhat in the nineteenth century. Table I below
illustrates this point.
In view of the disparities in the early part of the above table, it is of interest
that K.W. (1936-7, pp. 77-80) states or implies that all of the first seven Bakama
of the Babito Dynasty-Isingoma Rukidi, Ochaki Rwangira, Oyo Nyimba,
Winyi I, Olimi I, Nyabongo I and Winyi II-reigned a long time or died in old
age. This is recorded of only one of the early successors of Kimera in Buganda,
namely Kigala. Haddon (1957) draws attention to this point.
In any case we must not forget that three generations are normally in being
simultaneously. A man is living at the same time as his father and his son. Even
a fourth contemporary generation is by no means unusual. For example Edward
3 Gray, Uganda J., 2 (1934-5), 268, and 20 (1956), 70; and Haddon, Uganda J.,
21 (1957), 116.



(mostly as given by K.W. at

Uganda J., 3 (1935-6), 160, 4 (1936-7), 75-83 and 5 (1937-8),

Name of Mukama Name of contemporary Number of
of Bunyoro Kabaka of Buganda generations later

Isingoma Mpuga Rukidi Kato Kimera 0
Ochaki Rwangira Kigala 2
Oyo Nyimba Kiyimba 3
Winyi I Kayima 3
Olimi I Nakibinge 3
Winyi II Kimbugwe 3
Winyi II Katerega 4
Kyebambe I Kagulu 4
Winyi III Kagulu 3
Kyebambe II Mawanda 2
Olimi III Kyabagu 2
Duhaga I Junju 2
Kyebambe III Semakokiro 1
Kyebambe III Kamanya 2
Nyabongo II Suna II 2
Kyebambe IV Suna II 1
Kyebambe IV Mutesa I 2
Chwa II Mutesa I 1
Chwa II Mwanga II 2
Chwa II Chwa II 3
Winyi IV Chwa II 2
Winyi IV Mutesa II 3

Mutesa II is three generations later than Sir Tito Winyi IV and this is the same
interval as that between Nakibinge and Olimi I (see Table I above).
Nevertheless the gap of 71 years between our hypothetical dates for the deaths
of Olimi I (1523) and Nakibinge (1594) presents an awkward problem and, in
common with Sir John Gray, I doubt whether, with our present information, we
can arrive at any satisfactory solution of it. The only thing I can suggest is that
we again take an average, i.e. split the difference between the two dates, thus
getting 1558 as an approximate date for Olimi's death. Of the eclipses listed in
the Appendix he might then have seen those of 1520 (total) at 40 south or 1546
(annular) at 3 south, though not that of 1557 (annular) at 40 south as we have
noticed already that he lived for some time after the eclipse. Against the date
1558 is the fact that it involves the acceptance of an interval of 124 years
between the dates of death of Isingoma Rukidi in 1434 and Olimi, and an
average of 41-3 years per generation, which is rather a high figure notwithstand-
ing the alleged longevity of the early Babito Bakama (cf. Gray (1956), p. 70)
though not an impossible figure over a small number of generations (in this case

I have already stated that it is not possible to draw any firm conclusions from
the calculations set out above. All that can be said is that there seems a strong
probability that the date of the eclipse, which Olimi I saw at Biharwe, was
between 1492 and 1546 (both dates inclusive). In general, this conclusion is in
agreement with Sir John Gray's views.
I think it may be reasonably claimed, however, that our calculations
strengthen the case for early rather than late dating of the arrival and departure
of the Bachwezi and the beginning of the Babito Dynasty. On this point there
has so far been a definite conflict of opinions. In support of earlier dates we
(a) K.W. (1936-7, p. 77) who thinks that Isingoma Mpuga Rukidi, the first
Mubito Mukama, reigned between 1300 and 1400.
(b) Sir John Gray (1956, pp. 69-70) who is prepared to date Olimi I as early
as 1462.
(c) A. H. Cox (1950, p. 158) who puts Kimera at the end of the 14th century.
Those favouring later dates are:
(d) E. C. Lanning (1953, p. 58) who dates the Bachwezi period at "perhaps
some four to six hundred years ago".
(e) Father Crazzolara (1937-8, p. 21) who puts the Lwoo migration, with
which he associates the Bachwezi incursion into Bunyoro, at between 1600 and
(f) K. Ingham (1957, p. 132) who dates the Bachwezi at "possibly about four
hundred years ago".
(g) E. B. Haddon (1957, p. 115) whose tables, based on average reign
duration, suggest the dates 1575 or 1478 for Isingoma Rukidi.
Haddon's method of calculating by the average length of reigns is, I think,
much less satisfactory than the method of calculating by generations. The
results, compared with those obtained above by the generation method in the
cases of the Royal Family and the Erskine Family, are almost inevitably
chancy, being subject to a number of inconstant factors, as well as varying from
locality to locality and from period to period. At Table II are some statistics of
the reign method, as applied to various lines of rulers, which illustrate this fact.
There is one other matter on which I should like to offer some comment and
that is the conflicting Nyoro tradition recorded by Gray (1956, p. 70) that it was
four generations later during the reign of Mukama Chwa I (Chwamali), and not
during that of Olimi I, that the eclipse was seen at Biharwe. In this connexion
I would draw attention to Dr. J. M. Derscheid's (1934-5) letter in which he
excerpts four lists of the Babito Bakama of Bunyoro from the writings of Petero
Bikunya, Rev. J. Roscoe, R. P. Torelli and Bishop J. Gorju which respectively
give Chwamali or Chwa I as No. 5, No. 3, No. 4 and No. 4 in order of
succession, in all cases ante-dating the earliest Olimi for whom the correspond-
ing positions are Nos. 7, 6, 7 and 6. Sir Harry Johnston The Uganda Pro-
tectorate (1902), ii, 596-7) on the other hand gives Olimi as No. 2 and Chwa as
No. 5, and there is a further complication in that K.W. (1936-7, p. 79) gives
Chwa as one of the additional names of Nyabongo I, the successor of Olimi I.
K.W. (1935-6, p. 155) expressly repudiates all the four lists cited by Dr.
Derscheid, whereas Father Crazzolara (1937-8, pp. 19-20) is of opinion that short




(a) The Kings of Israel and Judah.
Period. From the division of Solomon's kingdom on his death c. 935 B.c. to the
destruction of the kingdom of Israel by Shalmaneser, King of Assyria in 722 B.C., 213 years.
Number of kings.
Israel 19. Judah 13.
Average length of reign.
Israel 11-2 years. Judah 16-4 years.
Remarks. There was no change of dynasty in Judah, but there were 8 changes in
Israel, two of which were very short-lived usurpations.
(b) The Achaemenid Kings of Persia.
Period. From the accession of Cyrus in 559 B.C. to the death of Darius III in 331 B.C.,
228 years.
Number of kings. 13.
Average length of reign. 17-5 years.
(c) The Kings of France and England.
Period. From the accession in France of Hugh Capet in A.D. 987 to the deposition of
Charles X in A.D. 1830, 843 years.
Number of kings.
France 35. England 40.
Average length of reign.
France 24-1 years. England 21-07 years.
(d) The Kings of France and England.
Period. From the accession of William the Conqueror in England in A.D. 1066 to the
death of George IV in A.D. 1830, 764 years.
Number of kings.
France 32. England 33.
Average length of reign.
France 23-87 years. England 23-15 years.
Remarks. The difference in average for England between (c) and (d) is caused by some
very short reigns in the Saxon-Danish period.
(e) The Holy Roman Emperors in Germany.
Period. From the coronation of Otto I as Emperor in A.D. 962 to the extinction of the
Holy Roman Empire by Napoleon I in A.D. 1806, 844 years.
Number of emperors. 44.
Average length of reign. 19-18 years.
Remarks. If we add the previous years of Otto I's reign as King in Germany from
A.D. 936 and the remainder of the reign of the last Holy Roman Emperor, Francis II, as
Emperor of Austria till A.D. 1835, it gives us 44 rulers in 899 years with an average reign
of 20-43 years.

lists of kings are more likely to be authentic than long ones. Without in any way
endorsing the latter opinion we may note (a) that, if Olimi I were antedated as
the immediate successor of Isingoma Rukidi (Johnston), the date of his death
would be approximately 1464, and, of the eclipses listed in the Appendix, he
might have seen that of 1462, though only within a short time of his death and
(b) that, if Chwamali were antedated at two generations after Isingoma Rukidi
(Roscoe), the approximate date of his death would be 1493, and he might have
seen the eclipses of 1462, 1474 and 1492 but the last also within only a year of
his death; if three generations (Torelli and Gorju), his approximate date of death
would be 1523 and possible eclipses would be those of 1492, 1506 and 1520 (the
last only three years before his death), the same in fact as those applying for

Olimi I; if four generations (Bikunya), the approximate date of death would be
1553, and corresponding eclipses those of 1520 and 1546.

Solar eclipses with paths crossing the Meridian of 3" east at or within 4* of the latitude
of the Equator. (Extracted from Ritter von Oppolzer, Th. Canon der Finsternisse. 1887.
(Vienna Academy of Science).

Date Character Latitude Page Plate Remarks
Ref. Ref.

21 Nov. 1462 Total 0 254 127 Not 10 Oct. and not annular
as Uganda J., 20 (1956), 69.
16 April 1474 Annular 1"N 256 128 Not previously mentioned in
Uganda J.
21 Oct. 1492 Annular 0" 258 129 Not total as Uganda J., 20
(1956), 69.
20 July 1506 Total 2*S 258 129 Not 1504 as Uganda J., 21
(1957), 116.
17 April 1520 Total 4*S 260 130
23 Nov. 1546 Annular 3"S 262 131
22 Oct. 1557 Annular 4*S 264 132
22 Aug. 1579 Annular 2*N 266 133
12 Sept. 1662 Annular 1S 274 137
30 Mar. 1680 Total 0* 276 138 Not April as Uganda J., 21
(1957), 116. See also ibid., 2
(1934-5), 268.
20 May 1947 Total 0* 302 151 See Uganda J., 12 (1948), 99-

NoTE.-The above list does not include three of the eclipses mentioned by Haddon at
Uganda J., 21 (1957), 116 with paths at 7* south of the latitude of the Equator.

Cox, A. H. (1950). The Growth and Expansion of Buganda, Uganda J., 14, 153-9.
Crabbe, J. R. (1948). The Total Eclipse of the Sun, Uganda, May 1947. Uganda J.,
12, 99-100.
Crazzolara, Rev. Fr. J. P. (1937-8). The Lwoo People, Uganda J., 5, 12-21.
Derscheid, Dr. J. M. (1934-5). The Bakama of Bunyoro better) Uganda J., 2,
Gray, Sir J. M. (1934-5). The Riddle of Biggo. Uganda J., 2, 230.
- (1934-5). Early History of Buganda, Uganda J., 2, 259-71.
- (1956). Kibuka. Uganda J., 20, 52-71.
Haddon, E. B. (1957). Kibuka. Uganda J., 21, 114-7.
Ingham, K. (1957). Some Aspects of the History of Western Uganda, Uganda J., 21,
K.W. (1935-6). Abakama ba Bunyoro-Kitara Pt. 1. Uganda J., 3, 149-60.
- (1936-7). Abakama ba Bunyoro-Kitara Pt. 2. Uganda J., 4, 65-83.
- (1937-8). Abakama ba Bunyoro-Kitara Pt. 3. Uganda J., 5, 53-84.
Lanning, E. C. (1953). Ancient Earthworks in Western Uganda, Uganda J., 17, 51-62.
Mayanja, A. M. K. (1952). The Chronology of Buganda, 1800-1907 (translated from
Kagwa's Ebika). Uganda J., 16, 148-58.
Morris, H. F. (1955). The Kingdom of Mpororo. Uganda J., 19,204-7.
- (1957). The Making of Ankole. Uganda J., 21, 1-3.



THE object of this inquiry was to follow up the topographical information
provided in the recently published summary of the traditional history of
Ankole-Abagabe b'Ankole by A. G. Katate and L. Kamugungunu, Eagle
Press, Kampala, 1955. This has the appearance of a careful and conscientious
piece of work. I understand that it is based in the first place upon information
compiled over a long period by the late Nuwa Mbaguta, who was Nganzi
of Ankole from 1895 until 1937. At his death in 1944, Mbaguta's papers passed
into the possession of his son-in-law and successor in office, L. Kamugungunu,
and the information contained in them is said to have been checked and supple-
mented by reference to a number of old men, all of whom were brought up
at the court of Ntare V, and all of whom were therefore taught the traditional
history of Ankole in the days before the coming of British rule. This supple-
mentary information was collected by Katate over a period of ten years. He
states that every item of it has been checked by Kamugungunu and his
advisers, and that the completed manuscript was then subjected to detailed
scrutiny before publication by an inter-racial committee of which H. F.
Morris was the Chairman. I have therefore assumed that it is as reliable a
survey of the oral tradition of Ankole as we are likely to get, and that the
information contained in it should not be lightly set aside. Katate states that,
while he encountered considerable conflict of evidence in matters of detail,
there were no significant differences in the names or sequence of the Bagabe,
and little conflict about the principal places where they built their capitals. If
this is so, the sites described in this report should be regarded as provisionally
and roughly dateable by reference to the genealogy of the Bagabe, which is
set out below. The figures in the left hand column show the number of genera-
tions back from the present. The figures in the right hand columns show the
approximate dates, calculated on a basis of 27 years to the generation. Brothers,
being of the same generation, are placed in the same line.
With regard to the dates, it should be realized that the average length of a
dynastic generation of this kind depends upon the system of succession. In
Ankole this system was not primogeniture. The heir was normally chosen from
among the sons of the official wives, therefore from among the sons who were
born after their father's accession to the throne. This would make for rather a
long generation, and I consider that 27 years is therefore a reasonable average.
Figures calculated on this basis should be regarded as liable to a margin of error
of two years, plus or minus, for every generation back from the present. The date
of a ruler of 20 generations back would thus be liable to an error of 40 years,
plus or minus. In addition, in estimating the outside dates of any particular
generation, it is necessary to make allowance for deviations from the average

caused by the proximity of exceptionally long or exceptionally short genera-
tions. For this I consider it is necessary to increase the margin of error by 20
years right through the list. Thus the total margin of error becomes 20 years,
plus 2 years for every generation back from the present.

