|UFDC Home||myUFDC Home | Help ||
|Table of Contents|
|The Status of the Nile Crocodile...|
|Mackay's Canoe Voyage along the...|
|Acholiland in 1897|
|Masaka Hill - An Ancient Centre...|
|The Baganda and the Bakonjo|
|African Performance on an Intelligence...|
|The Prehistory of the Entebbe...|
|The Northern Province Mountains:...|
|The Railway Extension to Western...|
ALL VOLUMES CITATION PAGE IMAGE ZOOMABLE
STANDARD VIEW MARC VIEW
|Table of Contents|
Front Cover 1
Front Cover 2
Title Page 1
Title Page 2
Table of Contents
Table of Contents 1
Table of Contents 2
The Status of the Nile Crocodile in Uganda
Mackay's Canoe Voyage along the Western Shore of Lake Victoria in 1883
Acholiland in 1897
Masaka Hill - An Ancient Centre of Worship
The Baganda and the Bakonjo
African Performance on an Intelligence Test
The Prehistory of the Entebbe Peninsula
The Northern Province Mountains: Speculations on Climate and Vegetation History
The Railway Extension to Western Uganda
Back Cover 1
Back Cover 2
Quy 40414 Imsm
how a t 1 '
Ctr f l,
mynn 7 t
beiQr4 t r i~ey
M r R B td ~ v
M r ,C Mtj im
Ro`orir; rl S e
"Mi W I ~
EdtivD ; ,
iij l ... ... ...........................
4Rii o -mm U~ j ri 'G a
THE JOURNAL OF THE UGANDA SOCIETY
VOLUME 18 No. 1
EDITORIAL COMMITTEE (1953-4)
Dr. W. H. R. LUMSDEN
Dr. A. W. SOUTHALL I Hon. Editors
MAJOR B. G. KINLOCH, M.C.
DR. AUDREY I. RICHARDS
DR. K. P. WACHSMANN
(assisted in England by H. B. THOMAS, O.B.E.)
PRINTED IN GREAT BRITAIN BY HEADLEY BROTHERS LTD
109 KINGSWAY LONDON WC2 AND ASHFORD KENT
The Status of the Nile Crocodile in Uganda DR. HUGH B. COTTr
Mackay's Canoe Voyage along the Western Shore of Lake Victoria in 1883
SIR JOHN MILNER GRAY
Acholiland in 1897 SIR JOHN MILNER GRAY
Masaka Hill-an Ancient Centre of Worship E. C. LANNING
The Baganda and the Bakonjo DR. ROLAND OLIVER
African Performance on an Intelligence Test DR. JOHN McFIE
The Prehistory of the Entebbe Peninsula K. MARSHALL
The Northern Province Mountains: Speculations on Climate and
Vegetation History H. C. DAWKINS
Jie Agriculture DR. P. H. GULLIVER
The Railway Extension to Western Uganda P. H. HICKS
Uganda Place Names -
Ancient Earthworks in Western Uganda -
African Highway (by Sir Malcolm Watson) DR. C. A. BOZMAN
THE STATUS OF THE NILE CROCODILE IN UGANDA
By HUGH B. Corr, D.Sc., Sc.D., F.R.P.S.
(University Museum of Zoology, Cambridge)
D URING 1952 I visited Uganda, with the primary object of investigating
the biology of crocodiles and fish-eating birds, with special reference to.
their feeding-habits and the relation of thesq animals to inland fisheries. The
scientific results of the expedition yet remain to be worked out in detail and
will be published elsewhere in due course. The following notes are therefore to
be regarded merely as a preliminary account, and in them I shall confine my
remarks to a brief survey of the status of the Nile Crocodile (Crocodilus-
niloticus L.) in Uganda, and to the supposedly conflicting interests of the;
fishery and crocodile-hide industries.
In the course of this work I received the most generous help from many
sources, and it is a pleasure to take this opportunity of expressing my gratitude:
to all concerned. In particular, I have to thank Dr. R. S. A. Beauchamp, for a
grant from East African Fisheries Research Organization in respect of air-
travel, for much kindness and hospitality, and for many splendid facilities.
readily made available at the Fisheries Laboratory at Jinja. Here also, both
Miss R. M. Lowe and Mr. P. H. Greenwood were always willing to examine:
fish recovered from crocodiles and waterbirds, and I gratefully acknowledge
their valuable assistance. Major B. G. Kinloch spared no pains to help in many
ways, giving valuable suggestions and introductions, and a financial grant from
the Game Department to cover the use of a launch in waters between Entebbe;
and Sese; as Acting Chairman of the Uganda National Parks Trustees he also.
generously gave permission for a stated number of crocodiles to be killed, for
scientific purposes, along the Magungo reach of the Victoria Nile. My thanks;
are also due to Fisheries Officers of the Game Department and Lake Victoria
Fisheries Service-Major R. E. P. Wyndham at Hoima, Mr. J. M. Warren at
Kichwamba, Mr. D. H. Rhodes at Serere, and Mr. H. T. Hayes at Entebbe-
for hospitality and help in connection with field work on Lakes Albert, George,.
Kyoga and Victoria; also to Mr. Norman H. Searle, of Bweramule, for similar
assistance on the Lower Semliki River.
During July to October I was joined by my friend, Mr. Jack Lester, Curator
of Reptiles at Regent's Park, and much practical help received from him added
materially to the results achieved. Finally, I am indebted to the Zoological
Society of London and to the Percy Sladen Trustees for financial assistance
towards general expenses incurred during a protracted and most profitable visit.
THE FOOD OF THE CROCODILE IN RELATION TO AGE
Two main charges brought against the Nile crocodile are that it is dangerous:
to human life and detrimental to fisheries. While in the days of its former-
abundance human casualties were numerous, today such accidents are rarely
2 UGANDA JOURNAL
reported. Depredations upon fish and damage to gear are more serious. But
losses from this source are likely to be exaggerated owing to the publicity which
they receive; and the view, often held, that the crocodile is primarily a fish-eating
animal requires qualification. Before relegating crocodiles to the class 'vermin'
and pronouncing the death penalty as sentence, it would be well to study the
evidence now available.
At the Anglo-Belgian Conference on Freshwater Biology and Fisheries in
1949, the need was stressed for research on the bionomics of these reptiles, and
especially on their food and feeding behaviour, about which there was inade-
quate information. During 1952 the writer had occasion to examine the stomachs
of crocodiles in various parts of Uganda-including Lake Victoria and the
Victoria Nile, Lake Kyoga, the Semliki River, Lake Albert and the Murchison
Nile-in the hope of throwing some fresh light on the ecological and economic
status of crocodiles in these waters.
Since Dr. E. B. Worthington's examination of a few stomachs in 1928 and
1931, it has been suspected that the diet of crocodiles may change with the
predator's age and size. In the past, few observers have recorded both the
alimentary content and the length of crocodiles examined. It is now clear from
recent work that striking changes do occur in the dietary of various age-groups.
For convenient comparison, five such groups may be considered, viz.: (1) below
1 metre; (2) 1 to 2 metres; (3) 2 to 3 metres; (4) 3 to 4 metres; (5) 4 to 5 metres,
in length from snout to tail-tip. Table 1 shows the number and percentage of
crocodile-stomachs (within these size-groups respectively) containing various
categories of prey, based upon data from 105 individuals examined during the
Size-group 1 2 3 4 5
Length in metres 0-1 1-2 2-3 3-4 4-5
Insects 43 (65-2%) 14 (30-4%) I ( 5-9%) -
Crustacea 2 ( 3-0%) 10 (21-7%) 2 (11-8%) -
Spiders 4 ( 61%) -
Molluscs 4 ( 6-1%) 4 ( 8-7%) 3 (17-6%) 2 (11-8%) -
Fish 1 ( 1.5%) 7 (15-2%) 9 (52.9%) 8 (47-1%) 1 (16-7%)
Anura 10 (15-2%) 4 ( 8-7%) -
Reptiles 1 ( 15%) 2 ( 4-3%) 1 ( 5-9%) 2 (11-8%) 3 (50-0%)
Birds 3 ( 6-5%) 1 ( 5.9%) 3 (17-6%) -
Mammals 1 ( 1-5%) 2 ( 4-3%) 1 ( 5-9%) 1 ( 5-9%) 2 (33-3%)
Number and percentage of stomachs containing various foods, by size-groups.
For purposes of comparison, and to present as complete a picture as possible,
in Table 2 are given the same data, to which are added such other records as are
available (i.e. where both stomach-content and crocodile-length are known).
These additional records incorporate those given by Jardine (1929), Welman
and Worthington (1943) (incorporating earlier data from Graham (1929)),
Swynnerton (1946), Pitman (1948), van Ingen (in Pitman (1950)), and Kinloch
(1951). A series of Nile crocodiles from West Africa, given by Welman (Welman
and Worthington, ibid.) as measuring from 6 to 11 feet, are here included in
THE STATUS OF THE NILE CROCODILE IN UGANDA 3
Size-group 1 2 3 4 5
Length in metres 0-1 1-2 2-3 3-4 4-5
Insects 46 (66-7%) 14 (26-9%) 1 (3-1%) -
Crustacea 2 ( 2-9%) 14 (26-9%) 7 (15-2%) -
Spiders 4 ( 5-8%) -
Molluscs 4 ( 5-8%) 6 (11-5%) 7 (15-2%) 3 ( 94%) 1 ( 2-5%)
Fish 1 ( 1-4%) 7 (13-5%) 21 (45-6%) 16 (50-0%) 15 (37-5%)
Anura 10 (14-5%) 4 ( 7-7%) -
Reptiles I ( 1-4%) 2 ( 3-9%) 3 ( 6-5%) 5 (15-6%) 13 (32-5%)
Birds 3 ( 5-8%) 4 ( 8-7%) 4 (12-5%) 1 ( 2-5%)
Mammals 1 ( 1-4%) 2 ( 3-9%) 4 ( 8-7%) 3 ( 9-4%) 10 (25-0%)
Number and percentage of stomachs containing various foods, by size-groups.
size-group 3 though there would be some overlap into groups 2 and 4. I have
added unpublished records for two crocodiles killed on the Tana River by
Mr. Tony Henley. Finally, I have added data from a fine series of twenty-three
specimens examined during 1950 from Kadem Bay, Lake Victoria, by Mr. F.
