Front Cover
 Title Page
 Table of Contents
 The geographical background of...
 Some thoughts on the Bacwezi
 The Bacwezi in Ankole
 An agricultural survey of Moruita...
 Rock paintings in Teso and...
 Uganda's old eastern province and...
 Some aspects of Runyankore
 West Nile District: some soil and...
 Notes on contributors
 Back Cover

The Uganda journal
External Link ( Journal issues, 1999-2002 )
Full Citation
Permanent Link: http://ufdc.ufl.edu/UF00080855/00064
 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: 1958
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: UF00080855:00064
 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
    The geographical background of Western Uganda
        Page 1
        Page 2
        Page 3
        Page 4
        Page 5
        Page 6
        Page 7
        Page 8
        Page 9
        Page 10
    Some thoughts on the Bacwezi
        Page 11
        Page 12
        Page 13
        Page 14
        Page 15
        Page 16
        Page 17
    The Bacwezi in Ankole
        Page 18
        Page 19
        Page 20
        Page 21
    An agricultural survey of Moruita Erony, Teso
        Page 22
        Page 23
        Page 24
        Page 25
        Page 26
        Page 27
        Page 28
        Page 29
        Page 30
        Page 31
        Page 32
        Page 33
        Page 34
        Page 35
        Page 36
        Page 37
        Page 38
    Rock paintings in Teso and Bukedi
        Page 39
        Page 40
        Page 41
        Page 42
    Uganda's old eastern province and East Africa's federal capital
        Page 43
        Page 44
        Page 45
        Page 46
        Page 47
        Page 48
        Page 49
        Page 50
        Page 51
        Page 52
        Page 53
    Some aspects of Runyankore
        Page 54
        Page 55
        Page 56
        Page 57
        Page 58
        Page 59
        Page 60
        Page 61
        Page 62
        Page 63
    West Nile District: some soil and vegetation types
        Page 64
        Page 65
        Page 66
        Page 66a
        Page 67
        Page 68
        Page 69
        Page 70
        Page 71
        Page 72
        Page 73
        Page 74
        Page 74a
        Page 75
        Page 76
        Page 77
        Page 78
        Page 79
        Page 80
        Page 81
        Page 82
        Page 83
        Page 84
        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
        Page 100
        Page 101
        Page 102
        Page 103
        Page 104
    Notes on contributors
        Page 105
        Page 106
    Back Cover
        Back Cover 1
        Back Cover 2
Full Text

The Uganda Journal


MARCH 1958

Contents Mw
The Mutiny Memorial at Bukaleba H. B. THOhuA
Arabs on Lake Victoria. Some Revisions SIR J. M. GRAY
On Paying a Visit in Bunyoro SARAH NYENDWOHA
The Suicide of Musinga T. MALONEY
Early Days in Kampala J. SYES.
Conurbation H. B. THoMAs
The Suspension of the Ankole Agreement in 1905 SIR J. M. GRAY
Bantu Bureaucracy (by L. A. Fallers) T. R. F. Cox
The Iteso (by J. C. D. Lawrance) A. C. A. WRIGHT
Sir Harry Johnston and the Scramble for Africa (by Roland Oliver)
Modern Cofee Production (by A. L. Haarer) J. F. ARCHIBALD
Introduction to the Botany of Tropical Crops (by Leslie S. Cobley)
African Tapestry (by Margaret Trowell) MRS. M. C. USHER-WILSON
The Fountain of the Sun (by Douglas Busk) R. N. POSNErr
Michael Moses, O.B.E. S J. M. GRAY
Mother Mary Kevin, C.B.E. SIR J. M. GRAY
Index to The Uganda Journal, Vols. 1 to 20.











Published by
Price Shs. 15 (15s.)


His Excellency Sir Frederick Crawford, K.C.M.G., O.B.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:

Hon. Librarian:
Hon. Auditor

Corresponding Secretary at linja:
Corresponding Secretary at Mbale
Corresponding Secretary at Masaka
Corresponding Secretary at Tororo

Dr. A. W. Southall

Mr. B. Kirwan
Mr. S. W. Kulubya, c.n.E.
Mr. D. K. Marphatia
Mr. R. J. Mehta
Mr. C. N. Mukuye
Mr. P. Tamukedde
Dr. H. C. Trowell, O.B.E.
Mrs. K. M. Trowell, M.B.E.
Mr. M. Barrington Ward
Mr. 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. T. R. F. Cox, C.M.o.
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.e.p.
Sir E. F. Twining, G.C.M.O., M.B.E. Professor A. W. Williams
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.O.
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.G., 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.F.
Professor A. W. Williams
Sir J. B. Hutchinson,
C.M.G., FR.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. NM. Trowell, M.B.E.

Mr. B. K. Mulyanti

Mr. G. P. Saben

Secretary: Mrs. M. M. Wallis





Uganda Journal



No. 1

MARCH 1958

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

Published by




The Mutiny Memorial at Bukaleba H.B. THOMAS
Arabs on Lake Victoria. Some Revisions SIR J. M. GRAY
On Paying a Visit in Bunyoro SARAH NYENDWOHA

The Suicide of Musinga M. T. MALONEY
Early Days in Kampala J. SYKES
Conurbation H.B. THOMAS
The Suspension of the Ankole Agreement in 1905 SIR J. M. GRAY

Bantu Bureaucracy (by L. A. Fallers) T. R. F. Cox
The Iteso (by J. C. D. Lawrance) A. C. A. WRIGHT
Sir Harry Johnston and the Scramble for Africa (by Roland Oliver)
Modern Coffee Production (by A. L. Haarer) J. F. ARCHIBALD
Introduction to the Botany of Tropical Crops (by Leslie S. Cobley)
African Tapestry (by Margaret Trowell) MRS. M. C. USHER-WILSON
The Fountain of the Sun (by Douglas Busk) R. N. POSNETT

Michael Moses, O.B.E.
Mother Mary Kevin, C.B.E. -



Index to The Uganda Journal, Vols. 1 to 20.




30E 31E 320E

20 10 0 20 40

Queen EIIzobzth
Notional Park ........ /

FRESTT oMond, Porl

Ko ro

M /r 72 KAAD

Fot I Por )a l/H o -tA PAL /

K l emb I "-


K 3

SMbFortPao / K

;- ;L ) / o/ /1 VIC'r Io i..
< ( AE A"'VAL/ ( be

:U --O-, R-O ---
;<, /^p -NT J

.Ka.. .. 2

R FII-. u R UoNDI A.Serubre, 1957.

FIG. 1. Western Province Reference Map.

WESTERN Uganda, as part of a wider Western Rift Region, owes some
of its most important and most characteristic features to the processes
which have been responsible for the development of the rift-valleys and their
associated highlands. The rift-valley system is represented in Uganda by two
troughs, that of Lake Albert in the north and that of Lake Edward and Lake
George in the south. These relatively shallow lakes occupy a considerable
proportion of the surface area of their respective troughs, and in both cases
they rest upon the Quaternary sediments which have been brought down from
the adjacent highlands to cover the original floor of the trough. Faulted scarps
with a north-east to south-west alignment delimit the well-defined rift-valley
of Lake Albert, but in the relative downwarp marked by the basin of Lake
Edward concentrated faulting, although present, does not play so prominent
a part. The abruptness and freshness of most of the boundary scarps derives
from the geological recency of the earth movements which gave rise to them.
Behind the escarpments lie the uplifted plateaux; and although there may
be uncertainty as to whether the depression of the rift-valley floors was
absolute or only relative to the upward movement of the flanking plateaux,
there can be no room for doubt about the uplift of the latter. The regional
elevation is less in the north than in the south, and correspondingly the Lake
Albert flats are just over 2,000 feet above sea-level as compared with the
3,000 feet of the present-day floor of the Lake Edward depression. Between
these two downwarps rocks belonging to the geological basement of the
continent have in the highest parts of the Ruwenzori massif been raised to a
height of over 16,000 feet above sea-level. Leaving this outstanding exception
aside in its isolation, the heights of the plateaux in western Uganda are in
very general terms as follows: 3,000-4,000 feet in Bunyoro, 4,000-5,000 feet
in western Toro, 5,000-6,000 feet in western Ankole, and over 6,000 feet in
Kigezi with the ridges there reaching 7,000-8,000 feet. From western Toro and
Ankole the general slope is eastwards to the coast of Lake Victoria, the
waters of which have gathered in the gentle depression which is the counter-
part of the upwarping of the western rift zone. Especially in eastern Ankole,
south-eastern Toro and western Masaka the relief is gentle and the horizons
are correspondingly wide. The regional tilt has caused a reversal of drainage
along the valleys of the Kagera, Katonga and Kafu, which were formerly
occupied by streams flowing westward to the Congo basin.
Volcanic activity is a phenomenon associated with the processes of rift-
valley formation. This is exemplified in the Mufumbiro chain of volcanoes
1 A lecture delivered to the Uganda Society on 20 February 1957. This written version
of the paper should be read in conjunction with the 1:1,000,000 map of the Uganda
Protectorate or, if greater detail is required, with the 1:250,000 District maps. It has
been possible, however, to show most of the names mentioned in the text in the
accompanying reference map (Fig. 1).

which stretches for about fifty miles from east to west across the Lake
Edward-Lake Kivu downward. It includes the cones of Muhavura (13,547 feet),
Mgahinga (11,400 feet) and Sabinio (11,960 feet) on the south-western
boundary of Kigezi; whilst further west among the craters in the Belgian
Congo Nyamulagira (10,026 feet) and certain lesser vents have been active
in contemporary times. To the north in the Bunyaruguru county of north-
western Ankole there is a group of craters, with accompanying lakes, up on
the plateau to the south of the Kazinga channel. Deep and well-formed, these
craters present scenes of compelling beauty, and they yield a rich soil for the
cultivator. In contrast with the plateau craters are the shallow explosion craters,
with or without lakes, which occur among volcanic tuffs on the floor of the
Lake Edward basin, where some seventy-eight of them are to be found within
an area of forty or fifty square miles. The best known and most southerly
member of this group contains Lake Katwe, the salt of which is derived from
underground sources presumably of volcanic origin. The Katwe salt industry
is controlled by the Toro Native Government, and the scale of operations is
indicated by the fact that in the record year of 1955 the industry earned over
50,000. Another considerable group of craters occurs to the south of Fort
Portal, and there are a few craters immediately to the north-west of the town.
Hot springs are to be found in or near the rift-valley in each of the districts
of the Western Province; and the frequency and occasionally the intensity of
earthquake shocks serve to remind us that western Uganda has not yet
reached a phase of tectonic quiescence.
Other features of the present-day geography of western Uganda have their
origin in events at the other end of the geological time-scale. One thinks of the
hills and valleys of Kigezi with their north-north-west to south-south-east
trend, of the granite-floored 'arenas' of Ankole, the granite hills of Mubende,
and the central hills of Bunyoro, all of which owe the foundations of their
character to events in the pre-Cambrian era. The mineral wealth of western
Uganda, too, is found in its old rocks, for apart from some good quality
limestone and various occurrences of salt, the rift valley is lacking in minerals.
A quite impressive variety of minerals occurs in central and south Kigezi
and in the adjacent parts of southern Ankole; and there is another strongly
mineralized area in the county of Buhweju in north-western Ankole up on the
tangled plateau country above Lake George. But the quantity of the minerals
has not so far matched their variety, and in 1955 metalliferous ores accounted
for only 0-5 per cent of the total value of the domestic exports of Uganda.
These ores comprised mainly Kigezi wolfram, the south Ankole tin having
reached a stage of near-exhaustion. The production of copper and cobalt from
Kilembe mine, as the result of mining operations on a scale never previously
undertaken in Uganda, is in process of effecting a revolution in this situation.
Western Uganda presents three main climatic types. (i) Over most of the
high ground east of the rift-valley escarpment the mean annual rainfall totals
50 inches or more, falling according to locality on 130 to 170 days. The
rainfall map in The Way to the West shows a clear correspondence of high
rainfall with high relief; and the inter-tropical convergence zone plays its
part in producing the bi-seasonal rainfall; but another factor influencing the

rainfall is the entry of moist air from the Congo basin and perhaps ultimately
from the South Atlantic Ocean. A convergence zone between westerly air
streams and south-easterly or northerly streams lies normally over the eastern
part of the Congo basin; but, accompanied by a belt of rainfall, this frontal
zone moves fairly frequently over western Uganda and sometimes as far east
as the shores of Lake Victoria. Mean annual temperatures in this zone decrease
with altitude from 73 F. at Masindi, through 66F. at Fort Portal to
62F. at Kabale. (ii) The climates of the Lake Albert and Lake Edward
depressions are hot with strongly marked dry seasons and with annual
rainfall totals in the order of 30 to 40 inches falling on 80 to 100 days. The
Lake Albert section of the rift-valley zone is slightly hotter and drier than
the higher southern basin. (iii) Away to the east and south-east lies another
dry belt, which is interior to the well-watered Lake Victoria zone and is yet
unaffected by the special factors which bring abundant rainfall to the western
hill country. The belt runs from south-west to north-east and includes south-
eastern Kigezi, the whole of eastern Ankole and western Masaka, and the
adjacent parts of south-western Mengo, southern Mubende and southern Toro.
The mean annual rainfall is everywhere less than 40 inches per annum, and
in a small area in the north-east astride the Katonga valley it is below 30
inches. Mbarara provides a type station, with 35 inches of rainfall occurring
on 127 days in the year and with a well-marked dry season in June and July;
whilst its mean annual temperature is 69"F. with a daily range of 20 to
23 according to season. The rainier conditions of the west tend to be prolonged
eastward through the Mubende hill country, but beyond this moister area a
counterpart to the Ankole dry belt occurs in south-eastern Bunyoro and
northern Mengo.
Short grasses in an open, deciduous woodland cover this flat to undulating
country in the western part of the Lake Kyoga basin and yield useful grazing
lands if they are controlled by occasional burning or cutting. But the short-
grass area of greatest interest for the purpose of this paper is that corres-
ponding with the south-western dry belt, where the dominant grasses, Themeda
triandra and Hyparrhenia filipendula, grow from three to five feet high and
provide a very palatable, nutritious food for cattle. The traditional Runyankore
name for these grasses, which are not distinguished from one another, is
mbubu, from a Bantu root meaning "soft" and "sweet". Often the grassland
occurs in an open woodland landscape, dotted with low-growing acacias,
occasional candelabra euphorbias and the ubiquitous, thicket-covered termite
hill; but sometimes the trees and bushes are absent, thus leaving an open,
grassland sward. Occasional burning sweetens the pastures, but the tendency
has been to reduce the quality of the grasslands by too frequent burning.
Separating the two short grass areas there occurs a belt of elephant grass,
Pennisetum purpureum, and kindred grasses in central and northern Mubende
and the adjacent parts of Singo, and this is continuous westwards with the high
grass and forest formations of the western highlands. These grasses are on the
whole too tall and luxuriant to be considered as suitable for regular grazing
unless it is possible to keep them frequently cut; for it is only the young grass
which yields palatable and nutritious food. The greater part of Bunyoro

and Toro is covered with this vegetation formation, which provides a reliable
index to the presence of rich, deep soils; and after a break in the Lake Edward
basin the formation reappears in the north-western parts of Ankole and Kigezi.
This vegetation zone contains some of the major tracts of closed forest of
Uganda: Budongo and Bugoma in Bunyoro, Itwara and Kibale in Toro,
Kasyoha and Kalinzu in Ankole, and Maramagambo in northern Kigezi. These
forests contain a considerable proportion of the timber reserves of Uganda,
which is a country not notably rich in this resource; but except perhaps for the
Budongo forest, exploitation has hitherto been hampered by lack of transport
facilities. Classified as montane forest by the ecologists, these forests have
broadly the characteristics of the tropical rain forests; and muvule, mahogany
and ironwood are their chief commercial timbers. Beyond the Ruwenzori
mountains, situated in the Semliki valley at lower altitudes than the forests
already mentioned, the Bwamba forest is a part of the Ituri forest of the Congo
basin. On the Ruwenzori slopes themselves there are temperate forests which
contain no saleable timber, their role in the forest economy being entirely
Short grass with scattered trees and bushes reappears in the rift-valley de-
pressions, but it is of poorer quality than that of the plateaux; and a map
accompanying the Report of the Agricultural Productivity Committee draws
the pertinent contrast between the "warm, dry grassland" of the former land-
scape and the "cool pastoral" area of the latter. The Lake Albert flats, lower,
hotter and drier than those of Lake Edward and Lake George, are the habitat
of a vegetation which can be described by no other term than semi-arid.
Such are the environments of western Uganda, in which the historical
processes have been enacted and in which well-defined societies have emerged.
Before passing on from this treatment of environmental considerations,
reference should be made to the fact that in many instances man shares his
habitat with various kinds of game and tsetse fly, to an extent not easily
realized by those who dwell in the Lake Victoria zone of Buganda. The funda-
mental difference is in density of population. Except in certain limited areas,
such for example as southern Kigezi and south-western Ankole, the density
of population in western Uganda has not been high enough for the human
grouping to become the dominant element in the biological environment.
With increase in population density the balance will inevitably change, but
there would seem to be little doubt that for some time to come game and the
tsetse fly will remain as important elements in the geography of western
Uganda. The present position in regard to the tsetse fly is broadly that along
various consolidation lines the advance of the fly has been checked, although
perhaps somewhat precariously in Ankole, where in relation to the present
unintensive pastoral economy the problem is at one and the same time most
acute and most difficult to counter. The control of the fly seems now to be
scientifically possible but by no means always economically practicable. The
precise time for taking the initiative in driving back the tsetse fly must be
related to a sum total of economic and social factors among which the fore-
most must be the possibility of maintaining a sufficiently intensive utilization
to ensure that the land gained at considerable expense is held against a re-

advance of the fly. The Bunyoro ranching scheme is designed to provide just
such utilization for newly reclaimed territory adjacent to the Kafu river.
Turning now to the consideration of those associations of people and place
which are the primary focus of interest for the geographer, it is convenient to
take as a first example the Bakiga in the mountainous terrain of south-eastern
Kigezi. The landscape is deeply cut by valleys of which the steep sides would
formerly be extensively forested and of which the bottoms are to the present
day usually occupied by papyrus-filled swamps. Grazing was available on the
thin, often gravelly soils of the ridge tops, and prior to the entry of the Bakiga
cultivators in the second half of the nineteenth century these hill-top pastures
were used by 'Hamitic pastoralists of the Bahororo group. But the landscape
provided only poor opportunities for extensive pastoralism, and the hold of
the Bahororo seems to have been slight, for they were forced to move north-
wards into Ankole as the country became more fully occupied by the Bakiga,
themselves driven out of Ruanda by pressure of population and by the raids of
the Batusi pastoralists. To adapt a metaphor from the terminology of plant
geography, south-eastern Kigezi became a region of 'closed human association'
for the Bakiga cultivators. In south-western Kigezi under the more open
topographical conditions of the fertile lava plains at the foot of the Mufumbiro
chain the Banyaruanda of southern Bufumbira carry a Batusi, pastoralist
element among the preponderance of Bahutu cultivators. In the rolling grass-
lands of Rujumbura on the borders of Ankole the Bahororo are of Hamitic
extraction, being closely related to the Bahima; and they founded the transient
kingdom of Mpororo, which during the short period of its independent existence
in the late seventeenth and early eighteenth centuries comprised south-west
Ankole, most of the present Kigezi and a part of Ruanda. The Bakiga, outside
the organization of the Bahima kingdoms, remained in their loosely organized
clan groups without political cohesion until the institution of modern
administration on the Buganda model, which, in the early years, was run
with the help of Baganda agents.
The Bakiga occupy the counties of Rukiga and Ndorwa and much of
Kinkizi. According to the 1948 census the Bakiga who lived in Kigezi
numbered a quarter of a million, or more than half of the total population of
the district: and population densities rising to 500 or more per square mile
occurred in four sub-counties. This population is precariously supported by
the intensive cultivation of the now largely terraced valley sides, with sorghum,
sweet potatoes, peas and finger millet as the principal subsistence crops.
In addition, the peasant often keeps a few cattle, although the amount of
pasturage decreases yearly with the extension of the cultivated area. The
problem of finding a major cash crop has not yet been solved, but tobacco
cultivation has expanded considerably in recent years. As a response to the
pressure of population upon land and resources there are swamp reclamation
schemes in the more populous areas of Kigezi and resettlement schemes in
the Kinkizi and Rujumbura counties of northern Kigezi and in Ankole. The
various schemes, some of which are reaching saturation point in respect of
their capacity to absorb further immigrants, show no present signs of accom-
modating the natural increase of population of the district. In these

circumstances it is not surprising that on the 1948 census figures there should
be 22,000 Bakiga outside their own district, and that there should be an
annual exodus of 30,000 to 40,000 from Kigezi for work in other parts of
Uganda.2 Within western Uganda, most of the plantation and mine labour of
Toro is drawn from Kigezi.
With their drier climate and their open grasslands the undulating plateaux
of central and eastern Ankole provided an initially less favourable terrain for
cultivators and a correspondingly more promising environment for incoming
Bahima pastoralists. The Bahima may have become numerically dominant
in certain localities, but over the country as a whole they were in a decided
minority. According to figures for 1933, provided in that year by Mr. F. Lukyn
Williams, District Commissioner of Ankole, the counties may be placed in
three groups in respect of the proportions of Bairu and Bahima. An eastern
group, including Nyabushozi, Mitoma and Kashari, together with Kajara in
the west, yielded ratios varying from 5:1 to 7:1; a central group, comprising
Rwampara, Shema and Igara, yielded ratios varying from 19:1 to 28:1;
and the remaining counties showed an absence of any appreciable number
of Bahima. The last-mentioned group of counties included Buhweju and
Bunyaruguru under the different environmental conditions of the north-west,
and Isingiro in the south-east from whose tsetse-invaded pastures virtually
all pastoralists had been evacuated. Within the kingdom which the Bahima
founded in Ankole there was room for them whilst exercising their dominance
over the cultivators still to retain their physical identity and their own way
of life. Thus the members of the two ethnic groups which were represented
within the political unit had not become members of one body politic, nor had
they been welded into a real national unity. A social gulf was fixed between
Bahima and Bairu, and under such conditions social organization became
stereotyped and therefore static. Times have now changed, and with them
social relations: and we speak no longer of Bahima and Bairu but, more
happily, of Banyankore. Earlier attitudes, though, have had their repercussions
and have left their tensions within the population of Ankole, leaving problems
which it would be wrong either to minimize or in any way to exaggerate.
Features of contemporary interest in the geography of Ankole include the
war against the tsetse fly, with its concomitant resettlement schemes. In
parenthesis, it may be stated that operations against the tsetse fly were equally
of contemporary interest a quarter of a century ago, with the emphasis at
that time upon bush and woodland clearance schemes. Certainly, it is necessary
to win this war and to secure a more active participation of the pastoralists
in it, if the future of the Ankole stock industry is to be assured. As an
indication of the value of this industry, in a recent year 10,000 cattle were
sold in markets at an average price of Shs. 340 a head. Of the 208,000 cattle
in the district in 1955 rather less than a half were owned by small peasant
farmers, who, with no previous tradition of cattle herding to hamper them,
are said to be receptive to improved methods of animal husbandry. But
2 For the part played by the Bakiga and the Banyaruanda in the economy of Buganda,
see Richards, A. I. Economic development and tribal change: a study of immigrant
labour in Buganda. Cambridge: Heffer, 1954.

despite the importance of pastoralism in Ankole, the cattle population figures
for the district do not come near to those of Teso and Karamoja, respectively
with 671,000 and 587,000 cattle in 1955; and Mengo, with 316,000 cattle,
had considerably more than Ankole. On the agricultural front, what may be
described as Kigezi methods of terrace cultivation continue to spread into
the hill country of western Ankole and to change the appearance of the
countryside; and coffee has become the principal cash crop in western Ankole.
Finger millet, the traditional food crop of the Ankole agriculturists, is
nowadays accompanied by other millets, beans, sweet potatoes and, especially
in the west, to a continually increasing extent by bananas. Within the last
few years a small-scale but prosperous fishing industry has been created on
Lake Nakivali. There is, however, still a considerable emigration from the
district in search of paid employment, and it is estimated that out of a total
population of about half a million, 15,000 to 20,000 Banyankore leave the
district each year to work in Buganda.
Until the early years of the nineteenth century Toro was a part of Bunyoro,
and the Batoro and Banyoro speak the same language and are in other ways
closely akin. In addition, what are now the three counties of Mubende district
were originally within Bunyoro. The south-eastern county of Buwekula was
absorbed by Buganda early in the nineteenth century; but Buyaga and
Bugangadzi were lost during the 1893-94 campaign of Colonel Colvile against
Kabarega, with their subsequent incorporation in Buganda formalized in the
Uganda Agreement of 1900. According to the 1950 recount3 of the 1948 census.
Banyoro comprised 70 per cent of the total population of Mubende district
and Baganda 21 per cent, other ethnic groups accounting for the remaining
9 per cent. In Buwekula county the Baganda formed the majority of the mixed
population; whilst in Buyaga and Bugangadzi the Banyoro were in an over-
whelming majority in every sub-county.
Toro, Mubende and Bunyoro form a convenient geographical unit for the
purpose of this paper, in that elephant grass, giving place westward to con-
siderable stretches of closed forest, is their dominant vegetation formation.
Only on the margins of the area do drier vegetation types and more
natural pastures occur; and it is interesting in this connection that in
olden days the royal herds of Bunyoro were pastured in the short-grass
country which is now included in the northern part of the Mengo district of
Buganda. Under the geographical conditions of the high-grass country it was
not easy for the Bahima incomes to keep their pastoral traditions intact.
although especially in the royal household many of the ceremonies continued
to be based upon cattle. Even though the two-fold division of the nation
between dominant pastoralists and subject agriculturists did not by any means
disappear, fusion of physical traits and culture took place to an extent unknown
in Ankole. It is said that the name Banyoro is itself an indication of social
mobility, in that it was first applied to a class of 'freemen' who were allowed
to emerge from the preponderant cultivator group, one of their freedoms
being entitlement to inter-marry with the pastoralists. To the natural diffi-
3 The recount was taken, for this district alone, because "the census of 1948 showed
an unbelievably high proportion of Baganda and too few Banyoro".

culties of keeping cattle under an unintensive occupation of high-grass country
there were added the troubled conditions of the period immediately prior
to 1900 and the ravages of the tsetse fly, the latter continuing to the present
day. Thus in 1955 Toro, Mubende and Bunyoro, respectively with 51,000,
26,000 and 7,000 head of cattle, had fewer than any other districts in Uganda.
Therein lies the irony of a people without cattle whose customs are associated
with cattle keeping.
There are other landscapes and other peoples, belonging to the Ruwenzori
mountains and to the rift-valley zone, upon which it is not possible to do
more than touch. The energetic Bakonjo hillmen practise cultivation in the
forests of the Ruwenzori flanks up to a height of 7,000 feet or more, growing
cool zone crops such as wheat and potatoes. With a 1948 population of
72,000, as compared with 146,000 Batoro in the district, the Bakonjo are
already pressing hard upon the resources of their mountain slopes, and they
form the preponderant element in the successful resettlement schemes in the
foothill country of Busongora county. On the far side of the Ruwenzori
mountains the Baamba cultivate land for the most part below 4,000 feet
and bordering on the Semliki valley. They number about 30,000. Their
physical and cultural affinities are with the Ituri forest tribes of the Congo
basin; and they also show some physical relationship with the Batwa who
are to be found in very small numbers in the western forest country of Kigezi.
Finally, there are various fishing villages on the shores of Lake Albert,
including the salt working settlement of Kibiro; and these fishing villages
have their counterparts on Lake Edward and Lake George and on the Kazinga
Food crops in the Toro-Mubende-Bunyoro group of districts extend over
the wide range available in modern times in the high-grass area, and they
include the banana; but the traditional finger millet retains an important
place in the diet of the people. Bunyoro has been fortunate in finding two
cash crops, cotton and tobacco; and coffee is on the increase. In Mubende
cotton, coffee and tobacco are all grown, whilst there is also some cultivation
of the first two crops in those parts of Toro which are adjacent to Mubende.
In the vicinity of Fort Portal, in the higher western part of the Toro plateau,
there is a number of tea estates, but the altitude is here too great for cotton.
Cotton and coffee reappear in Bwamba and in southern Busongora. In the
southern section of the rift-valley the activities of TUFMAC (The Uganda
Fish Marketing Corporation Limited) resulted in 1955 in cash payments of
over 23,000 to the local fishermen, a considerable reduction upon previous
years, owing to a decline in the catch; and, in addition, the company employed
some four hundred men on the processes of preparing the fish for the market.
Toro continues to export its men to the Lake Victoria zone, whilst it is at
the same time using Bakiga in particular for the rougher types of work within
its own boundaries. Inevitably, as developments now in their early stages
mature, there will be increasing opportunity for Batoro to find all types of
work in their own district.
A final point upon which some leading thoughts may be apposite concerns
the effects on western Uganda of the 200-mile railway from Kampala to

Kasese, which was opened to traffic in November 1956. The construction of
the line was clearly an expression of the accelerated tempo of economic
development in western Uganda, and the line will in its turn stimulate new
enterprise and bring new geographical orientations to the areas which are
tributary to it. Bunyoro does not come within the potential hinterland of the
railway, and its goods will continue to leave the district either via Masindi
Port and Lake Kyoga or by road to Kampala. Mubende will be tributary
to the railway, but in some instances only after a considerable road haul. To
most of Toro the new line will bring a southward orientation as something of
a counter-attraction to the directly eastward outlook fostered by the road
to Kampala. An associated phenomenon is the presence of new focuses of
activity in southern Toro: Kilembe and Kasese, the twin centres of operation
of Kilembe Mines Limited: Kasenyi, the TUFMAC station on Lake George;
the Busongora resettlement schemes; and, a focus of a different kind, Mweya
Lodge in the Queen Elizabeth National Park.
From the south, Ankole and Kigezi will develop a northward outlook
towards the railway as a challenge to the north-eastward pull of road com-
munication with Kampala. The attraction of the railway will be exerted by
means of its feeder roads, one of which will presumably be the new road
proposed in The Way to the West from Mbarara to the cattle market of
Nkonge. Another such feeder will be the south to north road, a kind of
gigantic ring road, which connects the termini of the radial lines of communi-
cation at Kabale, Bushenyi, Kasese and Fort Portal. The bridge across the
Kazinga channel, opened in December 1954, to replace the former Katunguru
ferry, is an example of the current acceleration of communications along this
south-north route. Fertile Bunyaruguru, in the immediate hinterland of the
bridge, rescued from its earlier remoteness, is likely to receive a considerable
stimulus to economic development; and there is a proposal that a tea estate
and factory should be established a little further south in northern Igara.
Northern Ankole may well experience a certain 'pull' between the new
economic orientation northwards towards the railway and the older admini-
strative centralization on Mbarara. The general effect of the coming of the
railway is likely to be a quickening of activity and circulation in western
Uganda combined with the further integration of its various forms of economic
and social organization within the life of the larger political unit of Uganda.