Ntare I
Ntare II
Ntare III
Kitera and Rumongye
Ntare IV
Rwebirere, Karara, Karaiga and Kahaya I
Nyakashaija, Bwarenga and Rwebishengye
Kayungu, Rwanga and Gasyonga I
Mukwenda and Ntare V
Kahaya II
Gasyonga II

1528-1555 52
1555-1582 50
1582-1609 48
1609-1636 46
1636-1663 44
1663-1690 42
1690-1717 40
1717-1744 + 38
1744-1771 36
1771-1798 34
1798-1825 32
1825-1852 + 30
1852-1879 + 28
1879-1906 26
1906-1933 + 24
1933- 22

i.e. between
1476 and 1607
1505 and 1632
1534 and 1657
1563 and 1682
1592 and 1701
1621 and 1732
1650 and 1757
1679 and 1782
1708 and 1807
1739 and 1832
1766 and 1857
1795 and 1882
1834 and 1907
1853 and 1932
1882 and 1957

In the following classification of sites I have adhered strictly to Katate's
ascription of particular sites to particular Bagabe. In an article published in
Uganda Journal, 22 (1958), 11-17, C. C. Wrigley has argued that the Bach-
wezi (i.e. on the above list Wamara) were gods rather than men, and I under-
stand he holds that the same may apply to the first two or three names on
the list of the Bahinda rulers (i.e. on the above list Ruhinda, Nkuba and
Nyaika). If this is so, it follows that reasonable explanations have to be found
for the sites associated with these names. In the case of Ntusi, which is clearly
a large occupation site, this can only be done by saying that the attribution to
Wamara is spurious. On the other hand, Itaba is only a grove. So is Rurama
rwa Ruhira. It may well be, therefore, that these two were places of worship
rather than capital sites. Furthermore, subsequent archaeological investigation
may call for a considerable revision of the rest of this classification; but for
the moment I consider that this represents the nearest we can get to the truth
from traditional and topographical evidence alone. Two general points should
perhaps be made in this connection. First, that in nearly every case where I
have set out to find a site mentioned by Katate and Kamugungunu, it has been
possible to do so. Secondly, every site that I have so identified has been at
least topographically consistent with what we know of Ankole history. We
know, for example, that after the break-up of the Bachwezi kingdom, Ankole
started as a tiny state, comprising the present Saza of Isingiro and a portion
of eastern Rwampara. The location of the earliest Bahinda capitals bears this
out. It is only at the turn of the 16th and 17th centuries that the capitals
spread out from the hilly fastnesses of southern Isingiro into the open Masha
plain. And it is not until the nineteenth century that they move north of the
Mbarara-Masaka road into Kashari and central Nyabushozi. It seems to me

therefore that Katate's ascriptions should be accepted unless there is strong
evidence to the contrary-as there is in one or two cases, but not in many.
Sites of Wamara.
(i) ITABA. Wamara is said to have been the last ruler of the Bachwezi dynasty,
and his kingdom, centred in Bwera, is said to have extended over most of
southern Uganda. During the lifetime of his father Ndahura (whose capital
was on Mubende Hill) Wamara is said to have exercised a sort of provincial
governorship over the Ankole region; and during this period his headquarters
were at Itaba-a site marked by a very prominent and well-known grove of
trees on the summit of the Rwampara hills, south-west of, and easily visible
from, Mbarara. The grove is one of the most striking in Ankole, and almost
bears comparison with the Witch Tree grove on Mubende Hill. The trouble
about groves from the archaeological point of view is that they continued in
use as places of sacrifice right through the centuries, and it is therefore impos-
sible to date from internal evidence any objects found in them.
(ii) RUSHOZI. From Mbarara take the Masaka road to Mile 9; then turn left
and continue northwards for 8 miles; turn right at T junction to the Land Use
Experimental Area; from the Experimental Area office take the track leading
straight on to a dam in the valley below; cross the dam and stop I mile further
on. The hill immediately on the right is said to be the site occupied by Wamara
after he left Itaba. There are signs of occupation, including pottery, but no
(iii) NTUSI, in north Mawogola, Buganda. This is said by Ankole tradition
to have been Wamara's main capital, after his accession to the throne of his
father Ndahura. This, together with the near-by earthworks at Bigo bya
Mugenyi, is of course the best known and most important archaeological site
in Uganda, and (see U.J., 2 (1934-5), 26) needs no further description here. It
was from here that the Bachwezi disappeared, leaving the drum of Bwera,
known as Rusama, in the keeping of the Bamoli family, who continued
thereafter as the semi-independent rulers of Bwera, paying tribute first to the
Babito rulers of Bunyoro, later to the Bahinda of Ankole, and most recently
to the Kabakas of Buganda.
Sites of Ruhinda.
(i) RUKOMA-KASANA. Ruhinda, traditionally the illegitimate son of Wamara,
and founder of the Bahinda dynasty, is said to have been born at Itaba, and
to have moved with his father to Rushozi and then to Ntusi. From there he
carried out extensive cattle-raids in Karagwe, and he was engaged thus when
the Bachwezi finally fled. Hearing the news, he returned to Ntusi, and evacu-
ated the few survivors to the Ankole district. He is said to have built his first
capital at RUKOMA, which is marked on the District Map about 6 miles due
south of Mbarara, and which is approached by a motorable road, forking
right at Mile 3 on the Kikagati road from Mbarara. It was here that Ruhinda
took possession of the drum Bagyendanwa from the Omusita, Katuku, who
had kept it for Wamara. As will appear later, Rukoma Hill is covered with
ancient sites, but I have been unable to identify this one of Ruhinda.
(ii) MWERUKA. This is said by Katate to have been Ruhinda's second capital;


'140 .. :.

21* 17
9* .,; '


-." AREAS -BO.
0spoo 0 -*

[Drawn by Dr. M. Posnansky

FIG. 1






.a, .i-~

but neither Katate nor anyone else I questioned had any idea where this place
was to be found.
(iii) RURAMA RWA NTUNGU. This seems to have been the main site of Ruhinda
in Ankole. It is certainly the only one which is identifiable to-day. Access is
extremely difficult. At Mile 19 from Mbarara on the Kikagati road, turn right
and ascend by motorable track to Kyabinunga village near the summit of
Kagarama. From here a jeepable track continues southwards over Bungura,
and then turns east along the main watershed between the Rugaga and Chezo
water-systems. The track is jeepable for about 4 miles, after which it is neces-
sary to continue on foot for another 3 miles, keeping always to the ridge of
the watershed. Here, in a magnificent position commanding the head of the
Rugaga valley northwards and the head of the Kabiganda valley southwards,
is Rurama Rwa Ntungu. At the time of my visit grass was long and very
thick, and it was only possible to see that an area about I mile square was
covered by a number of low mounds and banks. After grass-burning in July
the picture would be much clearer; and despite its inaccessibility I consider
that this site should certainly be examined further, as it is probably the earliest
Bahinda site in Ankole, and it is particularly important to discover any signs
of continuity that there may be with the Bachwezi sites. It was from Rurama
that Ruhinda is said to have left Ankole for Karagwe, Ihangiro, etc., where
he founded other Bahinda dynasties.
Sites of Nkuba.
(i) RURAMA RWA RUHIRA. This, the first and principal site of Ruhinda's son
and successor, Nkuba, is also extremely inaccessible, and I was unable to get
there myself, though from Rurama Ntungu it was possible to see, about 6
miles away, the hill on the southern slopes of which it is situated. Best access
is by jeepable track running northwards from Kikagati, by which it is possible
to get within about 3 miles' walk. The site is well known to local inhabitants,
and is said to be similar to Rurama Ntungu, but with a grove of trees still
standing and said to be infested by dangerous snakes. All this country looks
straight out towards the Karagwe hills, and enables one to understand the
traditions concerning the comings and goings of Ruhinda.
(ii) MITOMA. This is said to be the site of Nkuba's second capital, and it is
supposed to be the Mitoma marked on the District Map about 5 miles south
of the western extremity of Lake Nakivali. Access to this hill is via Ngarama,
whence a motorable track leads round the hill-tops to within a mile of Mitoma
summit. The summit is to-day crowned by a small village and banana planta-
tion; and no ancient occupation site is visible, unless possibly on a hill-ridge
running eastwards just before one crosses the final shoulder onto the village
summit. Here there is a natural rocky edge, planted with a line of mitoma
(bark-cloth) trees, with faint signs of occupation on top of the ridge. What
lends support to the tradition that Nkuba had a site in this area, is that he
was the first Mugabe to be buried at Ishanje, which is only a few miles away
on the shores of Lake Nakivali.
Site of Nyaika.
KICHWEKANU. This is a site of outstanding interest, though I should say that

Katate's attribution of it to Nyaika may need to be checked. It lies on the
most north-easterly shoulder of Kagarama Hill. Best approach is to drive to
the top of Kagarama, and then walk round the top to the northwards and
descend the north-east shoulder. Kichwekanu is the first section of this shoulder
that one comes to. It forms an elliptical platform, about I mile long and 300
yards wide, between two quite narrow necks on the main ridge. Many hut
circles, and small enclosures are visible on this platform, and I picked up a
few sherds of old, decorated pottery. A natural rocky edge forms the northern
perimeter of the platform, which is planted with a prominent line of bark-cloth
trees, euphorbia and with the shrub called ekicuncu, which is used as a purga-
tive and is planted near all occupation sites. From the centre of this rocky
edge a sort of natural ramp descends in zigzags down the northward facing
slope of the ridge, on which, about 500 yards from the upper site, there is
another site apparently connected with it. Grass was high on this lower site
at the time of my visit, but local inhabitants said that in the dry season its
outlines were clearly visible, and that its main gateways and approaches
faced uphill towards the upper site. A small collection of pottery from the
lower site has been deposited in the Uganda Museum together with the few
sherds from the upper site. Local opinion was that the lower site was the main
cattle-boma, and perhaps the Mugabe's own normal residence; and that the
upper site was for women and children, and possibly a place of retreat in time
of danger. This whole site is clearly worth further examination.
Sites of Ntare I and Rushango. 16th and 15th generations. 1528-82 50.
For Ntare I Katate gives KAKUKURU. For Rushango he gives KIBARE. Until
near the end of my visit I could find no one who knew where they were to be
found. Finally Katate made further inquiries, which seemed to show that both
were in the Ntungu area, near the Rurama of Ruhinda. Owing to the very
inaccessible nature of this part of the country I was unable to get there again,
and to track down these places myself. Inquiries are, however, being made
through the Ankole Local Government, and results will be reported. It is
worth noting that the first of these two reigns coincides with the first great
onslaught of the Babito under Olimi I of Bunyoro. It would therefore be quite
consistent with these circumstances that the Abagabe of this period should
have lived more or less in retreat in the remote corner of Isingiro, where the
dynasty had first established itself at the end of the previous century.
Site of Ntare II. 14th generation back. 1582-1609 48.
BUHANDAGAZI. This is a well-known hill in the Masha plain; and with a jeep
it can be approached by a track turning left off the Kikagati road at Mile 6
from Mbarara. The track leads due east for 2 miles, then bears south, then
east again, and mounts the shoulder connecting Enjerika and Buhandagazi
hills. Buhandagazi therefore lies about 2 miles north-east of Enjerika-a com-
manding position, with extensive views over the Masha in all directions. On
the north and north-east the edge of the hill-top is marked by gigantic slabs of
rock, well adapted either for defence or prestige. The largest rock slab on the
north-eastern side has two rather striking holes or troughs, about 6 to 8 feet
in depth, which are evidently filled with water at certain times of year. The

occupation site was presumably on the flat top behind this rocky edge, but
apart from suggestive lines of trees, including the shrub ekicuncu, I could find
no surface evidence. It is to be noted that Buhandagazi is ascribed by Katate
both to Ntare II and to Kahaya I. I think the latter ascription is the more
probable, although both may be correct.

Site of Ntare III. 13th generation back. 1609-1636 46.
KATAMBA. I did not have time to visit this place; but it appears to be well-
known, and to lie to the south side of the Mbarara-Masaka road, on the
northern slopes of Kabarangira Hill, about 6 miles east of Mbarara. It is to
be noted in connection with this reign that the traditional history has most
probably telescoped the events of Ntare III into those of Ntare IV. If so, Ntare
III was probably the Mugabe who suffered the great attack by Mukama
Chwa Rumomamahanga of Bunyoro, which resulted in a prolonged occupa-
tion of Ankole by the Banyoro. At this rate Ntare III would probably have
spent the greater part of his reign in exile.