Wilson who has kindly given me permission to make this use of his data.
Table 2 thus summarizes our present knowledge of diet in relation to age, being
based upon 173 stomachs, of which 105 were examined by myself and 68 by
others. These data are expressed graphically in Fig. 1, which shows, for
crocodiles of different size-groups, the percentage of stomachs containing
(a) invertebrates, (b) fish, and (c) vertebrates other than fish.
CROCODILE LENGTH IN METRES
o, 0-1 1 1-2 2-3 3-4 4-5
Percentage of stomachs containing various foods, by
4 UGANDA JOURNAL
A striking feature of this survey is the progressive increase in vertebrate
prey, as opposed to invertebrates, with growth. Vertebrates are represented in
the smallest size-group by only 19 per cent-mostly frogs, toads and tree-frogs;
the corresponding figures for the remaining groups, in order of increasing
length, are 35, 69, 87 and 98 per cent. It will be noted that crocodiles feed to a
greater extent than is usually believed upon quite small animals, individuals
up to two metres in length taking mainly invertebrates of one kind or another.
Juveniles of group 1 are almost wholly insectivorous: indeed, of the forty-six
specimens which contained recognizable food, all included some insects. Insects
also make up an appreciable item in size-group 2, above which they virtually
drop out of the pabulum.
Among the various insects taken, Hemiptera head the list-waterbugs of the
family Belostomatidae being prevalent. Second in importance are Odonata
(dragon-flies)-both nymphs and adults. Beetles are also frequently taken. The
remainder are sporadic inclusions, comprising crickets, grasshoppers, earwigs,
termites, ants and caterpillars.
Other invertebrate prey includes fresh-water crabs (which form a substantial
element in the diet of group 2) and molluscs (especially in groups 2 and 3,
though they are recorded from all size-groups). Spiders have only been found
in group 1; they were being taken, for example, by small crocodiles feeding on
the Kaiso sand-spit, Lake Albert.
Of vertebrate prey, we may note that fish are hardly ever captured by young
crocodiles of size-group 1, and for size-group 2 fish are still a subsidiary element
in the diet. Indeed, Dr. K. K. Zwilling of the Department of Local Industries,
Northern Nigeria, stated (15 October 1952) that on Panyam Dam, Plateau
Province, Nigeria, where between December 1950 and February 1952 he had
examined about twenty-five crocodiles between one and two metres in length,
insects were the main food and he had never recovered any fish from the
stomachs. With subsequent growth up to four metres in length, fish become
important and are the main prey. Individuals above this size tend to turn to
other fare, notably reptiles and mammals, both of which become increasingly
important throughout the whole series of size-groups from smallest to largest.
In Uganda, reptiles may form the main food item of the oldest and most
decrepit animals, which take turtles, monitor lizards and cobras, as well as eggs
and young of their own species.
THE QUANTITY OF FOOD EATEN
Wholly extravagant views as to the quantity of food consumed by an indi-
vidual crocodile have been held in the past. For example, Mr. C. W. Chorley
expressed the view that each crocodile in Lake Victoria consumed from 40 to
60 pounds of fish daily. And Captain R. J. D. Salmon (1932), referring to
crocodiles feeding below Murchison Falls, mentions 50 pounds a night as an
estimate. On the other hand, Pitman has stated, in commenting upon the slow
growth of the crocodile, that "a meal of any size is a rarity and a full meal
exceptional". I have now examined the stomachs of crocodiles killed at most
hours of the day and night, and during most months of the year, and the data
thus obtained fully substantiate Pitman's observations.
Photo by Hugh B. Cott
Crocodile sun-bathing below Murchison Falls.
Photo by Hugh B. Cott
Crocodile snare, Buluba Bay, Lake Victoria.
[lace p. 4.
Copyright reserved Photo by Hugh B. Cott
Jaluo crocodile hunter with hooks and tackle.
THE STATUS OF THE NILE CROCODILE IN UGANDA 5
Crocodiles are usually credited with extraordinary powers of digestion, and
to this has been attributed the frequent finding, by some authors, of empty or
nearly empty stomachs. However, many such records undoubtedly refer to
females shot at the nest. Careful, as opposed to casual, inspection of the
stomachs of crocodiles killed under other circumstances reveals a high percen-
tage with more or less recognizable food, and this in animals of all age-groups
and irrespective of the time of death. Thus, of the 105 specimens examined in
1952, remains of previous meals were found in 91 (87 per cent) and only 14
(13 per cent) were quite empty.
Again, there does not seem to be any evidence to support the view that
crocodiles digest their meals with undue rapidity. On the contrary, the follow-
ing incident suggests rather the reverse. At Jinja a captive crocodile, about
one metre in length, attacked a smaller crocodile on 23 December, severing
its head. The head was eaten on the 24th, and the remaining parts on the follow-
ing day. On the 1 January, one week after its last meal, I killed this animal; its
stomach still contained intact the stomach and gall bladder of its victim,
together with a partially digested foot and many bones and scales, as evidence
of its cannibalism. In short, I believe that the relative scarcity of food recovered
from stomachs is more an indication of infrequent feeding than of rapid
digestion. The crocodile leads an idle life, expends little energy in a climate
ideal for a cold-blooded creature, and troubles to take only the little food he
needs. In this respect he differs absolutely from the warm-blooded and active
fish-eating birds such as darters and cormorants, which, especially during the
breeding season, have to spend much of the day in fishing and in travel between
feeding-grounds and nesting-places.
THE ECONOMIC STATUS OF THE CROCODILE
Turning to the economic aspects of the subject, we have first to note that-
taking account of all size-groups-food other thhn fish forms a large part of
the diet. Thus the data in Table 2 show that fish were recorded from 60, and
animals of all other groups from 179 stomachs, the frequency of fish occurrences
in stomachs, as compared with the other prey, being in the approximate ratio
of 1 to 3 for the series under review.
Fish-prey taken are likely to differ in different waters; and losses by no
means necessarily involve the most palatable and economically important
genera such as Tilapia. On Lakes Kyoga and Kwania it appears from Hippel's
extensive survey of crocodile stomach-contents (1946), that Protopterus and
Barbus together account for nearly two-thirds of the catch, the remaining
genera, in order of importance, comprising Tilapia, Clarias, Bagrus, Mormyrus
and Haplochromis. My own less adequate records indicate that in north Lake
Victoria and the Victoria Nile between Namasagali and Ripon Falls, Proto-
pterus and Mormyrus are most frequently taken, while Barbus and Haplochromis
may also be important. The crocodile's predilection for Mormyrus is also
suggested by Jaluo hookers' use of its flesh as bait whenever it can be obtained.
I was frequently told both by Jaluo and by local men from Jinja that these fish
provide the most successful bait, second only to hippopotamus meat. Records
from the Semliki River, Lake Albert, and the Murchison Nile show occurrences
6 UGANDA JOURNAL
in the following order: Synodontis and Alestes as most important; Aucheno-
glanis, Hydrocyon and Tilapia; Clarias, Lates, Labeo and Barbus.
In other waters, the Nile crocodile has been regarded as seriously detrimental
to essential fisheries. For example, in Lake Nakavali they were reported by
Pitman (1950) to be taking Tilapia from set-nets, and to be highly destructive
of gear. The same habit has been reported by Swynnerton (1946) in Lake
Rukwa, though the crocodiles examined by him were individuals shot in the act
of raiding seines for Tilapia, and thus it is not clear how far the stomach-contents
of these animals are representative of the general diet of crocodiles in that lake.
In any attempt to assess the harm done by crocodiles, account must be taken
of the complex food-chains involved. For example, several of the above-men-
tioned fish are themselves predatory on fish, fry or fish-eggs-Hydrocyon,
Barbus, Clarias, Bagrus, Synodontis, Lates, and several species of Haplo-
chromis-and it would thus appear that the effects of crocodile destruction are
by no means necessarily beneficial to fishery interests. Thus, in Hunyani Poort
Dam, near Salisbury, Southern Rhodesia, the wisdom of killing crocodiles was
recently questioned by Mr. L. H. Stewart, Secretary of the National Resources
Board of Southern Rhodesia (Anon., 1952): it was suggested that crocodiles
should be maintained to limit the numbers of fish-eating fishes such as barbel
(Heterobranchus). It has also been reported that in the Belgian Congo, barbel
rapidly multiplied after crocodiles had been reduced. Again, in Lake Mweru,
Northern Rhodesia, destruction of crocodiles was followed by an increase in
damage to netted-fish by otters, which are themselves normally preyed upon by
crocodiles. Mr. J. D. Kelsall also stated at the Conference on Freshwater Biology
and Fisheries at Entebbe, 1952, that the reduction or extermination of crocodiles
in Lake Victoria was unlikely, in his opinion, to affect Tilapia fisheries.