Baker, S. J. K. Bunyoro: a regional appreciation. Uganda J., 18 (1954), 101-12.
Combe, A. D. The Geology of south-west Ankole and adjacent territories, with
special reference to the tin deposits. Entebbe: Geol. Surv. of Uganda,
Memoir 2, 1932.
The Volcanic area of Bufumbira, south-west Uganda. Entebbe: Geol. Surv.
of Uganda, Memoir 3, 1933.
East African Meteorological Department. Collected climatological statistics for
East African stations. Nairobi: 1953.
East African Statistical Department. African population of Uganda Protectorate.
Nairobi: (Geographical and tribal studies), 1950.

Henderson, J. P. Some aspects of climate in Uganda, with special reference to
rainfall. Uganda J., 13 (1949), 154-70. (Also published as: East African
Meteorological Department, Memoirs, 2, 5, 1949.)
Morris, H. F. The Kingdom of Mpororo. Uganda J., 19 (1955), 204-7.
The Making of Ankole. Uganda J., 21 (1957), 1-15.
Purseglove, J. W. Kigezi resettlement. Uganda J., 14 (1950), 139-52.
Roscoe, J. The Bakitara. London: 1923
The Banyankole. London: 1923.
Immigrants and their influence in the lake region of central Africa.
London: Cambridge University Press, 1924. (The Frazer lecture in social
anthropology, 1923.)
Snowden, J. D. The grass communities and mountain vegetation of Uganda.
London: Crown Agents, for the Government of Uganda, 1953.
Stamp, L. D., Editor. Natural resources, food and population in inter-tropical
Africa: report of a symposium held at Makerere College, September 1955.
London: Geographical Publications Limited, 1956, pp. 91-104.
Thomas, H. B. and Scott, R. Uganda. London: 1935.
Uganda Protectorate. Annual reports on the Eastern Province, Western Province
and Northern Province for the year ended 31 December 1954. Entebbe:
Government Printer, 1955.
Annual reports on the kingdom of Buganda, Eastern Province, Western
Province and Northern Province for the year ended 31 December 1955.
Entebbe: Government Printer, 1956.
Report of the Agricultural Productivity Committee. Entebbe: Government
Printer, 1954.
.Report of the committee of inquiry into land tenure and the kibanja system
in Bunyoro, 1931. Entebbe: Government Printer, 1932.
The Way to the west: being an economic and railway traffic survey of
certain areas of western Uganda, together with recommendations based
thereon. Entebbe: published by authority of the Government of Uganda.
Willis, B. East African plateaus and rift valleys. Washington: Carnegie Institution
of Washington, 1936.
Winter, E. H. Bwamba economy. Kampala: East African Institute of Social
Research, 1955.



THIS article is written in the belief that in recent years the study of
Uganda's prehistory has been tending to follow a false trail, in that far
more has been demanded of tribal tradition than it is actually capable of
performing. Dr. Middleton has argued that this is true of investigations into
the traditions of the northern tribes.' I believe that it is also true of theories
about the origins and antecedents of the inter-lacustrine kingdoms.
Let us consider first the story of Isaza, the paternal grandfather of the
Cwezi king Ndahura. The substance of the story, as told by the Mukama
Andereya Duhaga II (through Mrs Fisher),2 by Petero Bikunya3 and, with
slight variations, by Roscoe's informants,4 is as follows. Nyamiyonga (Old
Sooty, or the Lord of Ashes), ruler of the land of ghosts, sent an embassy to
Isaza the king, asking him to make blood-brotherhood. But Isaza, being
afraid to enter into so close a relationship with a being from the underworld,
deputed his katikiro, Kwezi, to perform the rite. Nyamiyonga was very angry
when he found out that he had been thus fobbed off with the blood of a
servant, and he resolved to lay hands on Isaza and punish him for his decep-
tion. To this end he chose the most beautiful of his daughters and sent her
to seduce the king. This plan failed, for though Isaza made the girl his wife
he was not sufficiently besotted to follow her when she went back to her own
country to bear her child. She had, however, discovered where his heart
really lay; had he not bluntly told her that he loved his cattle more than
her? So Nyamiyonga tried again, this time sending a pair of his finest beasts,
which became the apples of Isaza's eye. When the time came for the cow
to calve it too escaped and made its way back to its native land. And this
time Isaza did follow, even into the nethermost pit, and so found himself
trapped in Nyamiyonga's realm. The prince of darkness was, it seems, a
gentleman, for, having upbraided Isaza for his shabby behaviour, he did him
no harm but entertained him most hospitably. He would not, however, allow
him to return to the upper world, or rather, he told him to go in peace, "but
this he said to mock him, for when Isaza set out to return he could find no
gateway and no road. All day he wandered about in search of them, but at
night he found himself back at the same spot". On the other hand Nyamiyonga
showed great favour to his actual blood-brother, Kwezi, bidding him go up
aloft, where all men would do him honour.
Kwezi means 'moon' and there can hardly be much doubt about the identity
of Isaza, who was detained below the earth while his deputy the moon rose

1 Middleton, John. Myth, History and Mourning Taboos in Lugbara, Uganda J., 19
(1955), 194.
2 Fisher, Mrs. A. B. Twilight Tales of the Black Baganda, London: 1911, pp. 78-83.
3 Bikunya, P. Ky'Abakama ba Bunyoro, London and Kampala: 1927. pp. 9-15.
4 Roscoe, J. The Bakitara or Banyoro, Cambridge: 1923, pp. 323-5.

into the sky. What we have here, evidently, is not a romantic presentation of
a historical episode but an elemental myth, a splendid piece of fantasy woven
round the setting of the sun. This incidentally gives point to an otherwise
puzzling detail of the story. Nyamiyonga prefaced his offer of blood-
brotherhood with a series of riddles, which were interpreted by Isaza's
maidservant. Thus he asked for 'ignorance' (a baby which soiled the royal
couch), for 'the rope which binds water' (flour), and for 'that which turns Isaza'.
This last was construed as a cow which lowed in the courtyard so that the
king turned round to look at it. This seems very feeble, but when Isaza is
recognized as the sun it at once acquires poetic meaning: the lowing of the
cattle towards milking-time is conceived as causing the sun to turn down
towards the horizon.
My interpretation of the story of Isaza is not essentially new. Mrs. Fisher
herself had no hesitation in describing this phase of Bunyoro's 'history' as the
'reign of the gods', and the succeeding 'period' as the 'reign of the Bachwezi
or demigods'. But more recently the Greek theorist Euhemerus, who advanced
the proposition that the Olympian deities had originally been human kings,
seems to have found numerous disciples in Uganda.s For, although it is
admitted that the Bacwezi were actually worshipped as gods, it is very widely
assumed that they nevertheless had a real historical existence, being the
ancient rulers of a great Hima empire, who appeared so wonderful to their
Bantu subjects that they were deified after their deaths. And a similar
approach has led Mr. A. C. A. Wright6 and Dr. Roland Oliver7 to describe
Isaza as the last ruler of an earlier kingdom or chiefdom which was sub-
merged beneath the tide of Cwezi conquest.
It is very doubtful whether this kind of view would ever have been taken
of Nyoro and Nkore traditions if it had not been for the idee fixe which has
bedevilled the study of the subject ever since Speke sat down in his hut to
write the 'History of the Wahuma'--namely, the belief that the unusually
complex political organization of the inter-lacustrine Bantu is explicable only
in terms of the influence of a 'superior' immigrant people. (The reasons for
the popularity of this Hamitic hypothesis, as Dr. Fallers has drily remarked,
'would themselves be interesting objects for study'.9) The Bacwezi appeared
to fill precisely the required role, as a strange and formidable race to whom
the foundation of the Lake kingdoms and many features of their culture were
traditionally ascribed. And this view has received powerful support from the
present generation of African writers, "K.W."10 and Mr. John Nyakatura"1 in
Bunyoro, Messrs. Katate and Kamugungunu in Ankole, who for reasons of

5 Thus H. F. Morris and M. C. King say in their introduction to the recently published
history of Ankole (Katate, A. G. and Kamugungunu, L., Abagabe b'Ankole, Kampala:
1955) that "the Banyankole of old, like the ancient Greeks and most other peoples of
antiquity, turned their heroes into gods". The truth, both of the Greeks and the
Banyankole, is surely the opposite: they portrayed their gods as quasi-human heroes.
6 Uganda J., 13 (1949), 116.
7 Uganda J., 17 (1953), 135.
8 Speke, J. H. Journal of the Discovery of the Source of the Nile, 1863, p. 246.
9 Fallers, L. A. Bantu Bureaucracy, 1956, p. 27.
10 K. W. Abakama ba Bunyoro Kitara, Uganda J., 3 (1936), 149.
11 Nyakatura, J. W. Abakama ba Bunyoio-Kitara, St. Justin, Canada: 1947.

their own (partly because they wish to lengthen the lists of their kings, partly
because there is no place for gods in what are now Christian societies) have
emphasized the seemingly human and played down the more evidently divine
attributes of the Bacwezi and their predecessors.
It is instructive to compare the older Nyoro account of Isaza with the version
put forward more recently by Mr. Nyakatura.12 Generally, the story is the
same, but there are certain crucial differences. The land of ghosts has become
"a land which is unknown to us"; and Kwezi has disappeared from the scene,
his place as the donor of blood being taken by Bukuku, the Keeper of the
Queen's Gate, a personage whose mythological significance is not so apparent.
The historians of Ankole have taken the process a stage further." For them,
Nyamiyonga, lord of the underworld, has become Ruyonga, Mugabe of
Ankole. In other words, the original version has now been purged of precisely
those features which served as clues to its real meaning. In place of a
beautiful African variation on a universal mythological theme, we are presented
with a story which purports to be the narrative of an incident in the political
history of south-western Uganda but which, as such, seems curiously pointless
and implausible.
This tendency for myth to be converted into history as sophistication
increases appears to be a very general one. A striking example may be taken
from the well-known Ganda story of Kintu, the snake-king Bemba and
Wanfudu the tortoise. This, clearly, is in part an ordinary animal fable, one
of the many which celebrate the cunning of the tortoise, and in part an ancient
myth of rejuvenation, of which the snake, with its ability to renew its skin,
is a widely used symbol. It would, however, be quite impossible to deduce
any of this from the version of the story which was written by the late Mr.
James Miti shortly before the last war.14 For here Bemba has shed all his
reptilian characteristics and has become a mere brigand chief, while Wanfudu
is disguised as 'Kintu's best warrior, Nfudu'.
It may be thought that I am being somewhat arbitrary in accepting the
earlier versions of these stories and rejecting the later. There are, however,
grounds for believing that the traditional accounts reported by Mrs. Fisher,
Roscoe and Bikunya were, so to speak, much more nearly virginal than those
which have appeared more recently. Mr. Nyakatura is an educated man who,
we are told, served an apprenticeship to tribal history under Sir John Gray
and was clearly quite familiar with European speculations on this subject.
He is to be regarded, I think, as an able historian with special advantages
for the collection and interpretation of oral traditions, but hardly as the
authentic voice of the tribal ancestors. Moreover, whereas there seems no
good reason why an earthly kingdom, in Ankole or elsewhere, should ever
have been represented as the land of ghosts, there are powerful motives for
the transformation of the land of ghosts into an earthly kingdom. Modern
Banyoro and Banyankore are aware that kings who descend into the under-
world are not going to be taken very seriously as historical characters. Such
12 Op. cit. pp. 19-27.
13 Katate and Kamugungunu, op. cit. pp. 4-9.
14 Miti, J. K. History of Buganda (unpublished MS in the library of the School of
Oriental and African Studies, London).

features of the received tradition must therefore be discarded or explained
away. Similarly, Mr. Miti was clearly aware that Ganda tradition had to be
purged of nonsense about snakes and tortoises if the historicity of king Kintu
were to be established.
The stories told of the Bacwezi proper are mostly very similar in character
to the story of Isaza, and it is therefore reasonable to suppose that these
were beings of a similar type. Let us take a look at them in this light.
Long ago there was an African who, watching the lightning as it stabbed
through the night sky, saw it as a swift spear flying from the hand of a
celestial warrior; and so was born Kagoro, called the Fire of the Night, the
hero of the Bacwezi. This seems to me infinitely more plausible than the
reverse explanation, that a warrior called Kagoro had so keen a spear-thrust
that his ghost came to be propitiated as the god of lightning. Few people,
after all, suppose that Thor or Indra was ever the name of a man. Mugasa
(or Mukasa), the Rainmaker, the beneficent, is both the storm-cloud and its
source, or apparent source, Lake Victoria. It is to my mind fairly clear that
he is the Lake, and not the deceased ruler of an island in the Lake-in other
words, that he is a nature spirit and not an ancestor spirit. He was thought
to have been grievously offended when canoes were hauled over dry land,
because his property had thus been taken out of his watery domain." More
clearly still, the formation of Lake Nabugabo was regarded as the result of
a war between him and 'Buddu'.16 Wamala, similarly, is the lake of that
name, which the Banyoro ask us to believe that he dug and which the Baganda
say that he produced by upsetting his water-skin. (This is a small sheet of
water to have given rise to so great a god, but it is the only one actually within
the area to which the Bacwezi particularly belong, that is, in the country to
the north and south of the middle Katonga.)
Kyoma, the governor of Buganda under the Bacwezi kings, is a hill in the
Singo highlands,7 which from a Nyoro point of view could be regarded as
standing over Buganda.
King Ndahura, the terrible, belongs to a slightly different category. In
one aspect he is the malign spirit of small-pox and other spectacular afflictions,
and his 'conquests' probably represent the ravages of epidemic disease amongst
the peoples of the region. There are also, however, clear indications that he
was a celestial deity. (There is no real contradiction here; the Greeks attributed
plague to the arrows of Apollo.) While campaigning in the west he was
overtaken by his son Kiro (Night), who had already subjugated the countries
of the east.18 Like his grandfather Isaza he descended into the underworld,
but unlike him he was able to return thence (on the third day), though he never
resumed the throne. Even Bikunya was uneasy about this episode.19 He

15 Kagwa, Sir Apolo. Basekabaka be Buganda, Kampala: 1901. 4th edition, 1951.
16 Nsimbi, M. B. Waggumbulizi. Kampala: 1950, p. 41.
17 See Watson, J. M. Lake Mutukula. Uganda J., 8 (1940), 33. This note, together
with Mr. G. W. Nye s Legend of Some Hills in Bulemezi (Uganda J., 7 (1939-40), 140),
gives an indication of the kind of raw material out of which the Bacwezi were probably
18 Fisher, op. cit. p. 91.
19 Bikunya, p. 23.

prefaced his account of it with the words "The story goes that ." and
offered the explanation that Ndahura had really just gone away suddenly on
a journey. Later writers have of course improved on this lame theory by
providing the land of ghosts with a geographical location.20 They tell us that
he was taken prisoner on an unsuccessful expedition against Bwirebutakya or
Bwirebutaitsya, king of Ihangiro, from whose clutches he was later rescued
by his kinsmen. But Bwirebutaitsya is Darkness-without-end, which makes
it clearer than ever that the return of Ndahura is the resurrection of the
Unconquered Sun.
There is one point in particular which, given the historical existence of the
Bacwezi, has always seemed perplexing. How was it that so short-lived a
dynasty should have made so profound an impact? For there are only two
Bacwezi, Ndahura and Wamala, who are generally admitted to have reigned
as kings. In certain versions this honour has been variously extended to
Mulindwa, Kyomya and Kagoro, but in no case is there record of more than
three generations from first to last. If the Bacwezi were gods, however, there
is no problem. The sequence Isimbwa-Ndahura-Wamala and his brethren
is paralleled by the sequence Kronos-Zeus-Apollo, Ares and the rest. The
immortals are grouped together in the image of the human family, with its
three coexistent generations. Grandfather has been superannuated, and there-
fore Kronos and Isimbwa are not considered as regnant, while still further
in the background there is the celestial progenitor (Uranus, Isaza) who perhaps
stands for great-grandfather, now deceased, or for the clan ancestor.
One reason for the general acceptance of the Bacwezi as a historical people
is undoubtedly the circumstantial character of the stories which are told to
account for their downfall and departure. But when these stories are looked
at closely they, too, dissolve. Take, for example, the story of the soothsayer
who prophesied doom to the Bacwezi and who came, according to Nyoro
sources,21 from beyond the Nile. So far, so good; this fits well with the belief
that the Bacwezi were superseded by a new dynasty of Acholi origin. But when
one turns to the Nkore version doubts begin to arise. It is true that here, too,
the prophet from Bukidi is briefly mentioned,22 but this is probably a literary
borrowing. The authentic Nkore tradition appears to be that which was
reported at greater length a few pages earlier in the narrative,23 namely that
the prophet came from Karagwe, being himself a Muha. Clearly, there could
hardly have been two soothsayers who came from opposite directions to enact
the same part. But there is no need to discuss which of these versions is
'correct'; we are dealing with a myth whose details could be adapted to local
circumstance. The best prophets are always those who come from a far country.
For those who lived in the northern part of the area in which the legends of
the Bacwezi were current, the far country was naturally Bukidi, for those who
lived further south it was Karagwe.
Then again there is the story that the Bacwezi were much troubled by a
certain Misango, who nearly succeeded in depriving them of their herds. In
20 Nyakatura, p. 41. Katate and Kamugungunu, p. 16.
21 Bikunya, pp. 30-31 and 34-36. Nyakatura, pp. 53-56.
22 Katate and Kamugungunu, p. 31.
23 Ibid., p. 20.

the earliest version Misango is represented as the emissary of Kantu,24 the
eldest son of Nkya and nephew of Ruhanga, the Creator. In the later versions
he is said to have been an invader from Burundi,25 while Kantu has become
the central figure of a separate story. Misango means 'crimes' and Kantu
signifies 'mankind. It was the wickedness of men which caused the gods to
withdraw from the world, taking with them truth and love and leaving behind
lying and hatred;26 the rest is romantic elaboration. The belief that the Bacwezi
vanished into the lakes recalls the tradition that the Tuatha de Danaan, the
gods of the Gaels, after living for a while on the surface of Ireland, took
up their abode underground.
The reign of Bacwezi, then, does not belong to the fourteenth or any other
century but to the morning of the world, when the gods walked the earth.
In these latter days you will not see Mugenyi driving his red cattle to the
saltings, or come upon Mugasa fishing in the lake. For these are (and always
have been) the latter days, and the gods have departed long ago, and their
power can now be invoked only by a privileged few-that is to say by the
human Bacwezi, the members of the priestly clans or corporations.
Thus I suggest that the stories told of the Bacwezi cannot throw any direct
light on the history of central Africa, because the Bacwezi never existed
except in the imaginations of men. Their supernatural attributes are not
secondary but original and inherent. They are the lakes and hills of south-
western Uganda, the features and forces of the natural world, unusually
sharply personified, and converted into kings and princes in the image of a
later monarchical society. I say 'later' because, in spite of a superficial attribu-
tion of royalty to certain of the Bacwezi, the background of the stories told
about them-which is to say the background of the people who imagined them
-does not, on the whole, suggest a great kingdom, but rather a simple,
small-scale, pastoral community. So far from being the provincial governors
of a far-flung empire, the lesser Bacwezi are portrayed as the heads of a
group of related households, living close enough together to be able to join
in a dance, to hear one another's alarm call and to assist one another in
the recapture of stolen cattle.
There are indeed grounds for believing that kingship is ancient in this
region. There is the immense elaboration of ritual connected with the kings,
and there are the great earthworks in the valley of the Katonga. I suspect,
however, that the makers of the Bacwezi, who were almost certainly the
Bahima, had very little to do with the making of Biggo or Ntusi. Certainly
their descendants seem to have been able to make very little of these relics
of a largely vanished culture-less even than the barbarous Dorians were able
to make of Tiryns and Mycenae. Moreover, a pastoral people, whatever
its race or provenance, can carry only very light cultural luggage, and is
surely the last kind of community to be credited with the original foundation
of an elaborate system of administration. My own conjecture, for what it is
worth, would be that there was once, in this region, a large, loose-jointed
24 Bikunya, p. 27.
25 Nyakatura, p. 52. Katate and Kamugungunu, p. 28.
26 Roscoe, op. cit., p. 327.

Bantu kingdom of the same general type as those of Congo, Lunda and
Monomotapa; that this system broke down many centuries ago, leaving its
central area to be occupied by small groups of herd-folk; that later on, not
earlier than the sixteenth century, there arose in the north, under Lwoo
leadership, a new expansionist power, which owed its successes mainly to
control of the best supplies of iron; and that the quasi-feudal Hima and
Tusi states developed largely in response to the necessities of defence against
these marauders.
But these are speculations based in the main on the general probabilities
of the case. My main point is that oral tradition can tell us little or nothing
of events which took place more than, at most, four centuries ago. There
were kings in Greece, we have been assured, before Agamemnon; and it is
very probable that there were dynasties in Uganda before the Babito and the
Bahinda. But whatever dynasties there may have been, they lie beyond the
reach of tribal memory.
I should like to end with a word of explanation and apology to any Banyoro
and Banyankore who may have read this. It is perhaps an impertinence for
an outsider to tell people that their traditions, or some of them, are not 'true'.
I would suggest, however, that it is only very recently, and under European
influence, that Africans have come to think of such traditions as being
historically true in this sense. Previously, they were nfumo, 'legends'; they
helped to make the world intelligible; whether the events described had
actually happened was a question which did not arise. Moreover, to describe
a story as a myth is not to belittle it. If the stories referred to here tell us
little about the history of Uganda, that is because they are concerned with
more important matters-with the nature of God and Man and Death.
Banyoro and Banyankore have good reason to be proud of their ancestors,
not because they came from Abyssinia or because they were particularly
light-skinned (why should Africans want to have light-skinned ancestors?)
but because they are seen to have had poetic imagination of a high order.
Finally, it is not only African kingdoms whose histories begin with myths
of this kind. The royal house of England, according to old traditions preserved
in the Anglo-Saxon Chronicle, was descended from a certain Wotan. But
we know quite well that Wotan never really existed. He was a god worshipped
by the Germanic tribes, a being very similar to Isaza or Ndahura.