Reputed site of Kasasira. 12th generation back. 1636-1663 44.
BWEYORERE. This is perhaps the most impressive, and certainly the best
preserved, site in the series. It should however be noted at the start that this
is the most seriously disputed of all Katate's ascriptions, in that both the
present Mugabe and the present Keeper of Bagyendanwa are of opinion that
this is a site of Machwa, a full century later. I have thought it best, however, to
stick to my rule and to describe it under Katate's ascription, to Kasasira.
Bweyorere is extremely easy of access. At Mile 17 from Mbarara on the
Kikagati road turn left down the sign-posted track to Omubare bore-hole.
Leave car at the bore-hole I mile from the main road. From here Bweyorere
is only 1 mile in a north-easterly direction-a low, beautifully rounded hillock,
at the southern end of the Masha plain, with Gayaza Rest House on the high
ridge immediately above it to the south. Approaching from the bore-hole, that
is from the west, one comes first to a great U-shaped enclosure, measuring
220 feet by 150 feet, surrounded by earthwork banks 4 to 5 feet high. In the
middle of this enclosure is a mound with a very old finger cactus, which has
recently fallen down. Somewhere in this enclosure must lie the site of Rwemi-
hunda, the public reception house of the Mugabe. Behind the enclosure the
outlines of a large settlement can be clearly seen, stretching to the eastwards
for perhaps I mile. The line of a perimeter fence can be approximately dis-
tinguished, running in a series of flower-petal loops right round the site. For
the most part this perimeter is marked by low earthwork banks about one foot
high, but for two stretches, one at the north-east corner and the other in the
middle of the south side, these banks rise to 3 or 4 feet, much like those of
the main enclosure. Outside the main palace perimeter, on the slopes of the
hill are signs of levelling and terracing, which probably indicate the occupa-
tion sites of the pages, soldiers and chiefs who lived around the palace, but
not in it.
In other words, we have here at Bweyorere a compact town site, dateable
to somewhere between the mid 17th and the mid 18th century. It is therefore

a find of great importance; and it is to be hoped that steps will be taken
immediately to preserve it, both for scientific investigation and for its public
interest. The most immediate danger lies from the spread of cultivation, which
is fast enveloping the Masha plain. There are already plantations close to the
Omubare bore-hole, only J mile from the site.
First, therefore, legal protection of the site. Next, a detailed and accurate
survey, with which should be associated a general clearance of vegetation,
especially of the thorn trees, which at present obscure a clear vision of the
site. Thirdly, archaeological exploration, with which should be associated a
sociological inquiry conducted by a competent Runyankore linguist with the
aid of old men who were brought up at the palace of Ntare V, and directed
towards reconstructing the functions of various parts of the site and the tradi-
tional design. Finally, preservation. Bweyorere is so near to a main motor
road that it would probably attract a considerable number of visitors, if kept
properly cleared and sign-posted. This could probably be arranged for a very
small expense with the occupants of the neighboring plantations, who should
be encouraged to take a semi-proprietory interest in the site.
Lastly, if this site should turn out to be that of Machwa rather than
Kasasira, it will remain to discover the capital of Kasasira. In his article on
Historic Sites in Ankole, Uganda J., 20 (1956), 178, H. F. Morris reports a
site on Kagarama Hill which is said to have been the birthplace of Mugabe
Rumongye. This site is described below, under Rumongye and Mirindi. But
if Morris's information is correct, this site must have been occupied also by
Kasasira, since Kasasira was the father of Rumongye.
Site of Kitera. llth generation back. 1663-1690 42.
BUNGURA. I was shown a small site on the west side of Bungura Hill, just
to the north of the track leading to Ntungu, which may be that attributed by
Katate to Kitera. Kitera, however, was not an important ruler, as he was
quickly displaced by his brother Rumongye.
Sites of Rumongye and Mirindi. llth and 10th generations back. 1663-1717
KAGARAMA. These are the important sites, briefly described by H. F. Morris
in the Uganda J., 20 (1956), 178. The main occupation site is just to the south
of the trigonometrical point marking the summit of Kagarama, and a stout
motor-car can be driven to within a few yards of it. At Mile 19 from Mbarara
on the Kikagati Road, take the right hand track to Kyabinunga village, and
after a steep climb, bear right and right again, and continue to where the track
peters out in the middle of Rwaitirimba's mailo. In the banana plantation
immediately to the right of this point there are clear signs of considerable
earthworks, though now alas seriously marred by cultivation. A great saucer-
shaped depression about 120 feet in diameter is still, however, clearly visible,
and the earthwork banks which surround it must originally have been strictly
comparable with those at Bweyorere. On the opposite side of the track from
this site there is, as H. F. Morris says, 'a huge mutoma', which may mark the
site of another and still larger enclosure. The ground surface around this tree
certainly has a concave appearance, but neither Dr. Posnansky, the Curator

of the Uganda Museum, who was my companion, nor I were convinced that it
was not a natural formation. Here again cultivation has ruined the surface
evidence. Morris further describes a bathing-pool where the Bagabe washed.
This pool is to be found just beyond the trigonometrical point from the two
former sites. I think it is almost certainly artificial, or at least partly so. It
traps the head-waters of a little stream, and there is an artificial embankment
at the lower end. Moreover, about 50 yards below it there are traces of a dam,
which probably formed a drinking-pool for cattle. Finally, about mile further
back, towards Kyabinunga, just to the east of the track, there is a grove of
mitoma trees, not mentioned by Morris, which is said to have been used
for the ceremony known as Enkoko Enjeru, at which a white cow, a white
sheep and a white hen were periodically sacrificed by the Bagabe to the spirits
of the Bachwezi. Local inhabitants go in great fear of this grove, access to
which is stoutly opposed by large colonies of angry bees.
All in all, there must certainly be much valuable information locked away
in this most important series of sites. It seems from the traditional evidence
that it was in continuous occupation as the capital of the kingdom for the two
long reigns of Rumongye and Mirindi, and possibly also for the reign of
Kasasira which immediately preceded them. The great obstacle to further
inquiry is the intensive cultivation which has covered the whole area. Never-
theless, an accurate survey would certainly yield valuable results. Much could
also be achieved by persuading the local inhabitants to keep and hand in old
pottery which comes to light while they are cultivating; particularly if they
could also be persuaded to mark with stakes the exact sites of their finds. Trial
trenches could then be sunk at these points. I have no doubt that the extremely
able Muruka Chief, Mr. Shiringi, would be ready to co-operate in this matter.

Sites of Ntare IV-Kitabanyoro. 9th generation back. 1717-1744 + 38.
General Note. As explained above in connection with Ntare III, I believe
that Ankole tradition has considerably confused the reigns of Ntare III and
Ntare IV. I believe that it was Ntare III who suffered the devastating attacks of
Chwa Rumomamahanga of Bunyoro. I think the Bunyoro attack suffered by
Ntare IV was the singularly ineffective one ascribed by Bunyoro tradition to
the reign of Olimi III. This expedition was led by Olimi's son, Mali, who later
succeeded to the throne of Bunyoro as Chwa Duhaga I, thus explaining why
he is remembered in Ankole tradition as Chwamali. Bunyoro tradition states
that during the reign of Olimi III the southern frontier of Bunyoro was driven
back to the line of the Katonga. I believe therefore that the correct picture of
Ntare IV is that of a conquering monarch, who stood at the head of an expan-
sive and increasingly ambitious state of Ankole. It is for example highly signi-
ficant that, before the Bunyoro attack, Ntare IV contracted the first external
dynastic marriage, with the two daughters of Kamurari of Mpororo, the sequel
to which was to be the annexation of Shema to Isingiro. It may have been
Bunyoro which first attacked Ankole, but if so, it was an Ankole which was
able to hit back harder still, and to end the war by the annexation of Nyabu-
shozi from Bunyoro. I believe that it is in this light that the capital sites of
Ntare IV have to be considered.

(i) BYANGANGA. This is the site attributed to Ntare before the invasion of the
Banyoro. I was unfortunately unable to visit it, but, significantly, it proves to
be on the high hill Kakunyu, about 15 miles due east of Mbarara. Thus, apart
from the site of Ntare III at Katamba, it is the earliest capital of Ankole which
was built to the north of the River Rwizi. It was from Byanganga that Ntare
visited the capital of Mpororo to negotiate for the daughters of Kamurari. It
therefore seems likely that this may prove to be an important site, and it is
desirable that its precise location should be discovered.
(ii) KAGARAMA-BUNGURA-KICHWEKANU. After the Bunyoro attack, Ntare is
said to have returned from his temporary exile to a capital at Bungura. This
may be the site described under Kitera, above. But the big site on Kagarama
Hill is also associated by local inhabitants with Ntare IV, and so is Kichwekanu,
described above in connection with the reign of Nyaika. With all respect to the
opinions of Katate, I consider Kichwekanu to be very probably a site of Ntare

Sites of Machwa. 8th generation back. 1744-1771 36.
(i) BWEYORERE. For a description of this important site, see under Kasasira,
above, to whom it is attributed by Katate. From information given by the
present Mugabe and the Keeper of Bagendanwa, I consider that this site should
probably be attributed to Machwa.
(ii) BIRERE. This is the site attributed by Katate to Machwa; and in view of
the fact that one of Machwa's sons was called Rwabirere, I think that this
attribution is probably correct. This site does not appear to be at the present
village of Birere, which stands at Mile 8 on the Kikagati road; but rather
at the southern end of the summit of Rukoma Hill, where there is a large,
level platform, several acres in size, which is known to the local inhabitants
as Birere. Unfortunately the whole area is now under cultivation, and the
only surface landmarks are large euphorbia trees which stand at either end
of the site.

Sites of Kahaya I. 7th generation back. 1771-1798 34.
Note. Between Machwa and Kahaya I there were three pretenders. (a)
Rwabirere, said to have lived at Birere, (b) Karara, said to have lived at
Bweyorere, and (c) Karaiga, said to have lived at MABARE. This latter is a
small hill immediately to the north of Buhandagazi, which is remembered by
local inhabitants only as a site of Bagyendanwa at the time when Kahaya was
living at Buhandagazi, and possibly also later, in the time of Ntare V. Two
small earthworks are visible on the top of this hill, which are said to be the
kraals of Bagyendanwa's herd.
(i) BUHANDAGAZI, said by Katate to have been the first capital of Kahaya I.
For a description of this site, see under Ntare II.
(ii) NYABIKIRI. Said by Katate to be the second capital of Kahaya I. This is
a very prettily situated little hill, on the western edge of the Masha, which is
approached by a footpath of about 4 miles in length, which leaves the Kika-
gati road at Mile 12 from Mbarara. Nyabikiri is to the west of this road. The
whole hill is covered with signs of occupation, in the shape of hut-sites terraced

into the sides of the hill. The top of the hill seems rather too small for the
palace of a Mugabe, as it can measure little more than 100 feet in diameter.
Nevertheless there are undoubted signs of earthwork banks, comprising
apparently two levelled house-sites with spoil thrown backwards so as to
afford a clear view outwards; and a small banked enclosure about 30 feet in
diameter. Local inhabitants were of the opinion that it was the site of Bagyen-
danwa rather than of the Mugabe, and this should perhaps be borne in mind
in further investigations. But, taking into account the terracing on the hill
slopes, the size of the settlement must have been considerable.

Sites of Rwebishengye. 6th generation back. 1798-1825 32.
Note. Between Kahaya I and Rwebishengye there were two pretenders:
Nyakashaija, said to have lived at Mabare, for which see under Kahaya I,
above; and Bwarenga, said to have lived at Kikakarabwa, which place I have
not been able to identify.
(i) BWARA. This is a big conical hill on the eastern side of the Masha plain,
about 4 miles due east of Mabare hill. Owing to its inaccessibility I was not
able to visit it, to search for the site of the capital.
(ii) KAKUKURU. Not apparently the same Kakukuru as that of Ntare I, but
a hill in southern Nyabushozi, near Lake Kachira, with a site known to the
local Gombolola Chief. I was not able to visit this site.
(iii) NYAMIRIMA. This site is not mentioned by Katate, but it appears to be
a well authenticated site of Rwebishengye. It is to be found on the north-
eastern slopes of Rukoma Hill. Turn right at Mile 3 on the Kikagati road,
and when approaching Rukoma hill, bear left at a fork in the track and cross
over to the eastern side of the hill. 200 yards from this fork there is another,
and here, on the left hand side, a series of embankments are clearly visible.
Cultivation has come right up to the edge of this site, and immediate steps
should be taken to protect it from destruction. Though nothing like so complex
as Bweyorere, it is a good site, worthy of further investigation and survey.

Sites of Gasyonga I. 5th generation back. 1825-1852 30.
Note. Between Rwebishengye and Gasyonga I there were two pretenders:
Kayungu, for whom no site is mentioned; and Rwanga, said to have lived at
Kyangabukama, somewhere in the neighbourhood of Gayaza.
(i) RUKOMA-KASANA. This is one of the best sites in the series. It is easily
found, on the west side of Rukoma Hill, just below the road. The earthwork
mounds are large and clear, and the perimeter is further marked by bark-
cloth trees, which may originally have been the posts of a fence. The site
covers about 3 acres, and the internal divisions stand out well. A detailed
survey here would be comparatively easy, and, supplemented by information
from old palace retainers still living, it would probably be the best point at
which to begin a study of the structure and lay-out of the outstanding older
sites, such as Bweyorere. This site is also in need of immediate protection
from the encroaching cultivator.
(ii) KAGUGU. This appears to be the site at Rushozi, close by the Land Use
Experimental Area house, the outlines of which can be clearly seen in the dry

season. There is plenty of pottery, but no considerable embankments-only a
few low mounds. A bigger site than the first, however.

Sites of Mutambuka and Ntare V. c. 1850-1900.
I had no time to visit any of these sites, most of which are to be found in
the triangle formed by Mbarara, Rushozi and Bwizibwera. Perhaps as a result
of raids by the Baganda, these two Bagabe seem to have adopted the semi-
nomadic habits of the important Bahima, shifting their residences every three
or four years. For Mutambuka, seven sites are mentioned, mostly in the
Rushozi area; and for Ntare there are at least eight sites. It seems likely that
this peripatetic life produced a serious decline in building standards, and the
sites of this period would not therefore be very representative of the series as
a whole.