THE ECOLOGICAL STATUS OF THE CROCODILE
Outside the immediate interests of inland-fisheries, crocodile destruction
may disturb the balance of nature in various unexpected ways. For instance,
when the crocodile-skin boom was at its height in Ceylon in the early thirties,
parties of hunters wiped out a large part of the crocodile population over wide
areas. One result was that larger predatory fishes flourished and destroyed smaller
insectivorous species which then failed to keep down mosquito larvae, so that
malaria greatly increased.
From Madagascar, Professor J. Millot reported at the above-mentioned
Conference on one of the peculiar repercussions of crocodile diminution (result-
ing from trade in skins): crocodiles are particularly fond of dogs; they caught
sick animals more readily than healthy ones; and when this natural control had
been removed, there was a rise in the incidence of rabies. The crocodile is now
accorded some protection in Madagascar, there being in force a restriction on
the size below which animals may be destroyed.
The relation between crocodiles and sleeping sickness is not yet understood.
However, the Tsetse Control Department in Uganda has in the past supported
the destruction of crocodiles, being of the opinion that these reptiles and monitor
lizards are in some way connected with riverine tsetse. Dr. G. Marlier, of Uvira,
THE STATUS OF THE NILE CROCODILE IN UGANDA 7
Costermansville, referred at the same Conference (1952) to research which
indicated that in areas where crocodiles were scarce, tsetse were also less
numerous and smaller, but more ready to attack the hunter; this indicated that
the tsetse had a preference for crocodile blood.
In the present state of knowledge, it is impossible to predict what results will
ensue if crocodile stock is further diminished in Uganda. The complexity of the
inter-relations between these reptiles and other animals may be illustrated by
the following examples, which however give only a partial and much over-
simplified picture of the feeding-ecology.
As regards its relationship to other reptiles: nests of the crocodile are com-
monly despoiled by monitor lizards (Varanus niloticus), which also dig out eggs
of the turtle Pelusios. But the adult crocodile itself eats the monitor. Crocodiles
also prey upon eggs and young of their own species, and upon the turtles. These
latter include fish in their diet. And darters, themselves exclusively fish-eaters,
again form an item in the crocodile's menu. The Black-lipped Cobra (Naia
melanoleuca) is another reptile taken by the crocodile: this snake also includes
fish in its pabulum, and feeds extensively on toads which are important as eaters
of fresh-water and shore-dwelling insects.
Again, as regards its relationship, directly or indirectly, to insects: The feed-
ing habits of juvenile crocodiles reveal a complex network of relations with
other members of the shore fauna. In early life crocodiles include giant waterbugs
(Belostomatidae), as well as toads (Bufo spp.), frogs (Rana spp.) and tree-frogs
(Hyperolius spp.) in their bill of fare. Belostomatids in turn prey upon these
Anura, as well as upon other insects which also form prey for the juvenile
crocodile, while of course the Anura are themselves insectivorous in their turn.
Further, young crocodiles feed upon shore-spiders which prey upon emerging
dragonflies: these insects are themselves predatory upon other members of the
insect-fauna; and the situation is still further complicated by the crocodile's
penchant both for larval and adult dragonflies.
EXPLOITATION AND DECLINE OF THE CROCODILE
War waged against the crocodile has in the past completely driven the reptile
from the Nile below Khartoum, while in historic times it was found to the
shores of the Mediterranean and in the lakes that now form part of the Suez
Canal. During recent years it has suffered a serious decline in many other parts
of its range.
In Lake Victoria the work began about 1935 with Captain C. R. S. Pitman's
vigorous and successful campaign against breeding females on the islands and
lake shores between Sese and Entebbe. Even in 1940 Pitman was able to report
that crocodile numbers had been appreciably reduced along the northern shore
from Katebo to Damba. Subsequently the demands of the belly-skin industry
have accelerated the destruction. By 1946 exploitation both in the Victoria Nile
and Lake Kyoga was well under way: leather shortage during the war had led
to a booming trade in which were engaged Europeans, Indians, Arabs and
Africans. In the peak period on Lake Kyoga, one syndicate marketed over
1,500 skins in one month. About this time Pitman reported (1948) that crocodile
8 UGANDA JOURNAL
catching had become so profitable that many native fishermen, operating snares
"had abandoned their usual livelihood to the benefit of their pockets and to the
detriment of the fisheries".
Today, individuals and companies are engaged in ruthless exploitation, and,
variously in different areas, the crocodile is looked upon as something to be
clubbed, netted, snared, hooked, harpooned or shot. The result is not surprising;
everywhere one hears accounts, and sees evidence, of a marked decline in the
crocodile population. The extent of its changed status may be illustrated by a
few examples. From the Kavirondo Gulf, Mr. Boundford tells me that a few
years ago it was possible to go out any week-end and secure crocodiles without
difficulty in places where today none may be encountered. I am informed by
Mr. H. T. Hayes (15 October 1952) that in a bay near Entebbe where thirty
crocodiles had formerly been hooked in one week, recently in the same place
only six were taken in a month. Crocodile hooking is now prevalent every-
where along the north shores and islands of Lake Victoria, from Sese to Grant
Bay and beyond, and in the head-waters of the Nile. During a tour in company
with Mr. Jack Lester in November 1952, evidence of the destruction was every-
where apparent. For example, the once populous sand-spit at Buvu was almost
deserted; one breeding ground there contained a single nest, with a sun-bleached
crocodile-skull lying nearby. At a second ground a mile distant we found one
deserted nest-site, with another skeleton nearby: within a few yards of this place
was a recently vacated camp.
At Fumve we located a party of hookers working in the Sese group. During
an eight-hour launch trip among these southern islands, only three or four
crocodiles were sighted. At Damba we had been forestalled by the Jaluo, and
local fishermen said they had "killed all the crocodiles". Inspection of the shores
from a dinghy by torchlight after dark revealed no crocodiles. Such negative
evidence is significant, because-as is known to everyone who has hunted in this
manner-the crocodile's eye is easily picked up in the light beam. Night work
along the shores on both Lulamba and Buvu Islands (north Sese) gave the
Most striking was the change that had taken place on the grounds in the south
of Bulago Island. At the long-established nursery on a sandy promontory con-
cealed from the water by a belt of dense bush, where in former years Captain
Pitman had found some thirty nests in a season-some being so close that the
rims almost touched-there was but one deserted nest of eggs which monitors
had been raiding; litter of discarded hooking gear was found at the nearby
It was the same at Dyavodemi, north of the Damba Channel, where we saw
no trace of living crocodiles, but evidence of the recent visit of men who had
left behind them ambatch floats, ropes and litter of carcases. On the neigh-
bouring islet of Masovwi lay the rotting remains of a huge crocodile which the
men had not troubled to return to the water.
Nor are the prohibited tsetse areas exempt from the attentions of these
African poachers. Thus, during an expedition with Mr. J. Hinchcliffe in June,
two camps were found on Igwe Island, east of Buvuma. On this occasion the
men escaped, with some of their firearms, but abandoned a carbine and a
THE STATUS OF THE NILE CROCODILE IN UGANDA 9
quantity of -303 and 12-bore ball ammunition, skinning knives, pangas, ropes,
double-hooks, salt and other gear. Salted skins were found pegged out under
the trees, together with crocodile carcases, and a quantity of smoking water-
buck, bushpig and hippopotamus meat.
Recent disturbance is again reflected in the increased difficulty of approach-
ing crocodiles in all waters outside the National Parks. For example, Mr. Edward
Wilson informed me (29 October 1952) that on the islets of the Nile a few
miles above Namasagali, it was possible until five years ago to get within ten
yards of them. Hooking began in 1948-9, and has recently been intensified,
together with shooting. Today, as a result of harrying, the remaining animals
are increasingly wary, and will leave their basking places when a launch is three
hundred yards away; and it is difficult to come within fifty yards of them.
In 1929 Lieutenant-Colonel Gordon Brown reported an unprecedented num-
ber of crocodiles in the Semliki Delta, exceeding in number those below
Murchison Falls. Where are these crocodiles today? Three or four years ago
they were still fairly plentiful, and it was then possible to shoot over two dozen
in a morning; now it would be difficult to bag one-tenth of that number.
Mr. Norman Searle tells me that in 1949 crocodiles of large size were still in
evidence along the Semliki southwards to the forest. Now, except near the
Delta, large specimens have been eliminated from the river, and the industry
here has since become unprofitable.
This rapid decline in the crocodile's status is of course taking place elsewhere
than in Uganda. Thus, Mr. J. D. Kelsall reported (15 October 1952) that croco-
diles had been fantastically reduced during the past few years in Tanganyikan
waters of Lake Victoria. He informs me that four years ago at Mwanza, thirty
or more crocodiles could be seen along certain stretches of coast, the animals
being heaped together in places; today one may go there and see two or three
or none at all.
Again, I am informed by Mr. R. E. Taylor, Fisheries Officer, Sudan, that on
a 75-mile stretch of the Nile northwards of Malakal, where, about 1947, fifty
crocodiles might be seen in a day's passage, one may now do this journey and
see none. A regrettable feature of exploitation in these waters is that the
market, for export to France, includes small crocodiles about one metre in
length. Being much in demand, these young animals are clubbed, speared or
foul-hooked on lines with no chance of replenishing the stock. This industry
has been so profitable that purchasers of skins could afford air-freight charges
The trappers and agents are themselves well aware of the change brought
about by their own activities. Thus, in a letter dated 12 August 1952, one of the
Kampala agents trading in hides reported that "trappers complain that certain
areas round the lake shore are nearly fished out and they go further and
further afield". Again, I was told last year by an official of one of the concerns
engaged in the crocodile-hide industry: "Our chief difficulty is to find new
hunting grounds." Mr. Kelsall also told me that one European trapper who had
for a year been operating in Emin Pasha Gulf, recently gave up; his quarry were
getting smaller and so scarce that it no longer paid him to work there.