WHILE it is true to say that the Banyankole knew God in a certain sense,
they were not a very religious people. There was very little in the way
of formulated religion. A vague idea, however, existed, and still exists in the
minds of the pagans, that there is a God, Ruhanga, who was known as the
Creator, or Powerful One. He was thought to have lived in the heavens but
no prayers were offered to him; there were no priests, and no temples were
built for him, though his name was used in ejaculations such as Tata Ruhanga,
an exclamation used in joy when a cow is heard calling its calves. He has
the power to benefit man as he wills, but is above being influenced by pro-
pitiary offerings and is altogether beyond the sphere of human supplications.
He will act as he wishes apart from any human consideration. In the most
irrelevant manner the heathens bring the name of God into their conversation.
A man will escape just punishment 'if God wills'; a sick cow will recover 'if
God wills'; 'if God wills', a man will succeed in litigation even though his
defence is one sheet of lies. All this does not indicate trust in God but in a
fatalism that exempts men from all responsibility.
Ruhanga had three sons for whom he wanted to find names. It was evening
and the time the cows were being milked. When his sons had gone to bed,
Ruhanga entered their hut with three milk-pots which he commanded them
to guard for him until the morning, strictly warning them not to drink the
milk or spill any of it. At midnight the youngest son grew heavy with sleep
and some of his milk was spilt as he dozed. When he discovered this he was
afraid and turned to his brothers and begged some milk from them so that
his milk-pot might be full. This they gave. At cockcrow the eldest son upset
all his milk; but when he asked the others to pour milk from their bowls
into his they refused, saying that he would need too much to fill his empty
bowl. At dawn Ruhanga came and told each son to uncover his milk-pot.
When he looked into the eldest son's bowl he found it empty; passing on
to the second son he saw that a little had gone from it and he asked if the
boy had drunk it. The boy replied: "No, O God! I did not drink it; I filled
my brother's pot, for he spilled some of his." Then Ruhanga passed to the
third son and saw that his milk-pot was full to the brim. Ruhanga remon-
strated with them. The eldest boy he named Kairu-'little servant', for he
had proved himself faithless in his watch and henceforth he would be the
servant of his brothers. The second he named Kahima-'little herdsman' for
he would act as herdsman to the one to whom he had given milk. To the
youngest he said: "Your name is Kakama ('little king'), you shall reign over
all men; all shall fear and worship you and your word shall be law to them."
To this son was given the title Rugaba and the kings of Ankole have taken
1 Part of a lecture entitled The Peoples of Ankole, delivered at a joint meeting of
the Uganda Society and the Extra-Mural Department of Makerere College.

the title Omugabe from that day. Thus Ruhanga divided up mankind into
three classes-the kings-Abagabe, the herdsmen-Abahima, and the
peasants-Abairu, and this was the origin of the three classes in Ankole.
Ruhanga created a woman called Nyamate to be Rugaba's wife, and sent
them to rule the people of the earth.
Rugaba and his wife were not ordinary mortals for they had no mother,
but were both created by Ruhanga. They had a son called Isimbwa who was
the first of a dynasty of kings who were known as the Bacwezi, demi-gods.
These Bacwezi were not like other men but were gods, for although they were
born of women they had unending life and knew neither sickness nor death.
During their reign on earth these Bacwezi conquered and ruled the countries
of Ankole, Toro, Bunyoro, Buganda, Karagwe, Kiziba, Busoga and Bukedi.
They eventually decided to leave the kingdom of the world because they
thought it had been defiled. This is how it came about.
The Bacwezi owned many cattle while on earth. Among their cattle was
a cow named Bihogo bya Mpuga, which was of rare value and was regarded
by the Bacwezi as the most beautiful of all their cows. One of the Bacwezi,
named Mugenyi, loved this cow so much, that on one occasion he swore
that whatever evil afflicted it should afflict him also; if water was the cause
of its death, he would never again drink a drop; if salt was the cause, none of
the cattle should ever again taste salt; if it should die naturally, he would
on that day kill himself.
One day when Mugenyi was sitting in the house with his friends, a boy
came running and breathlessly announced that Bihogo had been seized with
a fit and was at the point of death. Without a moment's hesitation Mugenyi
gripped his spear to kill himself, but his brothers restrained him by force.
They sought to dissuade him from his purpose by offering gifts of other cows.
Although everyone sympathized with Mugenyi he was inconsolable; his
brothers stayed night and day to cheer him, but all in vain.
Then King Wamara, Mugenyi's uncle, sent for the witch-doctors to examine
the dead cow and to divulge to him the future as portrayed in its entrails.
In the morning the cow was brought and all the witch-doctors came together
to examine it and to read its buried secrets, but when they cut open the body
they found it quite empty; not one organ remained. This greatly perplexed
the witch-doctors and they were at a loss to understand its meaning. They
went to the king and told him of the great wonder, but they assured him that
the carcase of the cow was perfectly clean, which indicated that he should
possess great wealth and have many children born to him. As they were
speaking, a stranger, Kakara-Kashagama, crossed the courtyard and stood at
the entrance of the house. He was dressed in the skins of wild animals; his
neck, head and arms were laden with strange charms. As the king and witch-
doctors looked up, he told them that he was the wisest man in the world
and that he held intercourse with gods, men and devils. He claimed that
heaven, the world and hell were open to him. He said, "If you would under-
stand the mystery of Bihogo, take me to it. But first, give me a token of blood,
a seal of brotherhood, between us both, assuring me that no prophecy of mine,
either of evil or of good, shall jeopardize my life", Wamara called Mugenyi

and ordered him to make blood-brotherhood with the stranger, who readily
consented. When their contract was completed the witch-doctor ordered every
man to withdraw, except the Bacwezi, while he examined the cow. Then,
approaching Wamara, he held out his wand for him to touch and spoke thus:
"My master, I foresee evil only. The body of the cow, being empty, signifies
that the rule of the Bacwezi is over and the land is void. But I also see that
you will still exercise power over mankind and that you will wander con-
tinually over the earth. The smut is a black man, a barbarian, who will come
and usurp the kingdom; he will recognize no caste, will enforce no obedience;
in his time a servant will not respond, women will be ungovernable, cattle
will not heed the voice of the herdsman, dogs will not answer to the call of
their masters. The drum of the gods will be beaten by a savage and others
of his kin will possess it after him."
After hearing these words, the Bacwezi went into their houses and at night
they conferred together and decided to kill the witch-doctor, as he had pene-
trated into the secret councils of the gods and nothing was hidden from him.
But at night, when all men slept, one of the king's wives had a dream and in
the morning she hastened into the king's presence and warned him, saying:
"My master and my king, in my dream I saw the Bacwezi taking a long
journey and they became lost; they reached a land where they were unknown
and unrecognized and this happened to them because they killed the prophet
who had told them the truth." Meanwhile, Mugenyi went to the prophet and
told him to depart. "Arise," he said. "You are my blood-brother with whom
I made a covenant which I cannot break. We, the Bacwezi, have agreed to
kill you; therefore, return quickly to your own land and people." Mugenyi
then gave him some food for the journey and the prophet went back to
his home.
Meanwhile, Wamara the King had commanded his brothers to remain
with Mugenyi to guard him against self-destruction until such time as he
should recover from the death of his cow, Bihogo. One evening they prepared
a feast and they asked Mugenyi to join them. The brothers drank freely,
sang and danced and became very excited; in their songs they sang of their
greatness and their might, but scoffed at Mugenyi for wishing to cast away
his life-the life of a god for that of a cow. Their jibes at last prevailed,
and casting aside his mourning Mugenyi joined them in their revellings. In
the morning the Bacwezi departed to their own homes, for they saw that
Mugenyi had recovered from his grief. But when they had left him, Mugenyi
lay on his bed and thought over the happenings of the previous night and,
as he cogitated, his old aunt came in and mocked him, saying: "Did you not
swear that if Bihogo the faultless, the beloved cow died you would kill
yourself? This night you have been merry, you have feasted and drunk while
the cow lies dead." The words pierced Mugenyi like a spear and he swore by
Isimbwa these words: "I am despised and jeered at by a woman, therefore
will I leave the world; it is corrupt and no place for the gods." So he packed
up his belongings and went to wish his brethren farewell. When Wamara
heard that his nephew was determined to depart, he called together all the
Bacwezi and said to them: "Let us leave the kingdom of this world for it

is defiled. When women despise us who will respect us?" So they collected
their herds, their wives and their goods and departed, and were never seen
Before the Bacwezi disappeared they had divided up what was then the
Ankole Kingdom into various districts, each of which was ruled by a person
who had been appointed by the Bacwezi, from their friends and faithful
followers at the time. The division of these districts at the time was as follows:
In the north there was the district of Mawogola (called Kikokwa). The
district was ruled by one of the brothers of Wamara's mother. In the east
there were the counties of Kabula (or Bwera) and Nyabushozi (or Butaka).
The former was ruled by Karara and the latter by Rwanyabiju. In the south
and south-east, the counties of Nshara, Bukanga and Rwampara which were
ruled by Ishemurinda, Kinuma and Katuku. Finally in the north the Saza
of Kashiri or Kikyenkye was ruled by Nyawera. In later years the Kings
of Ankole enlarged their kingdom which included the counties of Mitoma
(called Bunyoro), Buzimba, Kitagwenda, Busongora, Bunyaruguru and Shema.
But Ruhinda, the son of Wamara, had shown interest in the affairs of the
people of the earth and had bought the royal drums of Ankole from their
caretaker, to whom food had been given in payment. After the Bacwezi had
gone he was to remain and rule the Ankole Kingdom and thus became the
first king of the Bahinda dynasty. Similarly, two of the sons of Kyonya, namely
Mpuga-Lukidi and Kato-Kimera-Kintu, did not go with their fellow Bacwezi
when they left this world. The former ruled Bunyoro and the latter the
kingdom of Buganda.



INTRODUCTION. Very little detailed information is available concerning
systems of farming on peasant holdings in the different areas of East Africa.
Tothill (1938) published a report on nineteen surveys conducted in various
agricultural areas in Uganda to ascertain the state of soil deterioration. These
surveys were all conducted on a similar plan; the results enabled comparisons
to be drawn about the relative productivity of the land in different parts of
the Protectorate, and initial observations were made on the effect of the
various farming systems on soil fertility and crop production. These surveys
were later used by Loewenthal (1937 and 1939) to relate different agricultural
systems and different levels of nutrition to the general health of the peasant
population. No follow-up work in the shape of published re-surveys of areas
originally investigated took place until Wilson and Watson (1956) published
the results of a re-survey of one of the areas in Teso District previously
studied by Watson in 1937. The changes which had taken place during the
period 1937/1953 were discussed in detail, and a full account of the agricul-
tural operations associated with the growing of each crop was given.
Tothill (1938) conducted the pre-war surveys on the basis of selecting in
each district three areas of sparse, medium, and dense population respectively.
In Teso District, however, a total of seven surveys were conducted, one in
an underpopulated area (Opami), three in balanced areas (Kasilang, Wera
and Kakusi), and three in over-populated areas (Ajuluku, Aputon and
Kachumbala'). In 1955 it was decided to conduct another survey in an over-
populated area of Teso and the village2 of Moruita in Kumi County was
chosen on the advice of the Teso District Team.
The survey was carried out in two parts. A first period was spent in the
field in late June 1955, when particular attention was given to mapping the
Erony and obtaining information on the first rain crops. A second period of
investigation was carried out in late August 1955 when the chief object of
study was the cropping programme for second rain crops and the cultivation
practices connected with the cotton crop.
The survey was carried out by 11 Makerere agricultural students, assisted
by a team of interpreters loaned by various Government departments. Each
student was made responsible for surveying one small portion of the Erony.
An average number of four taxpayers was interviewed by each student each day.

between lat. 1 28' and 10 25' N. and long. 330 49' and 33 53' E. The
1 Then in Bugwere District.
2 Erony or village is the term used for the administrative area of a 4th grade chief,
the lowest grade in the official administrative hierarchy.

Erony occupies an area of 4-2 square miles and lies south of the main Soroti-
Mbale trunk road. The eastern and western boundaries consist of streams
which are running with water for about nine months in each year. The
south-west boundary is formed by a dry valley, so that the Erony is delineated
on three sides by natural features. A detailed map of the Erony is shown in
Fig. 1. The Erony is situated in a part of Teso which is characterized by the
presence of large granitic outcrops on the higher parts of the land. Moruita
contains 42 such outcrops totalling 66 acres in area. These outcrops are used
as natural drying and threshing floors by the peasants. The mean annual
rainfall is not recorded in Moruita, the nearest gauge being at Ngora, 5 miles
W.N.W. The total rainfall at Ngora for the years 1953, 1954, 1955 and 1956
was 40-8, 44-3, 46-7 and 45-2 inches respectively.

VEGETATION OF THE AREA. The Erony has been extensively cleared of trees
and bushes and very little of the vegetation can be described as original. In
the rough grazing areas the common trees and shrubs found are Combretum
spp., Terminalia torulosa, Tamarindus indica, Acacia spp., Albizzia spp.,
Erythrina abyssinica, Ficus spp. and the Shea butter-nut tree, Butyrospermum
parkii. The grass communities are very similar to those recorded in detail at
Kasilang by Wilson and Watson (1955), with Eragrostis superba, Hyparrhenia
fillipendula, Imperata cylindrica, Pennisetum polystachyum and Rhynchely-
trum roseum abundant. Echinochloa pyramidalis is the most frequent grass
found in the swamps bordering the boundary streams, and in places the sugar
cane, Saccharum officinarum, has been successfully established and is spreading

SOILS OF THE AREA. The soils at Moruita have not been chemically analysed,
nor have soil profiles been systematically examined. The soil textures, however,
have been obtained by the American method described by Brade-Birks (1946)
and the soils of the Erony can be clearly divided in this manner into three
distinct types:
(1) Fine sandy loam. This type is found on the higher land on the better
managed farms.
(2) Loamy medium sand. This texture is typical of the shallow soils on
overworked farms, generally overlying murram.
(3) Silty clay loam. These soils are generally found on rough grazing land
near swamps. During the rains these soils are waterlogged, and they
set like cement when dry.

MINOR DIVISIONS OF THE ERONY. The area surveyed is the smallest admini-
strative unit recognized by the Protectorate Government. It is sub-divided by
the inhabitants into four hamlets (for a detailed definition of the terminology
employed see Wilson and Watson (1956)) These four hamlets are Moruita
('Pointed rock'), Omusio ('Where the Emus trees grow. (Emus =Mystroxyton
aethiopicum)), Ocutai ('the place of execution') and Aputiputi ('the place of
great rainfall'). These hamlets are marked on the map shown in Fig. 1. Each
hamlet is closely associated with a single Ateker chief or Clan Leader, and it

was these four Ateker chiefs, together with the Erony chief, who acted as
guides during the course of the survey.
AGRICULTURAL DATA. The data are shown in Table I and are compared with
the data for Kasilang (a 'medium density' village) as at the 1953 survey. These
data show that about 75 per cent of the total area of land is capable of culti-
vation, compared to 67 per cent at Kasilang. Of the land which is deemed
suitable for cultivation, 81-5 per cent was in fact tilled in 1955, compared to
45-5 per cent at Kasilang.
Land Utilization at Moruita (1955) and Kasilang (1953).

Item Moruita Kasilang
Acres Acres
(a) Total area of Erony .. .. .. .. .. .. 2,683 3,763
(b) Total area of cultivable land .. .. .. .. .2,014 2,529
(c) Total area of non-cultivable land .. .. .. .. 669 1,234
Cultivable Land (b)
(d) Maximum total area cultivated at any one time .. .. .. 1,641 1,151
(e) Total area resting at any one time 373 622
(f) Minimum total area of cultivable land not cultivated and not
resting .. .. .. .. .. .. 0 756
2,014 2,529

Non-cultivable Land (c)
(g) Total area under forest or suitable for forest, including rough
grazing .. .. .. .. .. .. .. .. 317 350
(h) Total area of rocky outcrops .. .. .. .. .. 66 48
(i) Total area of paths, roads, tracks .. .. .. .. .. 66 55
(j) Total area of semi-swamps .. .. .. .. .. .. 220 781
669 1,234

Table I shows that the crop/rest ratio at Moruita (i.e. the proportion of
(d) to (e)) was 4-4 : 1, a ratio which must definitely be regarded as insufficient
to maintain soil fertility. Tothill (1938) reported that the indigenous farming
system practised in many parts of Uganda was to crop land for 3 or 4 years
and then to rest, by abandoning to bush, for about 8 years. Under this system
the crop/rest ratio would be approximately 1:2. It is generally stated that,
where improved systems of farming are practised, incorporating crop rotations
and grazing of the resting land, a ratio of 1:1 is sufficient to maintain soil
fertility and this is achieved by a system of 3 years cropping to 3 years resting.
With the somewhat primitive crop rotations practised and with no adequate
provision made for grazing and manuring the resting land, it would appear
that the present system is likely to lead to an increasing rate of decline of soil
The crop/rest ratio, although a useful index, does not itself provide informa-
tion on the actual length of the resting period which many workers, especially


9qq9 BUSH

FIG. 1. Moruita Erony, Teso

N < E---

0 1/2 1

100 00

90- 90

80- ,0-- 80

70- x x-x- 70

60 x -60
50 SCx--/ 50

40 ^ 40

30- / -30

20 x- 20

10- D- 10

0 0
0 I 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20


FIG. 2

Regression of percentage of taxpayers owning farms of different size and percentage of
total farm land in farms of different size divided into discrete class intervals of one acre.


Martin (1944), have shown to be of importance. In general, a resting period
of two years under tumble-down fallow is thought to be the minimum neces-
sary for maintaining soil fertility in this environment, and the results of the
Soil Fertility Experiment at Serere Agricultural Experimental Station have
shown that there is an approximately linear relationship between length of
the resting period and subsequent crop yields. Table II shows that at Moruita
82-7 per cent of the plots were rested for 2 years or less. Only 17-3 per cent
being rested for 3 years or more. Therefore not only was the crop/rest ratio
inadequate, but the length of the resting period itself was extremely short.
Unfortunately the length of the resting period was not studied in the pre-war
surveys, so the changes in the length of this period cannot be determined.

Length of the Resting Period at Moruita.

Number of Years Resting Number of Plots Percentage
1 521 42-7
2 486 40-0
3 128 10-6
4 years or more 82 6-7
Total 1,217 100-0

(This table necessarily omits those plots which were never rested. It is known that
such plots exist at Moruita.)

The foregoing paragraphs are based on the study of average conditions
prevailing at Moruita. What is required, however, is some appreciation of the
circumstances of the average peasant, not the average circumstances of the
total population. This difference is clearly brought out by the graphs shown
in Fig. 2. Due to the very large number of small peasant holdings and the
relatively small number of reasonably large farms, the average peasant had a
holding totalling only 7 acres3 in area whereas the mean farm size is just
over 9 acres. 50 per cent of the total arable land was in the hands of the
wealthier farmers with holdings of more than 11 acres, and 25 per cent of
the total cultivable land was divided into large farms of more than 20 acres.
Perhaps the most significant fact revealed by the graphs in Fig. 2 is that 25 per
cent of the taxpayers were farming less than 3J acres of land each. These small
holdings were being farmed on a system closely approaching continuous
cropping, as may be seen from the data presented in Table III which shows
the regression of the crop/rest ratio on the size of the farm. Reasonably
satisfactory crop/rest ratios were only found in the cases of farms with a
minimum total size of 14 acres, that is with farms at least twice the size of
those owned by the average peasant at Moruita. The 'average peasant' at
Moruita was farming on a system with a crop/rest ratio of approximately 18:1.

3 i.e. the median farm size was 7 acres. 113 farms were smaller and 113 farms were
larger than 7 acres.

Regression of crop/rest ratio on the size of farm (Farm size taken in class intervals of
1 acre).

Farm size Crop/rest ratio Farm size Crop/rest ratio
0- 1 73 1 10-11 9 1
1- 2 79 1 11-12 10 1
2- 3 50 1 12-13 7 1
3- 4 63 1 13-14 7 1
4- 5 25 1 14-15 3 1
5- 6 27 1 15-16 4:1
6-7 23 1 16-17 2 1
7- 8 14 1 17-18 3 1
8- 9 12 1 18-19 2 1
9-10 12 1 19-20 1 1
Over 20 3 1

DEMOGRAPHIC INFORMATION. The overcropping of land at Moruita was due
to the population pressure on the available land, much of which was of low
agricultural potential. The population data are shown in Table IV.

Demographic Data for Moruita (1955) and Kasilang (1953).

Item Moruita Kasilang
(a) Total number of taxpayers .. .. 226 222
(b) Total number of families .. .. 190 173
(c) Total number of adult males .. .. 297 258
(d) Total number of adult females .. .. 345 311
(e) Total number of children living at home 322 311
(f) Total population .. .. .. 964 880

(g) Density of population per square mile .. 230 1 149-7
(h) Mean number of children per family living at home 1-69 1-80
(i) Mean area of cultivable land per head of population 2 09 acres 2-87 acres
(Counting children and adults equally.)

The population density at Kasilang of 150 per square mile could easily be
supported by the available natural resources, and Wilson and Watson (1956)
concluded that there could be a slight expansion of the population in this area
without prejudicing soil fertility. The population density at Moruita of 230 per
square mile, however, was exerting a markedly deleterious effect on the agricul-
tural land of the area. The lighter, sandier soils at Moruita were clearly being
over-cropped. By an ad hoc comparison of farms where soil fertility was being
maintained with farms where soil fertility was declining it was estimated that
a total of about 2-7 acres of cultivable land per head of population was the
minimum required on the existing systems of peasant farming. Over the whole
of Moruita only 2-09 acres were in fact available.

The demographic data are very similar to those obtained from the survey
of Kasilang. The mean total number of children born per family was 3-6,
compared with 3-4 at Kasilang, and the mean number of live children per
family, taken as at the time of the surveys, was 2-5 in both cases. It is debatable
whether this family size is sufficient to maintain or increase the population.
Wilson and Watson (1956) argued that it was sufficient, although the greater
part of the reported population increase between 1937 and 1953 was attributed
to immigration of adult males to Kasilang at about the age of marriage (18-25).
McFie (1956) studying the population age-group structure at Kasilang in the
course of a nutrition survey, came to the conclusion that the population was
neither increasing nor even maintaining itself. The small sample size (McFie
only took one quarter of the Kasilang Erony for his survey) and migrational
tendencies among the Iteso consequent upon marriage render elucidation of
the important problems of population stability extremely difficult. Information
was obtained on the place of origin of all the Moruita taxpayers and the
results show that 92 per cent of the taxpayers were born either in Moruita
Erony itself, or in other parts of Nyero sub-county (i.e. in adjacent villages).
The actual figures were: born in Moruita 177; born in other parts of Nyero
sub-county 31; born in Kumi County but not in Nyero sub-county 2; born
in Teso District but not in Kumi County 14; born in Uganda but not in Teso
District 2. The data show that there has been very little immigration into
Moruita and that most of the taxpayers were born on or near their present
The figures for the dispersal of adult children of Moruita taxpayers indicate
that emigration was regularly taking place, more especially with the women
than with the men. 39 per cent of all adult children have left Moruita to live
in a different sub-county and 30-3 per cent have left the overcrowded area
of Kumi County altogether to live in other parts of Teso District. The main
migratory stream was to Usuku and Amuria counties in north Teso, which
are relatively sparsely populated, due mainly to lack of adequate water
Of the Moruita taxpayers 98-2 per cent were Iteso, the others being
Baganda (3) and a Musoga. The absence of any Bahima, who are generally
found working as cattle herdsmen in most Teso villages, was significant.
Enquiries showed that, when pressure on the land became acute, there was
a tendency for Bahima herdsmen to be dispensed with and their work taken
over by the Iteso themselves. Three instances were found of Bahima being
employed as herdsmen at Moruita, but they lived in a neighboring Erony
and were not included in the survey. At Kasilang, 12 Bahima were permanently
resident in the Erony, their houses being adjacent to the cattle kraals.
Of the 226 taxpayers 39 were engaged in enterprises in addition to agricul-
ture. This percentage 17-3 per cent, approximates very closely to 18-9 per
cent, the percentage, of taxpayers engaged in non-agricultural enterprises at
Kasilang. The occupations represented at Moruita were: general labourers (17);
paid chiefs (8); carpenters (5); school-masters (3); shopkeepers (2). Roman
Catholic catechist (1); fishmonger (1); granary maker (1); and clerk (1).
The majority of the taxpayers were married and 24 per cent of all marriages

were polygamous. The actual data for the marital status of the taxpayers
were: Single (30); married with one wife (148); married with two wives (31);
married with three wives (12); married with four wives (3) and married with
five wives or more (2). These figures refer only to living wives, resident at the
time of the survey. In addition, the taxpayers claimed 244 additional wives,
either dead or divorced at the time of the survey. The total number of women
connected by some form of marriage with the 226 taxpayers was, therefore,
513, an average of 2-3 women per taxpayer.
The general level of literacy in Moruita was found to be very high, probably
due to the presence of a large number of schools, both primary and secondary,
in Kumi and Ngora counties within walking distance of the Erony. 47-8 per
cent of the taxpayers were literate in Ateso, as judged by proven ability to read
and write a short sentence in that language. 16-4 per cent were literate in
Luganda, 22-1 per cent in Kiswahili and 5-3 per cent in English. The total
school attendance percentage of 54-8 per cent was also extremely high by
Uganda standards, but there was nevertheless a pronounced difference in the
rate of school attendance between the sexes. Almost twice as many boys as
girls were attending some form of educational institution.
To conclude this section on the human population of Moruita, an attempt
has been made to estimate the number of people capable of performing
manual farm work. The data in Table V must be regarded as approximate,
since it has been assumed that all 'elderly dependants' of Moruita taxpayers
are non-workers. Clearly many, if not most, of these old people assist in the
farm work to some extent, and the estimates are therefore biased on the
low side.
Estimate of the Number of Workers and Non-workers at Moruita.

Item Males Females Total
(a) Non-workers
Children .. .. .. .. 120 79 199
Adults .. .. 71 77 148
191 156 347

(b) Workers:
Children .. .. .. 82 41 123
Adults .. .. .. .. .. 226 268 494
308 309 617

(c) Labour units per family and per taxpayer:
Labour units per family .. .. 1-4 1-5 2-9 3-0
Labour units per taxpayer .. .. 1-2 1-3 2-5 2-3
Labour units per acre of cultivable
farm land .. .. .. .. 0-13 0-14 0-27 020
(Children are counted as half labour units.)

CROP PRODUCTION. Tables VI and VII present details of the acreage of land
and the number of plots under each crop. These data enable the mean plot
sizes to be calculated, and at Moruita the overall mean plot size was 0-50 of
an acre, compared to the slightly larger mean plot size of 0-63 at Kasilang.
The more intensive crop production practised at Moruita is therefore reflected
in the mean plot size, for when land is in short supply the smallest discards
and irregular areas are turned into agricultural plots. It will be noted that
the chief opening crops (cotton, finger millet and groundnuts) have larger
mean plot sizes than the crops which follow them. There was a tendency for
relatively large areas of land to be prepared for opening crops grown in pure
stand, but in subsequent years these areas were subdivided into smaller plots
for the minor food crops. Sweet potatoes and beans, for example, were never
grown extensively but were generally confined to very small sub-plots. Hence
as the sequence of cropping continued, there was a noticeable tendency for
the plot size to become smaller.
The ratio of cash crops to food crops was fairly constant, 1:2-1 at Moruita
and 1:2-5 at Kasilang. The chief difference between the two places was in the
proportions of the various food crops grown. At Kasilang simsim formed an
important item in the diet, and the percentage of land sown to this crop did
not appear to vary greatly from year to year (8-1 per cent in 1937 and 8-6 per
cent in 1953). At Moruita the growing of simsim was uncommon, perhaps
due to the fact that simsim is a comparatively unproductive crop when
yield of food per acre is considered. For similar reasons, maize was grown to
a limited extent at Moruita but not at Kasilang. Maize was seldom grown in
pure stand but usually mixed with finger millet or sorghum; the reason given
was that it acted as an additional insurance that some cereal grains would be
Acreage of Land Under Each Crop.

Total Number of Mean plot Mean plot
Crop acreage plots acreage acreage
(Moruita) (Kasilang)
Cotton .. .. 636-3 895 0-71 0-75
Finger Millet .. 397-8 838 0-47 0-71
Groundnuts .. 199-8 393 0-51 0-66
Cassava .. .. .. .. 199-6 414 0-48 0.50
Sorghum .. 245-2 661 0-37 0-53
Cow Peas ..152-1 294 0-52 0-60
Sweet Potatoes .. .. 64-6 179 0-36 0-31
Maize .. .. .. 434 160 0-27 0-25
Bananas .. 3-7 10 0.37 0.30
Beans and Grams and Rice 1-9 7 0-27 -
Simsim .. .. .. .. 1-4 10 0-14 0-57
Total Cash Crops for the year 636-3 895 0.71 0-75
Total Food Crops for the year 1,309-5 2,966 0-44 0-59
Grand Total .. .. 1,945-8 3,861 0-50 0-63

(Total acreages of crops include all crops grown during the year. In some cases two crops are
grown on the same plot during the year and the total acreage of crops in the table, therefore,
exceeds the maximum acreage of crops grown at any one time, which is presented in Table I.)

obtained from the land in the event of bad weather, pests or diseases suppressing
the yield of the staple grain crop. Mixed cropping of millet and sorghum was
exceedingly common at Moruita for the same reason. In calculating the
acreages of each crop the proportion of the two crops in each mixed plot was
estimated and the theoretical acreage of each crop in pure stand was recorded.

Cropping Data per Taxpayer and per Family.

Percentage of arable
land under crop Acreage per taxpayer Acreage per family
Moruita Kasilang Moruita Kasilang Moruita Kasilang
Cotton .. .. 32-7 28-2 2-8 1-7 3-4 2-2
Finger Millet and
Sorghum .. .. 33-0 44-6 2-9 2-7 3-4 3-5
Cassava .. .. 10-3 5-3 0-9 0-3 1-1 0-4
Groundnuts .. 10-3 4-4 0-9 0-3 1-1 0-3
Cow Peas .. 7-8 3-8 0-7 0-2 0-8 0-3
Sweet Potatoes .. 33 4-3 0-3 0-3 0-3 0-3
Maize .. 2-2 0-4 0-2 0-0 0-2 0-0
Bananas .. 0-2 0-4 0-0 0-0 0-0 0-0
Simsim .. 0-1 8-6 0-0 0-5 0-0 0-7
Total Cash Crops .. 32-7 28-2 2-8 1-7 3-4 2-2
Total Food Crops .. 67-3 71-8 5-9 4-3 6-9 5-5
Total.. .. .. 100-0 100-0 8-7 6-0 10-3 7-7

Cassava and groundnuts were grown more frequently at Moruita than at
Kasilang. Cassava ensures a greater yield of bulk food per acre than any
other crop and, although solely a carbo-hydrate foodstuff, is probably grown
more in overpopulated areas for this reason. The explanation for the difference
in the proportion of groundnuts is that the light sandy soils of Moruita favour
this crop, whereas the heavier soils at Kasilang act as a deterrent to planting
groundnuts, a crop generally highly favoured by the Iteso. Groundnuts
appeared to replace simsim as a subsidiary food crop at Moruita.
The most interesting point arising from a study of the cropping data is that
overcrowded Moruita has a larger mean acreage of crops per taxpayer and
per family than the 'balanced' village of Kasilang. This difference must be
attributed to greatly lowered soil fertility, and consequent poorer crop yields
at Moruita which necessitate a larger mean area of both cash and food crops
per head of the population to meet basic living requirements. The Moruita
peasants in fact appeared worse off, both in regard to available food and annual
cash surplus, than the Kasilang peasants, and this point was emphasized to
the survey team members in private discussion, when the older inhabitants
made it clear that it required more land to feed and clothe their families in
1955 than it had done in their youth. Family sizes and family needs have
undoubtedly increased over the years, but there was no doubt in the minds
of the people themselves that the soil fertility and crop yields at Moruita were
steadily declining.