This all too brief inquiry has conclusively proved that oral tradition, when
as carefully collected as it has been by Katate and Kamugungunu, can point
the way to historical sites of the last four or five hundred years; and that
wherever it exists, traditional history is the obvious starting-point for an
archaeological survey, of the kind which a Director of Antiquities or any
similar officer might be expected to carry out during the first year or so of his
tenure of office.
In a purely surface survey it was only to be expected that some of the sites
mentioned by tradition should yield no surface evidence, and that others
should have been seriously disturbed by cultivation. Nevertheless, in Kaga-
rama, Kichwekanu, Bweyorere, Nyabikiri, Nyamirima and Rukoma a series
of sites has been discovered, which are capable of adding very considerably
to our knowledge of Ankole history. These sites probably cover most of the
period from the mid-seventeenth to the mid-nineteenth century; and even from
surface evidence it is clear that Roscoe statements about the Mugabe's kraal
(The Banyankole, p. 36) stand in need of modification so far as this period
is concerned. Roscoe says that "the Mugabe's kraal differed from that of his
chiefs only in size" and also that "the site was changed at least every second
year and often every year". On the contrary, it would seem to me that the
frequent changing of sites was a late nineteenth century innovation, and that
before that time the size and complexity of the sites, and in particular the size
of the earthwork embankments surrounding them, indicate that the Mugabe's
orurembo was a town, not a kraal; that even in the nineteenth century it was
intended to last a decade rather than a year or two; and that in earlier centuries
a single site was occupied for a whole reign. Accurate survey would
undoubtedly add greatly to the picture; and archaeological exploration,
directed primarily to discovering the original design and construction, and the
extent of 'creep' and erosion, would add more still. It is certain that pottery
could be discovered in large quantities, and it would be of great value to see
whether such pottery formed a sequence corresponding with the traditional
chronology of the sites. Finally, there is always the hope of discovering a date-
able import, such as beads.

My own view is that Katate's information about the number and sequence
of the Bagabe is probably correct at least as far as the sixteenth generation
back; and that this preliminary survey of the sites has produced nothing which
need throw doubt upon it. I think that his attribution of particular sites to
particular Bagabe, though frequently correct, is sometimes at fault. Most of
these faults are probably capable of being cleared up by further inquiry.

IN 1902

NO one has yet written a history of the Great Lukiko. It may soon be
possible to give some account of its development during the eighteen nine-
ties. But in the meanwhile one document, which is of obvious importance,
deserves printing in full. This sets out the first complete list of members of the
Lukiko, following its reconstitution by the Uganda Agreement of 1900.
Before 1900 the membership of the Great Lukiko had never, it seems, been
closely defined. Article 11 of the 1900 Agreement stated, however, that in
"The Lukiko or native council, shall be constituted as follows:
In addition to the three native ministers, who shall be ex officio senior members of the
council, each chief of a county (twenty in all) shall be ex officio a member of the Council.
Also each chief of a county shall be permitted to appoint a person to act as his lieutenant
in this respect to attend the meetings of the council during his absence, and to speak and
vote in his name. The chief of a county, however, and his lieutenant may not both appear
simultaneously at the council. In addition, the Kabaka shall select from each county three
notables, whom he shall appoint during his pleasure, to be members of the Lukiko or
native council. The Kabaka may also, in addition to the foregoing, appoint six other
persons of importance in the country to be members of the native council ... "
Yet it was not until over two years after the Agreement had been signed that
a complete list of members of the Lukiko, in accordance with Article 11, was
produced. It was then drawn up by George Wilson (who by 1902 was Deputy
Commissioner of the Protectorate, and the British official in charge of Buganda)
who obtained the necessary information from the Regents and the Saza Chiefs.
There was evidently some difficulty in completing the list, primarily, it seems,
because some of the Saza Chiefs' Mumyukas were objecting to their appoint-
ment as 'lieutenants' in the Lukiko (in accordance with their traditional position
as the Saza Chiefs' deputies) unless they were paid for their services.' But
eventually on 26 August 1902 Wilson was able to forward to Colonel Hayes
Sadler, Her Majesty's Commissioner for the Uganda Protectorate, at Entebbe,
the first complete return of members of the Great Lukiko.2 It was headed
'Confidential Uganda Lukiko', and ran as follows:
1 Kago. P 1 Mugimma. P 1 Luwekula. C
2 Sekyoya. P 2 Burisi. P 2 K. Muganwa. C
3 Musarosaro. P 3 Sekezi. P 3 Kabuzi. C
4 Kibale. P 4 Sabaragira. P 4 Kanya. C
5 Sebalija. P 5 Sebugwa. P 5 Katologa. C
6 S. Omutegomba. P 6 Muwanika. P 6 M. Kilevu. C
7 A. Gabunga. P
1 Wilson to Sadler, 8 May, 26 August 1902, Entebbe Secretariat Archives, A8/2; Wilson
to Sadler, 7 June 1902, Buganda Residency Archives, General 1902-1907; Wilson to
Sadler, 21 August 1902, BRA. General, 1902.
2 Enclosure in Wilson to Sadler, 26 August 1902, E.S.A., A8/2.


1 Mukwenda. P
2 Mwemba. P
3 Kajongolo. P
4 Omuyonjo. P
5 Munakulya. P
6 E. Kyakulese. P

1 Sekibobo. P
2 Kauta. P
3 Nanfumbambi. P
4 Muchwa. P
5 Omukabya. P
6 J. Kajugujwa. P

1 Kaima. C
2 Omulwanya. C
3 Kawaola. C
4 Lutuga. C
5 D. Wamala. C

1 Kangao. P
2 Yoana-Nsenge. P
3 Kisubika. P
4 Nachama. P
5 Masika. P
6 Muwanguzi. P
7 Omulwaza. I

1 Matayo Mugerere. P
2 Lutwana. P
3 Mariki. P
4 T. Lukanika. P

1 Omutesa. P
2 Lutakome. P
3 Erisa Aridiki. C
4 M. Luimbazi. P

1 Kitunzi. P
2 B. Omwaziza. P
3 Mawaranti. P
4 Gidioni-Ntanda. P

1 Pokino. C
2 Kagola. C
3 Katabalua. C
4 Bugana. C
5 Namwenda. C
6 Kasilye. C
7 Serego. C
8 Mukude. C

1 Katambara. I
2 Sekyeru. I
3 Muyanja. I
4 Sebalu. I

1 Kasuju. C
2 Banja. C
3 P. Kasabu. C
4 A. Kajarero. P

1 Kamswaga. P
2 Kabumbuli. P
3 Namgine. C
4 P. Sempango. C

1 Kiimba. C
2 MugimbaOmumuzi.C
3 Lwamubogi. C
4 Kauwa. C
5 Wakubirwa. C

1 Kyambalangu. C
2 Kamyuga. C
3 Semeo Nsubuga. C
4 Mutagoluwa. C

1 Kimbugwe. P
2 Mude. P
3 Sebuta. P
4 Semukata. I
5 W. Omuzimba. P

1 Kweba. P
2 T. Seluma. C
3 Nkazi. C
4 Muwambi. P
5 Kyakwambara. C
6 Tarika. P
7 W. Sewaya. C

1 Mbubi. C
2 Z. Kalinabiri. C
3 Sematimba. P
4 Mpembi. P
5 Kezala. C

1 Lumama. P
2 Lwanga. P
3 Luwarira. C
4 K. Mutatya. P

Byakwoyamba Lazaro. P
Total 106

NOTE.-Ex officio senior members of Lukiko.
Apolo Kagwa, Gulemye, Katikiro.
Sitanisirasi Mugwanya, Omulamuzi.
Zakaliya Kisingiri, Omuwanika.

The list would seem to be authoritative. But it contains a few anomalies. The
Saza Chief, the 'lieutenant' and one of the 'notables' of Mawogola are shown
as Protestants, when in fact the Saza Chief, Andereya Kiwanuka, was almost
certainly still a Roman Catholic and, from his first name which was Lazalo,
probably his 'lieutenant' also. Had, as the list suggests, the Saza Chief really
been a Protestant, this would have given the Protestants twelve of the twenty
Sazas into which Buganda was divided, which seems on other grounds more
than in fact the Protestants ever enjoyed.
More interestingly the list differentiates between the importance, or at all
events the size, of different Sazas, by ascribing more members to some Sazas
than to others, although the Agreement definitely stated that "the Kabaka shall
select from each county three notables". The list shows 8 Sazas with only
2 'notables', four with 3 (the proper number), four with 4, Busiro, and Sese
and Bulemezi, with 5, and Buddu with 6. This was almost certainly in contra-
vention of the Agreement.
At the same time the full number of members of the Lukiko is made up
without there being any special reference to the Kabaka's nominees, as they
have since been called. Upon this point, however, the list was not contravening
the Agreement, for according to Article 11 all those who were neither ex officio
members of the Lukiko, nor the Saza Chiefs' 'lieutenants', were to be selected
by the Kabaka. The Agreement merely stated that besides appointing three
notables from each Saza "the Kabaka may also, in addition to the foregoing,
appoint six other persons of importance in the country to be members of the
native council". It may well be that Lazaro Byakwoyamba was one of these.4
The total of 66 'ordinary' members (i.e. those who were neither ex officio
members, nor 'lieutenants') is made up by including him, and is the right total
of 'notables' and 'other persons', as laid down in the Agreement. It seems,
therefore, that, to begin with, there was, for the most part, no differentiation
between the 'other persons' and the 'notables'.
It will be seen too that the Saza Chiefs are all named in the list simply by
their titles; except Mugerere, who for some reason has been allowed his
Christian name, Matayo. The remaining members are listed, however-it
appears indiscriminately-either by their personal names, or by their titles,
where they had such.
The second name on each Saza roster-that is the italicized name-is the
Saza Chief's 'lieutenant'. Fifteen of the Saza Chiefs' Mumyukas are to be found
3 Windham to Wilson, 24 October 1902 B.R.A. General 1902-1907.
4 He may well be the son or successor of Yafeti Byakweyamba (though report has it
that Yafeti had been mutilated in his youth). Yafeti was the elder kinsman of Kasagama,
Mukama of Toro, and had died in 1897 (U.J., 12 (1948), 38 note). Lugard (Rise, ii, 151)
notes that Yafeti as a refugee at the court of Kabaka Mutesa had "obtained some office
about the person of the king, who showed him great favour". Lazaro was perhaps regarded
as one of the inner court circle.

in the list as a whole; but ten of these appear amongst the 'ordinary' members,
and only five as 'lieutenants'. Even so, every 'lieutenant' was a senior Mukungu
chief (if not Mumyuka, then Sabadu, Sabagabo, Sabawali, Mutuba omukulu, or
Mutuba omuto of his Saza). Every 'lieutenant' was of sufficient importance to
be allotted eight square miles or more under the land allotment; and with the
exception of the 'lieutenant' from Sese,5 they were all of the same religion as
their Saza Chief.
At least 37, moreover, of the 66 'ordinary' members of the Lukiko were
senior Bakungu chiefs (that is, either Mumyuka, Sabadu, Sabagabo, Sabawali,
Mutuba omukulu or Mutuba omuto of their Sazas); so that, if one includes the
23 ex officio members (the three Ministers and the 20 Saza Chiefs) at least 60
of the 89 members of the Lukiko were senior members of the chiefly hierarchy.
And a comparison with the land allotment lists under the terms of the 1900
Agreement, and under the terms of the supplementary memorandum of 13
October 1900, signed by J. F. Cunningham, the Secretary to the Special Com-
missioner, Sir Harry Johnston, yields some further information.6 For it appears
that of the full list of 89 members, at least 69, including, of course (though, as it
happened, with one exception) all these senior chiefs, were granted eight or more
square miles of land under the land allotment. There was a heavy preponder-
ance, therefore, in the Lukiko, of the leading men in the country, whether
'leading men' be defined in terms of chieftainships held, or by the amount of
land owned.
At the same time it is worth noting that of the further seven names from this
Lukiko list which can be traced in the land distribution lists, six do not appear
at all in the first, and most important, of the allotment lists, but do appear in
one of the two supplementary lists which were drawn up before 1905. This
suggests that there was after all an element in the Lukiko, though a small one,
of lesser men; but that these were very probably the close personal followers
of Saza Chiefs, not qualified to benefit from the main distribution of the land,
but men whom the Saza Chiefs were able to reward with small grants of mailo
land, when they had a supplementary distribution to complete later on, and
whom some Saza Chiefs occasionally suggested for nomination to the Lukiko.
Equally significant, of course, are the religious alignments which the list
reveals. It seems, to begin with, that although a few Protestants were occasion-
ally appointed for a Saza with a Roman Catholic Saza Chief, and vice versa,
only two of the 'ordinary' members appointed for the ten pre-European Sazas
of Buganda did not belong to the religion of their Saza Chief. This indicates
something of the strength of religious distinctions at this time in the heart of
Buganda. The fact, however, that these alignments were not quite so marked
in the newly attached and peripheral Sazas suggests that other considerations
applied there such as, perhaps, that Baganda overlords should pull together;
there was, for example, the remarkable combination in the newly annexed, and
5 He was a Roman Catholic, though the Saza Chief was a Protestant. This may well be
in consequence of the Roman Catholic hold over the islands between 1893 and 1900, when
it became subject to Protestant authority once again, as it had been once before in the
period before 1893.
6 The volume entitled 'Lukiko Allotment Lists, Nos. 1, 2, and 3' is in the possession of
the Land Office, Kampala.

still difficult, Saza of Buvuma, of Nova Naluswa, the former Roman Catholic
Gabunga, and his Mumyuka, Mika Sematimba, the first baptized Protestant
whom the Rev. R. H. Walker of the C.M.S. had taken on a visit to England in
But the most revealing figures of all come from the Lukiko list as a whole. It
will have been noticed that each name is followed by a letter, either P (for
Protestant), C (for Roman Catholic), or I (presumably for 'Islaimu'). At the
bottom of the list it has these totals:
P. 43
C. 60
M. 6

109 including Regents.