In October 1952, I fell in with a party of professional hookers, Jaluo from
10 UGANDA JOURNAL
Kenya, working the Victoria Nile above Namasagali. This gang of some eight
men had camped in the neighbourhood for a fortnight, hooking from three to
eight crocodiles each night, and intended to stay until the area was fished out.
I asked the head boy about his future plans: his comment was, "When we finish
them here, we go to another place."
THE NEED FOR RATIONAL CROCODILE CONTROL
It has sometimes been argued that a widespread and prolific species can be
in no real danger of extermination. But diminution of a wild animal popula-
tion is rarely dramatic and sudden: more often the drain is spread over a period
of years, and the resulting change may be scarcely heeded. In the early stages
of over-hunting, an added danger lies in the very abundance and apparent welfare
of a wild species. This may lead to the complacency shown, for example, in the
Report of a Select Committee of the Senate of Ohio. When dealing with a
proposed bill to protect the Passenger Pigeon (Ectopistes migratoria) in 1857,
this committee concluded: "The Passenger Pigeon needs no protection. Wonder-
fully prolific, having the vast forests of the North as its breeding grounds,
travelling hundreds of miles in search of food, it is here today and elsewhere
tomorrow, and no ordinary destruction can lessen them, or be missed from the
myriads that are yearly produced" (Hornaday, 1913). Within a few years, this
bird was reduced from millions to total extinction.
But for timely Government control, the koala of Australia would have met a
like fate. Although subject to exploitation from about 1880, it was still toler-
ably abundant in 1895. The demand for its warm and almost indestructible fur
led to an increasing slaughter; the number of skins exported rose from 300,000
per annum at the end of the century to some 2,000,000 skins (chiefly under the
false name of Wombat) in 1924. Koalas are now confined to the eastern fringe
of Australia, and the total population a few years ago was estimated at less than
one thousand individuals.
Other well-known examples of the effect of unrestricted exploitation are
furnished by the Biscayan Right Whale (Balaena glacialis) fishery of the twelfth
to sixteenth centuries, and the whaling industry in the Arctic around Spitz-
bergen and Greenland, which flourished during the seventeenth and eighteenth
centuries until the Greenland Whale (Balaena mysticetus) had been almost
extirpated: whaling was then no longer profitable in the Northern Hemisphere.
The Southern Fur Seal (Otaria pusilla) affords a further example of former
wasteful exploitation. Once stupendously numerous, this animal resorted for
breeding to various islands in the Southern Ocean. During the two seasons
following the discovery of the South Shetlands in 1819, about 320,000 seals
were clubbed and more than 100,000 pups died. By 1823 the colonies had been
virtually exterminated. A few years previously, 400,000 skins were taken from
Antipodes Island off New South Wales; on arrival in port, one spoiled cargo
of 100,000 skins had to be dug out of the hold and sold as manure. Similar
destruction took place in the Falkland Islands and South Georgia; in the latter
island alone, Weddell, writing in 1825, estimated that 1,200,000 skins had been
taken in fifty years, and the species was then nearly extinct. After fifty years of
THE STATUS OF THE NILE CROCODILE IN UGANDA 11
slow recovery, this seal was once again reduced to a handful of survivors by
the sealers; and as Laws remarks in hiis recent report (1953), "It has taken
seventy years for the survivors to rebuild the population to the few thousands
now existing. The wholesale destruction of the Southern Fur Seal between
the 1780s and 1822 was the most extravagant and wasteful exploitation of
biological resources that has ever occurred "
In striking contrast, is the history (recently summarized by Bertram, 1950)
of the related Northern Fur Seal (Otaria ursina), which provides an outstanding
example of rational and successful management of a natural resource, with a
well-balanced regard both for business and conservation of stock. A vast herd
was discovered breeding in the Pribilof Islands by a Russian expedition in
1786. By 1910, mainly as a result of pelagic sealing, numbers had diminished
from several millions to 200,000 survivors. In 1911 an International Fur Seal
Convention signed by Great Britain (acting for Canada), the United States,
Japan and Russia, agreed to prevent pelagic sealing; the herd was given a
complete rest for several years; and now the annual harvest amounts to about
70,000 skins taken from a herd which has increased tenfold, to more than
2,000,000 in forty years. Thus, this happy result of enlightened management
(of an animal which like the crocodile is in part a fish-eater) has been achieved
despite serious conflict with fishery interests and difficulties inherent in obtain-
ing co-operation of the different nations concerned.
It is not my intention in this brief report to offer any detailed recommenda-
tions directed towards crocodile conservation, but rather to suggest that the
time has surely come to reconsider crocodile policy in UgAnda, having regard
to such control of the cropping rate as will afford the highest sustained yield of
hides. Effective measures of control will not be easy to formulate or enforce.
If the belly-skin industry is to be saved some thought must be given to breeding
stock. This may require restriction or local prohibition of hooking and snaring
-methods of capture which allow of little control over the size of the quarry
taken. Again, the species is slow to come to maturity. Crocodiles begin their
reproductive life when about eight feet in length, or perhaps in about their
eighth year. Thus, areas that have been overworked would need a prolonged
period of rest from further disturbance. A close season at important breeding
grounds during the two annual breeding seasons could be envisaged, together
with the prohibition of egg destruction: such measures would of course involve
a reversal of the former policy of the Game Department.
All such restrictive measures would be highly unpopular with the concerns
and individuals who have come to look upon every live crocodile as a potential
bank-note, and whose incentive is greed for immediate gain and quick returns
in cash, without thought for the future of the crocodile itself, or even for the
trade in its skin.
It is the verdict of history that an industry based on a wild animal resource,
if uncontrolled, soon becomes uneconomical. One might here mention the
widely recognized need to control the world's whaling effort, as now achieved
12 UGANDA JOURNAL
through the International Whaling Commission. The former belief that nature
is inexhaustible is no longer held by enlightened administrators; yet the present
need for an effective policy of crocodile control is evident. The war of extermina-
tion cannot last many more years; and those principally concerned will be the
first to repent the virtual disappearance of an animal that was once a valuable
source of profit.
Anon. (1952). Crocs, are useful in food fish dams. South Afr. Shipping News and
Fishing Industry, 7, 75.
Bertram, G. C. L. (1950). Pribilof fur seals. Arctic, 3, 75-85.
Graham, M. (1929). The Victoria Nyanza and its fisheries. A report on the fishing
survey of Lake Victoria, 1927-1928. London.
Hippel, E. V. (1946). Stomach contents of crocodiles. Uganda J., 10, 148-9.
Hornaday, W. T. (1913). Our vanishing birdlife. New York.
Jardine, J. (1929). A note by the Hon. Game Ranger; 17 July 1929.
Kinloch, B. G. (1951). Annual report of the Game Department for the year ended
31st December 1950, 1-110. Entebbe: Govt. Printer, Uganda.
Laws, R. M. (1953). The seals of the Falkland Islands and Dependencies. Oryx, 2,
Pitman, C. R. S. (1940, 1948, 1950). Annual reports of the Game Department for the
years ended 31st December 1939, 1-28; 1946, 1-82; 1948, 1-74. Entebbe: Govt.
Salmon, R. J. D. (1932). Annual report of the Game Department for the year ended
31st December, 1931, 1-11. Entebbe: Govt. Printer, Uganda.
Swynnerton, G. H. (1946). Report of an investigation of the fisheries of Lake Rukwa.
Weddell, J. (1825). A voyage towards the South Pole-in the years 1822-3. London.
Welman, J. B., and Worthington, E. B. (1943). The food of the crocodile (Crocodilus
niloticus L.). Proc. Zool. Soc. London, 113, 108-112.
Worthington, E. B. (1929). A report on the fishing survey of Lakes Albert and Kioga.
Photo by Hugh B. Cott
Jaluo hunters with their catch, Namasagali.
Crocodile poachers' camp, Igwe Island, Lake Victoria.
Photo by Hugh B. Cott
[face p. 12.
MACKAY'S CANOE VOYAGE ALONG THE WESTERN
SHORE OF LAKE VICTORIA IN 1883
By SIR JOHN MILNER GRAY
THE following account of Mackay's boat voyage on Lake Victoria in 1883
originally appeared in the Proceedings of the Royal Geographical Society
for May 1884, and the writer wishes to acknowledge his gratitude to that Society
for permission to reprint a portion thereof in the Uganda Journal. It was sent
by Mackay to his father, who was a member of the Royal Geographical Society
and passed it on to the Society for publication.
The voyage was not nearly so exciting as Mackay's first voyage across the
Lake in 1878 in company with the Rev. C. T. Wilson, when the Daisy had to be
beached on the western shore in a storm and was so damaged that any Lloyd's
agent would have declared her to be a constructive total loss. Not so Mackay,
who, with Wilson's aid, set to work and in the course of two months cut out the
middle section of the boat and joined the two ends together. But the story of
the voyage of five years later is full of interest to modern readers and deals with
travelling under conditions which it is difficult in these days to imagine.