Crop Sequences Practised at Moruita

Number of plots Percentage of
Opening crop Rotation following total plots
sequence analysed
Cotton .. .Cotton, Millet, Cotton 44 8-1
Cotton, Millet, Groundnuts 32 6-0
Cotton, Millet, Cassava 24 4-4
Cotton, Groundnuts, Millet 17 3 1
Cotton, Millet, Millet 14 2-6
Cotton, Potatoes, Millet 10 1-9
Miscellaneous 45 8-1

186 34-2

Finger Millet .. Millet, Cotton, Millet 22 4-1
Millet, Groundnuts, Cotton 17 3-1
Millet, Cotton, Groundnuts 11 2-0
Millet, Cotton, Cotton 10 1.9
Miscellaneous 59 10-9

119 22-0

Groundnuts .. Groundnuts, Cotton, Millet 25 4-6
Groundnuts, Cassava (for 2 years) 24 4-4
Groundnuts, Millet, Cotton 10 1 9
Miscellaneous 27 5-1

86 16-0

Cassava .. .. Cassava, Cotton, Millet 14 2 6
Cassava (for 2 years), Cotton 14 2-6
Cassava (for 3 years) 10 1-9
Miscellaneous 31 5 7

69 12-8

Sweet Potatoes .. Potatoes, Cotton, Millet 16 3-0
Potatoes, Cassava (for 2 years) 10 1-9
Potatoes, Groundnuts, Millet 10 1-9
Miscellaneous 21 4 1

57 10.9

Cow Peas Cow Peas, Cotton, Millet 10 1.9
Miscellaneous 11 2-2

21 4-1

(Whenever finger millet is mentioned in the above table, it refers to the planting of finger
millet and/or sorghum as a first rains crop, followed in the same year by the planting of a
pulse crop, generally cow peas, in the second rains. Whenever cassava is mentioned it is
understood that the crop remains in the ground for a minimum period of 15 months before

CROP SEQUENCES. The crop sequences practised at Moruita were found by
asking the taxpayers the field history of each plot for the period 1952-1955
inclusive. In all, 67 different crop sequences were recorded. Table VIII shows
those sequences found in ten or more cases. The remainder are grouped together
under their common opening crop.
These sequences are of great agricultural interest, since it is generally loosely
assumed (Tothill 1938) that the main crop rotation in Teso is cotton(l); finger
millet followed by pulse(2); root crop(3) with the variation on the lighter soils
that groundnuts is the opening crop in the place of cotton. Table VIII shows
that the commonest sequence practised is Cotton(l); finger millet followed by
pulse(2); cotton(3). The 'accepted Teso rotation' was in fact practised only on
6-1 per cent of the recorded plots. Cotton is shown to be the commonest
opening crop (34-2 per cent of all plots are opened to cotton) but it is followed
by finger millet, which appears as the first crop on 22 per cent of the plots
examined. Groundnuts, grown on the lighter soils, are third (16 per cent),
and cassava (12-8 per cent), sweet potatoes (10-9 per cent) and cow peas (4-1
per cent) are all occasionally used as opening crops at Moruita.
The use of finger millet as an opening crop was unexpected, since this
crop requires land in good tilth. Kerkham (1953) writes, "Finger millet is not
grown on new land owing to its need for a fine seedbed. Normally grown in
the second year of the rotation Teso/Lango". Where finger millet was planted
in this place in the sequence it appeared that the land had originally been
prepared for cotton or groundnuts, but that when the time for planting arrived,
the farmer feared that he had through misjudgement sown an insufficient
acreage to the staple food crop and so finger millet was sown to make good
the deficiency.

CULTIVATING IMPLEMENTS. Table IX lists the various agricultural tools and
implements found in use at Moruita during the course of the survey.
There is a very marked similarity between the data for Moruita and for

Agricultural Implements in Use.

Item Number counted Mean number of implements per
Moruita Moruita Kasilang
Hoes-short handled 764 3-4 3-3
Axes .. .. .. .. 281 1-2 1-2
Slashing knives (pangas) 43 0.2 0.3
Ploughs-large (4-ox) 34
Ploughs-small (2-ox) .78
Total ploughs .. .. 112 0 5 0"5
Draught oxen .... 199 0-9 1 1
Bicycles .. .. 122 0 5 0 5
Draught oxen per plough .. 18 2"4
Arable acres per plough .. 146 11.0

Kasilang, the only difference being in the number of oxen available for
ox-ploughing. In both villages, it would appear that the number of oxen is
insufficient in relation to the ox-ploughs available. Assuming that two teams
of oxen are required to keep any given plough in maximum use during the
ploughing season, a total of 600 oxen would be required at Moruita, whereas
only 199 are available. The limiting factor for timely ploughing of land
would therefore seem to be the availability of working oxen, and not the
capital involved in the purchase of ploughs.
The bicycle is included in Table IX since it must be regarded as the main
means of transporting agricultural produce. No carts or wheelbarrows were
found at Moruita, nor were any ox-drawn implements other than mould-
board ploughs seen. No land at Moruita has ever been tilled by mechanical
means, although all peasants are familiar with agricultural tractors which are
frequently seen working on the murram roads.

TREES AND VEGETABLES. Almost every taxpayer possesses some pole trees
or fruit trees which he has planted near his homestead, and 7,865 such trees
were found growing at the time of the survey. Of the 226 taxpayers, 197
(87 per cent) had pole trees of some sort or other, mostly Melia azedarach.
These thin, spindly trees have been popularized by officers of the Forestry
Department, who encouraged the planting of seedlings in south Teso in the
1930s. They are said to provide stronger building poles than Cassia siamea,
the second most important pole tree grown, but they do not coppice well and
only a few are now planted. Other trees noted were Markhamia platycalyx,
Chlorophora excelsa (muvule) and Toona ciliata.
The commonest fruit tree found was the mango, Mangifera indica, said to
have been introduced by the Baganda. 159 taxpayers (70 per cent) together
owned 647 mango trees; of these the majority, 140 taxpayers, possessed pole
trees as well. Mangoes are mainly eaten by children and herdsmen. 454 citrus
trees were counted in the homesteads of 132 (58 per cent) taxpayers. The
commonest types were orange and lemon; where grapefruits were grown they
were rarely eaten by any Iteso. 120 taxpayers (53 per cent) grew both pole
trees and citrus trees on their land. Very few peasants planted the quick-
growing pawpaw tree, of which only 27 were found. The pawpaw fruits were
occasionally sold to Asian traders living in the neighboring Erony.
Vegetables appeared to be increasing in popularity in Moruita; cultivators
who grew vegetables generally consumed the whole crop themselves, incor-
porating them into soups and stews to embellish the staple finger millet porridge
(atap). 37 taxpayers had tomatoes growing on their farms, 15 had pumpkins,
8 had red peppers, 7 had coco yams (Colocasia antiquorum), 5 had cabbages
and 2 grew onions. The vegetable plots were never very extensive and were
generally sited near the homesteads, often in or adjoining the banana gardens.

NOTES ON THE COTTON CROP. Detailed notes on the husbandry of each of
the chief crops grown in Teso have been presented in a previous article
(Wilson and Watson, 1956), but a few additional facts on the cotton crop were
obtained in the course of the Moruita survey.

Manning (1949) has shown the great importance of early planting on the
yield of cotton. In Buganda Province, nearer to the Equator, the best sowing
date for cotton was shown to be early June, and it was demonstrated that in
years where the mean sowing date was very late, in August or September, the
yields were well below average. The mean sowing date in Teso is not known
with any accuracy, and so an attempt was made to estimate the 1955 mean
cotton planting times at Moruita. It was not possible for the peasants to
remember the actual date of cotton planting with any great accuracy, and
so sowing dates were given in the form of 'early', 'middle' or 'late' of the
month in question. The data have been summarized in Table X, which shows
that the mean sowing date is near the middle of June. The peasants prefer
not to plant cotton in May since this early planted crop requires more weeding
than later plantings. This extremely satisfactory mean sowing date is probably
due to the fact that cotton is very often given a whole year to itself and does
not often follow the planting of a food crop in the early rains. Although
Table VIII demonstrates that cotton is only planted as an opening crop in
34-2 per cent of the rotations practised in 1955, nevertheless in most of the
cases cotton is planted in land sown to another crop the preceding year and
left fallow February-May. Only the August planted cotton follows a food
crop, such as millet, planted during the same year and August was the planting
month only in 6-5 per cent of the plots examined.

Month of Cotton Planting.

Number of plots
Month planted in month Percentage
April .. .. 52 64
May .. .. 197 24.2
June .. .. 322 40.0
July .. .. .. 187 22 9
August ... 56 6 5
The number of plots recorded is less than the total number of cotton
plots at Moruita. Whenever there was any doubt about the planting
time, the plots in question were omitted from the table.

Detailed observations on 110 plots of April, May and June planted cotton
showed that 73 per cent of the plots were kept fairly clean and had been
weeded three or more times, and that 27 per cent of the plots were dirty, having
been weeded only once or twice since planting. The correct spacing of 3 feet
x 1 foot was found to be practised in only 13 per cent of the fields. 77 per cent
of the plots were too closely planted, the mean distance between plants being
less than 2 feet 10 inches, and 10 per cent were overspaced, the mean distance
between plants being greater than 3 feet 2 inches. Further, in 24 per cent of
the plots examined the sowing rate was moderate, averaging not more than
6 seedlings per hole, whereas in 76 per cent of the cases the crop had been
sown much too thickly, groups of twenty plants or more per hole being

common. This overplanting is difficult to control when the seeds are merely
dropped into holes by a woman walking along the row with a handful of seed,
and some form of simple ox-drawn planter would not only allow more efficient
control of the seed-rate but would enable a great saving to be made in the
use of cotton-seed for planting purposes, allowing more of the crop to be used
for the expression of oil.
As Moruita is situated near to a ginnery (3 miles), to co-operative societies
(in the same sub-county), and to cotton stores (on the main road adjacent to
the Erony), each taxpayer was asked by what means he usually disposed of
the crop of seed-cotton. 61-5 per cent stated that they took the crop them-
selves to the nearby cotton stores, 24-8 per cent marketed the crop through
one of the local producers' co-operatives, and 12-4 per cent transported the
crop by bicycle to the nearest ginnery. 1-3 per cent paid an agent to sell the
crop for them to save themselves trouble.

THE TESO FARMING YEAR. The information obtained from the Moruita
survey and the Kasilang survey enables an approximate calendar of agricul-
tural operations and activities to be drawn up. Most of the details given in
the calendar are of general application not only to Teso District but to parts
of Bukedi and Lango Districts as well, but the facts are most relevant and
accurate to central and south Teso, to the areas designated 'Zone 1, and
'Zone 2' by Watson (1941) in his paper on general aspects of Teso agriculture.

Month Expected Rainfall Agricultural Activities
January 0-2 in. Cotton picking.
If rain falls, finger millet with sorghum sown. If dry,
early burning of communal grazing land may com-

February 0-2 in. Cotton picking ends. Cotton sorting and selling.
Finger millet sown, whether month wet or dry.
Grass burning. In very dry years, many 'rain-making
ceremonies' will be conducted. Season for weddings
and many cattle change hands as a result.

March 3-9 in. Long rains generally commence. Ploughing season,
new land being broken for cotton. Late finger millet
and sorghum sown; simsim planted.

April 5-10 in. Ploughing continues. Ground broken for cotton
ploughed a second time. Cassava, maize and late
maturing groundnuts planted. Finger millet and
sorghum weeded for first time. Tree planting, citrus
and pole trees.

May 3-8 in. Early cotton planted and more land re-ploughed for
cotton. Sweet potatoes, cassava and maize planted.
Final weeding of finger millet and sorghum.

June 2-5 in. Main cotton crop planted. Early maturing ground-
nuts planted. Early finger millet and sorghum
harvested and fermented.
July 3-8 in. Cotton planted, and early cotton weeded. Finger
millet and sorghum, groundnuts, simsim harvested.
Finger millet and sorghum fermented and stored in
granaries unthreshed. Simsim dried on racks and
stored threshed.

August 5-9 in. Second rains generally commence. Late cotton
planted, and bulk of cotton crop weeded. Finger
millet and sorghum stubbles ploughed. Cassava,
sweet potatoes, cowpeas and late sorghum planted.
Remainder of groundnuts and simsim harvested.

September 3-8 in. Weeding cotton and other second rain crops. Sweet
potatoes and cassava planted. Agricultural shows held.

October 2-5 in. Early cotton picked; cowpeas weeded. Sweet potatoes
and cassava harvested, surplus being sliced and dried
for storing in granaries. Agricultural shows held.

November 1-3 in. Cowpeas harvested and sweet potatoes and cassava
harvested and prepared for storage. Cotton picking
begins in earnest.
December 0-3 in. Cotton picking and sorting.

The author's thanks are due to the Makerere College Research Grants
Committee for financing the project and to Professor Fergus Wilson for
criticism of the original manuscript and for a number of suggestions. The
author also wishes to acknowledge the help given by the District Commissioner,
Teso, and the members of the Teso District Team who were most generous in
placing Ateso-speaking field staff at the disposal of the non-Teso team of
Makerere undergraduates.

Brade-Birks, S. G. (1946). Good Soil. Eng.U.P.
Kerkham, R. K. (1953). Notes on the Principal Annual Food Crops. Govt. Printer,
Loewenthal, L. J. A. (1937). An Investigation into health and agriculture in Teso,
Uganda. Govt. Printer, Entebbe.
(1939). Abstract of a further survey of health in relation to agriculture in
Teso, Uganda. Govt. Printer, Entebbe.
Martin, W. S. (1944). Emp. J. Exp. Agric., 12, 21.
Mcfie, J. (1956). A comparison of the health of six villages consuming different
types of food. Govt. Printer, Entebbe.
Manning, H. C. (1949). Emp. J. Exp. Agric., 17, 245.

Tothill, J. D. (1938a). A report of nineteen surveys done in small agricultural
areas in Uganda. Govt. Printer, Entebbe.
(1938b). Fifteen agricultural surveys selected from the above nineteen
surveys. Govt. Printer, Entebbe.
Watson, J. M. (1937). Unpublished report on Kasilang, Teso.
-. (1941). E. A. Agric. J., 6, 207.
Wilson, P. N. and Watson, J. M. (1955). Uganda J., 19, 183.
(1956). Uganda J., 20, 182.



K AKORO Rock lies close to the sub-county headquarters of that name in
Pallisa County of Bukedi District. It is a large granite outcrop, typical
of those found over many parts of Bukedi, Teso and Lango Districts. It runs
approximately north to south, close to the Kakoro-Kachumbala road, and
viewed from that road it appears to consist of two separate hills joined by a
saddle. It is in the southernmost of the two hills, on which a trigonometrical
point is situated, that the rock paintings were found a year or two ago by
T. D. H. Morris, from whom news of their existence was obtained.
The approach to these paintings is by a well-defined track leading from the
Kakoro-Kachumbala road to the foot of the southernmost hill and thence up
the hill on its south face. Near the top of the hill is a ledge of rock on which
stands a large monolith, some thirty or more feet high. It is on the south and
west faces of this monolith that the first group of paintings is situated. On
its south side the monolith is approximately nine feet wide at the base; a
large boulder lies close against it. The paintings start on the monolith immedi-
ately above this boulder, that is about four feet from the base, and cover the
whole width of the monolith and extend upwards for about six feet on the
right-hand side and for about three feet on the left-hand side. On the west
face the paintings extend over an area approximately six feet square near
the base of the monolith.
The condition of these paintings is poor; those on the west face are almost
entirely obliterated and only traces remain to show that the designs were
geometric. The paintings on the south face are clearly visible on the right-hand
side, but become faint in the centre. Flaking of the rock may be responsible
for the damage, but it also appears that the elements have caused the paint
itself to deteriorate.
The designs include concentric circles which are a feature of the paintings
at Nyero, Ngora and Asuret.1 They also include less usual geometric designs
(see Fig. 1). In the left centre, where the paintings are least distinct, there are
lines which may possibly represent human figures similar to those found at
Nyero. The designs in the centre are jumbled and for this reason there may
have been some superposition, but it is difficult to be certain of this point
because the paintings are indistinct and are executed in identical colouring
throughout. This colour is red ochre. In technique there appears to be no
difference between these and the paintings at Nyero.
The second group of paintings is on a rock face immediately above a small
shelter a few yards to the north of the monolith. This group is also in red
ochre and is in a very poor state of preservation and no one design is com-
plete. It is, however, perhaps of interest to note that the concentric circle motif

I See Uganda J., 17 (1953), 8-13 and Uganda J., 19 (1955), 90.

is absent. The shelter itself is very small and provides little natural protection
from the elements. Nevertheless, a pot shard was found in the shallow deposit.
The third group of paintings is in the saddle between the two hills. Here
there is a low shelter penetrating far under the rock. The inner recess of this
shelter has been blocked with stones, presumably by hunters, leaving only
a narrow aperture. The shelter can only be entered by crawling on all fours
and the height between the floor and the ceiling in the inner recess is only



I foot

Fig. 1.
Group of paintings in red pigment on south face of monolith, Kakoro.
(From freehand drawings)
about two feet. The paintings are on the lip of the rock at the entrance to the
shelter and can only be viewed by crouching inside the shelter. The designs
are executed in white paint which shows up well against the dark colour of
the rock. (Fig. 2.) It appeared that the paint had been daubed, probably
with the fingers, in a noticeably more careless manner than in the first and
second group of paintings.


/ **

Fig. 2.
Paintings in white pigment in rock shelter, Kakoro.
(From freehand drawings. Faint markings are shown
by dotted lines.

Another rock painting site has recently been discovered by P. L. Carter in
Tira sub-county in Teso District. This site is approached by a track which
leaves the main Soroti-Kyere road at mile 8 from Soroti and runs roughly
westwards for a distance of some three miles to two granite outcrops. The
first outcrop seen after leaving the main road contains a large cave, which
is full of bats. A second outcrop, in which the paintings occur, lies about 200
yards north of the first.
Carter states that in the second outcrop a large slab has at some time
fallen from the rock face, thus creating a shelter some 40 feet long and 30 feet
wide. The paintings are on the inside of an overhanging piece of rock at the
entrance to this shelter. They now cover an area approximately 10 feet by
4 feet, but were originally more extensive, because traces of pigment occur
on other parts of the rock face.

Fig. 3.
Group of paintings in red pigment at Tira, (from freehand drawings.)
The paintings are executed in red pigment and contain geometric designs,
including the usual concentric circle motif. Two of the circles, however, consist
of dots instead of the usual lines. (Fig. 3.) It appears as though these dots
were applied with the fingers and not with a brush. The state of preservation
of all the paintings is poor.
The shelter gives good protection from the elements and has probably been
used as a dwelling-place, for the deposit on the floor yielded quartz debitage,
pottery and bones.
Eight groups of paintings at five different sites have now been discovered
in Uganda-there must undoubtedly be many more-and some comparison of
these sites may be of interest. When evidence of technique from further sites is
available it may prove possible to date the paintings. All that can be said
now with certainty is that the paintings are the work of a race of people
who occupied the country before the arrival of its present inhabitants, the
Iteso, some 200 to 300 years ago. Drawing styles, colour of pigment and
superposition of paintings give clues to dating. Excavation may also produce

evidence of pottery and other artifacts which can be linked to known cultures
elsewhere. At present, however, the available evidence is meagre.
The eight groups of paintings are at the following sites:
Latitude N. Longitude E.
1. Nyero (i) 1 29' 33 51'
2. Nyero (ii) 1 29' 33 51'
3. Ngora 1 30' 33 48'
4. Asuret 1 39' 33' 37'
5. Kakoro (i) 1 11' 340 04'
6. Kakoro (ii) 1 11' 34 04'
7. Kakoro (iii) 1 11' 34 04'
8. Tira 1 39' 330 35'
Groups 1-3 have been described in U.J., 17 (1953), 8-13, Group 4 in U.J., 19
(1955), 90, and groups 5 to 8 in this article. Red pigment only, with no dis-
cernible variation in shade, is found in groups 1, 3-6 and 8, and white pigment
only in groups 2 and 7. White and red nowhere occur in the same group and
no other colours have been found. Superposition of paintings occurs with
certainty only in group 1 but probably in group 5. Superposition at 1 shows
that concentric circle designs are later than the naturalistic drawings. Con-
centric circle designs, usually with five circles in the design but also with
three, four and six, occur in groups 1-5 and 8, although in 8 some of the
circles are executed in dots. Second in popularity is the 'pod' design which
occurs in groups 1, 2, 8 and possibly 5. Naturalistic paintings of canoes with
figures and line drawings of humans occur with certainty only in group 1.
The paintings in groups 1 and 8 occur on the wall of a large and inhabitable
shelter. Those in 2, 4, 6 and 7 are at the entrance to a small and often exposed
shelter. Those in 3 and 5 are on rocks of curious shape in exposed positions,
which could not possibly have provided any shelter. It is interesting to specu-
late that these rocks may have had some magical or religious significance.
The sites at 1, 4, 6 and 8 were occupied, as is evidenced by the discovery of
pot shards. Pottery has also been discovered in a number of other shelters
in the rocks at Ngora. Virtually no other artifacts have been found, although
up to the present no excavation has been carried out except at Nyero.



THE antagonism between Sir Harry Johnston and Sir Clement Hill, recently
described by Dr. Ingham and Dr. Roland Oliver' is a good example of
how the clash of personalities has affected the destinies of the East African
territories, and of the recurring differences between the man on the spot and
the man having the ear of the minister in London. The antipathy between
Johnston and Hill is apparent beneath the formal wording of their despatches,
and Hill's suspicion of Johnston's policy and actions is even more evident
in his minuted comments to the Foreign Secretary on the Special Commis-
sioner's proposals and reports.
According to Sir Harry's brother, Alex Johnston,2 this antagonism had its
origin in a somewhat trivial incident: "Lord Salisbury recalled Johnston to
England (from the Oil Rivers and Cameroons Vice-Consulate) in the middle
of 1888 and had a long talk with him in private on African affairs in general.
The next week-end he was asked to Hatfield, much to the chagrin of a highly
placed official at the Foreign Office, jealous of privilege shown to one so
young in the Service, whom he persisted thereafter in representing as an
'outsider', with the implication that my brother was in some way unworthy
to be admitted into the higher ranks of officialdom. This persistent critic and
opponent-Sir Clement Hill-somehow engendered the idea that Harry was
a rash, pushful person, whose very activity detracted from his dignity as an
It was against this background that Sir Harry was set to carry out his
Special Commission and to explain to the Foreign Secretary how some form
of union or federation of the two Protectorates could be effected. Events
which had occurred before real consideration had been given to the question
of any change in the relationships between the two territories had indicated
the differing attitudes adopted by the two men towards the many problems
inevitable in running and developing backward countries with insufficient
knowledge, staff and money. The two problems on which there was most
disagreement were those of the position of the local administrators relative
to the railway authorities, and the best and cheapest manner in which to main-
tain law and order.
Hill was a member of the Uganda Railway Committee, on which he seems
to have been rather under the influence of the managing member, Sir Francis
L. Callaghan. This tended to shape his thinking and, together with the fact
that the railway was the most important enterprise in the Protectorates and

1 K. Ingham. Uganda's old Eastern Province: The Transfer to East Africa Protectorate.
Uganda J., 21 (1957), 41; and Roland Oliver. Sir Harry Johnston and the Scramble for
Africa. 1957.
2 The Life and Letters of Sir Harry Johnston. 1929. p. 114.

the one which caused a continual barrage of embarrassing questions in the
House, resulted in Hill's backing the railway in almost every case when
differences occurred. He was always ready to defend the imperium in imperio
set up by the engineer-in-charge whenever his organization came into conflict
with the Commissioners and their staffs. The causes of this friction have been
set down elsewhere; they are so many and so sharp it was an amazing feat
that, despite the dissipation of effort and enthusiasm thus caused, the
railway was completed and the administration of the Protectorates carried
on. This leads one to think that there must have been a deal of good
will and give-and-take amongst the junior officers in the field, and that
it was the senior officials of both administrations who sought, pursued
and magnified grounds for complaint and bickering. Among the major causes
of friction were: delays on the railway telegraph and the poor construction
of the line; the provision of escorts for railway and telegraph parties; the
policing and protection of rail and telegraph camps, staff and stores; thefts
of material; the conditions in the coolie camps; the setting up of a separate
police force and courts having jurisdiction over the railway zone and staff;
rail facilities for officials on duty; difficulties over the movement of troops;
and the payment by the railway for the protection afforded to it by the civil
administration. Johnston naturally considered that the government of Uganda
as a whole was at least as important as the advance of the railway through
the Eastern Province, but Hill, always aware of the proddings of parliamen-
tary and press critics, thought otherwise. There was much to be said on both
sides and one wonders whether, if the antagonism between Hill and Johnston
had not clouded the issues, these two men (together with the Commissioner
of East Africa at Zanzibar and the Chief Engineer of the Railway) might
not have contrived a satisfactory compromise.3 Apart from the fact that Hill
and the Railway Committee would have to deal with one Commissioner
instead of two, there were many advantages in having the railway under
one administration. Worthy of consideration on this score was the fact that
the Lake formed a natural boundary and that the Government at Entebbe
had found itself unable to undertake much development of its outlying Eastern
On the maintenance of law and order, Johnston's policy was one of concilia-
tion whenever possible coupled with the use of adequate force quickly and
effectively whenever conciliation failed. Hill, on the other hand, was exceed-
ingly touchy on the subject of police measures and punitive expeditions and
several of his minutes imply that in some cases, where the men on the spot
considered such measures were necessary and urgent, he thought they could
have been avoided. It must be remembered that, both in Parliament and the
press, he had to face the criticism of the Liberals, who were only too willing
3 It was unfortunate that Johnston's first communication with Whitehouse, the Chief
Engineer, was a personal letter in which he gave his impressions of rail and telegraph
progress, construction methods, staff conduct, etc., and also, somewhat ungraciously as
he admitted in a later letter, his views on the use of Whitehouse's wife's name, Florence,
for the rail terminus on the lake. Johnston advocated either the local name, Kisumu,
or one connected with a celebrated African explorer. The railway authorities balanced
the account when one of the first steamers on Lake Victoria was named Winifred after
Lady Johnston.

to make political capital out of Uganda's difficulties and to infer that dis-
turbances had been created so that military action to quell them could be
undertaken. He was opposed in principle to the use of native levies, although
he appears not to have realized that without much larger military votes the
intermittent and temporary employment of Masai levies was inevitable. He
was always quick to compliment officials who were able to patch up a com-
promise, however manifestly unsatisfactory, insecurely based or short-lived.
During the respite afforded by these compromises, however, the line nosed
slowly forward towards the lake, and to that extent Hill's policy achieved its
object and was successful. The culmination of this policy came during his
visit to Uganda in 1900 at the time when Lieutenant-Colonel Evatt was
recovering from the severe mauling his force had received from the Nandi.
Hill repudiated Johnston's dictum, 'when you punish, punish severely', and
gave instructions for peace to be made at almost any cost, so that construction
of the railway could continue and thus hasten the long-desired announcement
in Parliament that the line had reached the Lake. The effects of this inter-
vention are well known and resulted in the continued insecurity of the railway
route and the expenditure of much money and effort before the Nandi were
eventually subdued six years later. Although Hill's action was regarded by
the angry local officials as inefficient meddling, he could claim that the line
was pushed through to Kisumu in 1901 and finally consolidated by the end
of 1903.
Hill's expensive and somewhat extraordinary visit to Uganda did nothing
to bring the two men closer together, as the impression was bound to be
created that the Superintendent had been sent out to make sure that Johnston
was carrying out his special commission within the limits set by the Foreign
Office, and that he was not, in his enthusiasm and exuberance, committing
the Home Government to a series of extravagant and costly ventures.
Johnston had been sent out only the year before to investigate and advise
on a list of specific problems, one of which was the question of the closer
union of the two sister Protectorates. It must have been irksome to one of
Johnston's calibre and seniority, with a fine record of achievement in other
parts of Africa, to have his policy and acts subjected so openly to inspection
by the administrator from Whitehall. It is probable that Hill's presence was
resented by the staff throughout the country as a tactless. insult to their chief.
The visit was certainly ill-timed as the period of Johnston's Special Com-
mission had already run half its short course, and as a result there was even
less chance of the two men settling their differences and viewing Uganda's
problems without personal prejudice.
Accustomed as we are to ever-increasing secretariats and a multiplicity
of departments, it is surprising to note how the transfer of the Eastern Pro-
vince was made with an absence of formalities, being treated rather as a
normal incident of administration. There seem to have been no special files
opened on the subject, there were a few letters and special despatches, and
occasional meetings, but nothing approaching a handing-over ceremony. The
military change-over was a more complicated operation than the civil transfer,
especially in connection with the location and payment of troops. This fact

in itself emphasizes the important role played by the military in maintaining
law and order in the troubled Province and in keeping open the rail and
road routes.
One of the reasons why the change-over was completed so easily was that
very little development had taken place in the Province. It was on the extreme
eastern fringe of Uganda Protectorate, and there were more than enough
problems to be tackled in Buganda and the west by the handful of staff
available. Apart from the civil stations at Kisumu (Hobley and Partington),
Mumias (Corbett), Nandi (Mayes), Ravine (Isaac), Baringo (Hyde Baker) and
a civil and military centre at Fort Ternan, there were only military, railway,
staging and store posts in the Nyando and Rift Valley and over the Mau,
and a Sub-Commissioner (Bagge) at Naivasha. These officials were trans-
ferred with their stations at the change-over and several of them benefited in
status and pay as a result. From time to time officers made safaris into their
districts, but for the most part administration was confined to the neighbour-
hood of the stations or bomas and to the country through which the connecting
roads ran. There was no post south of Kisumu, and the plains east of the
Gulf were known as South Kavirondo. A projected extension of the sphere
of influence towards Elgon had come to nothing owing to the need to concen-
trate in face of the Nandi and Kipsigis threat to the railway. Johnston wished
to establish a post north of Baringo to control ivory and other caravans, to
protect the wild life, and to keep an eye on the Suk and Turkana. This was
frowned on by the Foreign Office who wired Johnston on 7 March 1901:
"Report further before taking any definite action north of Baringo, since
H.M. Government do not at present desire to increase their responsibilities by
establishing a permanent post in that district."4
A summary of the state of the Province on transfer shows clearly how little
real administration had been undertaken, in face of the over-riding necessity
for keeping the Nyando Valley clear and the construction of the railway up
to schedule. A new administrative post had been sanctioned in September
1901 to house Hobley's headquarters at Kisumu and to take the place of the
Ugowe Bay station. The port was open for traffic, a market had been started,
and two European firms, Boustead, Ridley & Co. and a German company
were trading there. The settlement of Indian farmers along the railway line
from Kibigori to Kibos was being discussed in the early part of 1902, as a
means of safeguarding the line and of providing some traffic for the railway.
There was a small caravan trade from Mumias and Kavirondo with the
countries to the north, and also from Ravine and Baringo to the ivory
collecting tribes to the north and north-west. General trade consisted almost
entirely in the supply of food for the civil and military posts, for the expedi-
tions against the Nandi, and for porter caravans and railway construction
parties. This was centred mainly on the depots at Kitoto's (near present-day
Miwani), Kisumu, Kakumega and Mumia's. Indian traders were beginning
to appear beyond railheads with their head porters and donkeys. In January
1901 the ubiquitous Allidina Visram was able to advertise branches at
Naivasha, Shimoni (Ravine), Nandi and Port Ugowe.
4 Foreign Office Prints. January-June 1901.