This total, of course, includes the 'lieutenants'. If they are subtracted, it appears7
that the full members of the Lukiko in 1902 were divided thus:8
Protestants 49
Roman Catholics 35
Muhammadans 5

Total 89
It will be noticed that there were no 'pagans'. Every member of the Lukiko
adhered to one or other of the new religious creeds that had been introduced
into the country during the previous half century. This is some measure of the
extent of the political revolution which had occurred within Buganda during
that time. The figures also provide a guide to the balance of power between the
religious parties in the country at the turn of the century. As measured by the
religious adherence of members of the Lukiko, this does not appear to have
altered very greatly, in all the years which have supervened.
7 I have taken the list at its face value, and have made no allowance for the mistakes
(as I believe) in the Mawogola roster. They would make, in any event, little significant
8 The figures Wilson gave in May 1902, before the list was finally settled, were Protes-
tants 48, Roman Catholics 36, Muhammadans 5. Wilson to Sadler, 8 May 1902.
E.S.A. A8/2.



W HEN Uganda's Eastern Province was transferred to the East Africa
Protectorate on 1 April 19021 there went with the Province a number of
Uganda officials. Of these the most important were Frederick Jackson appointed
to be Deputy Commissioner in his new Protectorate with a special responsibility
for the up-country tribes, and C. W. Hobley, Sub-Commissioner in charge of
the transferred Province. Between these two men of long experience in Uganda
a difference developed over the future boundary between the two Protectorates.
They were to take with them more than the Eastern Province, though they
argued about how much this should be. This happened despite Sir Clement Hill
at the Foreign Office thinking of the new border as following "natural boundar-
ies" from the end of the railway north to Lake Rudolf. His "only very approxi-
mate" line he sketched on a map: he began from the mouth of the Yala and
swung away in a north-easterly direction until he reached the Elgeyo Escarp-
ment which he took as his main feature northwards. He thus left to Uganda the
north-western portion of the old Eastern Province.2
However, these were only cogitations in the Foreign Office. Sir Charles Eliot,
Commissioner in the East Africa Protectorate, and Jackson, who was Acting
Commissioner in Uganda during the interregnum between Sir Harry Johnston
and Colonel Hayes Sadler, did not receive any definite instructions as to the
frontier north of the lake. When, therefore, they met at Njoro in November
1901, it was they who laid down the principle of following tribal divisions. This
meant a line further to the west and raised the whole question of Elgon which
lay entirely within the Central Province of Uganda. They recommended that all
the Kavirondo should fall to East Africa, leaving to Uganda "all whose popula-
tion is more akin to the Baganda". They thus wished to begin from Berkeley
Bay on Lake Victoria, by way of the Sio river. About Elgon they had not
sufficient information to decide within which sphere it should be included, and
wished to consult Hobley who was on leave in England. North of the mountain,
they concluded, the boundary was then "of purely academic interest" since
they were not likely to extend effective administration there for some time to
Before Eliot's despatch reporting this reached the Foreign Office Hobley had
sent in a memorandum setting out the first draft of his line, the one that was
eventually to be effective. He pointed out that the western boundary of the
Eastern Province had many disadvantages since it cut right across the heart
of the Kavirondo country regardless of tribal boundaries, several "important

I See article by Kenneth Ingham in the Uganda Journal, 21 (1957), 41.
2 F.O. 2/519 Memoranda by Sir C. Hill, 25 and 27 July 1901, map in the second.
3 F.O. Confidential Prints: East Africa, Part 68. Eliot to Lansdowne, Njoro, 30 Novem-
ber 1901 (the original of this despatch is at present in the Colonial Office in an uncata-
logued volume).

tribes" having land on both banks of the Yala. This was bad for administration
as between Provinces but the situation would be worse between Protectorates.
Hobley, like Jackson and Eliot, preferred the division between Kavirondo and
Busoga but wished to start from Lake Victoria a little west of Berkeley Bay.
Thus far there was general agreement but controversy was to centre on Mount
Elgon. Hobley recommended that the boundary should reach the south-western
corner of the mountain and go diagonally across to join Sir Harry Johnston's
line of the western boundary of the Eastern Province.4
Hobley's memorandum was sent out by the Foreign Office to Eliot and
provoked an immediate telegraphic reply. This spoke of a "considerable differ-
ence of opinion" about Elgon. Sir Clement Hill realized that Eliot's telegram
had been sent before Hobley, returning from leave, had met Eliot and Jackson.5
The Foreign Office was willing to allow a delay for the discussion with Hobley,
even though they had intended to mention the new boundary in the Order in
Council then being drafted for Uganda.6
By the time Hobley arrived in Nairobi he had advanced his position. Jackson
was for leaving Elgon to Uganda, but Hobley now wanted the whole mountain.
He described the western and southern slopes as "a dense banana forest full of
rather truculent natives in a very low state of civilization whom he regarded as
akin to the Kavirondo". Eliot, feeling a need for more information, preferred
to leave the boundary as "an open question for the present" but, if compelled to
decide, he thought that they should follow Jackson. Whilst he acknowledged
Hobley as probably the best authority on the mountain tribes for scientific
purposes, Eliot could not believe that the political connection of the western
tribes with the Kavirondo was close. If, as it seemed, they occasionally raided
Busoga, then repression should be left to Uganda. Eliot again reiterated his and
Jackson's strong opinion that they should have "a very flexible boundary north
of Mount Elgon"-indeed, he had suggested in his telegram the week before
that power to modify the boundary should be reserved to them. Sir Clement Hill,
on reading the later despatch, was prepared to accept this view for the present
since the doubtful territory was not likely to become "an Alsatia for white
men", and if it did, they could summarily include it in one jurisdiction or the
other. He thought that "Mount Elgon should belong to Uganda if possible".7
Instructions in these terms were sent to Eliot.8
Before these arrived-or had even been penned-Eliot, hesitating because of
the dispute between Jackson and Hobley, found an excuse for further delay.
He wrote to the Foreign Office saying that he would await the arrival of Hayes
Sadler, the new Commissioner of Uganda, and discuss the matter with him,
Hobley and a Uganda official to be then chosen.a8
In the end the boundary was demarcated by a two-man Commission of

4 F.O. 2/523. Hobley to Sir C. Hill, Nuneaton, 10 December 1901.
5 F.O. 2/577. Eliot to Lansdowne, telegram, 18 January 1902.
6 F.O. 2/566. Lansdowne to Eliot, 24 January 1902.
7 F.O. 2/569. Eliot to Lansdowne, 25 January 1902.
8 F.O. 2/566. Lansdowne to Eliot, 19 March 1902.
8a F.O. 2/569. Eliot to Lansdowne, 14 February 1902.

Hobley and William Grant, the Sub-Commissioner of Busoga, who had accom-
panied Lugard to Uganda in 1890. They reported on 25 July 1902. Tribal
affiliations proved not to be the only factors, though the Commissioners began
by saying that the main features were that the Basoga and the Bakedi remained
in Uganda whilst the Kavirondo went to the East Africa Protectorate. Adminis-
trative considerations also played an important part. Thus several of the islands
near Berkeley Bay were allotted to Uganda though they were inhabited by
Kavirondo, because they traded with Busoga and could be administered from
there. The Sio river was accepted though found not to be such a good tribal
boundary as Jackson had expected: there were "a considerable number of
Kavirondo on the right bank" but as they had been administered for some time
from Busoga it was agreed to leave them to Uganda. The Commission pointed
out that on the southern flanks of Elgon were tribes which had been adminis-
tered from, and had paid a little hut tax at Mumias, whilst on the north side
were certain tribes of Nandi race. They, therefore, considered Hobley's sugges-
tion of drawing the boundary along the western flanks of the mountain leaving
"all the so-called Elgumi or Bakeddi tribes" in Uganda, and giving the mountain
to the East Africa Protectorate. Again administrative considerations prevailed:
they preferred to draw an imaginary line from the Namisindwa Bluff at the
south-western corner of Elgon diagonally across to where, from the crater, the
river Suam rose, then to follow it to its junction with the Turkwell and thence
by that river to Lake Rudolf. Jackson, Acting Commissioner of East Africa in
Eliot's absence in England, reported that Sadler had concurred but left the
final decision to the Foreign Office. There Eliot was shown the despatch and
commented: "I do not much like this boundary but it has been prepared by
local experts." He considered it of no importance if amalgamation of the two
Protectorates were still intended but, if not, it should be recognized as "merely
provisional". He noted that they had no evidence of the tribal divisions on the
Suam and the Turkwell.9 In November the Foreign Office approved provision-
ally10 but the matter only attained full consequence when administration reached
the boundary. Then an adjustment was made, some territory being transferred
from Uganda in 1910."
Thus the major transfer of 1902 consisted not only of the old Eastern Pro-
vince but also of the southern part of the Elgon district of the Central Province,
based around the famous station of Mumias, and the first bite out of Rudolf
Province, of which the remainder was to follow in 1926. It was the Foreign
Office intention that this vast area should be administered as one unit: Sir
Clement Hill himself christened it 'Lakes Province' and it was so designated in
the official notification of transfer.12 Eliot immediately replied that all his
advisers were agreed that the new province was too large a unit for the East
African system, both administratively and judicially. He suggested the formation
of two Provinces: Naivasha and Kisumu. The Foreign Office agreed, Lord

9 F.O. 2/573. Jackson to Lansdowne, 31 August 1902.
10 F.O. 2/567. Lansdowne to Eliot, 7 November 1902.
11 East Africa Protectorate Report, 1910-11 (Cd. 6007), p. 41.
12 F.O. 2/566. Lansdowne to Eliot, 6 March 1902.

Lansdowne accepting that they could not put themselves in opposition to the
local authorities on such a point, though Sir Clement Hill wryly minuted that
he gave up his name of Lakes Province "with some godfatherly regret".13

13 F.O. 2/577. Eliot to Lansdowne, telegram, 10 April 1902, but see also F.O. 2/570
Eliot to Lansdowne, 28 March 1902, where he had spoken of the natural division of the
two provinces in terms of Naivasha Province being suitable for white settlement and
Kisumu not so suited because of climate and "a thick native population which is not in
all cases friendly".


THE results which may follow the adoption, in the absence of adequate
topographical information, of a specified line of latitude as an international
frontier are instanced by the situation which formerly obtained at the mouth
of the Rovuma River (H. B. Thomas, 'The Kionga Triangle' Tanganyika
Notes and Records, No. 31 (1951) p. 47). The difficulties were there resolved
after World War I by a realistic concession to common sense; and the course
of the Rovuma River is now the boundary between Tanganyika Territory and
Portuguese East Africa.
A somewhat similar state of affairs is in existence at the other extremity
of Tanganyika Territory where around the lower reaches of the Kagera River
its frontier marches with the Uganda Protectorate. At a distance of about
55 miles to the south of Masaka, and while still some 20 miles short of the
Kagera River at the Kyaka Ferry, the traveller making for Bukoba and facing
a featureless plain framed by distant forests learns that he has crossed an
international frontier and is no longer in Uganda. Not a few of those who have
passed this way during the past half century must have wondered if the main-
tenance of so artificial a boundary serves the best interests of the development
of the isolated and sparsely populated tract of country which it bisects. Twenty
miles to the east, at the very mouth of the Kagera River, the Uganda authori-
ties have a comparable anomaly on their hands. Here they have granted to
Buganda chiefs documents of title to land-in ten parcels comprising some
1,321 acres-lying to the south of the river, which at this point enters Lake
Victoria a few miles north of the boundary-the parallel of 1 south latitude.
The present position is that a triangular area containing say 15 square miles
to the south of the Kagera, together with Busungwe Island and even the
northern extremity of Rubabu Point, all of which may for convenience be
referred to as the 'Kagera Triangle', lies within the Uganda Protectorate:
while a segment-shaped area, the 'Kagera Salient', containing some 660 square
miles and lying to the south of the parallel of 1 degree south latitude but
north of the Kagera River, forms part of Tanganyika Territory.
Burton, the explorer, when at Tabora in 1858 was the first European to
glean information regarding the Kagera or Kitangule River-later referred to
by Stanley as the Alexandra Nile. In The Lake Regions of Central Africa,
1860 (vol. ii, p. 177) he refers to "the kingdom of Karagwah, which is limited
on the north by the Kitangure River", and later he records that the first two
marches north from the Kitangule River traverse the territory of "dependent
Unyoro" so called because it had lately become subject to the Sultan of
Speke, who was the first European to see the Kagera River, crossed from
south to north on 16 January 1862 at or near the present Kyaka Ferry,
approximately at the most southerly point of the salient. The following quota-
tions are from his Journal of the Discovery of the Source of the Nile, 1863.


I 3 0 '3 o' 3 'd 3 1'3 0 -
MILrS S 0 5 t0 Is 20 25 30 35 MILES

FIG. 1
The Kagera Frontier.