Thus we learn from Mackay's Journal that seventy years ago the Lake was
used by Arab dhows for the transport of slaves and that Mutesa's canoes raided
as far as the southern shores of the Lake.
We also learn that to him must be given the credit of discovering not only the
bay at Bukoba, but also Emin Pasha Gulf, the credit for which Stanley, some-
what typically, made a point of ascribing to himself.
Mackay's object in making the journey in 1883 from Buganda to the south
end of Lake Victoria was to assemble the Eleanor which had been brought up
in sections to Msalala in Smith Sound. The Daisy had by that date become
unseaworthy and the building and the launching of the Eleanor was a matter of
vital importance. Numerous obstacles and disappointments held up the building
of the Eleanor, but eventually she was launched in the early days of Decem-
This was not Mackay's last effort at ship-building. In 1887, whilst at Usambiro,
he was commissioned to assemble the sections of what he hoped would be the
first steam vessel to sail on Lake Victoria. But once again the obstacles and
disappointments were innumerable. Many of the sections had been more than a
dozen years coming up from the coast to the Lake and thanks to rough treatment
and exposure were almost hopelessly out of shape. In August and September
1889, Stanley and Emin Pasha spent close on three weeks with Mackay at
Usambiro on their way to the coast. According to Emin, Stanley once saw fit to
reproach Mackay for the slow progress which he had made with the steamer
(Uganda J., 14 (1950), p. 219). But Mackay the missionary had other things to
do besides boat-building, many of which Stanley the explorer could hardly
14 UGANDA JOURNAL
appreciate. A few weeks after Stanley and Emin left Usambiro, Peres Schynze
and Girault of the White Fathers' Mission, stayed with Mackay for one night on
their way to the coast. Pere Schynze tells us that they arrived on 5 October 1889,
and were received by Mackay in the most friendly manner. His companion
Deekes was stricken down with fever and Mackay had to do the honours.
Mackay showed them all the work which he was undertaking, including the
steamer which he told them he hoped to launch in eight months' time. Next
day he accompanied his guests part of the way on the next stage of their
journey (Schynze. Mit Stanley und Emin Pascha durch Deutsch Ostafrika,
pp. 9-10). Except for his companion Deekes these were the last Europeans he
was to see. For four months more the work went on. Then, on 3 February 1890.
Mackay was suddenly stricken down with fever and five days later the end came.
He was buried not far from the workshop where he had been working only six
days before. In 1926 his earthly remains were taken to Uganda and reinterred
at Namirembe, but the ship's boiler, upon which he had been working until the
day before his last fatal illness, still marks the place where first he was laid to
rest (S. Napier Bax. Early C.M.S. Missions ... in the Mwanza District. Tangan-
yika N. and R., No. 7 (June 1939), pp. 54-5).
THE BOAT VOYAGE'
June 27, 1883.-I left the Mission Station at Natete (1), in Bu-ganda (Uganda).
Slept for the night at Kyikibezi, only 10 miles distant and near Murchison Bay.
June 28.-Marched another good dozen miles to the plantation of Sebagoya
(in Sebukule's country) near Naambwa Hill.
June 29.-Reached Mugula's (2) capital, Ntebe, after some eight miles' march.
June 30.-Got men and loads distributed among the canoes, and embarked
in fair weather. After a long pull reached Bunjako, at a point near Sale Island,
late in the day.
July 1.-We were late in starting. The lake was rough and the weather very
hazy, but it became smoother as we got under the lee of Sese. Put in for dinner
at my old camp in Sese, among wild palm-trees. Re-embarking at 4 p.m. we
paddled till after dark, hoping to reach Bujaju, but put into a cove in a small
lumpy island between Sese and mainland, and found here Sungura's boat at
anchor, en route for Usukuma (3). The crew were in huts ashore; they have a
cargo of ivory and slaves in stocks (4). The sailing boat had been eight days
from Ntebe to this, while we have covered the distance in two. We cut brush-
wood to clear a space for my tent which we pitched by the light of my lantern.
The outer awning of my tent I rig up separately for my men to sleep under, as
it is cold.
July 2.-At dawn it blew a cold north wind, and the dhow took advantage
of it by hoisting sail and getting under way. My tent was alive with biting brown
ants, which only fire and hot ashes will drive away. We embarked at length,
intending to make for the opposite shore of Bujaju, where we must buy some
earthen pots for cooking. Wind unfavourable, so we strike south, while the
I Reprinted from: A. M. Mackay, Boat Voyage along the Western Shores of Victoria
Nyanza, from Uganda to Kageye; and Exploration of Jordans Nullah. Proc. R. Geog.
Soc., New Series, Vol. VI, May 1884, p. 273.
MACKAY'S CANOE VOYAGE ON LAKE VICTORIA IN 1883 15
dhow held on her way to Dumo. By 3 p.m. it got rough. We soon landed on an
open beach near Mbroyaga, the country seat of Mungobya (5), who has just
returned from a war in Karagwe, whither he was sent with an army, to put a
certain grandson of Rumanyika's on the throne, and plunder some other
claimant; he had returned with a large booty of cattle, women, and slaves as is
the custom in Bu-ganda.
Soon after encamping I set off with some men to see the "general", whom I
knew well, and expecting at least to get some plantains from him. His place is
three miles from the shore on the face of a high hill. He received me in state,
with beating of drums and "present arms". After he had rehearsed his exploits,
and I had complimented him, he presented me with two good head of cattle, a
gourd of beer, and as many bunches of plantains as my men could carry.
July 3.-Put off in calm and dense fog; soon were off Dumo, where we met
Nambigya returning from Kyigaju's (6) with four canoes and a present of ivory
from Kyigaju to Mtesa.
In a wide bay south of Dumo it became so hazy that we lost sight of land. The
boatmen could not steer as the sun was invisible. I took out my compass and
directed them. They took it for a charm or idol, and the chief of my canoe
begged for it that he might carry it to a sick child of his whom he had left at
home. He would recover at once! They followed the direction of the compass
more from their faith in its virtues as a charm of divination than as a scientific
instrument. Finally, we reached the promontory of Sango, where we camped
and made huge fires to dry our beef to preserve it for the voyage. Half-roasted
and half-smoked in large pieces, it keeps from ten to twenty days. The extreme
moisture of the air in Bu-ganda and Bu-zongora causes meat to smell in two,
or at most three days, while in U-sukuma the air is generally so dry that a
carcase will keep a week if hung up in the open.
July 4.-Passed the mouth of the Kagera. All this part, and in fact the whole
coast of the lake, needs correct mapping, which I hope to be able to do when
our boat is afloat. Stanley's map is merely a sketch, and very, very far from
accurate. It will be a matter of many months to map the whole lake with any
degree of accuracy. But it must be done, even for our own sake in sailing,
especially at night.
Crocodiles of enormous size guard the mouth of the Kagera, and are regarded
by boatmen as possessed of the spirit of the river-god. This word Kagera must
be distinguished from Kagera, which is the name which the Ba-ganda give to
Smith Sound, between Mwanza and Rwoma's.
We put ashore in Kayoza's territory (7) and set off to call on Mugula (8), who
is encamped a few miles from here on his way back with thirty canoes from
Jangiro (9). As we mounted the hill it came on to rain very heavily, and in spite
of my umbrella I got rather wet, whilst those with me were drenched. It was
very cold the day afterwards, and the poor fellows got severe catarrh. After the
weather cleared up we stumbled up and down on a rocky slope in a plantation
for more than an hour, looking for Mugula's quarters. We found him at last: he
received me cordially, and provided me with abundance of plantains and some
fowls, besides huts in which to sleep. But we had arranged to return to our
canoes, which we had ordered to join us round the point. They rounded two
16 UGANDA JOURNAL
points, however, and were well nigh inaccessible by land. I sent messengers to
find them, and after dark one canoe returned for me. We reached camp in
Kayoza's bay at a late hour.
July 5.-Day broke with thunder, and it rained most of the forenoon. The
rocks here have a very stalactitic appearance. A stream of semi-fluid lava has
metamorphosed all the strata, forcing its way between them in most cases and
tilting them to the vertical in many places. In other lower parts it has flowed
over the surface, and where it reaches the lake has been scooped out by the
waves, thus forming caves of which the roof has been bored through by the
rain in many cases. These holes are dangerous, for, being hidden by the long
grass, they form so many pitfalls in walking. It is the uplifted metamorphosed
strata that have been weathered away to points that present the stalactitic appear-
ance. These are often in groups on knolls, and look not unlike village church-
yards, with various sized dilapidated tombstones.
We embarked at 3 p.m. hoping to reach Makongo, but did not succeed. Put
in at sundown by low rocks in a bay next to one where Wilson and myself
spent two months repairing the Daisy nearly four years ago (10). It blew cold
all the early part of the night, but, happily, there were no mosquitoes. Almost
no firewood to be had.