Communications were generally poor and the roads were in fact only
caravan tracks, with the exception of Sclater's road, which had, however, been
virtually abandoned from the Mau to Lessos, and of the service road which
in 1900 had been pushed ahead of railway construction in the Nyando Valley.
Although there were a few private contractors, most of the carriage was done
by the Uganda Transport and all manner of experiments had been tried-
oxen, mules, donkeys, camels, carts-with varying success. The railway was
coming into use, although it was not finally completed until 1903, and Johnston
had taken over the telegraph from the railway in 1901.
There were military posts at railhead and along the line while small garrisons
of soldiers and police were stationed at the various bomas. Their duties were
confined to the containment of the local tribes rather than to attempting
any extension of the area under effective control, and most of the companies
were split into small detachments engaged solely on escort duties. Concentra-
tion only took place to meet emergencies, such as the Nandi outrages of 1900
which were followed by Evatt's expedition. When the immediate object was
achieved by the concentrated field force, or more often when it was only
partially achieved, the troops were dispersed and resumed their normal escort
and police duties.
Apart from the lack of staff and funds, there were many reasons why
development of the Eastern Province had lagged behind the progress made
in Uganda proper. The revenue produced east of the Lake was insignificant,
although Johnston had instituted the Rs. 3 hut tax in some of the districts;
apart from the areas around Kisumu and Mumia's, however, the tax was
scarcely collected, as it was not Johnston's practice to levy tax on tribes
which were not really controlled and protected by the Government. Also,
despite the valiant efforts of Hobley and his small band, little was known
of the tribal organization of the Province; there seemed to be no chiefs and,
apart from the Wanga dynasty in the person of the young Mumia, there was
nothing approaching the centralized, quasi-feudal government of the Uganda
kingdoms. There had been very little missionary endeavour and scarcely
anything was known of the numerous tribal tongues.
It was understandable that the Eastern Province tended to be looked on in
Entebbe and London as a somewhat tiresome stretch of country on the
eastern edge of the Protectorate which had to be crossed in order to reach
Uganda proper. It had always been an unsettled area and more than once
the need to divert men and money to the Province from development in
Uganda had exasperated the Entebbe officials. There had been operations
against the Nandi, Kamasia, Suk and Kipsigis and occasional small affrays
with the Kitosh, Kakamega, Kikelelwa, Nyangori, Luo and other tribes. These
were not conducive to peaceful administration and development, and it was
in fact not until 1906, when the Nandi were finally subjugated, that any
real progress was made and the way opened up for settlement of the 'waste
lands' of the area, and for the improvement in the lot of the Nyanza tribes
under peaceful rule.
The main emphasis had to be on protecting the communications between
the Coast and the Lake and most of the stations had been sited and estab-

lished with this in mind. The Guaso Masa post had been built in 1895 to
protect the then all-important northern caravan route from the Nandi warriors,
and, shortly after the transfer of the Province, a station was opened at
Kericho to attempt to contain the Kipsigis and to prevent any large scale
combination of the Nandi and Kipsigis against the railway. For the past ten
years administrative and military effort had been concentrated in the first place
on securing the caravan routes and later on making roads, organizing and
protecting the transport service across the Province, trying to give some pro-
tection to those tribes who had accepted Hobley's regime, and, finally, on
pushing through the telegraph and railway with the least possible interference
from the neighboring tribes. With all this on their hands, it is not surprising
that the few officers found little time to organize their districts, to get to know
the people, their customs and languages, or to extend the area of effective
jurisdiction. Whatever other factors may have engendered the idea, there can
be no question that the demands of the Bantu Kavirondo to return to Uganda
are prompted by nostalgic memories of the well-ordered and peaceful days
their ancestors had spent under the Uganda Administration before the transfer.
It is now proposed to detail the steps by which the transfer was effected.
Johnston had reorganized the western part of the Province in January 1901,
and at the same time had appointed Hobley as Sub-Commissioner for the
Eastern Province. His headquarters were at Port Ugowe, which was also the
centre for the Nandi district, and the districts coming directly under his
control were those of Mau, Baringo, Suk and Nandi. Very little is heard at
this time of the eastern end of the Province, since it was mainly in the hands
of the railway with a Sub-Commissioner, Bagge or J. P. Wilson, looking after
civil matters from Naivasha. Johnston reported in February that he intended
occupying the site for the proposed capital near Londiani and, to that end,
he was transferring most of the Ravine staff there, since, whatever the out-
come of his proposals for the joint capital, he considered it to be the place for
the headquarters, administrative and military, of the Eastern Province and
probably for the departmental capital of the Nandi District.5
Not until after his return to England in June 1901 did Johnston learn
that the Foreign Office was having second thoughts regarding federation, and
that the transfer of Uganda's Eastern Province to the East Africa Protectorate
was being considered. In a letter to the Foreign Secretary (9 August 1901)
he opposed any extension of the East Africa Protectorate to the Lake, and
recommended that the sway of John Ainsworth, now Sub-Commissioner
at Nairobi, should not be extended to include those sections of the Masai
living in the Mau, Baringo and Nandi districts. But this arrived too late to
alter the decision, and in the following month Hill minuted a letter from
Jackson, who was Acting Commissioner of Uganda since Johnston's departure:
"I want Jackson to be Deputy Commissioner of East Africa Protectorate so
that he can have all Masai and Eastern Province under him."6 Jackson was,
of course, a natural choice as he had been stationed at Ravine from the end
of 1895 until he took part and was wounded in the Mutiny operations, and
5 Johnston to Foreign Office, 4 February 1901.
6 F.O. 2/453 (Public Record Office).

he had been sent to the Eastern Province again with special powers during
the 1900 Nandi expedition, while Johnston was touring the western marches
of Uganda.7 The decision was duly transmitted to East Africa, and on 17
November Jackson sent a message to the Collector at Fort Ternan that all
officers should meet Sir Charles Eliot, Commissioner of the East Africa Pro-
tectorate, when he passed through Fort Ternan on the 22nd. On the 18th
Isaac, who was in charge of the Nandi District with headquarters at Ravine,
left his station to meet Eliot at Londiani, and from there the two men went
on to Njoro. At the end of November Jackson and Eliot met at Njoro to
discuss the estimates for the Eastern Province.8 The only question of any
difficulty was that of deciding what troops should be retained in the Nyando
Valley and what contribution Uganda should make to their upkeep. Major
Gorges, O.C. Eastern Military District, was called in and it was decided that
the two companies of Uganda's Indian Contingent should remain in the
Nyando Valley and that they should be controlled by the Uganda Protec-
torate, which would continue to be responsible for their pay and upkeep. This
detachment was essential as a reserve for the protection of the railway, to keep
open the road to Uganda, to overawe the truculent Nandi and Kipsigis and
to inspire the coolies with confidence.9 It was agreed that in the event of
hostilities taking place the two Indian companies would, while the operations
lasted, form part of the Field Force and as such be under the control of the
O.C. Troops, East Africa Protectorate. As far as local troops were con-
cerned, Eliot insisted on retaining two companies of 4th (Uganda) Battalion
K.A.R. in the Eastern Province, and he made provision for these in his draft
estimates. Jackson who was still Acting Commissioner of Uganda was naturally
not anxious to lose the Sudanese and the mixed Somali and Swahili companies
stationed there. He particularly wished to return the Sudanese to Unyoro
but, as the Uganda Protectorate was quiet, he agreed to leave them on loan
until Eliot could recruit two full companies to take over the protection of
the railway.'0
On 5 January 1902, Eliot was in the Nandi country and held a meeting with
the so-called chiefs, and on the 7th a mass meeting of over 1,000 tribesmen
was with Jackson's consent, arranged at Muhoroni. Eliot warned the Nandi
that, if thefts and raids by certain sections did not stop, severe punitive action
would be taken against the tribe as a whole. As Hobley went to Entebbe
in January, presumably to confer with the Acting Commissioner, Jackson, on
the subject of the transfer, it is probable that Hobley did not meet his new
chief, when Eliot visited the Kisumu district in that month. On 25 January
1902, Eliot informed the Foreign Office that Jackson and Hobley had dis-
cussed the inter-protectorate boundary question and had agreed on the River
7 Jackson was already Deputy Commissioner of Uganda. He was transferred to the
East Africa Protectorate service with the same rank from 1 April 1902.
8 For some details of this meeting see H. B. Thomas. A Federal Capital for East
Africa-Some Early Proposals, Uganda J., 2 (1934-5), 247-9.
9 From Uganda's point of view, it was convenient to keep the healthy Nyando
stations, as it allowed the troops of the 5th (Uganda) Battalion (formerly the Indian
Contingent of the Uganda Rifles) to have some months of their service away from the
unhealthy areas of Uganda.
10 Jackson to Foreign Office, 24 December 1901.

Sio. Tindi's and Mangiki's were to be transferred to the East Africa Protec-
torate, but no agreement was reached over the boundary near Elgon.
On 3 February 1902, Jackson wrote to inform Hobley that, as he would
be in charge of the whole of Kavirondo, Nandi and Lumbwa from 1 April
1902, he should assume control over these districts now. Foaker at Mumias
was ordered to send all correspondence dealing with the Elgon district to
Hobley and not to William Grant in charge of the Central Province at Jinja.
Grant and J. P. Wilson were also instructed to send any letters they had in
their files concerning the Eastern province to Hobley. Jackson noted that the
north-western and eastern boundaries of Hobley's charge had not yet been
settled, but, in the case of the former, the Sio river and the southern slopes
of Elgon would be sufficient for the moment and, for the latter, there was
such a large extent of no man's land between Nandi and the Mau that there
should be no difficulty in arriving at a workable boundary later on." On
14 February, Eliot stated that he thought it essential that any tribe likely to
cause trouble should be under one administration. This referred, of course,
to the Masai, who were considered by Eliot at that time to be the great danger
to the security of East Africa.12
On 15 March 1902, the following notice signed by Eliot as H.M. Com-
missioner appeared in the official Gazette: "Notice is hereby given that after
the 31st instant the borders of the East Africa Protectorate will be extended
to Lake Victoria. The territory taken over from Uganda includes the following
stations: Naivasha. Ravine, Baringo, Fort Ternan, Kisumu, Nandi and
Mumias." Before this, on 5 March, the Foreign Secretary (the Marquis of
Lansdowne) had signed the Order-in-Council transferring the former Eastern
Province to the East Africa Protectorate, under the title of Kisumu and
Naivasha provinces.
On 28 March, Eliot gave his views on the division of the area into two
provinces, and suggested that one province should comprise Mau and Rift
Valley, as these districts were scarcely peopled at all and the land was con-
sidered suitable for European settlement; and the other should include
Lumbwa, Nandi and Kavirondo, districts with a tropical climate unsuited to
Europeans, and with a dense native population, not in all cases friendly.
Hobley would take charge of both provinces as Assistant Deputy Com-
missioner, and he and Bagge would also act as Sub-Commissioner, Kisumu
and Naivasha, respectively.1 This arrangement was notified in the Gazette
on 1 April 1902, when the districts in the Kisumu Province were set out as
Kisumu, Mumias, Fort Nandi and Fort Ternan; and those in the Naivasha
Province as Naivasha, Nakuru, the Ravine and Baringo. The same issue of
the Gazette promulgated an amendment of the Hut Tax Regulations main-
taining the tax at Rs. 3 per hut for the two new provinces. The final delineation
of the boundary between the two provinces was approved by the Foreign
Office on 8 July 1902; the line was described as running along the summit of
the Elgeyo escarpment and cutting the railway line at mile 501, to the west

11 Entebbe Archives file A/19.
12 F.O. 2/569 (Public Record Office).
13 F.O. 2/570 (Public Record Office).

of Londiani station, and thence continuing along the top of the Mau.14
Various results affecting the two provinces stemmed from the transfer,
which also naturally affected policy and progress in Uganda. The most impor-
tant of those taking effect immediately was the separation of the Uganda
Railway from the Uganda Protectorate and the absorption into the East
Africa police force of the 39 officers and 259 men of the Uganda Railway
police.1 This undoubtedly hastened the breakdown of the railway's imperium
and led to a more healthy relationship between the railway and the administra-
tion. An incidental and not very important result was that free travel along
the Sclater's and Nyando Valley roads was restored, as there was no East
African legislation authorizing the tax on caravans, carts and animals pre-
viously levied by the Uganda authorities.
Of the long term results, the most important was probably that the 'waste
lands' of the two new provinces were made available for white settlement,
just at the time when Eliot was formulating and pressing his settlement
proposals. In general the Uganda Government was doing little to further
white settlement, and it may be that, had the Eastern Province remained under
Entebbe, the Rift Valley, Mau, Kericho, Nandi and Plateau farms would
not have been alienated.
The addition of this huge tract meant that the centre of gravity of the
enlarged East Africa Protectorate had shifted to the west, with the result that
in a few years Zanzibar was to lose much of its importance and the removal
of the administration from Mombasa to some more central position was
For a long time the western half of the transferred territories was to cause
trouble to its new rulers, and to divert men, money and effort from the more
rewarding enterprises being undertaken in the east. In fact hardly had the
transfer been effected when, on 5 April 1902, the Nandi raided Kibigori
station and carried off 30 steel sleepers. This was quickly followed in the same
month by 500 Kipsigis warriors attacking a rail camp at mile 524 and, for
a time, it looked as if the long-dreaded combination of Nandi and Kipsigis
was about to take place. Troops were rushed up from Nairobi, General
Manning, Inspector-General of the K.A.R., visited the area and drew up a
scheme for defeating a combined attack, all railway camps from mile 516 to
mile 560 (Lumbwa to Kibigori) were fortified, the line through the Nyando
Valley was patrolled and permission was given for the establishment of a
station at Kericho under Major Gorges. In the following year there was trouble
with the Kamelilo Nandi which resulted in an expedition and the setting up
of the Soba sub-station below Tinderet. In June, 1905, there was an expedition
to Sotik, in September a column visited Kisii, and finally the large expedition
to Nandi took place from late 1905 to early 1906. The subjugation of the
Nandi brought peace at last to the troubled area and allowed the administra-
tion to consolidate and to begin positive government. The other side of the
picture is, of course, that with this weight off their shoulders, the Uganda
administrators and soldiers were able to push out from the central kingdoms
14 Foreign Office Prints, 1902.
15 It is often erroneously stated that the Uganda Railway never reached Uganda.

and for the 4th K.A.R. to bring security and ordered government into the
hitherto lawless and largely unexplored countries to the north.
The question of a capital for the two protectorates, a project which was so
near to Johnston's heart, was at first associated with the greater question of
union or federation and later with the fate of the Eastern Province. As early
as January 1900, Johnston informed Whitehouse he had selected a site just
outside the Uganda Railway mile limit and not far from Mount Blackett.
He did not let the grass grow under his feet and in April, Bagnall, who was
then in charge at Ravine, asked for an assistant so that he could devote more
time to supervising the construction of the new road from Ravine to the
capital. The following month, James Wallace applied for a land grant of
3,000 acres for stock grazing and agriculture near to the new capital, and
was prepared to surrender the grant already made to him of land near Ravine.
In October 1900, Johnston was advocating to Hardinge, the Commissioner
and Consul-General of East Africa, that the capital of the two merged Pro-
tectorates should be on the Mau, but Hardinge, in far-away Zanzibar, does
not seem to have been drawn.
The project really began to take shape after Hill, Colonel Gracey who was
engaged on an inspection of the railway, and Whitehouse visited the site and
the railway authorities decided to build a station some 1 miles from the
centre of the area selected by Johnston. They also advised laying the site
out immediately as the railway inevitably brought irregular settlers, squatters
and other undesirables. Johnston had already taken possession on behalf of
the Uganda Government because, as he put it, of its importance in connection
with the railway, the water supply and the Uganda transport, A telegraph
station had been set up for the use of the military, the railway and the
administration and Johnston thought the time was opportune to transfer the
Ravine staff to the new site. He proposed that the Government should sell
some of the Ravine buildings and land to missionaries for a settlement and
for a sanatorium.
The desirability of establishing a sanatorium in this healthy area was one
of the many advantages Johnston saw in his scheme for the new capital,
and one which was to be pressed several times in the next few years. In June
1901, Drs. Mackinnon and Macpherson recommended that a 10 to 12-bedded
hospital, with a doctor and English nurses should be built on the Mau. In
August that year, for some reason not given in the correspondence, the Foreign
Office was trying to get money from the Treasury for a similar joint institution
in the Kikuyu country. However, on 18 January 1902, a small joint committee
chose a new site for the joint sanatorium at Njoro.
At no time did the Foreign Office show much enthusiasm about Johnston's
plans for the new capital, despite the fact that Hill, on his visit to the site,
gave the impression that the project had his approval. However that may
be, by April 1901, Downing Street had decided to adopt a cautious attitude
towards the suggested bestowal of the name of King Edward's Town to the
new capital, and proposed to defer seeking His Majesty's pleasure in the
matter until it was more certain that the site would prove in every way a

But Johnston was not to be deterred and on the 18th of the next month he
appointed F. W. Isaac as Collector of the Nandi District to reside at the
site of King Edward's Town. Isaac was instructed to leave Ravine as soon
as possible to establish the new station; he was to refer all magisterial questions
to J. P. Wilson, who was acting Sub-Commissioner in Kisumu while Hobley
was on leave; and finally all matters regarding the sale or lease of land in
the Nandi district and the allotment of the various divisions of the new
capital to officials, European settlers, Indians and natives were to be decided
by Jackson, who was to act as Commissioner when Johnston left for home.
These orders were soon to be countermanded and three days later Johnston
wrote to Isaac: "I have been examining the site of the new capital with the
surveyors and Mr. Jackson. We have found parts of it very marshy. Of course
this is the result of the abnormally heavy rains; still I have decided to wait
a little longer before opening a new station at the site, until such time as
Mr. Jackson has definitely decided in favour of the present site or has
sanctioned the site being laid out a little higher up or lower down (as the
case may be) to avoid proximity to swampy down. You had better therefore
wait before moving from Ravine until Mr. Jackson definitely authorizes you
to take the step."16 The letter ends with an injunction to Isaac to have the
road from the Ravine to the new capital put into a good state of repair. It must
have been most galling to Johnston to leave the country before he had seen
this cherished dream put into visible shape on the ground. It is evident from
these letters, which were some of the last he wrote before leaving for home,
that he had no idea there had been any change of policy, and that the Nandi
Plateau, about which he always spoke in lyrical terms, was about to be
severed from Uganda and, with it, the concept of federation or union was
to cease to have any practical significance for many years."

16 Entebbe Archives file A/19.
17 Some clamour for federation arose locally in 1905 when the Foreign Office handed
over to the Colonial Office its responsibilities in East Africa. Struggling European
settlers voiced, in letters to the East African Standard, the hope that under the new
regime they would be treated more like their fellows in Canada, Australia and South
Africa: and Lugard was suggested as an experienced first High Commissioner who
would give effect to their wishful thinking. There is no evidence that this proposal made
any impression on the Colonial Office.



RUNYANKORE belongs to what is usually termed the 'Runyoro group'
of languages which are spoken in the Western Province of Uganda and
in Bukoba District of Tanganyika. It has, however, been recognized for some
years that Runyankore is a separate language from Runyoro and not merely
a dialect of it, there being wide differences phonetically, grammatically and
in matters of vocabulary. In this essay I propose to deal chiefly with the
phonetic structure of Runyankore, several aspects of which are of considerable
interest when examined in relation to the allied Bantu languages of Uganda.
First it must be explained that 'Runyankore' is a word which was applied
in this century to the language spoken by the inhabitants of the administrative
district of Ankole-itself a creation, in its present form and extent, of the
20th and last decade of the 19th centuries. In fact, however, there are various
dialects spoken in different parts of the district and though the grammatical
differences are slight there are considerable differences in pronunciation.
Similarly, in Kigezi, dialects of the same language are spoken (save in Bufum-
bira where Runyarwanda prevails), but because Kigezi was a separate
administrative district, the language was known by a different name, being
called Rukiga after the dominant Bakiga tribe of Ndorwa. The principal
differences in pronunciation lie between the speech of the semi-Hamitic
Bahima and that of the Bantu agriculturalists, whether they be the Bairu of
Ankole or the Bakiga of Kigezi, and this has been the principal consideration
in the task of evolving a common written language for the two districts. In
Ankole this has not, in the main, presented any great difficulty since the
Bairu have accepted without question that the speech of the Bahima, and in
particular that of the Omugabe and his circle, is the standard to be followed
and that the written language should, as far as possible, represent this pro-
nunciation. There is, therefore, in Ankole, a form of 'received pronunciation',
to use a useful piece of linguistic jargon, which the bulk of the population do
not necessarily try to follow, but which is recognized as the ideal of an
educated Munyankore and which is commonly referred to as 'the language
of the court'. Throughout this essay, when wishing to distinguish between
this and other variations of pronunciation, I have called it Kihinda speech
from the name of Omugabe's clan, the Bahinda.
As might be expected, the Bakiga have been less ready to accept Kihinda
speech as a determining factor in settling problems of how the common
language should be written. At the Orthography Conference of 1954 the Kigezi
representatives objected strongly to the adoption, for instance, of the is which,
though used by the Bahororo of Rujumbura, as well as by their fellow Bahima
of Ankole, is unknown to the Bairu and the Bakiga. However, agreement was
I From the Uganda Government Essay Competition 1956.

finally reached and there is now one written language, officially recognized
in both districts, though called Runyankore in Ankole and Rukiga in Kigezi.
The Bahima when they first entered Ankole, at least six centuries ago, may
well have spoken at that time an Hamitic language, but, if they did, no
vestige of it survives and it is profitless to try to trace the peculiarities of
Bahima speech to such an origin. On the other hand, it can be shown that
certain of these peculiarities represent an older form of the Bantu language
than that now spoken by the Bairu and neighboring people. A striking
example of this is the use of the single consonant sound ts. The Bahima, like
the Batusi of Ruanda, have both a ts and an s sound whereas the Bantu
speakers of Uganda have only an s. A word such as empitsi (a hyena) in
Runyankore, appears in Runyoro and Luganda and in the speech of the Bairu
and Bakiga as empisi. The Ur-Bantu stem of this noun is, however, -piti 2 and
in Lubwisi, a primitive Bantu language of the Ruwenzori mountains, the noun
is mpiti. A sound change from the explosive consonant t to the affricative ts
and so to the fricative s is not an uncommon linguistic phenomenon and it
is reasonable to suppose that such a process has taken place in the Bantu
languages of the Lacustrine area, and that at the time when the Bahima
entered Ankole the process had reached the intermediate affricative stage.
Living a life apart from the people whose language they had adopted, the
Bahima retained this consonant sound as it was when they first began to speak
the language, whilst among the Bantu the process has proceeded and the
sound has become a fricative.