"To reach Unyoro, the party would have to cross a portion of Uddu, which
the late King Sunna, on annexing that country to Uganda, had divided by
alternate bands running transversely from Nkole to the Victoria Nyanza"
(p. 243). "The kingdom of Kittara (i.e. the kingdom of which the nucleus and
only remnant is now Bunyoro) bounded by the Kitangule Kagera on the south
and the kingdoms of Utumbi and Nkole on the west" (p. 249); "Uddu beyond
the river" (p. 262); "Sunna, after annexing this part of Uddu to Uganda, gave
Rumanika certain bands of territory in it as a means of security against the
possibility of its being wrested out of his hands again by the future kings of
Unyoro" (p. 265).
This last suggestion, that Suna gave certain bands of this area to Rumanika
of Karagwe, is the only foundation which has come to light for any claim to
the ownership of this adverse to that of Unyoro or Uganda. But a careful
reading of the contemporary literature shows that this 'gift' was at the highest
not more than a delegation by Suna, the suzerain, to Rumanika, his tributary
(who had indeed only retained his throne with Suna's help) of a share of the
responsibility for the protection of what was regarded as Buganda territory.
Subsequent research has established the fact that Speke was in error in
placing the wresting of Buddu from Bunyoro so recently as the reign of King
Suna. This conquest was made by King Junju of Buganda towards the end of

the eighteenth century (Roscoe, The Baganda, 1911, p. 225). Buddu signifies
the country of Bairu, Baidu, Ba'du (Lunyoro-Luganda) or serfs.
Grant, Speke's companion, crossed the Kagera in May 1862. His Walk
Across Africa, 1864 (p. 204), records that "Some of the Wazeewa had
migrated from the right to the left bank of the Kitangule, and were now
cultivators under the King of Uganda". The Wazeewa were the inhabitants
of the Buziba country occupying a territory north-eastwards from Karagwe
to Lake Victoria. As evidence of the mixed elements of the population is
Grant's statement that the country between the Kitangule and Katonga Rivers
was occupied by Wanyambo helott aborigines from Karagwe), Wanyoro,
Wazeewa (i.e. Baziba) and Waganda.
Stanley was in this neighbourhood in 1875-6. In his Through the Dark
Continent, 1879, he states "The Alexandra Nile constitutes a natural boundary
between the sovereignty of Uganda and its subject kingdoms of Karagwe
and Uzongora (sc. Buziba) which begin south of the river" (Chap. X). "The
great war-drum of Masaka (the headquarters of Buddu) sounds the call to
war, and the natives from the banks of the Alexandra Nile to the Katonga
respond to it by thousands" (Chap. XV).
Of this period preceding the British occupation, Roscoe (The Baganda,
p. 255) records that the Pokino (District Chief of Buddu) was the overlord of
the Kiziba country; and Stanley, while in Karagwe on his return to the coast
with Emin in August 1889, evidences the hold still maintained by Uganda far
south of the Kagera (In Darkest Africa, 1890, Vol. ii, p. 378).
The origin of the adoption here of the parallel of 1 degree south latitude
as a boundary is of interest. The 'hinterland' doctrine that each European
power should have the exclusive right to influence in the regions of the
African interior subtended by the section of the coastline which it controlled,
found expression in the Anglo-German Agreement of 29 October/1 November
1886 (Hertslet, Map of Africa by Treaty, 1909, No. 264), when the line of
demarcation between the respective spheres was laid down as running from
the mouth of the River Wanga or Umbe by various courses "to the point on
the eastern side of Lake Victoria Nyanza which is intersected by the 1st
degree of south latitude".
This left the control of Uganda and the west side of Lake Victoria inde-
terminate, but the Agreement of July 1887 (Hertslet, No. 267) left Great Britain
with the impression that Germany was satisfied with the recognition in its
(Germany's) favour of a free hand "in the territories south of the Victoria-
Nyanza Lake". In September 1888, however, the initiation of Carl Peters'
project for leading a German expedition for the relief of Emin Pasha at
Wadelai became known, and Sir William Mackinnon on behalf of the recently
chartered Imperial British East Africa Company hastened to point out to the
British Government that the utterances of the German promoters made it
clear that they planned further to acquire a portion of territory to the west of
Lake Victoria, thus trespassing beyond the limit (that is the south of Lake
Victoria) to which it was understood that Germany had undertaken to confine
herself by the Agreement of 1887. Sir William Mackinnon urged the need for a
formal delimitation of the boundary west of Lake Victoria on the lines of the

Agreement of 1887, that is to say "by drawing a line due westwards from the
southernmost point of Victoria Nyanza till it meets the eastern boundary of
the Congo Free State". (See McDermott, British East Africa or IBEA (2nd ed.
1895), p. 15.)
In October 1889 German chauvinism was again evidenced by the declaration
of a protectorate over the coast between Witu and Kismayu (Hertslet, No. 210,
p. 689) which, in accordance with the 'hinterland' doctrine would afford a
basis for claims in the interior to the north even of 1 degree south latitude.
A. M. Mackay, the pioneer missionary of Uganda, must have been aware of
Sir William Mackinnon's representations when he wrote to H. M. Stanley the
letter, dated 5 January 1890 which is printed in the latter's In Darkest Africa
(vol. ii, p. 392). In this he says "I have written to Sir William Mackinnon's
agents in Zanzibar, explaining the absurdity of their acceding to Germany's
wish to draw the boundary-line west of the Lake, along the 1st parallel of
S. Lat., as that would cut the kingdom of Buganda into two halves: for
Karagwe, Usui, and Usinja, as far south as Serombo, are actually part of
Buganda, being tributary to it. No paper delimitation, made in Berlin or
London, can ever remove these states from their allegiance to Buganda."
Nevertheless the exigencies of political compromise were embodied in the
Anglo-German Agreement of 1 July 1890 (Hertslet, No. 270) by which the
parallel of 1 degree south latitude was adopted as the boundary from the west
shore of Lake Victoria to its intersection with the frontier of the Congo Free
State. It is this boundary, determined without consideration of anything but the
political expediencies of two European powers, which has, in the section
between Lake Victoria and the crossing of the Kagera River by this parallel
at a point some three miles east of Nsongezi, been rigidly maintained as the
international boundary to this day.
Captain (later Lord) Lugard was satisfied that the Kingdom of Buganda
had legitimate claims on territory to the south of the Kagera River. In his
report of 24 December 1890 (Blue Book, Africa No. 4 (1892), p. 102) he says
"The districts of Karagwe, Kiziba, Buzinga, Serembe, Usui, etc. below the
first degree of latitude, which are now ceded to Germany, have for very many
years (since the time of Kamanya, predecessor of King Suna) been considered
an integral part of Uganda, and have paid tribute to the King. If it appears
to you feasible to obtain for the King, who is now under British protection,
any compensation in lieu of this tribute, which is presumably his legal right,
I have the honour to ask that you will take such steps as may appear to you (to)
be right in communication with the German Company."
The Baziba, who according to Grant (1862) had been permitted by King
Suna to settle north of the Kagera River, maintained their identity. They were
one of the elements of the population of Buddu whose status called for con-
sideration in the territorial re-arrangements following the religious distur-
bances of 1892 in which, despite much temptation, they declined to join the
Roman Catholic factions against the Company. The settlement is mentioned
in Lugard's Report No. 4 of March-August 1892 (Blue Book, Africa, No. 2
(1893), p. 68). "The Buziba-These people have always been friendly to us.
I understand that they have been settled in Buddu for a very long time and

that their fields and gardens have been reclaimed from the waste lands by
themselves, and are therefore entirely their own. It is therefore only just that
they should be supported and protected . Their estates are mixed up with
those of Buddu chiefs . ". The substance of this arrangement was that they
were not to be disturbed when Buddu was evacuated by the Protestants to
make room for the Roman Catholics.
The local opinion in 1895 of the artificial boundary provided by the parallel
1 degree south latitude is given in informal terms by Sir Frederick Jackson in
his Early Days in East Africa, 1930 (p. 283). "Among other things which
von Kalben (sc. officer-in-charge of the German station at Bukoba) and I
discussed and agreed to represent in the strongest terms to our respective
Governments, were the disadvantages of the one degree of South Latitude as
our frontier, in place of the Kagera River, which the former crossed and
recrossed in several places.
"The small areas within the various bends were mostly marsh and of no
value but were regarded as a harbour of refuge to all sorts and conditions of
'wanted by the Police' who, when hard pressed, found it easier and safer to
step across an imaginary line and defy their pursuers than to face a deep river
full of crocodiles. However, no notice was taken of our pleadings and nothing
was done. I have always understood that politicians at home like to have a
little bit of land to play with, particularly if it is an annoyance to a neighbour;
it may some day come in handy as a useful bit to exchange for a bit elsewhere.
In the meantime local annoyances, that might lead to unpleasantness, are of no
Jackson's official reference to this matter is to be found in State Papers-
East Africa Section No. 627/1895. No. 29 in a letter to Sir Clement Hill dated
7 June 1895. "A short time ago the German officer at Bukoba, Von Kalben
by name, paid me a visit. I had a long talk with him on many points and,
amongst other things, he proposed that the Kagera River should be the
boundary between the two spheres."
Von Kalben had in fact been sent to Uganda by his senior officer, Captain
Langheld, commanding the German Nyanza Province, in order to discuss
outstanding problems (May 1895). Langheld who had first reached Bukoba
in time to take over from Emin Pasha in February 1891 must have had an
unrivalled knowledge of the Kagera region at this period. His book Zwanzig
Jahre in deutschen Kolonien (Berlin, 1909) gives no suggestion that he sought
control in any territory north of the Kagera. Not until August 1895, following
von Kalben's visit to Entebbe, does he record a short tour into German
Nevertheless the Anglo-German Boundary Commission, 1902-04, ultimately
demarcated the line of 1 degree south latitude, and demonstrated, in territory
hitherto almost unknown, its unsuitability as a boundary. But Imperial
Germany in the first years of the twentieth century was in no mood to abate the

I Von Kalben died at Bukoba 13 February 1896. Lieutenant-Colonel Wilhelm
Langheld died of wounds at Kurzcany on the Polish front on 9 July 1917. Langheld's
relations with the British authorities in Uganda from Lugard until he left the Lake
region at the end of 1895 were always on a most cordial footing.

plenitude of its legal claims;2 and this demarcated line became, for a distance of
some 70 miles west from Lake Victoria, the International frontier. It does not
seem however, that these accepted boundaries between Uganda and German
East Africa as they existed in 1914 were ever done into a comprehensive Anglo-
German convention. (Thomas and Spencer, History of Uganda Land and Sur-
veys, Entebbe, 1938, p. 39.)
There is much evidence that the Germans themselves did not regard the
Kagera Salient as part of their Buziba district, and it is constantly referred to
as German Buddu. For example Paul Kollmann who served in this district with
the Imperial troops, in The Victoria Nyanza, 1899, devotes a chapter to the
crafts of Uganda, his observations regarding which appear to have been made
at the village of a Muganda chief in German Buddu; and he accords separate
treatment to the Kiziba country.
Hermann Rehse in Kiziba Land und Leute (Stuttgart, 1910) remarks "Die
Kagera bildet die nordwestliche Grenze des Landes". His map, on which the
Kagera Salient is shown as "Buiru (Buddu)" is evidence that the German
authorities did not regard this area as part of their Kiziba Landschaft. Again
the Duke of Mecklenburg (In the Heart of Africa, 1910) mentions that on
crossing (in 1907) the torrential Kagera at the Kifumbiro outpost (i.e. near the
Kyaka Ferry) "we left Kisiba to enter into Buddu territory".
So matters stood at the beginning of the 1914-18 War; and so they remained
until such time as the Principal Allied Powers and the League of Nations came
to decide the future of the ceded German territories.
The British Government can hardly have been unaware of the unsatisfactory
nature of this rectilineal boundary. And the League of Nations demonstrated
its readiness to countenance, in the interests of administrative convenience,
small variations of the frontiers of the new Tanganyika Territory from those of
the former German East Africa. Not only was there the adjustment in favour of
Portugal of the Kionga Triangle; but by the Milner-Orts Agreement of 1919
Great Britain and Belgium agreed that a corridor on the left bank of the upper
reaches of the Kagera River should be included in the British sphere to form
the Lusaka sub-district of the Tanganyika mandated territory. And when later
this division was found to cut through the territory of the native ruler of
Ruanda, Britain and Belgium agreed, with the approval of the League of
Nations to the retrocession of this area to Belgium as at 31 December 1923. As
The Times (21 January 1924) reported, the Belgians "expressed their deep
gratitude for the spirit of equity and the sincere desire to respond to the wishes
of the native population manifested by Great Britain in these negotiations".
Nevertheless Great Britain refrained from representing the case for an
exchange of the Kagera Triangle for the Kagera Salient; and in this decision, it
may be assumed that other considerations than the interests and convenience of
the local peoples prevailed.

2 Indeed in 1900 the Germans had established a post at Kakuto, in Buddu, some
9 miles north of latitude 1 S. (see, Thomas and Spencer, History of Uganda Land and
Surveys, p. 12).