July 6.-We got early afloat in calm. Soon heavy rain appeared at sea to the
east, moving southward. As it was getting rough, we put ashore on the lee side
of the island Musira (11), off Makongo. Here was a very large camp, probably
made by Mugula and his fleet. A sharp storm of rain came from the north-east
With a struggle we got the tent up and the goods under cover. When it cleared
up, the wind as usual changed to south, and a high sea rolled in between our
channel and the mainland. But we could not spend the night in this filthy camp,
so we risked crossing the channel, and by the mercy of God reached a cove on
the mainland safely, although the waves were very high, and all but swamped
our frail canoes. This place lies between Makongo and Bu-bembe, and is called
July 7.-It was calm overnight, but at dawn came on a squall with sparkling
rain. It was very rough outside, the sea running high all day. We lived on
spur-winged geese, of which I shot a brace yesterday at Musira Island. The
Wa-ngwana eat these, as also the Ba-sese canoe-men (who eat in fact every-
thing), but the Ba-ganda decline. The captain of our fleet, Kabona, here fell
sick. I sent my man to the neighboring villages to buy a stock of food for the
days ahead when we camp only on uninhabited islands all the way from
Bu-bembe (Kaitaba's) to Kageye. They brought in a few bundles of beans,
exactly like French beans; also coffee-berries, fowls, and plaintains, but every-
thing dear, as the Ba-ganda warriors have devoured much of the natives'
July 8.-It was calm this morning; so we embarked, hoping to fetch Bu-bembe
promontory. When we got near Kisaka Island, just opposite the promontory,
it began to blow hard, and we had to go ashore at the point at the north side of
the first bay. This was a vile place for camping in, but we got some shade from
a large tree. My boy, Sambo, brought me some resin from a tree in the neigh-
bourhood, exactly like red sealing-wax, only somewhat soft.
MACKAY'S CANOE VOYAGE ON LAKE VICTORIA IN 1883 17
July 9.-The night was calm until dawn when a gale arose from the south:
with thunder at sea. Day bright, but sea very high. Took a round of bearings
from Alice Island to Bu-bembe, with prismatic compass (13).
July 10.-It was calm again at night, but about 3 a.m. it began to blow hard
from the south-east. During the day it was slightly overcast, wind and sea high.
We planned to start in the afternoon if the wind fell, as our supply of food was
getting low, and the young moon would serve us for several hours after sun-
down. Many of the Ba-bumbire are here. Their canoes are rough models of
those of the Ba-sese, but workmanship very inferior. They venture out in very
rough weather with them, which no Mu-sese would risk on any account. Wind
and sea gradually fell towards afternoon, and by 4 p.m. we made a start. Pulled
along the channel and coasted along the west side of Bumbire (14) hoping to
reach the small wooded island, Lubili. The moon set, however, by the time we
got as far as Mayiga Island. We put in at a wretched place of bush and rocks;
and well it was that we did so, as soon after the wind rose and blew hard all
night. We set fire to the dry grass on the top of the islet, and all night our camp
was brightly illuminated. On the mainland they doubtless thought us to be of
July 11.-Bright all day, but high wind till afternoon. Our cove faces the open
sea to the east; hence the breakers did not subside enough to let us launch the
canoes, the place being besides very rocky. In the evening I climbed to the highest
point in the centre of the island, and took a round of bearings with the prismatic
compass. The summit of Soswa was barely visible with the binocular and quite
invisible with the naked eye. There is a deep bay to the south-west with many
islands in it, which require care in surveying (15). All the mainland opposite
this, as also Bumbire, Iroba, Muzinga Islands, present bare bluff ranges of rocks
and stones, but not so craggy as the coast from Bu-bembe north to the Kagera
mouth. Between that again and Bu-ganda the shore is low and even wooded.
July 12.-We are still prisoners here. Night again calm, but the wind rose as
usual with the sun; only today it is more moderate, and slightly more easterly,
being S.E.E. In the evening we succeeded in launching the canoes, empty,
through the breakers, and sent them round to the lee side of the island, where
we found a recess among the rocks just large enough to let us load the canoes
one by one. Made for Lubiri Island, almost due south, reaching camp about
7 p.m. Felt drowsy and out of sorts, as if an attack of fever was near.
July 13.-Soon after midday, the weather being moderate, we embarked for
the pull across the wide reach to Soswa, off to east, but not visible till we got
about half way across. A bilious attack rendered me unfit to sit up or read.
Thank God for fair weather in this dangerous reach! We arrived at Soswa in
dead calm about 9 p.m.; found plenty of huts, but some hippopotami about the
landing place, among ambatch trees, made us active in our movements.
July 14.-Sea rough outside. The boatmen were too fatigued to start today.
Before crossing the bay of yesterday, the Ba-ganda and Ba-sese called into
service all their most potent charms. On embarking, they put some bananas on
a paddle, and throwing them into the water, offered up a prayer to the lake god,
Mukasa (16), "Oh, Lubare, come and take this offering to thee, and grant us that
we may reach the other side in safety!" Not a few of them perish, however/
18 UGANDA JOURNAL
Their canoes cannot stand a heavy sea, and are easily swamped. Whole cargoes
of ivory are often thrown overboard. The boatmen themselves can seldom
swim, and are, besides, terribly afraid of crocodiles, although very fond of the
flesh. One crocodile's tail alone will purchase two goats! I believe the tail is
nearly all fat, something like the white layer under the hippopotamus's skin.
The Arabs are just as superstitious as these poor heathen. Before crossing Soswa
Bay they select a small tusk of ivory, and, holding it up that all the boatmen
may see, throw it into the lake to propitiate the deity.
July 15.-Still down with fever. We embarked at dawn, and halted at noon
on a low rocky island for breakfast. I took an emetic which did not do me much
good. At sundown reached the now uninhabited island of Kulu, off Kyigaju's
coast (17). The island with its neighbour, To, as also Soswa, Lubiri, Irioba, etc.,
have been all devastated by the Ba-ganda (18).
July 16.-Our food all but exhausted. Crossed over to mainland, at present
friendly, but formerly very hostile, in the hopes of being able to get or buy
plantains. After much palaver with the natives (Wa-zinja), the local Mulangira
(prince) made a present of two bunches of plantains and a pot of beer, which
latter delighted the hearts of the canoe-men. The natives themselves were all
drunk. Their beer, made of banana juice with the ferment of bulo (or bulezi), a
very small grain like canary-seed, is much stronger than that ordinarily drunk
by the Ba-ganda. We soon left; but the sea was far from calm, while the boat-
men paddled lazily. We encamped for the night on a deserted rocky island
(Lwa Mulangira), off the coast of Rwoma's country. Rwoma is at present hostile
to Bu-ganda (19), and the natives on the mainland seeing our fires, and fearing
that we might be a war-party, beat their drums all night. This was perhaps the
most wretched camp I have ever seen. I could find a level spot nowhere large
enough to pitch the tent upon. The land was all a ruin of pointed rocks and
stones, half-hid with grass. I spent a sleepless night, taking two full doses of
July 17.-Feeling most wretched and sick all day. Put ashore at a low bamboo
island called by the Ba-ganda Maua, hoping to find firewood, but saw none. At
dark we drew up our canoes near the petty village, in the country of Mwanza,
where Sungura has established himself. As this place only some two hours from
Kageye, I wrote a note to our brethren there, sending it over by two Wa-ngwana.
We obtained here some milk and gruel of matama which revived me not a little,
being the first food I had been able to take for a week.
July 18.-Early up, and soon reach Kageye.
(On 26 July Mackay and the Rev. Cyril Gordon set off for Msalala by land.
On 31 July Mackay notes that there is "no milk, as Mirambo carried off all the
cattle some six years since".
On 9 August they reached the end of open water in Smith Sound. "This is
evidently the furthest point reached by Smith in the Daisy.")
A letter from Mackay dated "Head of Smith's Creek, Victoria Nyanza, S. lat.
3", 10 December 1883, and addressed to the Royal Geographical Society,
MACKAY'S CANOE VOYAGE ON LAKE VICTORIA IN 1883 19
appears on pp. 282-3 of the Society's Proceedings for 1884. It contains the
"You will be interested to hear that the Church Missionary Society have
launched their vessel the Eleanor on Victoria Nyanza. I expect to start across
the lake with her to Bu-ganda in a day or two, and hope by her means to be able
to make an accurate survey of the whole coast. I have already surveyed much
of the western side. Stanley's charts are wonderful for the short time he had at
his disposal, but extremely inaccurate so far as I have been able to test them.
Judging by the west side, the survey made by him of the east.shore must be very
"The only accurate survey I have seen of any portion of Victoria Nyanza is
that made by Lieut. Shergold Smith, of the Church Missionary Society, of
U-kererwe, Speke Gulf, and what is called Jordans Nullah."
(1) Natete was the C.M.S. headquarters until expulsion of the missionaries by the
Muhammadans in 1888, when the buildings built by Mackay were destroyed.
(2) Mugula of the Mamba clan had charge of the canoe Direnjogera. He lived at
Buira near Entebbe (Kagwa. Mpisa, p. 292). He was a sub-chief of Makamba, Sabadu of
Busiro (Roscoe. Baganda, p. 253).
(3) Sungura alias Songoro (the Rabbit) was a half caste Arab and was found by
Stanley at Kageyi, when he arrived there in February 1875. His dhow was seen in a half-
finished condition by C. T. Wilson of the C.M.S. on Ukerewe Island in March 1877. It
was then said to have been five years building. Shergold Smith, the leader of the expedi-
tion, purchased it with a view to employing it, when complete, for transporting the
Mission's heavier goods (The Victoria Nyanza Mission, p. 45). Eventually Lukonge, chief
of Ukerewe, claimed a lien on the dhow for the price of the timber used by Sungura in
constructing it. In the dispute which followed Shergold Smith, his companion O'Neill, and
Sungura were attacked and killed by the inhabitants of Ukerewe. After this tragedy the
Mission seems to have abandoned all claim to the dhow. Thereafter Sungura's kinsfolk
appear to have resumed possession of it. It was always known as "Sungura's dhow" in
contradistinction to a larger dhow built by Said bin Seif, which was known as "Kiponda's
dhow". Both dhows were destroyed near Entebbe on 2 September 1889 by the Baganda
Christians in their war against the Muhammadans (Uganda J., 14 (1950), pp. 38-40).