Runyankore, like Runyoro and Luganda, is a five vowel language. These
vowels, a, e, i, o, and u may be either long or short. When they are long their
pronunciation is substantially the same as that of the Luganda long vowels.
In the case of short vowels, however, there are certain peculiarities of pro-
nunciation which have no counterpart in Luganda.
The short vowel i has a centralized version, the articulation of which lies
midway between that of a normal i and that of a u. This has led certain
linguists to suppose that the Runyoro group of languages of which this is a
common feature, is a seven vowel language group.3 In fact, however, the
different versions of the vowel all belong to one phoneme, that is to say the
use of a particular version is governed simply by phonetic and not by gram-
matical or etymological criteria. A centralized i is only found in medial
positions in a word and then only between certain consonants as, for example,
the words empikye (an ant hill) and okuzina (to dance). There are varying
degrees to which this centralization can be taken and when the following
syllable contains a back vowel, that is to say o or u, the centralized i may be
articulated so far back in the mouth as to be indistinguishable from a u. The
word harimu (there is within) and tindikumanya (I do not know) are pro-
nounced as though they were harumu and tindukumanya and indeed, in the
2 The circumflex indicates that the vowel is close.
3 Malcolm Guthrie The Classification of Bantu Languages. Oxford University Press,

past, they were usually spelt in this way. According to the present ortho-
graphy rules, the grammatical structure of the word must determine the
spelling in such cases, but sometimes it is not possible to come to a definite
conclusion. The Runyankore word for 'completely', for example, is pronounced
kyarumwe but it is doubtful whether its spelling should be kyarumwe or
kyarimwe. If the -rumwe or -rimwe (once) is in agreement with orubu (a turn)
then it contains a real u, but if it is in agreement with eizooba (a day) then
it contains an i articulated so far back as to be indistinguishable from a u. No
Munyankore can answer this question.
There are also variations of the vowel a. Between b or m and h it may
become close and approximate to the centralized i mentioned above, as for
example in Abahima. When a back vowel follows, it is heard as a u, as for
example in amahuri (eggs) which is pronounced amuhuri.
In Kihinda speech there is a strong tendency to devoice or whisper short
vowels. This characteristic is not only unknown among the Baganda and the
Banyoro but also among the Bakiga. When in a final position in a phrase,
a vowel is almost always whispered except after r and n. In medial positions
this tendency is most marked between two voiceless consonants as, for example,
in ekitabo (a book) in which the i and the o are whispered. The initial vowel
of a phrase and a long vowel are never whispered.
In Kihinda speech not only are short vowels often devoiced but they are
sometimes elided altogether. This occurs when a short vowel separates two
homorganic consonants as, for example, in ekikyere (a frog), ekyanira (a
white eagle) and omuriro (a fire) which are pronounced ekkyere, ekyanda and
omurro. This elision is also found between certain other consonants as in the
case of okushutama (to sit) and amajuta (oil) which are pronounced okushtama
and amajta. Although this is a Kihinda characteristic, the elision of a short
vowel between two r's is also the general rule in Runyoro and is indeed recog-
nized in the spelling of that language.
There are twenty-one consonants and two semi-vowels and, in the case of
five of these, two letters of the alphabet are used in writing to represent single
consonant sounds. The consonants are b, c, d, f, g, gy, h, j, k, ky, m, n, ny,
p, r, s, sh, t, ts, v and z, and the semi-vowels are w and y. Table I below

Labio Post
Labial Dental Dental Alveolar Alveolar Palatal Velar Glottal
U. V. U. V. U. V. U. V. U. V. U. V. U. V. U.
Explosive p t d ky gy k g
Fricative b f v s z sh j h
Affricative -- ts c -
Nasal m n -- ny -
Liquid r -
Semi-vowel w y -

U = Unvoiced. V. = Voiced

shows the manner in which each of the consonants and semi-vowels is
The pronunciation of these consonants in many ways differs considerably
from Luganda and the following are notes on certain points of interest.
b As in Luganda and in most Bantu languages of this part of Africa, this
consonant is normally a fricative, half-way between an English b and an
English v. In the nasal compound mb, however, it becomes an explosive
similar to the English b.
c This consonant is pronounced like the ch in the English word church.
d In Runyankore, this consonant only occurs in the nasal compound nd
and in words of foreign origin.
g, gy, k, ky The representation of these consonant sounds has been the
greatest single problem in the task of working out a practical orthography.
In Runyankore, as in Luganda and Runyoro, the prefixes gi and ki immedi-
ately before a vowel produce the single palatal consonants gy and ky with
consequent lengthening of the subsequent vowel, as for example eki-aro
which appears as ekyaro, the a being long. So, too, in all three languages
g and k when followed by i are palatals and the Luganda and Runyankore
words for a door, oluggi and orwigi, are pronounced with a soft palatal
final syllable. It might be argued that, for consistency, they should be written
oluggyi and orwigyi and indeed in Ankole this spelling prevailed until recently;
but since g and k must always be palatal before i it was decided at the Ortho-
graphy Conference that the y was unnecessary. The real complication arises
from the fact that in Runyankore, unlike Luganda and Runyoro, g and k are
almost always also palatal before e and hence we have okugenda (to go) in
Luganda and Runyoro but okugyenda in Runyankore. Were the g and k always
palatal in Runyankore in such cases, then it would simply be a matter of
following the convention for gi and ki and omitting the y. There are, however,
cases where, because there has been contraction of an a, the g and k are velar
before e, as for example in okugenda ( okukenda ( insisted on a distinction being maintained by the use of gye and kye where
the consonants are palatal and ge and ke where they are velar. In the present
orthography, therefore, g and k represent palatal consonants if followed by i
but velar consonants if followed by any other vowel, whereas gy and ky can
only represent palatal consonants. This, however, gravely affects the general
orthographic rules for the representation of long vowels. In general, the com-
binations CyV and CwV represent a contraction of CV + V and the result
is a lengthening of the vowel.4 Since this is a rule with few exceptions, it is
unnecessary to write the vowel double in such cases and this is the accepted
practice in Luganda, Runyoro and Runyankore. In Runyankore, however,
kyV and gyV do not now necessarily represent the result of kV +V and
gV +V with an inevitable lengthening of the vowel, and the rule no longer
holds in the case of these combinations. For example, in okukyeta ( eta, to call it) the e must be long but in okukyeta (from the verb stem -kyeta,
to examine) the e is short. Since a vowel is invariably written single after
4 C=consonant; V=vowel.

gy and ky, as after all Cy combinations, there is, therefore, under the present
orthography rules, no way of showing whether this vowel is long or short.
To make matters even more complicated, the palatal gy and ky of Kihinda
speech are not palatal at all in the speech of the Bairu and the Bakiga who
pronounce these consonants even further forward in the mouth and make
them post-alveolar, approximating to the English j and ch; hence the common
spelling of Bachiga.
j In Runyankore, unlike Luganda and Runyoro, this consonant is pro-
nounced like the French j and until recently was usually represented by zh.
There is a tendency among the Bairu to replace it by z.
n In the compound nt the n is dental and not alveolar and in the com-
pounds ng and nk it is velar. Since, however, in Runyankore, unlike Luganda,
the velar nasal never appears except in ng and nk compounds, there is no
need, as in Luganda, to use the long-tailed r. There are a few words in which
n is followed by a semi-vowel. This combination cannot be written ny as this
is used to represent a single consonant sound, the palatal nasal. This problem
also exists in Luganda where ny is used. In Runyankore, however, this com-
bination is represented by ni, e.g. eizinio (the hindquarters of a cow).
p This consonant very rarely occurs except in the compound mp and in
words of foreign origin.
r Though the Bakiga and some Banyankore use an I in their speech, this
consonant is not accepted as a true Runyankore sound and it has been agreed
that r alone shall be written except in foreign words including the name
of the district which is a European corruption of Nkore.
s, sh The Banyankore have both the consonant s and sh but there is a
tendency among the Bairu to replace sh with s.
t Throughout the Runyoro group of languages, though not in Luganda,
this consonant is dental and is pronounced as in French.
ts As has already been mentioned, this consonant is a characteristic of
Kihinda speech and the Bairu and Bakiga replace it with s.
The double consonant, which is an outstanding characteristic of Luganda,
does not occur in the Runyoro group in which, in similar circumstances, the
consonant is single and is preceded by i, as for example omusajja (a man)
in Luganda and omushaija in Runyankore. The only consonants which are
written double in Runyankore are n and m in cases where the first personal
prefix n precedes a stem or another prefix beginning with a nasal, as for
example, in ninnaga (I am throwing away) and mmanya (I know) ( The reason for this spelling is merely to give clarity of meaning; in pro-
nunciation the nasal is not doubled though it has the effect of lengthening
the preceding vowel and ninnaga is actually pronounced as though it were
written niinaga. In speech, however, double consonants will occur through
the elision of a short vowel which has already been referred to, as for
example in ek(i)kyere (a frog).
The assimilation of a nasal to the following consonant in compounds is
a common feature of Bantu languages and in Runyankore the nasal appears

as m before bilabial consonants (e.g. embuzi before t, though this is not shown in the spelling (e.g. ente as a palatal n before a palatal consonant though this also is not shown in
the spelling (e.g. enkima consonant, again not shown in the spelling (e.g. enkoko and as an alveolar n before any other consonant. In such compounds the
nasal also produces a variety of sound changes which are as follows:
n + b (fricative) > mb (explosive) embuzi (stem -buzi) goat
m +h > mp empunu (stem -hunu) pig
n +r > nd endwara (stem -rwara) illness
n +n > n enana (stem -nana) sickles
n + m > m emanzi (stem -manzi) hero
These changes correspond to similar changes in Luganda and other Bantu
languages. In Luganda, however, there is another nasal compound change
n + y > nj. This sound change, with the exception of one word, does not
operate in Runyankore where a regular formation is found such as nyoora
< n-yoora (I scoop up). The one exception is the word njeru (a white cow or
animal for sacrifice) which is formed from the stem yeru (white).
In Luganda there is, in addition to the sound changes referred to above,
the operation of what is called the Ganda Law.5 According to this law, when
a nasal is prefixed to a stem beginning with 1(r), b, y or g, the second syllable
of which also begins with a nasal, the first consonant of the stem is dropped
and the nasal, which is assimilated to the dropped consonant, is doubled:
e.g. n-bamba > mmamba I peg out: n-linda > nninda I wait for. In Runyan-
kore this law has no general application and 'I peg out' and 'I wait for' are
mbamba and ndinda.
There is, however, a small group of nouns in the RU-N class with stems
beginning with b which have plurals which conform to the Ganda Law except
that the nasal is not doubled; these are:
orubanja court case plural emanja
orubango spear shaft ,, emango
orubingo elephant grass ,, emingo
orubungo dung heap ,, emungo
orubambo peg ,, emambo
orubengo lower grind stone ,, emengo
orubanga hill side ,, emanga
Other nouns in this class with stems beginning with b the second syllable
of which begins with a nasal do not follow this law, e.g. orubindi (pot) plural
embindi. There is also the noun emandwa (a medium). This noun is clearly
derived from the verb okubandwa (to consult an oracle) and in its formation
the Ganda Law has been observed, but here there is no longer a conscious
process and if a Munyankore is asked to put this noun into a diminutive
class, he will say akamandwa, not akabandwa. In normal speech a diminu-
tive of emandwa would not be used, but since in constructing this rather
artificial noun he does not restore the b, although there is no longer a nasal
5 The Ganda Law which was first so named by Meinhof is also found in varying
degrees in other Bantu languages.

compound to bring the Ganda Law into operation, it is clear that the fact
that emandwa is derived from the stem -bandwa has been lost sight of. In
Runyoro there are no such vestiges of the Ganda Law when a nasal precedes
a stem beginning with b and in that language 'a medium' is embandwa. In
both Runyankore and Runyoro there is, however, one vestige of the operation
of the law (again without the doubling of the nasal) when a nasal precedes
a stem beginning with r. This is found when the adjectival stem -rungi (good)
is in agreement with a noun in the N-N class or the plural of the RU-N
class: e.g. n-rungi > nungi not ndungi.
Another linguistic law which, though not in general operation in Runyan-
kore, appears to have had some effect upon the language, is Dahl's Law.
This law, which has had no influence on Luganda, operates in many Bantu
languages and is found operating to its full extent in Runyarwanda and
Kikuyu. According to this law, if two succeeding syllables would otherwise
contain an unvoiced explosive consonant, the first of these consonants is
voiced and hence in Runyarwanda: e.g. iki-tabo > igitabo (a book); tata
>data (father). There are few vestiges of this in Runyankore and those which
exist are contained within the stems of words and not in prefixes. In Runyan-
kore, for example, the word for middle is ahagati, whereas the Ur-Bantu
stem is -kati, as it is in Luganda. Slight though the traces of the influence of
this law are in Runyankore, there is evidence that its operation is still a
conscious process and the Runyankore word for a coat, taken from the
English, is sometimes pronounced egooti instead of ekooti.
Among the Bakiga there is a tendency to insert an extra consonant in
consonant plus semi-vowel combinations. This is rarely heard in Runyan-
kore and is unknown in Kihinda speech. It is clearly a result of the influence
of Runyarwanda in which language there is a series of such consonant
changes which invariably take place, although they are not shown in the
spelling of the language; they are as follows:
bw ndw pw rw sw tw by ry sy ty are pronounced
bg ndgw pk rghw skw tkw bgy rgy sky tky
The degree to which the Bakiga are influenced by this phonetic rule varies
and the whole range of these consonant changes is never found.

Runyankore, like Luganda but unlike Runyoro, is a tonal language. That
is to say, to use the phraseology of the linguists, tone in Runyankore has
a semantic value and one word can be distinguished from another with a
different meaning merely by the difference in its tone:
e.g. Omugongo a ridge
OmugSngo a back
Tibarigura they will not buy
Tibarigura they do not buy (habitual present) it (RI-MA class).6
A knowledge of tone is, therefore, a help, though not essential, to a foreigner

6 High tone syllables are marked '; falling tone syllables are marked low tone
syllables are unmarked.

learning the language and, of course, it is only by its correct use that he can
hope to 'speak like a native'.
A European on first hearing Luganda spoken is immediately struck by the
cadence in a Muganda's speech and the early missionaries made frequent
reference to the 'musical quality' of Luganda. The wide range between high
and low tones in Luganda is not found in Runyankore, the tonal aspect of
which is far less noticeable. Here the range is narrow and, unlike Luganda,
a low tone is a far more common occurrence than a high one.7 Furthermore,
as has already been mentioned, there is a marked tendency to whisper vowels
and if, therefore, a whispered vowel follows a voiceless consonant, the syllable
can have no perceptible tonal quality and the tone it possesses is purely
notional. That it does possess a tonal value can, however, be discovered by
asking a Munyankore to whistle the word and observing that he will then
give the syllable the tone which the grammar requires.
I do not propose here to make more than the briefest reference to the
tonal grammar of Runyankore, a most complex subject, the study and under-
standing of which is in any case, as yet, far from complete. Nouns are
divisible into three classes. Nouns of the first class have all their syllables
on a low tone, e.g. Omukungu a chief.
Nouns of the second class have a high tone on the first syllable of the stem
if this syllable contains a short vowel, and a falling tone if it contains a long
vowel: e.g. Omukama a king; Omushdija a man; Omutambi a corpse.
Nouns of the third class consist
(a) of those which have a high tone on the second syllable of the stem:
e.g. Omushomisa a school teacher;
(b) of those which, having a long vowel in the first syllable, have a high
tone on this syllable: e.g. Ruhdnga the Creator.
Verbs are divisible into two classes, those which, in the infinitive, have all
their syllables on a low tone and those which, in the infinitive, have a high
or falling tone on the first syllable of the stem:
e.g. Okugura to buy Class I
Okuk6ra to do Class II
Okurdeba to see J
Each class has its own distinctive conjugation which differs tonally, not
merely according to the tense and whether it is affirmative or negative but
also according to whether or not there is an object prefix. On the other hand,
the phonetic nature of the prefix does not alter the tonal behaviour as it
does in Luganda:
e.g. Nddgura I bought (this morning) Class I
Nddk6ra I worked (this morning) Class II
Ndydkigura I shall buy it Class I
Ndydkik6ra I shall do it Class II
The tonal behaviour of words is quite different when they are in isolation or
the last word of a phrase (as in the examples given above) from when they
occur in a phrase with a word following. In the latter case a word will usually
7 For this reason high tone syllables are marked and low tone syllables are not;
in Luganda the opposite convention is used.

lose such high tones as it possesses in isolation and the initial vowel of the
following word will take a high tone instead:
e.g. Nddkigura I bought it ahondatho immediately
Nddkikdra I did it ahondaho
become in sequence
Naakigura dhondaho I bought it immediately
Naakikora dhondtho I did it immediately
This is another factor in the reduction of the number of rises in tone in a
Munyankore's speech.
There are variations in the tone employed by different speakers attribu-
table to some extent, but not entirely, to dialectic differences in various parts
of the district. There is a very marked tendency for certain speakers to ignore
high tones in certain verb forms8 and some will, for example, pronounce the
praise-name meaning 'he who gambols like a calf' Rukina and some Rukina.
It is possible that tone, too, is a survival of an older form of the Runyoro
language group and that Runyankore is in the process of following Runyoro
in dispensing with it as an essential part of the language.

Runyankore is a particularly interesting language to the Bantuist. On the
one hand, he has the strange idiosyncrasies of pronunciation which the Bahima
seem to have superimposed upon a Bantu language and which are without
parallel in neighboring languages. On the other hand, he has what appears
to be a comparatively primitive language, the vocabulary of which reflects,
to a greater extent than does, for example, Luganda, the stems of Ur-Bantu.
Furthermore, this vocabulary contains a wealth of words connected with the
keeping of cattle which again are without parallel in neighboring languages.
Much of this peculiar and vivid vocabulary is now in danger of extinction,
being known only to the illiterate Bahima. The reason for this is that the
early missionaries came mainly into contact with the Bairu who formed the
vast majority of the population and to whom this vocabulary was unknown
and it has, therefore, been their more limited vocabulary which has prevailed
in the schools and which is now alone known among the educated, even
those who are Bahima.
To the European learning the language, Runyankore is very difficult. Gram-
matically it is, like Runyoro, far more complicated than Luganda in its verbal
conjugation. Luganda, though it has its own problems resulting from its use
of double consonants and the sound change complications to which they give
rise, has what appears to be a more developed, and certainly a very much
simpler, verbal conjugation. For example, in Luganda all that is required to
form the negative of a tense is, with very few exceptions, to prefix si or te
to the affirmative form, whilst to obtain the relative form of a verb, the initial
vowel is prefixed to the simple form. In Runyankore, however, to obtain the

8 C. Taylor in the introduction to his Runyankore Dictionary (in the course of
publication) refers to this and calls such Banyankore "low tone speakers" remarking that
they are largely found in the towns. There does not, however, appear to be any
complete consistency in this respect in the speech of any one speaker.

negative and relative, completely different forms of most tenses have to be
e.g. They worked they did not work they who worked
Luganda: Baakola tebaakola abaakola
Runyankore: Bakakora tibarakozire abaakozire
The principal difficulty, however, which the European encounters is in
understanding the language when it is spoken on account of the elision and
devoicing of vowels. This makes words, which are familiar when written,
often unrecognizable when spoken. As has already been mentioned, this
peculiarity of Bahima speech is not found in Rukiga, in which each syllable
is fully pronounced, and a European wanting to learn the language would
be well advised to start with that dialect.



THE West Nile District of Uganda forms the north-west corner of the
Protectorate, bordering the Belgian Congo in the west and the Sudan in
the north (Fig. 1). The Albert Nile forms the eastern boundary for all the
District except for East Madi, with which the present paper is not concerned.

Fig. 1.
West Nile District.

The western boundary, shared with the Congo, is the Nile-Congo watershed.
The Nile valley here lies at an altitude of 1,800 to 2,000 feet above sea level,
and extends for 10 to 20 miles on either side of the river. On the western

/0 0 /0 20

side of the valley an ill-defined escarpment runs from north to south of the
District, rising from the level of the valley floor to about 5,000 feet at the
southern end of the District and to about 3,500 feet in the north. Beyond the
escarpment, apart from the extreme northern and southern ends, there is a
plateau rising gently to the watershed. Rainfall over much of the District is
40 to 60 inches per annum and is, for the tropics, well distributed, the only
marked dry season being from December to February. A break usually occurs
in the rains about June. Along the Nile valley the rainfall is rather less, and the
effect of this on the vegetation is accentuated by the higher temperatures which
prevail there. The rainfall distribution map (Fig. 2) was compiled by H. L.
Manning on the basis of rainfall probability, and observation of vegetation

Fig. 2.
Rainfall Distribution in West Nile District.

types was used to assist in marking boundaries where actual data were scanty.
Apart from rainfall very little meteorological data is available, but tempera-
tures in the Nile Valley probably seldom fall outside the range 70-105F.,
whilst on the plateau it is generally much cooler, night temperatures often
being as low as 50F. and maximum temperatures very seldom exceeding
90F. Some more general climatic data are given by Henderson (1949.)
The plateau and highland areas of the District are underlain by rocks of
the Ancient Basement Complex. The main plateau area in the centre of the
District is an old erosion surface, interlaced by a great number of tiny streams
which unite further from the watershed, and eventually reach the Nile as
six main rivers. On this part of the plateau there are frequent outcrops of a
grey granitic gneiss composed mostly of feldspar and quartz. In the south of

the District the soils are more commonly derived from various schists; actino-
lite, mica, kyanite and chlorite schists are all found (Harris, 1941). The more
rapid weathering of these softer rocks has given rise to the steep hills and
deeply cut river valleys which are typical of this area. In the extreme north
of the District there is again a change in topography. A massive quartzite
ridge, terminating in the peak of Otze dominates the landscape. Besides this
ridge which rises from an almost level plain, other rocky hills are present,
formed from later intrusions into the more general rocks of the Basement
Complex. These intruded rocks include granites, pegmatites and an augite
rock (Dawkins, 1944). The Nile Flats are deeply overlain by alluvial material.
The topographical variations in the area are accompanied by ecological
differences, so that the District may be conveniently divided for the purpose
of ecological description into the four areas of (1) the Nile Valley; (2) the
Central Plateau, where in addition to the factors already mentioned a high
concentration of population exerts an extremely important ecological influence;
(3) the Alur Hills at the southern end of the District; (4) the Madi Mountains
in the extreme north.
In the following account we shall not attempt to describe all types of soils
and vegetation occurring in the District, but only some typical associations of
the three highland areas mentioned above. In these areas the accounts given
refer primarily to those parts centred on Maraca and Omugo in the Central
Plateau, Paidha in the Alur Hills, and Metu in the Madi Mountains. Although
efforts were made to cover as much ground as possible from these centres
it was not possible in the limited time spent in the District to conduct any-
thing approaching a complete survey of soils and vegetation.
Previous accounts of the vegetation of the District have been confined to
travellers' reports, e.g. Emin Pasha (1888), and incidental interests, e.g.
Lumsden and Buxton (1951), who describe the tree and shrub vegetation in
some detail as a cover for insect vectors of yellow fever. No previous descrip-
tion of the soils of the District has appeared. They are referred to by Tothill
(1950) as belonging to the Bukalasa Catena. In fact only the soils of the Central
Plateau area bear any resemblance to this catenary association, and even here
the soils are very much the poor relations of the rich red loams and deep
black clays typical of the Bukalasa Catena in central Uganda. Griffith (1948)
mentions the soils of the West Nile Highlands in the stony phase of the
Buganda Catena. However a closer relationship exists between the soils of the
Central Plateau area and those of the adjoining areas of the Sudan, described
by Greene (1948), and the soils found in the Alur Hills are closely related
to the soils found in the immediately adjacent parts of the Belgian Congo,
described by Holowaychuk et al. (1954).

This large area, which includes the major part of the district, contains two
main types of vegetation, typified by the area of intense cultivation around
Arua, and the rather lower northern and eastern part of the plateau which is
less densely settled. Although these areas will be described separately it must

ItfroIO: J. f. M. Middleton
FIG. 3
Open-cultivated grassland on the Central Plateau area near Maraca, West Nile.

Irtroo: n. J. nlanceuor
FIG. 4
Riverine forest on the Alur Hills near Paidha, with typical hillside covered
with Hyparrhenia grasses and bushes in the background.

be remembered that the differentiation is only a convenience and that they
The intensely cultivated part of the plateau around Arua is divided into
small convex hills by many streams. The soils on these hills conform to a
catenary pattern, although only the central member of the catena is widespread.
This generally consists of clearly divided creep and sedentary horizons (using
Nye's terminology, 1954). The creep (Cr) horizon is formed of about 8 inches
to 1 foot of a dark, brownish-grey loamy sand, and the sedentary (S) horizon
of 2 to 3 feet of a uniform, compact red-clay-loam, which overlies fragmentary
quartzite. This soil is generally acid (c.pH 5-5). The subsoil often contains
orange and mauve mottles, which are sometimes found to have hardened to
form a pisolitic ironstone. Nowhere in this part of the plateau however are
there areas of indurated laterite.
The grey topsoil has a loose, powdery consistency and is very liable to
erode if exposed during the rainy season. Where this has happened the red
subsoil appears at the surface, and a moderately hard crust is formed, but
beneath this the subsoil remains soft. In one place gulleying was observed
to have followed the loss of topsoil. Such eroded areas support no vegetation
apart from an occasional thorn-bush. They are not very common and at the
moment not extensive.
The upper slope member of the catena is often absent. Where it does occur
it consists of a shallow layer of coarse, dark-grey sand, often mixed with
quartzite fragments and lying directly on the parent rock. The depth of soil
is seldom as much as one foot and over much of the area the tops of the slopes
have been completely denuded. The rocky outcrops so formed are a noticeable
feature of the landscape. It is likely that the gradual erosion of the higher
land has taken place by a process of soil creep, and that the grey sandy topsoil
of the midslope member of the catena is to be attributed primarily to this
The valley bottom soils generally consist of a layer of dark grey coarse
sand, one to two feet thick, which merges into a sticky, very plastic blue-grey
clay, used for pot-making, i.e. a kaolinitic and not a montmorillonitic clay.
These soils are slightly less acid and more fertile than the others of the catena.
They are however very limited in extent, seldom covering more than a few
yards on either side of the stream.
On all the soils in this area there is a very considerable response to fertiliza-
tion by kraal manure, ashes and household refuse, which gives the impression
that all the soils are low in fertility. This is confirmed by figures given by the
local Agricultural Officer, but it is very difficult to obtain an accurate estimate
of actual crop yields, as the normal practice is to sow in mixed stands. With
the abundant rainfall and short dry season it would be expected that adequate
fertilization would produce much higher yields.
The vegetation associated with these soils is very poor. All the land surface
is or has recently been cultivated (Fig. 3). There are virtually no trees except
those preserved for ritual reasons and the recent Eucalyptus plantations of
the Forest Department. On grazing and eroded areas Acacia bushes (A. hockii
De Wild and A. macrothyrsa Harms), sometimes associated with other species,

are found. The grasses of the uncultivated land rarely exceed 2 to 3 feet in
height, and the number of species is small. Setaria is by far the commonest
genus. It is usually associated with one or more of Hyparrhenia spp.,
Brachiaria brizantha (Hochst. ex A. Rich.) Stapf, Sporobolus pyramidalis
Beauv., Chloris gayana Kunth, Pennisetum polystachyon (L.) Schult., Era-
grostis spp. and others. Although common these species rarely approach the
frequency even locally of Setaria spp. In wetter situations Echinochloa
pyramidalis (Lam.) Hitchc. and Chase, although infrequent when compared
with other marshy plants, is very noticeable because of its large size. Imperata
cylindrica Beauv. and Brachiaria spp. are also common on wetter ground.
To the north and east of this densely settled area a more typical open
savannah woodland type of vegetation is found. Although the change in vege-
tation is clearly determined by the decrease in population density and not
by factors of climate or soil, there is also a difference in the soils of this
area compared with those of the densely settled area. A variant of the typical
catena is commonly found on the lower and more gently undulating parts
of the plateau, the variant being obviously related to Greene's Ironstone Catena
of the Southern Sudan, and the typical catena to his Red Loam Catena. The
upper slope soil of the Ironstone Catena is a dark to buff-grey loamy sand
overlying a mottled red-brown clay loam. Down the slope from this soil
an area of exposed or thinly-covered indurated ironstone is found. Where there
is a soil covering to the 'cuirasse' it usually consists of a few inches of pinkish-
grey sand containing much pisolitic ironstone. Below this soil on the slope
a heavier topsoil occurs, usually a grey-brown clay-loam, which merges into
a tough, compact and frequently mottled loamy clay. Drainage is often impeded
by the compactness of the subsoil, and the water-table is often near the
surface. The valley-bottom soils associated with the Ironstone Catena consist
of black loamy clays. They are rather more extensive than in the Arua area.
The underlying rock is the same grey granitic gneiss as at Arua. Hence it
is considered that the soil differences are due to differences in topography
and drainage conditions, although a slight change in parent material may have
resulted in the formation of the rather heavier soils which are typical of this
region. It is known (D'Hoore, 1954) that the formation of indurated lateritic
soils is closely related to the movements of soil water, and relatively indepen-
dent of soil composition. Thus it seems most likely that the formation of
sheet ironstone in this area is to be associated with the particular type of
slowly undulating topography.
The density of tree growth here varies very considerably, in places achieving
a patchily closed canopy and in others having an open parklike appearance.
The variation is probably related to previous cultivation, although open grass-
land may be somewhat commoner on the shallower soils of the Ironstone
Catena than on the other soils. The grass flora is similar to that in the more
densely populated area, with Setaria spp. as the most common dominants, but
Hyparrhenia spp. are more frequent generally and Sporobolus pyramidalis
Beauv. on the Ironstone soils. On poorly-drained soils of this catena Imperata
cylindrica Beauv. often forms pure stands.
In the more open grassland on the Ironstone Catena soils Borassus aethio-

pum Mart. is common, the other prominent trees being Ficus spp., and
Butyrospermum parkii Kotschy. Where the woodland is more dense Acacia
spp. are common, together with Combretum spp. and Terminalia spp. Tree
height seldom exceeds 20 to 30 feet.
In the more broken country around Mount Wati and on the escarpment deeper
more fertile soils are found, doubtless due to the more considerable additions
of soil material from higher ground. There is a corresponding change in the
vegetation. Pennisetum purpureum Schumach., occasionally appears among
the grasses, and the sizes of trees, shrubs and crops all indicate more vigorous