THE craft of bark-cloth making is known to a greater or lesser degree in most
parts of Uganda. Three species of fig trees are used for its production. The
principal source for this malleable bark is Ficus natalensis, known in Luganda
as mutuba. Another related tree, Antiaris toxicaria, sometimes known as the
False Mvule, yields an inferior type of cloth. In the past the craft was mainly
practised in the areas now known as Buganda, Bunyoro, Ankole, Toro, and
We must turn to Buganda for the story of the evolution of the mallet now
used for the preparation of the cloth. This, J. Kakoza relates, evolved from
the nkonyo, a wooden block originally used by women for beating the core of
the banana stem into a pulp, which was used to scrub down the body. From
the nkonyo there came to be made a hammer having a single flat but grooved
surface with which to beat the bark.' As the demand for bark-cloth for such uses
as wrappings, coverings, shrouds, or curtains increased, the craft and the tools
used in it developed. Eventually the implements which came to be commonly
used consisted of three different types of wooden hammer, each having lateral
grooves encircling a cylindrical head. In Buganda these hammers or mallets,
made of hard wood (usually Teclea species), are known as:
ensamu esaka-a hammer with large grooves.
ensamu etenga-a hammer with fine grooves.
ensamu enzituze--a hammer with even finer grooves.
Other types of hammer which appear to have been used in the past were
made of bone. In the making of miniature hammers, purely for ceremonial
purposes, ivory has been used as well as wood.
According to tradition the craft of bark-cloth making is said to have been
practised before the arrival of Kintu, the legendary founder of Buganda King-
dom. There is also extant a tale amongst elder craftsmen that Kintu brought
an ivory hammer with him when he came to the country of his adoption. Other
versions say that the craft came into its own in Buganda in the 16th century
during the reign of Kabaka Kimera (Kakoza 1949). Tradition also tells of
another ruler, Kabaka Kigala, who was particularly fond of wearing bark-cloth.
The royal bark-cloth makers of this time were, and have been ever since,
members of the Otter (Ngonge) clan. It is related that to show his appreciation
of their work Kigala ordered two small symbolic hammers to be carved out of
ivory to signify the clan's duty to make such cloth for the royal household.
Since those days it has been the practice for the bark-cloth makers of the Otter
clan to hand two ivory hammers to the new ruler at the time of his enthronement.
He in turn hands the miniatures to the chief maker of bark-cloth (mukomazi)
who retains them for as long as he holds office.
Such tokens, also made for other craftsmen accredited to the royal household,
served in particular as a means of identification in case of apprehension by the

king's officials (Roscoe 1911). One such miniature ivory hammer is in the
regalia of H.H. the Kabaka of Buganda. This is the Ganda type of hammer,
being mallet-shaped with the cylindrical head encircled by lateral grooves.
Little is known of this heirloom which is probably of some antiquity. Two
other such miniatures, but made of wood, are also in the regalia of the Nakaima,
Priestess of Mubende Hill (Uganda Museum E.53117/8).
It was not until the reign of Kabaka Semakokiro towards the end of the
18th century that the use of bark-cloth for wearing apparel became general in
Buganda as a substitute for skins (Thompson 1934). Cloth was, however, made
for and worn exclusively by members of royal families or notables for genera-
tions before that period.
It was not generally known that bone was used for bark-cloth hammers in
Uganda until the discovery in 1953 of two hammer-like objects within the
Munsa Earthworks in Mubende District (Lanning 1955; 1957). These hammers
are shaped differently from their wooden counterparts and have the grooves on
only one flat beating surface in the one case (Uganda Museum No. E.5403) and
on two flat beating surfaces in the other (Uganda Museum No. E.5609).
As a result of enquiries it has been learned that in Mawogola County of
Buganda Kingdom, in that part which was originally the independent chiefdom
of Bwera, bone hammers were used at the beginning of this century. Two or
three such bone implements are said to have been in the possession of the
Muhima overlord of the area. It is recalled, however, that none of these par-
ticular hammers was incised, as is required for bruising bark in its early stages
of preparation; they were smooth and rounded. Probably they played a part in
executions and were used for beating victims to death. My informant does
recollect having seen a bone hammer with grooves similar to one of the Munsa
hammers, which he has examined.
As has already been stated, both the common wooden hammer and the
known examples of wooden and ivory miniatures have a cylindrical head with
numerous encircling grooves. The two bone hammers from Munsa, however,
show a different technique.2 They are not mallet-shaped. One has a single beating
surface composed of lateral grooves on one side of the head only, whilst the
other has two beating surfaces, one of lateral grooves and one of rectangular
cross-hatching. Both hammers are incomplete having their reverse sides formed
by the marrow cavity of the bone from which they have been carved. The
workmanship of these hammers is of a good standard. The quartering or
cross-hatching is reasonably even and precise. All grooves have been cut deep
and firmly. This cross-hatching is not known in other bark-cloth hammers used
in Uganda. Because of their flat beating surfaces they might well be relics of
the early nkonyo beater type, for no guess can be made of their age.
These two hammers are not unique, for certain similarities are to be found
elsewhere in Africa and particularly further afield. Mrs. Trowell has observed
parallels with the Indonesian bark-cloth tapa, made of wood or stone, which
are similarly shaped and are often covered with incised patterns, including the
rectangular cross-hatching.3 Dr. Ad6 has also commented on the identical shape
of the Congo pygmies' hammer, mulumba, to the Polynesian tapa.4 The cross-
hatching pattern occurs throughout Central Africa, but is usually found only

Modern Bark-cloth Hammer from Buganda.

[Photo: E. S. Adams

FIG. 2 [Photo: Dr. G. Timmis
Bone Bark-cloth Hammer from the Munsa Earthworks. Mubende District, Buganda.

on the bases of beaters. However, hammers from West Africa, without handles,
have cross-hatching on the side and not on the base (Mauny 1954).
The incidence of hammers used for various purposes of beating is wide
throughout the world. Although in some cases their use might have been for
grinding and not for bark-cloth making, the similarity which can be seen in the
earliest beaters to those definitely known to be employed in the bark-cloth craft
deserves attention.
Stone, as may be expected, has been made to serve for hammering from
the earliest right up to modern times. Specimens of neolithic stone beaters
have been found in Western Sudan (Hu6 1913), French Equatorial Africa
(Bequaert 1947) and in French Guinea (Joire 1952; Mauny 1954). As a point
of interest the last-named beaters had cross-hatching on their bases (Lutten
1944). Bequaert has, in fact, drawn the conclusion that the implements, gener-
ally of stone but sometimes of bone, used by the neolithic peoples of Western
Sudan and of the north-east of the Belgian Congo, were typical of those
required for the preparation of bark-cloth. Later examples of the stone beater
have been found in the Congo. Travellers there have recorded having spoken
with native elders who recalled the use of stone hammers at the beginning
of this century.5 Dr. Desmond Clark has observed that the curious, long,
conical pestle stones from the Orange Free State might have been used for
some such purpose as bark-cloth beating.6
Ivory has also served as a beater at least since neolithic times. An ancient
piece cut from an elephant tusk with one end incised with cross-hatching was
recovered from Yebu in the Belgian Congo (Bequaert 1947). Likewise there
are from the Njoro River Cave in Kenya ivory (?) pestles which were perhaps
used for grinding grain or even, as has been suggested, for pounding bark.7
Cross-hatching on the base used as the beating surface is evidently common
and has been found on implements of stone, ivory and horn from Central
Africa as well as from Indonesia. Some Congo specimens are of undoubted
antiquity. Similarly worked wooden hammers have also been reported from
Northern Rhodesia and are still in use in West Africa. Beyond the continent
of Africa the wooden, stone and shell Indonesian bark-cloth beaters have
either lateral grooves or even elaborate designs apart from cross-hatching.
Stone implements so marked have also been recovered from the Malay
Peninsula (Evans 1930). It is interesting to note that the wooden beater,
whether square, round or oval, is usually scored with longitudinal or lateral
grooves. It is known in Borneo, parts of America, Oceania and New Guinea,
where again stone is sometimes employed (Horniman Museum 1924).
Whilst wood is the material most commonly used for making hammers in
forested countries where bark-cloth making is practised, as in Uganda, use has
also been made from time to time, of stone, bone, ivory, horn and other
materials. The following summary gives a brief picture of the use of materials,
other than wood, for making bark-cloth hammers in neighboring African
In the Sudan an ivory hammer has been recovered from Equatoria Province.8
In the Belgian Congo hammers of bone or ivory were recorded as being in use
when that territory was first opened up. Specimens of ivory and cow-horn

hammers have been collected.9 Ivory hammers have been found in use amongst
the people of Bondo, Uele.10 In the Ituri Forest on the Epulu River, pigmies
still use ivory hammers for beating bark-cloth. These are invariably made
from the end of the short tusk of a young elephant. The wider end of the
broken-off tusk is that used for pounding and is either cross-hatched or incised
with lateral cuts or lines. The ivory head is tied to a wooden handle with a
piece of vine." In Ruanda-Urundi hammers of cow-horn are said to have
been in general use in former times. A specimen has cross-hatching on the
beating surface. Ivory hammers, with cross-hatching on the base are also found
in Mbuba.12 An ivory hammer has been recorded from southern Tanganyika.
As is thus amply evident the craft of making bark-cloth is widespread over
Africa. Whilst hammers vary from tribe to tribe in type, material, shape,
position of beating surfaces, or patterns of grooves, there are nevertheless
many similarities. Outside influences there must be as a result of migration,
conquest, or trade; these influences must sooner or later have affected the
development of the tools of so common a craft. Nevertheless the variety of
hammers throughout the African continent alone suggests that each area may
well have developed its own particular technique in the fashioning of its own
implements. Types of hammers would be dictated not only by the limitations
of tribal skill but also by the availability of materials and the requirements
demanded of the implements; for some were used to produce rough sheets or
inferior wrappings, others to produce cloth suitable for wearing apparel, while
others were used to produce nothing more practical than strips of binding

Bequaert, M. (1947). Maillets a ecorce de pierre et d'os de l'Oubangui (A.E.F. et
Congo Belge). Bulletin de la Sociedt Royale Beige d'Anthropologie et de
Prdhistoire, T. LVIII.
Doke, C. M. (1931). The Lambas of Northern Rhodesia.
Evans, I. N. (1930). Notes on two types of stone implements from the Malay
Peninsula. Man, 124.
Joire, J. (1952). La Prdhistoire de Guin6e Frangaise. Conferencia international dos
Africanistas Ocidentais. LV.
Kakoza, J. (1949). The Making of Bark-cloth (in Luganda). E.A. Literature Bureau.
Horniman Museum publication (1924). The Evolution of Domestic Arts. Part II.
Second Edition.
Hud, E. (1913). L'Age de la Pierre au Fouta Djallon. Le Mans; Monnoyer.
Lanning, E. C. (1955). The Munsa Earthworks. Uganda J., 19.
- (1956). A Note on Two Bone Hammers from Munsa, Uganda. Unpublished.
Uganda Museum, Kampala.
- (1957). Two Bone Bark-cloth Hammers from Mubende, Uganda. Man, 37.
Lutten, E. (1944). Inventaire sommaire de materiel prehistorique de Guinde
frangaise. Notes Africaines, 22.
Mauny, R. (1954). A Note on two Neolithic specimens from French Guinea. Notes
Africaines, 61.
Roscoe, J. (1911). The Baganda.
Thompson, A. D. F. (1934). Bark-cloth Making in Buganda. Uganda J., 1.


I J. Kakoza (1954): Personal communication.
2 For full data see Lanning (1956 : 1957).
3 Mrs. M. Trowell (1954): Personal communication (see specimen 'Tapa or Bark
Beaters', University Museum of Archeology and Ethnography, Cambridge).
4 B. Add (1955): Personal communication.
5 B. Add (1955): Personal communication.
6 J. D. Clark (1956): Personal communication.
7 J. D. Clark (1956): Personal communication.
8 Antiquities Service (1956): Sudan Government.
9 Musde Royal du Congo Belge, Tervuren, Belgium.
10 B. Ad6 (1955): Personal communication (see specimen hammers ivory and wood,
Uganda Museum Nos. E.56.10-16).
11 Mrs. Anne Putnam (1955): Personal communication.
12 Mme. E. Maquet (1954 : 1957): Personal communication.


THERE is considerable evidence of ancient contact by sea and land between
south Asia and Africa. Despite this, there have been few serious attempts
to determine the nature and extent of the cultural affinities between the two
regions. Such attempts would involve, among other things, a careful catalogue
of artifacts shared by the peoples of the two areas. One artifact which is found
in both regions but which, to my knowledge, has not been noted previously, is a
rain cape of singular form (Fig. 1). I saw this type of rain cape in many places
in northern Ethiopia. It is used, too, by the Tutsi, Hutu, and Kiga Bantu tribes
north-west of Lake Victoria in Uganda (Trowell and Wachsmann (1953), 201,
208) and by the Mulera Pygmies of the East Kivu region of the Congo
(Schumacher (1943), Tf. III, Abb. 2). A strikingly similar cape is used in South
India (Cipriani (1936), 506), and I have seen another pictured for Indonesia.
In north-west Ethiopia, the rain cape mentioned above is commonly used by
boy herders when they are watching their cattle in the field. The boys make the
capes (Amharic: gessa) of a certain tall wild grass, gidcha. They tie the long
pieces of gidcha together with strings of twisted grass to form a mat. Then,
with twisted bark rope, they fasten a double thickness of the mat to a pole

FIG. 1
North Ethiopian Rain Cape. (Measurements in centimetres.)



FIG. 2
Amhara Boy with his Rain Cape.

which forms the centre support of the cape. When not in use, the cape is carried
by means of a leather thong attached to the pole. When in use, the cape rests
on the boy's head, covering his head, shoulders, and his back to the waist
(Fig. 2).
Wherever this type of rain cape is found outside of Ethiopia, it is constructed
of fibrous plant materials; in some places the principal construction material is
grass, and in others it is leaves or bark. Regardless of these differences in con-
struction materials, the rain cape discussed above is so unique in form that it
probably had a single origin. Since it is not used exclusively by related ethnic
or occupational groups, but is found among Pygmies, Semitic Amhara, Bantus,
and Indians, among farmers and herdsmen as well as hunters it may already
have been diffused widely in ancient times.
Cipriani, Lidio. Case su gli alberi. L'Universo, vol. 17 (1936), pp. 491-514.
Schumacher, P. Die Kivu-Pygmiien und Ihre Soziale Umwelt im Albert-National-
park. Brussels: Institut des Parcs Nationaux du Congo Belge, 1943.
Trowell, Margaret and K. P. Wachsmann. Tribal Crafts of Uganda. London:
Oxford University Press, 1953.