(4) In the Story of the Life of Mackay of Uganda, p. 245, Mackay's sister says: "The
poor victims, having endeavoured in the darkness of the night to break loose, were
re-secured by their owners, which caused much bitter wailing."
(5) Rumanyika, the ruler of Karagwe, died in 1878. The eldest of his sons, Kakoko,
was debarred from the kingship because he was left-handed. He managed, however, to
put certain of his brothers as his puppets on the throne. Rumanyika's brother, Rwegira,
managed to drive out Kakoko and his nominees; Kakoko thereupon appealed to Buganda
for aid. Mutesa sent Ngobya, who was Kajerero (Sabagabo) of Buddu, to his assistance
(Kagwa. Basekabaka, p. 153). Ngobya defeated Rwegira at Kabwera, near the bend of
the River Kagera, and drove him out (Ford and Hall. History of Karagwe, Tanganyika
N. and R., No. 24 (Dec. 1947), pp. 11-12).
(6) Kyigaju was the chief of the island of Kome lying off the southern shore of Lake
Victoria. He appears as "Kijagu" in Stanley. Through the Dark Continent, i, pp. 246, 268-9.
His island must be distinguished from that of the same name in the Sese archipelago.
See also Notes (15) and (17).
(7) Kayoza was chief of the district of Bugabo just to the south of the Kagera in what
is now Tanganyika. He was of Muhinda descent (Kollmann. Victoria Nyanza, p. 17;
Stuhlmann. Mit Emin Pascha ins Herz von Afrika, p. 711; Rehse. Kiziba Land und Leute,
(8) See Note (2).
(9) Ihangiro is a district of Kiziba. For the history of Ihangiro cf. Rehse. op. cit.,
pp. 285-92, and Stanley. Through the Dark Continent, i, pp. 277-95.
(10) It was more nearly five years ago. When proceeding from Kageyi to Buganda in
the Daisy, Mackay and the Rev. C. T. Wilson were forced into this bay by stress of
weather on 28 August 1878. They had to beach the Daisy to save her from foundering and
as a result the boat was badly damaged. They repaired the damage by cutting out the
middle section of the boat and joining the two end sections together. It was not. until
22 October that they were able to launch the vessel again. During their enforced stay
Kaitaba of Busongora treated them most hospitably (Wilson and Felkin. Uganda and
Egyptian Soudan, i, pp. 242-5; Mackay of Uganda by his sister, pp. 94-7).
(11) For an account of this island see Sillery. Musira and its Burial Caves, Tanganyika
N. and R., No. 13 (June 1942), pp. 57-8.
(12) I think Mackay must be given the credit for discovering this bay.
(13) During his enforced stay in this bay Mackay wrote the letter to his father which
appears in Mackay of Uganda by his sister, pp. 238-40.
(14) In a letter to Dr. Kirk, dated Rubaga, 9 April 1881, the Rev. P. O'Flaherty men-
tioned that in the course of his voyage from Kageyi to Buganda, "the people of Bumbire
Island, where Stanley fought, refused to allow my people to land, driving them off with
poisoned arrows, being afraid of the Waganda who rob and plunder them" (Kirk to Earl
Granville, 27 July 1881. Slave Trade. No. 1 (1882), pp. 195-6).
(15) To Mackay must therefore be given the credit of having discovered the Emin
Pasha Gulf. Writing on 11 November 1889, to the British Consul-General at Zanzibar,
Stanley gave himself the credit for the discovery in the following words: "We have made
an unexpected discovery-of real value in Africa-of a considerable extension of the
Victoria Nyanza to the south-west. I was so certain in my mind that this fact was
known through the many voyages of the Church Missionary Society missionaries that I
did not feel particularly moved by it. Mackay, however, showed me the latest maps
published by the Society and I saw that no one had even a suspicion of it." In dealing with,
and effectively disposing of, Stanley's claim to be the first European to set eyes on the
snow peaks of Ruwenzori (Uganda J., 2 (1934-5), pp. 249-50), H. B. Thomas says "the
incident is typical of Stanley, whose abilities as a publicist were hardly surpassed by his
great achievements as an explorer. When Stanley was in command he saw to it there was
little limelight for his assistants."
(16) For a full account of Mukasa and his cult see Roscoe, The Baganda, pp. 290-301.
For the custom of propitiating water spirits before venturing across the water see Frazer.
Folklore of the Old Testament, ii, pp. 414-20.
(17) Kulu is to the north of the island of Kome, of which Kyigaju was chief.
(18) Writing to Kirk from Rubaga on 9 April 1881, at the conclusion of his voyage
from Kageyi to Buganda, the Rev. P. O'Flaherty of the C.M.S. said: "The beautiful and
numerous islands have had their inhabitants all massacred or taken into slavery by Mutesa's
forces for assaulting his men when they wished to land. We used to encamp on these
beautiful, but now desolate, islands. Certainly there are now no inhabitants to molest the
progress of the traveller or stranger. The Waganda never buy plantains, which are their
constant food. They cut down bunches with their matchets. The inhabitants flee from their
houses for woe to the women, if they remain" (Slave Trade. No. 1 (1882), p. 196).
(19) Rwoma was paramount chief of Buzinja on the southern shores of Lake Victoria.
Kollman (Victoria Nyanza, pp. 108-9) tells us that "Rwoma was not a popular ruler, and
injured the country by almost incredible measures. For example, as a sign of mourning
for the death of his son Rwikama, the cultivation of the fields was forbidden for six years.
Later on he league himself against the German rule with several other chiefs on the Lake,
and fell in battle against Lieutenant von Kolben in the autumn of 1895." Kyigaju's present
of ivory to Mutesa (see entry of 3 July 1883) was evidently made with the object of obtain-
ing the assistance of the Baganda against Rwoma. Early in January 1884, Gabunga Kaya
was sent with a fleet of canoes to assist Kyigaju in a rebellion against Rwoma. According
to Mackay the Gabunga "found his force insufficient, and is on his way back, having
fought only some islands off Roma's coast". According to accounts given to Stuhlmann
the Baganda carried off 8,000 cattle. Sir Apolo Kagwa tells us that the Gabunga died on
the return voyage. In 1887, however, the tables were turned and the Baganda assisted
Rwoma in an attack on Kyigaju (Mackay of Uganda, pp. 246-7; Stuhlmann. Mit Emin
Pascha, p. 125; Kagwa. Basekabaka be Buganda, p. 123).
ACHOLILAND IN 1897
By SIR JOHN MILNER GRAY
A S an addendum to my Acholi History, 1860-1901, I send a copy of a report
written in 1897 by Mr. Michael Moses. It is interesting not only because it
gives an indication of the state of affairs in the Acholi country at that date, but
also because it is written by a prominent and highly respected member of the.
community, who is happily still with us.
When Mr. Moses wrote his report, there were no fixed rules of orthography
in regard to local names. The spelling of them was very much a matter of the
personal choice of the writer. I have therefore ventured to insert in brackets.
after certain names as written in the letter the accepted modern spelling, but
let me hasten to say that at that date Mr. Moses' spelling of the names came
far nearer to the mark than did that of his superior officers.
A few prefatory words regarding the situation in the Acholi country in 1897
may not be out of place.
Kabarega, Mukama of Bunyoro, had been driven out of his own country by
Lieutenant H. J. Madocks at the end of 1895. In 1897 he was known to be
moving about in the Acholi and Lango countries. Some, but by no means all,
of the local chiefs were friendly to him.
A military post had been established at Mahagi in April 1894 by Captain
A. B. Thruston, and a treaty made with Tukwenda (Tukenda), the nearest
important local chief. Later, Mahagi and the whole west shore of Lake Albert
was ceded to the Congo State.
Yet another treaty had been made by Major Ternan in 1896 with the local
chief, Abura, who lived at Patira on the east bank of the Nile, a few miles to the
south of Pakwach. A year later Ternan persuaded Abura to pay a visit to'
Kampala and Entebbe. The following report was made with regard to that visit
by Captain (afterwards Lieutenant-General Sir) William Pulteney on 22 Novem-
ber 1896: "The Shulis (Acholi) state that Abura, their chief, who went with
Major Ternan to Kampala, told them not to bring food (sc. for the Sudanese
garrisons) until he returned, that if we treated him well, and he was satisfied
the English had plenty of guns, they meant to throw in their lot with us, drive
out Kabarega, and trade."
"Await" in "Mr. Moses' letter is the Acholi chief, Awich of Payera, whose
story has been told by Bere (1947) and Anywar (1948), but I think the statement
in Mr. Moses' report that in 1897 Awich endeavoured to enter into friendly
relations with the British is entirely new. Reports received at about this date by
the officer commanding the troops in Bunyoro suggested that Kabarega had
made himself unpopular with the Acholi and it may well be that Awich made
this approach to the British because he wanted to be rid of an inconvenient
guest. Apart from this he may well have thought that the times indicated the
desirability of having some sort of foothold in both camps.
On 31 March 1897, Lieutenant W. R. Dugmore, who was in command of the
troops in Bunyoro was informed by the Sudanese garrison at Mahagi that they
22 UGANDA JOURNAL
had had to go to the aid of Tukwenda, who had been attacked by a combined
party of Acholi and Madi raiders, led by Abura. The Sudanese claimed to have
driven the raiders off with the loss of 150 men. Dugmore sent Mr. Moses to the
fort at Fajao, near the Murchison Falls, to get in touch with Abura, and obtain
his explanation of the affair. The following is Mr. Moses' report:
LIEUTENANT W. R. DUGMORE, 16 April 1897.