This is the highest part of West Nile, and also has the highest rainfall.
The general level of the area is c. 5,000 feet with hills rising to 5,600 feet. The
rocks are chiefly schists, with some gneisses and quartzites. The more highly
metamorphosed rocks are commoner near the escarpment. The river valleys
are mostly sharply cut and lined with 'gallery forest'. Not only is the landscape
more sharply sculpted than around Arua, but also the rocks are more deeply
weathered, so that on more level ground very deep soils are often found.
However the commonest soil type, which covers almost all the steeper hill-
sides, is a free-draining very shallow grey sandy loam, seldom more than 1 foot
thick, overlying fragmentary quartzite and decomposing rock. The loss of soil
from these hillsides has undoubtedly been accelerated by their cultivation,
with at best very inadequate anti-erosion measures. This soil type clearly
corresponds to the Niagaki Series described by Holowaychuk et al. (1954) in
the adjoining part of the Belgian Congo. The erosion of the hillsides has
contributed to the formation of deep, rich soils in the valley bottoms. These
are black loams, sometimes mixed with coarse sand, and overlying a grey
clay. The recent cutting of drainage ditches through these soils in settled areas
has enabled their stored fertility to be used, and they are now largely
employed for continuous maize or vegetable cultivation.
The deep soils generally found on the more level ground usually consist
of a brownish-red loamy horizon overlying a very uniform red clay loam,
sometimes more than 10 feet deep, usually with gritty quartzite fragments
in the lower portion and occasionally containing quartz veins. These soils
merge into soft decomposing rock. Drainage is generally free throughout the
profile. No mottling was observed, but mica flakes are occasionally found.
They are rather more fertile than other soils in West Nile, being less acid
(pH c. 6-5) and better supplied with nutrients. They correspond to the Zeu
Series of Holowaychuk et al. (1954).
The higher rainfall, lower population density and higher soil fertility all
contribute to the notable differences in the vegetation compared with the rest
of the District. Thus along the valley bottoms riverine or gallery forest is
found, characterized by tall trees, reaching 100 to 120 feet in height (Fig. 4).
The commonest of these are probably Syzygium spp. On the shallow hillside
soils which have not been too recently cultivated, Hyparrhenia spp. are domi-
nant, associated with Brachiaria brizantha (Hochst. ex A. Rich.) Stapf as

undergrass. These grasses usually occur with considerable small tree and
bush cover. Common species in this cover include Acacia hockii De Wild.,
Acacia eggelingii Bak. f., Entada abyssinica Steud. ex A. Rich., Erythrina
abyssinica Lam, and Ficus sp. (natalensis?). On more recently cultivated land
less bush and tree growth is found, and the grass Loudetia arundinacea Steud.
var. tricantha C. E. Hubbard ex Hutch. is often dominant. Other grasses found
on cultivated or recently cultivated land in this area include Andropogon
schirensis Hochst. ex A. Rich., Imperata cylindrica Beauv., Panicum maximum
Jacq., Pennisetum purpureum Schumach., Eragrostis aspera (Jacq.) Nees,
E. tenuifolia Hochst. ex Steud., Chloris pynothrix Trin., Digitaria velutina
(Forsk.) Beauv., Eleusine indica (L.) Gaertn. (said to be an indicator of poor
land) and less commonly Hackelochloa granularis 0. Kuntze and Melinis
tenuissima Stapf.. Setaria spp. are only rarely found, those identified being
S. longiseta Beauv., and S. kagerensis Mez which grows mainly in the shade of
trees. Liben and Sperry (1954) have described the savannah vegetation of the
adjoining area of the Congo, but although they also distinguish grasslands
dominated by Loudetia arundinacea and Hyparrhenia spp., they ascribe the
two grassland types to different soils. In agreement with the findings reported
here the Hyparrhenia spp. are reported to be dominant on the better soils.
The floristic composition of the grasslands given by Liben and Sperry differs
considerably from that found in the Alur Hills in West Nile. This grassland
flora also differs very considerably from that of the Central Plateau area. In
addition to the differences in climate and soil already mentioned it is probable
that the different grazing systems of the Alur and Lugbara are important in
determining the respective floras. Although the Lugbara keep relatively few
cattle, these are traditionally grazed near the homestead on the older fallows.
The present density of population means that most of the available grassland
is heavily grazed. The Alur on the other hand usually claim a greater ownership
of cattle than the Lugbara, but their cattle are not to be seen near the home-
steads as there are areas reserved for grazing purposes where the cattle are
herded and grazed together.

This area in the north-east corner of the District consists of a broad
plateau at c. 3,000 feet which is overshadowed by the massive quartzite ridge
of Otze-Urudri. To the west of the area there is a series of small granite hills
probably originating from a single intrusion into the biotite gneisses and
schists of the Basement Complex which form the underlying rock of the
plateau. The soils on the quartzite ridge are mostly brown or grey-brown sands.
These sands have crept or drifted to cover all the neighboring area to a
depth of 2 feet or more. Below the sand on the plateau there is the usual
sedentary horizon of compact red-brown clay-loam, in this region often
mottled and sometimes containing pisolitic ironstone. Indurated laterite was
seen in only one limited locality. Even along many of the streams the expected
dark clay soils are overlain by the brown sandy soil.
The granite hills to the west of the area are largely denuded but at their

foot and where soil does remain on the hills there is a grey sandy soil over-
lying a grey-brown loamy clay. The surface soil is generally mixed with rock
All the surface soils so far described are acid (pH 6-0-6-5) with the acidity
increasing with depth. As mentioned earlier however there are a number of
intrusive rocks besides the granites in this area. Of these a hill of basic augite
rock is prominent. This is now covered by only a very thin layer of soil mixed
with rocky fragments, but the remains of a rudimentary form of terracing
can be seen here, and it is evident that at one time the soils on the hill were
intensively cultivated, and were recognized as sufficiently superior to the other
soils of the area to merit an attempt at preservation by terracing.'
This region is much more afforested than others in the District. Cultivation
and grazing are at present mostly confined to the level areas of land between
the hills, and the hillsides carry a rich deciduous forest flora. Dawkins (1944)
has described this in some detail. The principal components of the tree flora,
which he describes as "very much an association" are:
Bridelia scleroneuroides Pax, Grewia mollis Juss.,
Butyrospermum parkii Kotschy, Khaya senegalensis (Desr.) A. Juss.,
Combretum binderanum Kotschy, Lannea kerstingii Engl. & Krause,
Combretum ghasalense Engl. & Lonchocarpus laxiflorus Guill. &
Diels, Perr.,
Ficus dicranostyla Mildbr., Oxytenanthera abyssinica (A. Rich.)
Ficus glumosa Del., Munro,
Ficus gnaphalocarpa (Miq.) A. Tamarindus indica L.,
Rich., Vitex doniana Sweet,
Ficus. sp.? Ziziphus mauritiana Lam.
In the valley bottoms narrow belts of high-forest are found. The tree flora
here is at present different from that of the hillsides, the characteristic species
being Khaya grandifoliola C.DC., Mitragyna stipulosa (DC.)O. Kuntze,
Phoenix reclinata Jacq., Syzygium guineense (Wild.) DC. and Pandanus sp.
Fallow and very recently cultivated land support associations of grasses
containing most of the species common in the Central Plateau area. There
are however some exceptions. Brachiaria comata (Hochst. ex A. Rich.) Stapf
is locally very common, apparently occurring very early in post-cultivation
series. It is often found co-dominant with Setaria sphacelata Stapf and C. E.
Hubbard ex M. B. Moss. Pennisetum polystachyon (L.) Schult. is also much
more common here than elsewhere. A species of Ctenium occurs in nearly
pure stands near the bases of the hills of the intruded granite, but was found
nowhere else in the District.
The climax vegetation of the West Nile District, apart from the Nile Valley,
must be presumed to be high forest, as the District enjoys a rainfall of
40 to 60 inches with only a short dry season. But this vegetation is now only
found as riverine or gallery forest in small areas of the Alur Hills and West
1 A similar type of rudimentary, indigenous, terracing is practised on soils derived
from the 'greenstones'-a metamorphosed basic intrusion-in the Northern Territories
of the Gold Coast. It does not occur in other soils of the area.

Madi. Elsewhere the vegetation has been modified very considerably by the
effects of cultivation, which have given rise to the Setaria dominated grass-
lands of the densely settled Central Plateau area. Grasslands are also found
in the less densely populated Alur Hills, but here the characteristic dominants
are Hyparrhenia spp. on the better soils and Loudetia arundinacea on the
poorer soils. These grasslands have probably also been formed under the
influence of cultivation. Their recolonization by woody species has been slow
because of the thinness of the hillside soils. The poor management of these
soils by the Alur has frequently produced considerable erosion (cf. Middleton
and Greenland, 1954), and this has doubtless contributed further to their slow
recolonization. Thus even where the population density is at present low,
the effects of previous settlement are evident in the vegetation, as Thomas
(1946) has emphasized for other parts of Uganda. Where no previous intensive
cultivation has occurred as in the northern part of the Central Plateau area,
where the typical vegetation is wooded savannah grassland, the effect of the
biotic factor is still evident in that the development of the vegetation is
restricted by annual burning. The biotic factor also influences the composition
of the flora, notably in the determination of the grassland dominant by different
grazing methods, and by the selective encouragement and introduction of
useful species.
It has been pointed out that the thin hillside soils exert an important effect
on the vegetation in delaying recolonization by woody species. The riverine
forests are also determined by the nature of the soil, as is the composition of
the flora of the hills of intruded basic rocks in Madi. A more intensive study
than this might have confirmed a difference in the floras associated with the
Red Loam and Ironstone soils of the Central Plateau, but on the whole the
effects of soil on vegetation are much less important than the effects of human
The most important effect of settlement on the soils of the region has been
to accelerate the erosion of soil from the steeper hillsides. The occasional
truncation of the midslope member of the Red Loam Catena must also be
attributed to the effects of cultivation, as well as the occasional exposure of
the indurated horizon of the Ironstone Catena. The formation of indurated
laterite in this area is however an entirely natural process. The laterites are
dense, relatively impermeable, indurated to a considerable depth and clearly
formed in situ. To use d'Hoore's classification (1954) they are laterites of
"absolute accumulation with negligible relative enrichment" i.e. they are
formed by the precipitation of sesquioxides after their mobilization in a soil
horizon nearer the surface or further up the slope. The upper horizons have
been lost at a later date. It is possible that a more intense dry season than
occurs in West Nile at the moment is required for the induration of the
lower soil horizons of sesquioxide accumulation. Such a climatic change is
by no means improbable, and while the savannah type vegetation is at present
biotically controlled, it may in the past have been climatically determined.

The field work on which this paper is based was carried out during a three-

month visit to the District in June to October 1953, made under the auspices
of the Oxford University Exploration Club. The expedition received financial
and other assistance from the Royal Geographical Society, various Trusts
and Colleges of the University of Oxford, the Colonial Office, Makerere
College, and the Government of Uganda, to whom grateful acknowledgement
is made. The expedition was led by Dr. J. F. M. Middleton. H. C. Dawkins,
Forest Ecologist, Uganda, and R. Milne-Redhead of The Herbarium, Kew,
rendered great assistance with the botanical work, and in particular arranged
for the identification of the plant collection. I. Langdale-Brown, Ecologist,
Uganda, gave very considerable assistance in the field. The authors are
indebted to the East African Herbarium, Nairobi, and the Agricultural Re-
search Station, Kawanda, for providing, collecting and sampling materials,
and to the Agricultural Research Station, Kawanda, and the East African
Agriculture and Forestry Research Organization for analyses of soil samples.
One author (D.J.G.) acknowledges permission to take part in this work during
tenure of an award from the Agricultural Research Council.

Dawkins, H. C. (1944). Unpublished report to the Chief Conservator of Forests,
Uganda, on the West Madi Sub-district of West Nile.
Emin Pasha (1888). Emin Pasha in Central Africa. George Philip and Son,
Greene, H. (1948). Soils of the Anglo-Egyptian Sudan, in Agriculture in the
Sudan, Ed. by J. D. Tothill. O.U.P., London.
Griffith, G. ap. (1948). Provisional Account of the Soils of Uganda. Comm. Bureau
of Soil Sci. Tech. Comm. No. 46, p. 16.
Harris, N. (1941). Ann. Rept. Geol. Survey Dept. Uganda, p. 17.
Henderson, J. P. (1949). Some Aspects of Climate in Uganda. Uganda J., 13, 154.
Holowaychuk, N., Denisoff, I., Gilson, P. and Croegaert, J. 1954. Publications de
I'I.N.E.A.C. Carte des Sols et de la Vigetation du Congo Belge et Ruanda-
Urundi. 4, Nioka.
D'Hoore, J. (1954). L'Accumulation des Sesquioxide libres dans les Sols Tropicaux.
Publications de l'I.N.E.A.C. S6rie Scientifique No. 62.
Liben, L. and Sperry, T. (1954). Publications de 1'I.N.E.A.C. Carte des Sols et de
la V6g6tation du Congo Belge et du Ruanda-Urundi. 4, Nioka.
Lumsden, W. H. R. and Buxton, A. P. (1951). A study of the Epidemiology of
Yellow Fever in West Nile District, Uganda. Trans. Royal Soc. Trop. Med.
and Hygiene, 45, 53.
Middleton, J. F. M. and Greenland, D. J. (1954). Land and Population in West
Nile District, Uganda. Geog. J., 120, 446.
Nye, P. H. (1954). Some Soil-forming Processes in the Humid Tropics I. J. Soil
Sci., 5, 7.
Thomas, A. S. (1946). Illustrations of Human Influence in Tropical Ecology.
J. Ecol., 33, 46 and 153.
Tothill, J. D. (1940). Agriculture in Uganda. O.U.P., London.

UST sixty years ago, at sundown on 19 October 1897, three Europeans,
held by the Sudanese mutineers in Luba's Fort at Bukaleba within sight
of what is now Jinja, were murdered. The circumstances are related in J. V.
Wild's The Uganda Mutiny 1897, published in 1954 in the 'Treasury of East
African History' series. A memorial cairn with three inscribed marble tablets
was shortly afterwards erected near the spot by relatives and friends of the
Since 1893 Luba's Fort, besides being the administrative centre for Busoga,
had become a busy staging post on the main caravan road from the Coast
and Kavirondo through southern Busoga to Buganda. For here was the
landing place for the canoes by which men and goods were ferried to and
from the Kyagwe shore.
After the mutiny, Luba's Fort was abandoned and a new Fort Thruston
was constructed hard-by. But in 1900 district headquarters were moved to
Iganga and in the following year to Jinja, and thereafter caravans tended to
cross the Nile at Jinja. A few years later the neighbourhood was depopulated
by sleeping sickness and became a closed area. Bukaleba was henceforward
far removed from any beaten track. Thus this cairn has been seen by very
few people and it will be fitting to put on record the inscriptions on the
three tablets. The accompanying photograph taken in 1901 by Mr. T. J.
Spooner, District Engineer, of the Uganda Railway Construction staff' shows
the cairn as it was soon after its erection.
A recent inspection shows that the memorial remains today in generally
good condition, though the area is still unpeopled, and is 'alive' with game-
buffalo, elephant and hippopotamus. In a recess below Norman Wilson's
tablet a number of cent pieces were noticed, placed there, it is understood,
by Africans who, from time to time pass the spot. This may be a parallel
to the urge-of piety or propitiation-to which are probably due the Cairns
of Koki, which are described at U.J., 21 (1957), 176.
Major Thruston is also commemorated in London, in the crypt of St. Paul's
Cathedral, by a white marble mural tablet framed by red marble columns.
This was erected by the officers of his old regiment, the Oxfordshire Light
Infantry. The inscription is printed at page 34 of Thruston's posthumously
published African Incidents, 1900. The numbers 43 and 52 within bugle
badges which are displayed at the foot of the tablet refer to the amalgamation
of the 43rd Foot (Oxfordshire and Buckinghamshire Light Infantry) with
the 52nd Regiment (Oxfordshire Light Infantry).

1 There is an account of some of his lion-hunting adventures in Lieut.-Col. J. H.
Patterson's The Man-eaters of Tsavo. The lettering of the tablets is somewhat inexpertly
executed. Perhaps they were engraved, not in England, but by a fundi in East Africa.

Fig. 1.
The Mutiny Memorial at Bukaleba in 1901.

[Photo : T. J. Spooner


On east side:
(Line of Arabic writing which translates)

On south side:


On the north side:




IN two articles in The Uganda Journal. The Year of the Three Kings of
Buganda (U.J., 14 (1950), 28-9), and Mackay's Canoe Voyage on Lake
Victoria in 1883 (U.J., 18 (1954), 19 note 3), I sought to equate "le petit
boutre de Songoura" (Nicq, Le Pare Simdon Lourdel, 480), which sank in
flames near Entebbe on 2 September 1889 with Songoro's dhow which had been
purchased in 1877 by Shergold Smith, the leader of the first C.M.S. expedition
to Lake Victoria; and I failed to distinguish the respective owners of these
craft. A re-examination of contemporary authorities indicates that there were
two distinct vessels with unrelated owners bearing the same name, or rather
nickname ("The Rabbit"), though variously spelt.
When Stanley reached Kagehyi (Kagei) at the south of the Lake on 27
February 1875 he was well received by "Sungoro Tarib, an Arab resident".
He had been for "a long time constructing a dhow" (Stanley, Through the
Dark Continent). There is no record of any such earlier vessel on Lake
Victoria. Songoro was not perhaps a full Arab, for the Rev. C. T. Wilson
who saw him early in 1877 speaks of him as "a negro trader" (Wilson and
Felkin, Uganda and the Egyptian Soudan, i. 85), and he had apparently been
trading round the Lake for some years. In my account of "Ahmed bin
Ibrahim" (U.J., 11 (1947), 85, note 7), I pointed out that neither Speke nor
Grant makes any mention of Arabs in Buganda at the time of their visit in
1862. It is interesting therefore to find in Shergold Smith's diary for 23
November 1877 that Songoro "had warmly eulogized Speke, who, he said
had bought cloth and beads of him in Uganda, and, when he presented the
paper at Zanzibar, it was immediately cashed" (Church Missionary Intelli-
gencer, July 1878, 424).
There would appear to be some corroboration of this in a remark by Wilson
that Songoro "had been to Uganda several times" (Wilson and Felkin,
ibid., i. 97). Further circumstantial evidence as to the truth of Songoro's
assertion is forthcoming from a file entitled "Captain Speke's Expedition
1863-4, vol. 3" which is now in the Peace Memorial Museum, Zanzibar (see
U.J., 17 (1953), 146). In that file is a paper, which is signed by Speke and dated
31 May 1863. It is headed "My account with Ladha Customs Master Zanzi-
bar by orders sent in I.O.U. from Central Africa." It gives copies of the
I.O.U.s which extend from 1 October 1860 to 29 May 1863. No particulars
are given in them beyond the dates and amounts of each debt, but one dated
6 July 1862, for fifty dollars almost certainly must have been given in favour
of Songoro for his cloth and beads whilst Speke was at Mutesa's capital
(Speke, Journal of the Discovery of the Sources of the Nile, 449). Another
dated 10 October 1862, for 210 dollars was evidently issued to Msalima when
Speke was in Bunyoro. ibidd., 548). These are the only I.O.U.s which according
to their dates, were issued by Speke during the period in which he was in
Buganda and Bunyoro.
O'Neill, one of the original C.M.S. party, noted immediately on their arrival

at Kagei on 29 January 1877, that "the king of this place, in conjunction with
Songoro, the slave trader is building a dhow on the island of Ukerewe. This
has been in progress for the last three years, and is not yet finished: this is the
same vessel referred to by Stanley" (C.M.I. Aug. 1877, 467).
At length in June 1877 Songoro agreed to sell the still unfinished dhow to
Shergold Smith for 100. O'Neill was left to complete it, while Smith and
Wilson set off to visit Mutesa, in the Daisy, a small cedar boat which had
been brought up in sections from the coast and assembled at Kagei.
After a month at Rubaga, Smith rejoined O'Neill on Ukerewe in August.
At length, on 14 November, the dhow-now named Chimosi, being bad
Swahili for 'The First' and also embodying the initials C.M.S. (C.M.I. July
1878, 422)-was successfully launched. O'Neill, who had also built a dinghy
dubbed the O'Neill, was on the best of terms with Lukonge, the "young
handsome-looking" but avaricious chief of Ukerewe.
At this juncture the latter interposed with a claim for payment for timber
which Songoro had used in the original construction of the dhow. Negotia-
tions followed, and on an undertaking by Songoro to pay three frasilas of
ivory against which he was to deposit pledges, Lukonge, with good grace
towards Smith and O'Neill, allowed the Chimosi to sail on 22 November.
Only two days later when lying off the pier at Kageyi she was driven on to the
rocks and became a total wreck. Her contents were salved, and some of her
timbers used to fence the mission burial ground.
Here, parenthetically, one may pause to remark that this pier built of stone
by Smith in the previous June (C.M.I. Dec. 1877, 754) must have been the
very first jetty to be constructed on Lake Victoria, and thus the first sign of
a revolution in the use of its waterways as a means of communication.
The record regarding the course of events which ended some three weeks
later with the death of Shergold Smith, O'Neill and Songoro on probably
13 December 1877-the date is discussed in an addendum to this note-has
been confused by the accidents of erratic communications; only gradually
was the story pieced together. Briefly on 5 December the Daisy with the
O'Neill in tow left Kagei for Uganda But they soon had to return on account
of the weather and the inability of the Daisy to tow the dinghy. On the
8th Songoro came to see Smith complaining that Lukonge was threatening
his establishment on Ukerewe. Smith agreed that provided Songoro would
faithfully redeem his pledges he would go to Ukerewe to see fair play.
Probably on the 12th Smith and O'Neill went in the Daisy with a small
party of six or seven men to Songoro's establishment at Bukindo north of the
Rugezi Passage. It seems that the ivory was paid to Lukonge and pledges
restored. But Lukonge was in a truculent mood; that this had been provoked
by Songoro's own shiftiness seems more than likely. Songoro prevailed upon
Smith to assist in the removal of his women and children who were sent
across the island to Nasso, a landing place on its south-western shore, while
the Daisy, under her native captain Hassani, went round to Nasso to
await orders.
The Daisy and the women arrived at Nasso at much the same time.
Hassani, without awaiting orders, shipped the women and children. Smith,

O'Neill and Songoro following, reached Nasso to find the Daisy in the offing
but refusing to return. Lukonge's men closed in; Smith and O'Neill decided
that they could not stand aside, and when their ammunition was exhausted all
were massacred, except two of Songoro's men who escaped and one C.M.S.
man whose life was spared. Hassani despatched two survivors from among
Songoro's men to Unyanyembe (Tabora) with messages for Mackay, and
himself sailed to Uganda to inform Wilson. It was some time before the facts
of Hassani's dereliction were realized, whereupon he was dismissed.
So ends the story of Songoro the first and his dhow.

Wilson, who was alone at Rubaga awaiting news of his fellow missionaries
from the south of the Lake states that, towards the end of December 1877
(when Songoro the first was already dead) "I was overtaken by a negro
trader Sungura, who lived in Uganda" and who told him that Emin had
arrived at Mutesa's court (Wilson and Felkin, op. cit., i. 112).
This Sungura continued in Buganda for some time. Emin encountered him
on 7 January and 19 March 1878 (Stuhlmann, Die Tagebiicher von Emin
Pascha, i. 374, 469) and Dr. Felkin's diary of 6 April 1879 refers to "Songoro,
a coast man" (C.M.I. Dec. 1879, 720). The entry in Emin's diary for 7 January
1878, translates as follows: "Sungura is much travelled and knows the land
from Zanzibar to Tanganyika (he was with Burton in Kirira) and from
Zanzibar to here quite well and I hope to get some information from him."
Kirira lies slightly off the direct route from Tabora to Ujiji. Burton and
Speke arrived there in December 1857, and it was there that they met Ahmed
bin Ibrahim el Amuri, who claimed to have reached Buganda in 1844 (U.J., 11
(1947), 83) Emin's diary unfortunately tells us no more about the past history
of Sungura the second. Burton does not mention Sungura in his Lake Regions
of Central Africa, i. 392-4-very probably because his hosts were Ahmed
bin Ibrahim and another full-blooded Arab while Sungura was, as Wilson
states, "a negro trader".
The next reference comes from A l'assaut des pays negres, the journal of
the White Fathers' first expedition to Uganda, where (p. 256) is an entry dated
at Kagei 10 March 1879. Pere Lourdel and Frere Amans had set off for
Uganda on 20 January, and the rest of the party was awaiting news from them.
"Towards mid-day the lake was covered with canoes from Uganda .
[thinking that these] had been sent to us by Pere Lourdel we ran to the shore.
We were bitterly deceived. The chief of the fleet told us that these craft had
been sent by Mr. Mackay to meet his colleagues [He said he had met our
brethren] on the lake close to Uganda Bad weather had prolonged their
voyage and his; it was already twenty days since he had set sail.
"This commodore is called Songoura and is a native of Zanzibar. He
left the coast some twenty years ago to take service under Mtesa. He told
us everything that was good about this king and his kingdom. We offered
him a trifling present of beads; but he politely thanked us, saying his illustrious
master had provided generously for all his needs .. Songoura was in a hurry
to finish his business with Kaduma [the chief of Kagei] and in company

with the English take his way back as soon as possible. The reverend
ministers [sc. the C.M.S. missionaries A. J. Copplestone and Charles Stokes]
came to say farewell and were, as always, most pleasant."
Mackay notes on 2 November 1879 that "Said-ben-Saif is now building a
dhow on the coast of Uzongora-just on l1S. lat." (C.M.I. July 1880, 411),
and later on 9 July 1880, that "Said bin Saif's dhow is almost ready for
launching He allowed that he meant to trade with the coast of Bugaya,
where Songoro [sc. Songoro the first] used to get slaves so cheaply Another
vessel, but rather smaller, is now being built at Kagei, by a coast man named
Sungura. This will also be a slaver." (C.M.I. Nov. 1880, 679). It is reasonable
to conclude that this latter is the man Wilson met in Rubaga in December 1877.
On 1 July 1883 Mackay, who was on his way by canoe to the south of the
Lake, states that, at a small island between Sese and Bujaju (Buddu), he came
upon "Sungura's boat at anchor, en route for Usukuma [with] a cargo of
ivory and slaves in stocks" (U.J., 18 (1954), 14).
Nothing more emerges from the record until Mackay writing from
Usambiro on 6 June 1889 remarks that "in the neighboring district of
Mwanza there has settled a mischievous coastman, named Sungura, whom I
know well, and who has already threatened to kill the white men at Nassa
[the near-by C.M.S. station] in revenge for his native town of Whindi (near
Saadani) having been bombarded by the Germans" (C.M.I. Jan. 1890, 24).
Three months later, 2 September 1889, the dhows of Said bin Saif and
Sungura came to their richly merited end off Entebbe as a result of well-timed
operations led by Gabulieri Kintu, one of Mwanga's chiefs. These are described
in detail in U.J., 14 (1950), 38-40; and from Nicq's Life of Lourdel (op. cit.), it
seems that Songoro himself was present and was killed.
The second Songoro and his dhow may thus also be written off.