By H. S. S. FEW

THAT there is still a strong belief in witchcraft in Buganda is well illus-
trated by the facts revealed in a case recently tried there by the High
Court. On the 22 May 1957, 'Dr.' Kimote Muwanga in company with his
mistress Esita Nakasula and a client called Paulo Sentamu went to drink
mwenge (Kiganda beer made of bananas) at the beer-shop of a woman called
Nakaiza near Mityana, Singo. They arrived about 7 p.m. and stayed an hour.
That Kimote and his lady left the beer-shop on good terms is evidenced by
his remarking that he was carrying off his kasolo, a Kiganda word meaning
'little animal' and regarded as a term of endearment.
From subsequent admissions made by Sentamu it appeared that Kimote
and Esita started to quarrel a few hundred yards down the track from the
beer-shop. Kimote complained about Esita refusing to marry him despite
the fact that he had spent so much money on her. Esita answered back, and
in the heat of the ensuing argument he struck her at the base of the neck and
she fell to the ground unconscious. It was never satisfactorily determined
whether she died from that blow, or as a result of strangulation from the
cloth which Kimote proceeded to tear from her clothes and tie around her
neck. Die however she did, almost immediately.
Kimote was a well known witch-doctor in his area and greatly feared. The
wretched Sentamu had come down from the wilds of north Singo for treatment
from Kimote, and was staying at Kimote's house at the time of the killing.
While the woman lay dying Kimote, a quick thinker, despatched Sentamu to
his home to bring coffee berries and some one cent pieces. When Sentamu
returned with the berries and coins Kimote "dedicated them to the deceased's

ghost and thereby made blood brotherhood with it". He then scattered both
berries and money on the sides of the path around the corpse. Sentamu was
then despatched to call Kimote's two sons, and when they arrived the body
was carried off and dropped in the bush half a mile away. Unfortunately it
was not found until some four days later, and by the time the body reached
Kampala and a post-mortem was held it was too decomposed for the patholo-
gist to say what had been the cause of death.
Meanwhile Kimote was not idle in summoning the powers of witchcraft to
his aid in his predicament. He did not return to his house until early the
following morning, and then he brought a tree root with him. He fastened this
across the front door of his house, and then he tied up his deck-chair with a
trailer of a plant that probably belonged to the convolvulus family. He then
pointed to the root and said "the case which might cause me to be imprisoned,
let it go against the owner of the house-there he is".
Kimote's next action was to send Sentamu to find a discarded banana fibre
head-pad (nkata) from the local well. When Sentamu brought it, Kimote dug
a hole and put the pad in it; Sentamu was instructed that whenever he wished
to urinate he must do so in this hole, and say each time "I have bound up
Semeo Dungu's case" (i.e. 'bewitched' it). (Semeo Dungu was the father of Esita
Nakasula, and in Kiganda eyes he would'be the 'owner' of any case concerning
her death.) This pad was later exhibited in Court.
Kimote then buried a small vaseline bottle containing medicine and blood
nearby the place where he had put the head-pad. Where the blood came from
is not known. The Government Chemist was able to confirm later that the
bottle did contain blood, but for various reasons was unable to say whether or
not the blood was human. On top of the place where he buried the bottle
Kimote planted a small tree stump (ekikonge). Here Sentamu was instructed
to wash himself and to say "I have distrained (seized) the case of Semeo
Dungu against Kimote and may he remain firm as a tree stump".
Subsequently Kimote and Sentamu were taken by the Mityana Police to
Mityana Police Station. They were held there for a day or two and then
released for lack of evidence. While they were still there-they were not in
custody-Kimote obtained a small packet of coffee beans. One of these he
buried in the earth after smearing it with blood from a small incision he made
on his stomach with a razor blade, and another bean he swallowed. Sentamu
was directed to follow suit and did so. The following morning Kimote bought
a packet of salt. He cut off some of his hair, made an incision on his finger,
and pared his nails, mixing these trimmings with the salt. The mixture was
scattered on the road by a confederate, who cannot be named, as he later
denied, not unnaturally, that he knew anything about this or any other matter
connected with the case. Kimote said that once thrown down the mixture
would 'go with the case'.
After being allowed to leave Mityana, Kimote and Sentamu returned to
Kimote's home. Kimote's stock as a witch-doctor must have been enormously
enhanced in his village following his and Sentamu's release, since there can
be no doubt that this would be attributed to his magic. Nevertheless he did
not underestimate the police and he continued to take precautions.

He cut down a banana plant and made Sentamu lie flat on his back. He
then laid the stem of the banana plant lengthways on top of Sentamu. He
then split it through saying he had handed over the case to this person. 'This
person' was presumably the banana stem which was left on the scene and later
produced in court. Kimote told Sentamu not to look behind when they walked
away home, and we can be pretty sure that the terrified Sentamu did not fall
into the error of Lot's wife. But, as was remarked in the course of the hearing,
was the purpose of the splitting of the banana stem merely to throw the case
off Kimote, or was Kimote making doubly sure, and putting it on to Sentamu?
If so it is unlikely that Sentamu thought so. He was a very dull fellow and the
perfect stooge.
Kimote's next step was to break a splinter of dried wood off a tree stump,
spit on it and say "The case of murdering Nakasula which Semeo Dungu is
taking charge of, let it go off me. I have put it off my body completely. For
this case to be put on me again, this tree (splinter) will have got leaves".
Despite all these precautions a summons came one day from the chief to
Kimote and Sentamu to come and gather with the others in the village. The
C.I.D. had after long delays been called in to do what the local police had
failed to do. It was now the turn of Inspector J. J. Kintu and Corporal Mukasa.
But before he answered the summons Kimote played his trump card. He dug
clay and made an image of himself. He made Sentamu cut his (Kimote's)
finger and toe nails, and hairs from his head. He kneaded these into the effigy
and then clothed it with a piece torn from Sentamu's kanzu. He then sent
Sentamu into the bush to build a little hut (sabo) for the effigy. The effigy was
about 8 inches long and obscene in character. It was exhibited in court and is
now in the Uganda Museum. The hut (sabo) was a foot or so high and made
as a replica of a beehive hut. Such huts seem to be commonly used in Uganda
for housing such effigies and also for storing 'medicines' and tools of the
witch-doctor's trade. When he had done this, Kimote told Sentamu to put the
effigy inside and to make a door to close the doorway of the hut. Sentamu was
told to say while doing this "Semeo Dungu's case of murdering a person for
which Kimote is being charged, let it go off him, and then go against the host
of this home".
But it was all no good. Whenever Sentamu was with Kimote he remained
silent, but whenever they were apart he babbled all he knew to the police. He
gave no evidence in court and therefore it was not possible to clear up in
cross-examination the meaning of some of the spells and some of the words
said. But it is doubtful if he would have known: he was a very dull fellow. As
it was, the order of the events described varied somewhat in his various
statements; but that the events took place is certainly true, since he was able
to lead the police officers to the various places where the articles, which were
later used as exhibits in court, had been placed or hidden.
In the result Kimote was convicted of manslaughter, and Sentamu of being
an accessory after the fact. The judge not unnaturally took a merciful view in
Sentamu's case in view of the circumstances revealed and sentenced him to
8 months. Not so in Kimote's case however; he got ten years.


Mr. C. C. Wrigley in Uganda J., 22 (1958), 11 and Mr. J. Middleton in
Uganda J., 19 (1955), 194, show themselves to be opposed to the arguments
advanced in The Lwoo.1 In that work, I sought to describe the Madi group
of tribes and their dispersal, which resulted in the occupation and domination
by the Madi of the greater part of Uganda. The Banyoro (Madi) were headed
by the Abatembuzi or Madi Ndri. I then proceeded to describe the origin and
rise to fame of the Lwoo, which were followed by their migrations which
resulted in the replacement of a Madi by a Lwoo domination. This was typi-
fied by Jo-Oyima or Ba-Hima (Lwoo tribes) headed by the Bachwezi and
Messrs. Wrigley and Middleton, however, are seemingly unconvinced and
are apparently inconvenienced by the historic events recorded in The Lwoo.
The book was, however, based on local traditions collected over the span of
40 years spent among different tribes in remote countries. Those who disagree
with my conclusions have chosen to attack these traditions by distorting or
ridiculing them. But in Africa, these traditions of the old people, passed on
from generation to generation, are the best part of a tribe's spiritual culture.
These traditions present the best available sources of knowledge about any
particular people. Historians of peoples who possess no literature, have from
the very earliest times until the present day, willingly spent years in collecting
all available traditions. Traditions are not, of course, infallible like records
from well-kept archives, and it must, of course, be left to the intelligence of
historians and their readers to value, sift and interpret them. Even natives
have difficulty in this respect.
One thing is certain, history based on the traditions of older people far sur-
passes history based on conjectures or mere theories. The two massive in-
vasions of Bunyoro and countries beyond it, which I have mentioned in The
Lwoo, are reflected in the country's traditions, and are certainly historic facts.
Students would be well advised to accept them and other Nyoro traditions as a
basis of real historical research. J. P. CRAZZOLARA.
Ngeta Catholic Mission, Lira.
October 1958.
1 The Lwoo by J. P. Crazzolara. Three volumes. Verona: Missioni Africane, 1950 to
1954. (Obtainable from the Catholic Church Procure, P.O. Box 200, Gulu; or The
Catholic Bookshop, P.O. Box 2615, Kampala.) These volumes have been reviewed in
Uganda J., 16 (1952), 17 (1953), 19 (1955).

I think that Mr. Carter in his communication to Mr. Lawrance at Uganda J.,
22 (1958), 41 has confused two different sites. The description given of the Tira
site exactly fits the site Mr. Carter discovered at Onyeri when he, Dr. Dean and I
were looking for the Asuret site.

I cannot fix the position of Onyeri with certainty, but it is about 1 32' N.,
and 330 35' E. The only route to it that I know is as follows: turn off the Soroti-
Kyere road 1-2 miles north of Brookes Corner (Sapir) and go 1-3 miles east to
the old Olupe school site. Then turn north (the corner is difficult to see, but
the track soon improves) and, after going 2-4 miles the large rock with the
bat-cave will be seen about two hundred yards to the east. I think the track
must run on into the 'back road' from Mile 70 to Kyere by Kakworo. The
people at Onyeri told me that a European had once camped there and taken
a great interest in the paintings; but as he did not publish anything about the
site, Mr. Carter should have the credit as discoverer. J. E. COMPTON.
Ngora, Teso.
25 July 1958.

I have read with great interest the article on rock paintings in Teso and
Bukedi in Uganda J., 22 (1958). On page 42 the author has made a list of
known paintings of that kind in Uganda. In 1954 I was stationed as a mis-
sionary at Kigarama in Bukoba District, Tanganyika, which is near the
border between Uganda and Tanganyika, not far from the place where the
River Kagera flows into Lake Victoria. On a rock just outside the mission
station, facing the swamp formed by the river, are several rock paintings of
the kind described in the article. I mention this fact as a possible line for
further investigation. Bukoba District is sociologically closely related to
Uganda. I am convinced that research work in Bukoba District would often
supplement investigations undertaken in Uganda. I have found many articles
in the Uganda Journal which are of direct relevance to Bukoba District, and I
have often thought that fuller understanding of the topics concerned could be
reached if further investigations in Bukoba District had been undertaken.
Urshult, Sweden. CARL J. HELLBERG.
9 September 1958.

Having spent four years at Butare in Buhweju and having made a study of
the historic sites of the county, I venture to make the following comments on
Dr. Morris's article, Historic Sites in Ankole (Uganda J., 20 (1956), 180).
One of the references to Buhweju sites states; "Kyamujumi. A mutoma and
migorora mark the grave of an unknown Mukama". There is a mutoma with
migorora at Kyamujumi but it does not mark a grave. It is an enjeru, that is to
say a mutoma planted by the Mukama when he came to the throne. Dr.
Morris has referred to this custom in an earlier note, The Balisa Bakama of
Buzimba (Uganda J., 17 (1953), 73); "It is said that each Mukama would plant
a mutoma (bark-cloth tree). A similar custom existed in Buhwezu." Such a
mutoma is called enjeru because white cows or white goats were sacrificed
there, the Runyankore for a white cow being ente enjeru and for a white goat
embuzi enjeru. The mutoma at Kyamujumi is the one planted by Kashoma
Kitonera, Ndagara's father, when he came to the throne, and is at the site of
Kashoma's first rurembo.

Dr. Morris also states in his article that at Butare Catholic Mission "a large
mutoma marks the site of Ndagara's rurembo". This tree is also an enjeru, and
was very probably planted by Ndagara when he came to the throne. Many
people say that this is the place where he built his first rurembo when, on his
father's death, he became Mukama.
The article also refers to the kraal of the Mugabe Ntare at Ntobora. Unfor-
tunately, many of the trees of this kraal, which are muko (Erythrina
bequaertii), were cut down last year in order to make a fence and it is now
difficult to make out the site. B. CLECHET.
Ibanda Mission, Ankole.

Dr. Morris comments:
There are in Buhweju two series of large mutoma trees, the first being the
njeru which each Omukama planted at or near his rurembo, and the second
marking the Abakama's graves. According to information given to me by an
old man, Rushaki, who died a couple of years ago at the age of over ninety
and who had been Ndagara's gate keeper, the following events took place on
an Omukama's death. A kraal would be chosen and in it the body would be
buried; the grave would be watched for two months, during which there was
mourning and the drum Mashaija was turned upside down. When the new
Omukama had been chosen, the kraal was demolished and a mutoma planted
over the grave, the owner of the kraal becoming the hereditary keeper of the
place. From then on, from time to time, a bull would be provided by the
Omukama and sacrificed at the tree. These sacrifices took place not only at
the grave of the Omukama's immediate predecessor but also of his ancestors,
provided they were not too remote. Four such trees associated with the names
of the last four Abakama survive and a keeper lives (or still did so a couple of
years ago) at Kitoha and was responsible not only for the grave there of
Rusharabaga but for the other three as well, and ceremonies have taken place
certainly within the fairly recent past. I am most interested that Father Clechet
has managed to identify the tree at Kyamujumi as belonging to the first series.
I could get no definite information about it and was led to believe it belonged
to the second series, the name of the Omukama concerned having been for-

In his interesting article on Runyankore (Uganda J., 22 (1958), 56), Dr.
Morris says that "the tendency to devoice or whisper short vowels ... is ...
unknown among the Baganda".
Some Baganda certainly unvoice final u after m, and final i after n. I came
across this very early in my study of the language, when a Muganda said to me
Weebal' obulim'. "Don't you say the u on the end of omulimu?" I asked. He
repeated the word to himself several times, then said, "No, I suppose not."
"Then why is it written there?" "Oh, we couldn't say it if it hadn't got a u
on the end." I have since decided that the muscular movement for u takes
place even when no sound escapes: some individuals pronounce the vowel