An interesting comment upon early Arab contacts with Uganda appears
in a letter from Shergold Smith dated 27 August 1877. "One of Songoro's men
here [Ukerewe] has traversed the Masai country twice. Eight years ago he
did the distance from the borders of the Waruri's country [roughly around
the present port of Musoma] to Tanga in twenty-four days, but says it has
been done in fifteen. The chief difficulty is the hatred of the Masai to any
stranger. Since his time, travelling became more hazardous and had to be
abandoned. Hamad bin Ibrahim, [see U.J., 11 (1947), 80-97], to go back
twenty-five years, tells me that when he first went to Uganda, an Arab caravan,
numerous and well armed, tried to force a passage through the Masai country.
They did so but arrived in Uganda with a loss of six Arabs killed and 300
men killed, captured, or rendered unable to proceed." (C.M.I. Mar. 1878, 159).
Disregarding a few years discrepancy in date this last might well be the
Arab caravan which "came across the Masai as far as Usoga" shortly after
the death of Suna, of which Speke learned from Mutesa's ambassador in
November 1861. (Speke, op. cit., 187.)

Conflicting rumours of the disaster on Ukerewe Island in December
1877 began to circulate in Zanzibar early in March 1878, and Smith,

Mackenzie and Co., as agents for the C.M.S. telegraphed the bare report
to London on 19 March, which was published in C.M.I. April 1878, 251.
Meanwhile Dr. Kirk, the Consul-General at Zanzibar, received a coherent
and creditably accurate account of the affair from Mr. Morton, an Englishman
who had left the coast in the previous March to take a caravan of supplies
up-country for the C.M.S. (He committed suicide later in 1878), and this
account Dr. Kirk transmitted to Lord Derby at the Foreign Office in a despatch
dated 6 March 1878 (a copy is in Zanzibar archives). Morton writing from
Unyanyembe on 16 January 1878, whence he was about to return to the
coast with a caravan of ivory from Mirambo, had his account from the two
survivors sent off to Unyanyembe by Hassani and he dated the affair about
7 December. He gave details of the final attack which was said to have lasted
from morning until afternoon when the defenders' ammunition ran out. Not
until next morning did the Daisy return, when three survivors who had
concealed themselves swam off and got on board. The bodies of Smith,
O'Neill and Songoro were seen naked and mutilated on the shore but could
not be recovered.
This despatch reached London on 1 April and is summarized in C.M.I.
May 1878, 313-4; and this is substantially the only account that was current
for some months.
As soon as the news reached C. T. Wilson in Uganda he set off for the
south of the Lake reaching Kagei on 12 January 1878. Here he made all
possible enquiries, and was thus the earliest and nearest investigator. The
report he then made reached London on 27 April (summarized in C.M.I.
June 1878, 380-1) and did not greatly modify or amplify Morton's under-
standing of the affair. At this date the last communication to hand from
Shergold Smith was dated 5 December (C.M.I. July 1878, 427).
But Wilson continued his enquiries as later opportunities offered, in the
course of which he learned of Hassani's desertion. Later a dispatch box of
Smith came to him in which he found an uncompleted letter from Smith dated
10/11 December 1877 which was published in C.M.I. November 1878, 701-4.
Only then was it recognized that the date of the disaster was probably 13 or
14 December 1877. But the earlier date had already gained wide currency
(e.g. in C.M.S. Gleaner July 1878, 73-4).
Wilson's considered conclusions are given in his book, Wilson and Felkin
op. cit. i, 120-5. This and C.M.I. for November 1878 should therefore be
referred to for the most authoritative statement of the whole affair.
In one particular only may Wilson's emphasis be questioned. He put the
major part of the blame on Lukonge. In July 1878, Mackay was waiting at
Kagei to cross to Uganda. In response to an invitation, alone and unarmed,
he visited Lukonge, and was satisfied that the latter, as he protested, had no
intention of killing Smith and O'Neill who died "only because they got mixed
up with a bad Arab-Songoro". (C.M.I. Feb. 1879, 88-91.)
Evidence to much the same effect comes from another direction. As has
been noted above, early in 1879 a party of White Fathers was waiting at
Kagei. The party's diarist, who is almost certainly its able and open-minded
leader Pere Livinhac, took the opportunity to enquire into the Ukerewe

disaster, and gives his account in A l'assaut des pays negres, 248-9. He was
satisfied that Lukonge's attack was precipitated by the bad faith of Songoro.1
The islanders who fell upon Songoro and his men gave Smith to understand
that they intended nothing against him (Smith), and that he had only to
stand aside. Smith replied that he could not abandon Songoro, the attack
then became general and only when their ammunition ran out were they
inevitably massacred.

In conclusion I wish to express my indebtedness to Mr. H. B. Thomas for
drawing my attention to the fact that there were two Songoro dhow-builders,
as also to the extracts from the Church Missionary Intelligencer which go
to prove this fact.
1 An extract from Shergold Smith's last letter provides an illuminating comment on
Songoro's conduct and helps to explain Lukonge's attitude.
"Two days ago, Songoro came to say that Lukonge was threatening his establishment
in Ukerewe I promised that, if he would faithfully redeem his pledge, I would
go over and settle it with him today [10th Dec.] he brought, as he said, the ivory;
but on weighing it, it was found to be only 70 Jb. [i.e. two frasilas instead of the
promised three]; when asked for more, he observed, 'Oh, those Washensi won't know
the difference'. Compelled to bring more, he reluctantly added another tusk, and made
up the amount."


F you had visited the home of a well-to-do Munyoro about forty years ago,
you would have been received in a grass hut with the women and children
peeping at you through the reed partition. If the family knew you well, the
children, their mothers and their father, if he happened to be at home, would
all rush out to greet you. If they did not know you well, they would see if
they were tidy and presentable before coming to welcome you. In addressing
you, they would use your 'pet name'. In Bunyoro the young greet the old
and the inferior the superior.
There are eleven pet names3 (empako) one of which is given to children
of either sex at the time they receive their other names. These names are of
Lwo origin and each empako is supposed to indicate something about its
holder's personality. Nothing is known of their origin, although they were
presumably introduced by the Babito, who were related to the Lwo-speaking
peoples. Empako are used in greetings and at times when in English terms of
endearment would be used. These are the names:
Aboki Apuli
Abwoli Arali
Acali Atenyi
2 Part of a lecture entitled 'The Peoples of Bunyoro' delivered at a joint meeting
of the Uganda Society and the Extra-Mural Department of Makerere College.
3 Beattie (Uganda J., 21 (1957), 99) refers to 12 empako which he translates as
'praise names'. He may, however, have included the name, Okali, among the twelve.
Okali is not an empako in the ordinary sense. It is used only for the Omukama and
with one greeting, Zona Okali, waitu.

Adyeti Atwoti
Akiiki Bala

When people who are closely related greet each other the younger sits on
the elder's lap and, among the Babito, the one who sits on the other's lap
touches his friend's forehead and chin with the tips of the fingers of his
right hand.
After greetings, coffee berries, specially kept for guests, would have been
offered to you in a small basket made of raffia and soft papyrus. Then would
come a pipe or tobacco for chewing and the neighbours who would have
been told about your visit, would perhaps call in to see you.
Meanwhile the women would be preparing a meal for you. A guest must
be given some kind of meal, no matter at what time he arrives-even if it
is in the middle of the night. If you were an honoured guest, you would be
given a cow or a goat, but if you were a close friend of the family you would
be given a chicken, because you would not be likely to take offence at such a
small present. The present would be brought to you and your host would
ask you to accept it, but you would beg your host to kill it. It would be a
very strange guest indeed who asked to take such a present home. Your meal
would consist of millet or bananas and meat. The millet would be served in
a small basket (endiiro) and the meat in a wooden bowl (orucuba). Potatoes
were never given to a guest except in times of scarcity.
If your host lived near the capital and he thought you were a presentable
guest he might take you to the Omukama's palace. The Omukama, when he
is in residence, must by custom sit in an advertised place for certain specified
hours so that any of his subjects can go and see him (okukurata). The guest,
if he knew Runyoro, would immediately notice that his host used different
words when addressing the Omukama. There are more than twenty different
ways of greeting the Omukama and appropriate greetings can be said to him
at appropriate times by any man. The Omukama is not expected to answer
these greetings verbally; indeed, he never does. Women do not take part in
this type of greeting (okuramya.). They merely kneel down and greet the
Omukama in the normal way and he is expected to answer them. When
addressing the Omukama the third person singular must be used. Nearly all
verbs and some nouns used in addressing the Omukama are different from
those used for commoners.
If this visit to the Omukama coincided with the new moon ceremonies, the
guest would have found hundreds of people assembled in the kikali, dancing
to music provided by the royal bandsmen. The people would have been
rejoicing because the King had outlived the moon. The royal band consisted
of about twenty men who served in relays playing drums, flutes and other
wind instruments. These festivities used to go on for only a few days in the
kikali. But an annual celebration used to go on for nine days-seven days in
the Omukama's palace, one day in King's mother's enclosure, and the ninth
day in the enclosure of the chief medicine-man. It was held in the dry season
(ekyanda), between the months of December and February. Today empango

is celebrated on the Omukama's accession celebrations on 12 April once in
three years, and only for two or three days.
Our guest would end his day sitting by the fire in the evening listening
perhaps to stories of the brave deeds of Banyoro warriors who used to raid
Buganda and the neighboring tribes and bring devastation in their train.
Or perhaps they would talk of the different activities carried on in Bunyoro
throughout the year. The year was divided up according to the activities of
the people, and these divisions come very near to a 12-month calendar.









or Katuruko

... January

... February
... March

... April

... May

... June

... July

... August

... September
... October
... November

... December

... the month for harvesting
... the month of white ants.
... the month of the rain which
comes after the dry season.
... the month for planting
... when the white ants (en-
seizere) appear.
... when the edible grasshoppers
(enkozikonzi) arrive.
... the month when women start
to prepare fields for plant-
ing millet.
... the month for burning grass
in the millet fields.
... the month for planting millet.
... the month for weeding millet.
... the month of edible grass-

... the month for rejoicing and
festivities because there is
little work to be done in
the fields.

Or perhaps our guest would be told the story of how Ruhanga the creator
gave his children their names and the work they and their descendants were
to do after them.
These evening discussions round the glowing fire were an important part
of the day's activities. The whole family would gather together after the day's
work to discuss all kinds of subjects and to tell stories for instruction and
amusement. Since there was no formal education this was the opportunity for
passing on information of historical interest to the next generation.



Mr. Morris in his interesting article, The Making of Ankole, (U.J., 21 (1957),
1-15) refers to Macallister and R. R. Racey as men of energy and determina-
tion, and later mentions the Mukama of Igara, Musinga, who committed
suicide by disembowelling himself in 1901. An interesting sidelight on this
statement is a letter written to H.E. the Governor of Uganda by R. R. Racey
from Winnipeg, Manitoba, on 12 August 1930, describing the suicide in more
detail and mentioning also the attempt on his own life at the same time. I
quote it in full.

May I be permitted the honour of placing the enclosed instrument in your
hands for disposal.
The explanation is the following given briefly. By Command of the late
Sir Harry Hamilton Johnston, when Her Majesty's Special Commissioner
and Commander in Chief for Uganda, given me on August 6 1900, at
Mbarara, Collectorate of Ankole, invitation was extended to Chief Msinga
of Egara to visit at Mbarara. This he promised to do at some favourable
opportunity. Our objective was to ultimately gather the lesser chiefs of the
collectorate in one baraza under Kahia (later King Solomon Edward, I
believe). Some time afterwards I was engaged in Egara in connection with the
settlement of a boundary dispute. This was on 3 January 1901. Chief
Mbaguta (Kahia's representative) and Chief Msinga accompanied me. In the
course of the work of demarcation my safari was to have passed a certain
point. The following morning Chief Mbaguta approached who informed me
an 'ambush' was lying in wait with the evident intention of trying to spear
me, whereupon I caused my men to deflect and saw what apparently was the
ambush party come out of hiding. I closed my work and invited Chief Msinga
to come with me to Mbarara suggesting it would be a favourable opportunity
to pay the promised visit. This he agreed to do. He was carried by bearers
in a large basket. We began the journey towards Mbarara and were passing
through a papyrus swamp, Msinga being ahead, when a cry was raised to
the effect Msinga had killed himself. Immediately I ran forward and by so
doing escaped a spear-thrust from Msinga's sub-chief who had come back and
who speared an unfortunate nafar [soldier] belonging to my escort who had
lingered behind contrary to orders. On coming up to Msinga he was found to be
disembowelled and the knife which is returned herewith was lying alongside
of his body in the basket. This knife I retained. It passed from my hands
and was subsequently returned to me. Owing to the gruesomeness of the
story attached to it and the unpleasantness of the memories connected with
it, it occurred to me to ask you if you will graciously cause it to be disposed
of, wherever it may seem best to place it. I have the honour, etc."

The letter and the knife were sent to the then District Commissioner of
Ankole, G. L. Maitland Warne, and there is a note on the covering letter
from the Chief Secretary that the knife was handed on to W. F. Poulton,
Director of Veterinary Services. It seems a pity that it could not have been
preserved in the Uganda Museum as a kind of modern deodand. The corre-
spondence is contained in the District records at Mbarara.
18 July 1957.

EDITORIAL NOTE: A letter received from Mr. F. Lukyn Williams confirms
that the knife was, at the instance of P. W. Perryman, the Chief Secretary,
sent to Mbarara and later passed into the possession of W. F. Poulton. He
died on 15 March 1941, and it has not been possible to ascertain what has
since become of it. At the time the Kampala Museum was in a moribund
condition and regarded as no fit repository for such a relic. It is recorded in
U.J., 2 (1934-5) that after 1920 practically no new specimens had been added
to its collections, for the reason that there was no space left in which to put
Another point of interest arises from Mr. Racey's letter. In his recent
life of Sir Harry Johnston, Dr. Oliver remarks that there is virtually no
direct record of Johnston's visit to Mbarara. "It is a matter about which less
is known than any other in Johnston's African life." (p. 323.) Here is first-
hand evidence of that visit.


Two points in Mr. Hoyle's interesting article in vol. 21 (1957) call for
p. 93. The fire at the Uganda Company Stores building occurred not in
1921 but in 1925.
p. 97. Sir Winston Churchill was Under Secretary of State for the Colonies,
not for Foreign Affairs. J. SYKES.
Crowthorne, Berks.
19 August 1957.


English is a living language, adapting itself to expanding needs. Twenty
odd years ago in the handbook Uganda (1935), Robert Scott and I, with
avant-garde daring, ventured to refer to Kampala and its environs collectively
as a conurbation. The Oxford University Press proof-readers questioned the
word which did not appear in any Oxford dictionary. We stood firm, for it
expressed what we had in mind; but the Press registered its disapproval by
ignoring the word's fairly obvious component parts, breaking it at the end
of a line into conur-bation (p. 426).
A distinguished reviewer in U.J., 3 (1935-6), 325, considered that "for the

use of such a word ... there is no excuse"; and later, in a letter to the Journal,.
a Makerere professor administered a magisterial reproof for this jeu d'esprit.
(U.J., 4 (1936-7), 98-9.)
During the 1940s the word was found to meet a definite need, as applying
to "a cluster of contiguous towns forming (for some purposes) one great
community". In 1948 it graduated as the title of a major report emanating
from the Ministry of Town and Country Planning. And when an article in
The Times (10 January 1949) recognized its formal adoption into the English
language, Stalky and Co's immortal but not very adult exclamation, "I gloat!"
came, somewhat improperly, to mind. The word is now common currency and
may be encountered in respectable company on any day in the week.

Referring to the murder of Mr. H. St. G. Galt in 'The History of
Western Uganda' (Uganda J., 21 (1957), 141), Dr. K. Ingham says "This
evil deed was the work of certain chiefs who superstitiously attributed the
difficulties which Ankole was undergoing to British intervention. The chiefs
were banished and a communal fine was paid by the people of Ankole."
In actual fact the two chiefs were convicted in the High Court of Uganda
of abetment of Mr. Galt's murder by one Lutaraka, who subsequently com-
mitted suicide. That conviction was set aside by the Eastern Africa Court
of Appeal on the ground that the evidence against the accused was "made
up of contradictions and inconsistencies incapable of reconciliation". That
acquittal must in the circumstances stand for all time. Not only so, but
having examined the record of the proceedings carefully and critically, I
may perhaps after thirty-six years' experience of trials of Africans in Africa
be allowed to say that in my opinion the conclusion reached by the Court
of Appeal was the only one which could be reached in accordance with
justice, reason and common sense.
The evidence at the trial showed that the chiefs in question had not
rendered assistance which they ought to have given in the investigation of
the offence. The order for their deportation on this ground met the justice
of the case.
The enquiry which provided the basis of the case for the prosecution had
been conducted by George Wilson, and there is extant an indignant letter
from him to the Colonial Office protesting at the Appeal Court's judgment.
But, not for the first nor last time, the standards of evidence acceptable to
the administrative officer failed to convince the judiciary.
Writing of Nuwa Mbaguta, Nganzi of Ankole, Mr. F. Lukyn Williams has
said (Uganda J., 10 (1946), 134) that "this crime was always a sore point with
Mbaguta, and he was unwilling to discuss the matter in later years. He shared
the suspicion and the opprobrium of the Government with the rest of the
Ankole chiefs and suffered with them." In justice to his memory, I think
it should be recorded that the evidence discloses that Mbaguta promptly and

to the best of his ability did everything that he was allowed to do to assist
in detecting the possible offenders.
One of the tragedies of the whole affair was that the evidence disclosed
that there was reason to believe that Lutaraka, the actual murderer, was
suffering from sleeping sickness and certainly had been behaving in a very
queer manner shortly before the murder. It is very possible that, if he had
been arrested and brought to trial, he might have been found "guilty, but
The greatest tragedy of all was that Gait, who was a capable administrative
officer with a promising career before him, was respected and extremely
popular amongst the majority of the people in the Western Province.


BANTU BUREAUCRACY. By L. A. FALLERS. Cambridge: W. Heffer & Sons
Ltd., 1956. xiv+283 pp., 13 plates, 8 maps. 30s.
Dr. Fallers set out to study the results of contact with western influence on
the Basoga over the last sixty or seventy years and with this end in view he
spent some twenty months living in the District. He divided his time more
or less equally between two small areas, one in the centre and one in the north,
with short visits to other parts, and consequently a certain amount of what
he has written is not generally applicable throughout Busoga. Much of what
he was told by the 'vested interests' in these two areas he did not take with
a sufficiently large grain of salt; as a result he does not give an accurate
picture of Busoga in the past. His references to 'kingdom states', 'rulers' and
their 'palaces', 'prime ministers' and 'princes of royal birth' give a false and
grandiose idea of some three hundred thousand people divided, at the end
of the nineteenth century, into not less than fifty-two major clans, each acting
more or less independently of the others and each having its ups and downs
in the scale of power. In fact, as is clear from past records including despatches
to the Foreign Office, the Basoga of those days were primitive, lazy and
apathetic, harried by Baganda raiders, and their only positive characteristic
which struck those who came in contact with them was their skill at thieving
from the caravans which passed through the country.
There are a number of inaccuracies in the book. For example on page 149
Dr. Fallers states that in 1919 one of the 'rulers' was chosen by the 'Council'
to be its permanent president. At that time there was no council, although
there was a sort of committee consisting of only the Saza Chiefs. In 1919 one
of the Saza Chiefs was appointed by Government to be president of the District
Native Court for a period of three years. His appointment was later renewed
and in due course, when a council was brought into being in 1938, he became
president of it. In 1939 he was given the title of Kyabazinga and by 1944 he
was claiming to be the hereditary Kyabazinga of Busoga. Dr. Fallers spent
much of his stay in the north of the district with this gentleman, now living
in retirement.
Another inaccuracy is the statement on page 249 that "though local loyalties
continue to exist, persons living within the District think of themselves first
as Basoga". The reverse is the case and one of the difficulties in administering
Busoga is that the people consider themselves first and foremost as Baisemenya,
Balamogi, Basiki, Badiope, etc. rather than as Basoga. A district commissioner
was told by the people in more than one area that they would prefer as a
chief a bad man from their own clan rather than a good man from elsewhere.
On page 60 Dr. Fallers makes the astonishing statement that the peasant
meets the European only in the context of formal authority relations carried
on across a desk in the central offices in Jinja, or at formal meetings held by
officers on tour. Nothing could be further from the truth, as will be borne
out by the scores of Administrative officers and the officers of the Agricultural,

Co-operative, Community Development, Veterinary and other departments
who have spent and are still spending so much of their time amongst the
peasants in their gardens, in their homes and on the field of sport.
Bantu Bureaucracy really needs two reviewers, a social scientist and a
layman with experience of the Basoga. The social scientific (if there is such
a term) sections are in a language foreign to the ordinary individual who
probably cannot understand even such a simple sentence as "the state
hierarchy was not a solidary group but rather consisted in chains of dyadic
relationships of subordination and superordination".
In a short review it is not possible to do full justice to Dr. Fallers's book.
Despite the criticisms above it is generally sound and shows a shrewd apprecia-
tion of the subject. It is a most valuable addition to the meagre supply of
literature on Busoga and will be a very useful guide to any future wider study
of the tribe. T. R. F. Cox.

UGANDA. By J. C. D. LAWRANCE. London: Oxford University Press, 1957.
xi+280 pp., 1 map, 22 diagrams, 11 plates. 36s.
This is a most useful contribution to Uganda literature, in that little
information has been available hitherto outside Uganda about the important
and numerous Teso tribe except in Government reports and in one or two
out-of-print missionary publications. Its chief value to the general reader will
be found in its first fifty pages, which comprise a summary of the annual
reports of Teso District for the period in question together with much historical
material from other sources. This condensation is a most useful piece of
work, the substance of which has already appeared in Uganda Journal, 19
(1955), 7-40.
As regards the history of administration, the author tends to give a slightly
anachronistic impression, in that he does not stress the remarkable change,
which has taken place in the attitude of Government (and of public opinion
as a whole) towards the Iteso in the last ten years. Until the late nineteen-
thirties they were hardly thought of as a Nilo-Hamitic tribe, but as a com-
munity of doubtful (query Bantu) origin. The discovery of their relationship
to the Karamojong cluster of tribes and to the Topotha in particular was due
to private inquiries made in 1936 and subsequently published in Uganda J., 9
(1941-2), 57.1 These inquiries showed that the Teso language and social struc-
ture so closely resembled these somewhat remote people that the practice
hitherto maintained by British officers of treating the Iteso as a kind of
second-rate Bantu tribe was obviously inappropriate. This discovery was
naturally unpopular since in twenty-five years of administration the Eastern
Province of Uganda had come to be regarded as a sacrosanct unit of accepted
social pattern. It was not, however, necessary to pay particular attention to

1 The priority of this claim appears to be disproved in favour of Seligman who, in
the first (1930) edition of The Races of Africa, p. 161, classifies the Iteso as Half-
Hamites and notes the close relationship between them and, among others, the Karamo-
jong and Topotha. ED.

this discovery in the flux of the war years, and it was only subsequent to
World War II that these facts began to have political relevance.
Another criticism which may be advanced is that, though the title refers
to the Teso tribe, the book deals only with the inhabitants of Teso district.
The important migration of Iteso down through Bukedea and Pallisa into
Bugwere and Budama and over the border into Kenya (Elgon Nyanza) forms
an essential piece of tribal history, which could advantageously have been
included in a book of this title. The aim of a monograph of this size must
surely be to give the general reader a broad conspectus of the whole tribal
community and of its social background. In these circumstances, Teso
political relationships with the Sebei and Bagisu tribes along the foothills
of Masaba (Mt. Elgon), with the Bagwere and Jo pa Dhola in the plains of
Mbale and Tororo Districts and with the Jo pa Owiny and the Baluhya, as
they spread over what is now the Kenya border, are of importance. Old
kraal sites are still known and the lines and approximate dating of the
successive migrations southward have been recorded by some of the older
Mill Hill fathers.
To a modern Etesot this book is flattering. It suggests by implication the
rather improbable thesis that the Iteso had a considerable cultural heritage
of their own. It thus makes them the obvious political leaders of 'the
Karamojong Cluster' from the point of view of future social and economic
As regards culture, though all the objects illustrated-granaries, baskets,
milk-pots, stools, hoes, knives, spears-can, or could, be found at one time
or another in one or other part of Teso district, it would be unwise to
suggest that they were mostly of Teso origin. The early European travellers
in this region were chiefly struck by the absence of culture and the techniques
which support it around Lake Salisbury and Lake Gedge.
In fact most of the ironwork (as the author states on p. 134) and pottery
were directly of Bunyoro origin-either the result of trade across Lake
Kyoga, or the work of immigrant Banyoro smiths and potters. By contrast,
in the north of the district, ironwork was generally of Labwor origin. The
cotton-reel or bobbin stool was an artifact of Hima origin and the three or
four-legged round stools were probably learnt from contact with the Lwoo.
The true Teso stools, where they existed, were unsymmetrical objects made
of a Y-shaped branch. The fishing apparatus was mostly derived from the
lacustrine Bakenyi or Baruli. Much the same can be said of the musical
instruments; but the Iteso certainly developed a speciality in their huge
atenus drum and the tiny itelele. They had also an arched harp adeudeu
which was closer than that of any contemporary tribe to the old Egyptian
prototype, and some special wind instruments. One important male ornament
I do not see recorded here is the carved ivory armlet (of which the modern
wooden beaded apel is a development). This was functionally related to the
armlet still worn by the Masai moran and also to a Meroitic ivory armlet
dug up in Nubia some years ago. A specimen of this old type of armlet is in
the Uganda Museum, and several were in the possession of the Foster family
of Kaptagat.

As regards the anthropological and legal sections, ne supra crepidam sutor
is the cry. The author has fallen into this difficulty, that both law and anthro-
pology have now become such highly technical studies that to publish literary
material without 'submitting to accepted disciplines and definitions is to
invite criticism; (the sections dealing with 'user' pp. 249-50 are a case in
point). H. Cory's legislative compendia of tribal law for the Bahaya, Basu-
kuma, Banyamwezi and others are probably not the best practical models in
a rapidly changing economy. The customary detail given so soon becomes
obsolete with the change in social sanctions and ideas, and the method omits
the essential social sanctions which support the verdicts. Such handbooks of
tribal law indeed rapidly become extinct reference books, which are only
used in appeal courts to support claims, which the common-sense de facto
judgments of the lower courts have often already dismissed as absurd.
One final observation. Somehow from all these pages of detailed facts
emerges a feeling that the character of the Iteso people themselves has escaped
description; for example, their record of a persistently high murder rate all
through the nineteen thirties and (it is presumed) up till the present time,
the tempo of their most typical dances, which rise like a breaking wave and
crash in the same way; their tendency to personal hysteria associated either
with murder or suicide; their odd inconsequential folk tales (partly recorded
by Father Schut) which read like a queer dream. All these suggest that the
developing mind of the 'simple savage' was not as simple a process as might
be gathered from the bureaucratic outlines of development here related.
Instead one would presume that there are real and persistent anxieties and
repressions disturbing this communal life. The tale of the development of
Teso District as here related sounds too easy. A stranger might think that
this district was really managed by an ordinary local government; but 'local
governments', as the term is used in England, do not spend hours discussing
whether adultery is implied if a woman sits on a man's cycle carrier. The
Teso Council does, and finds it stimulating. It remains therefore a 'council'
which is still far more 'tribal' than 'local'.
Nevertheless Mr. Lawrance must be credited with a valuable achievement
in getting this material collected and printed, while still occupying the busy
post of an administrative officer. No other contemporary administrative officer
in East Africa has achieved as much. The book is not perfect; but how much
better than none at all! Its publication at this stage will place considerable
ammunition in the hands of those Iteso who see no reason why their affairs
should be dominated from Kampala or Entebbe. It may indeed be that Mr.
Lawrance's book will find its great significance as a weapon in the post-British
stage of political relationships. For this, its weighty format may well be a
positive advantage. A. C. A. WRIGHT.

SIR HARRY JOHNSTON and the Scramble for Africa. By ROLAND OLIVER.
London: Chatto & Windus, 1957. xv+368 pp., 30s.
'Prancing Proconsul', the sobriquet bestowed by his contemporaries on
Sir Harry Johnston, is, when shorn of its more frivolous insinuations, a
remarkably appropriate label for a truly remarkable man.