In compliance with your instructions I proceeded to Fadjao (Fajau), paid
and rationed the troops there up to the end of this month, and obtained the
On my arrival at Fadjao I immediately sent for Aburra to meet me there.
He was away at the time and was a couple of marches away, but after a reason-
able time he came over and made the following statements.
A few days ago the Shuli (Acholi) under himself, Pagwan the chief of Pakwas
(Pakwach), late Ayera's son, and the Wadelei under Shk. Ali's son went on a
raiding expedition to Allaro to the north west of Tickenda's (Tukwenda's)
village to punish the people for having killed some Shuli (Acholi) that had gone
there for trading purposes. The report that they had gone for the purpose of
raiding Tickenda is entirely false, as they would not harm any of the chiefs that
are friendly to the English. Tickenda, seeing them encamped near his village,
at once ran to the fort at Mahagi, telling the Effendy there that he was attacked.
Thereupon the Effendy sent out some men that were fired at by his party by
mistake, not knowing of their being soldiers until it was too late. The loss on
their side was six killed and one wounded, whilst the soldiers on our side lost
two killed and one Snider rifle with 30 rounds captured. He says that they are
all very sorry for what has happened, and the Shuli (Acholi) king Await (Awich)
has ordered the chief Rabeja (Lubega) in whose hands the Snider has fallen to
at once return it to the Yuzbashi at Fajao.
Kheiralli, in handing me over two tusks of ivory, said that one of them was
presented by Await (Awich), the big Shuli (Acholi) chief, as a mark of friend-
ship (about a week before the attack) with a particular desire of being called to
Fadjao (Fajau) when a European officer happened to go there, when he would
give him a good many more tusks.
I am sorry to say that I could not see Await (Awich) as he lives (far) off (from)
Fadjao (Fajao) and would require at least two weeks notice to present himself.
The Effendies at Fadjao (Fajao) speak well of Await (Awich) and other Shuli
(Acholi) chiefs and are of the opinion that they are friendly.
The Belgians are said to be in great numbers and have already established
themselves in four forts, namely, Muggi, Madi, Abbo, and Lutuka.
Aburra also reports that Kabrega, mistaking the Belgians for Arabs or
Dervishes, sent messengers to them offering ivory and cattle to help him fight
the English, but the Belgians imprisoned his messengers and confiscated their
ACHOLILAND IN 1897 23
Kheirelli Effendy says the Belgians are raiding right and left, showing little
Kabrega is reported to be at present at Atiga between Dufileh (Dufile) and
Fatiko (Patiko) and living with a chief called Abbusale.1
I am, Sir,
Your most obedient and humble servant,
Shortly afterwards Captain Madocks was ordered to proceed on a boat
expedition to Dufile in order to ascertain what he could about Congolese move-
ments on the west bank of the Nile. Messengers were sent to Awich to meet
Madocks at that place. Later instructions were given to Madocks not to pro-
ceed farther downstream than Wadelai. Unfortunately there was no opportunity
to notify Awich in time of this change of plan. How far-if at all-this failure
to make personal contact with Awich at this date affected his subsequent atti-
tude to the British Government it is difficult to say, but it would appear very
unfortunate that the proposed meeting never took place.
Anywar, R. S. (1948). The life of Rwot Iburaim Awich (translated by J. V. Wild).
Uganda J., 12, 72-81.
Bere, R. M. (1947). An outline of Acholi history. Uganda J., 11, 1-8.
Gray, Sir J. M. (1951-2). Acholi history, 1860-1901. Parts 1-III. Uganda J., 15, 121-43;
16, 32-50, 132-44.
I "Abusala" was the name by which the Sudanese called the Acholi chief Ogwok of
Padibe, but the report here given does not receive corroboration from any other source.
MASAKA HILL-AN ANCIENT CENTRE OF WORSHIP
By E. C. PLANNING
M ASAKA HILL (1) (not of course to be confused with the District station
of Masaka) is in Buganda on the north bank of the Katonga River, close
to the border with Toro. It is 15 miles distant in a straight line from the ancient
settlement of Ntusi (2) which lies south of the river. It is probably as old as
Ntusi. Eleven miles downstream, also on the south bank, are the earthworks of
Kagago, Bigo and Kasonko (3).
The summit of Masaka Hill (4,210 ft.) is crowned with a large copse including
fig and dracaena trees, and affords an excellent view over the papyrus-filled
river, over undulating hills to the west and north, and over plains of grassland
and swamps to the east. To the south the hill's rocky slopes fall away steeply
to the river's edge. In the surrounding countryside buffalo are plentiful; smaller
game, even zebras, roam close to the river bank. The general area, known as
Nabubale, is sparsely populated by pastoral people, the rulers being of a hybrid
Munyoro-Hamite stock; there are also a few groups of agricultural people of
Not surprisingly the legends and traditions of this area, the Nabubale of old
which was once a part of Bwera (an important principality and at times even
an independent kingdom), have their roots deep in pre-Babito days, in the
times of the Bachwezi incursion (4) of many centuries ago. The pastoral over-
lords of the area, on both sides of the river, though grazing their cattle widely,
have dwelt there for many generations (5). It is from them that the ancient
legends can still be heard, legends wrapped around the deified invaders of old.
Certain hills, places and earthworks are related to-have been built or occu-
pied by-Bachwezi heroes such as Ndahura, Wamala, Mugenyi and others.
Masaka Hill, in the first place, is spoken of as a residence of, at least, the three
above-named leaders (6). Secondly, in legends on both sides of the river, it
comes into prominence when the country is said to have been evacuated by
the Bachwezi. Then they left the area in the care of a man called Kihesi of the
Bamoli (bushbuck) clan who some say was (as the name implies) a smith, others
a chief, but who, most story tellers agree, hailed from the region of Bunyoro.
To this man the Bachwezi are said to have given a drum made from the skin
of a waterbuck; this drum was called Rusama (7). Kihesi, with the drum,
remained in Nabubale close to the foot of Masaka Hill (8).
The advent of new rulers, the Babito, found Kihesi at Masaka, His connec-
tions with the departed and, so it is said, much admired Bachwezi, and his
claims to the territory entrusted to him, resulted in his independence being
respected (9). The hill, because of its associations, soon became a place of
reverence and after the death of Kihesi it became more and more a place of
worship. As the leaders who it was believed had once resided there became
deified, pilgrimages to the summit of the hill began to be frequent (10). From
MASAKA HILL-AN ANCIENT CENTRE OF WORSHIP 25
then onwards, Masaka Hill has continued to be a place of worship of individual
Traditionally the hill has always been solely the responsibility of the Hima
head of the bushbuck clan (who is also the ruling prince, Omulangira, of Bwera);
and the Bairu, who dwell at its foot, have always on appointment by the clan-
head, provided guardians and servants for the enclosure at the summit.
DESCRIPTION OF THE SITE
The flat top of the hill is clothed with a heavy growth of trees and shrubs
which suggests some past occupation. Careful examination soon reveals portions
of an earth rampart, about 6 feet wide, enclosing a circular area about 180 feet
in diameter (Fig. 1). Further search reveals a second earth rampart which is
concentric with the outer one on the west and joins it on the east and south-east.
There are six 'openings' or gates in the outer circle. When visited in 1952, pots
of some antiquity lay broken in a shaded grove in the centre of the whole
A reconstruction of the site has been made possible with the aid of the grand-
son (12) of Nanzigombe, the last head of the bushbuck clan to hold an active
interest in the hill, and of the present-day guardian (named Wamala) of the
enclosure and his assistants.
In Nanzigombe's day the enclosure was still kept in good repair. In a large
hut at its centre was housed the ancient drum of Kihesi, Rusama. There also
were kept spears, smaller drums, and two ceremonial tusks of ivory, of which
one is said to have been of great antiquity. Between the inner and outer enclo-
sures, on the river side, was a hut in which dwelt the only two keepers who were
allowed to remain permanently on the site. The enclosure could only be visited
with the permission of the clan-head. Each of the six gates had its name and
special use (Fig. 1):
(1) Mugabente (The Giver of Cows).-The main gate (facing east); the
entrance for gifts of cattle, and for pilgrims and others not from the
(2) Mununuzi (The Saviour).-A small gate for use by people who came to
seek guidance in matters of a judicial nature.
(3) Kangabo (The Gate of Many People).-The entrance for local people and
(4) Kabona (The Sacred Gate).-The entrance for the keepers.
(5) Mugabuzi (The Giver).-A general exit only.
(6) Rwensinga (The Gate of Spears).-Both entrance and exit for the keepers
and other attendants.
Outside the main enclosure, beyond the gate Rwensinga and close to the edge
of the hill-side, was a small hut (the hut-circle, 6 feet in diameter, is still dis-
cernible). Here were kept about eight small drums. These drums would be taken
inside the enclosure and beaten at the end of ceremonies.
The drum Rusama is said to have been beaten last by Nanzigombe himself
26 UGANDA JOURNAL
in the 1880's as a necessary adjunct to the ceremonies which took place within
At times one of the tusks lay across the threshold of the hut in the centre,
facing the gate Mugabente. It played as essential a part in ceremonies as the
drum Rusama. The second tusk rested inside close to the treasured drum,
flanked by the spears and other drums.
Enclosure on Masaka Hill, Buwekula County, Buganda.
MASAKA HILL-AN ANCIENT CENTRE OF WORSHIP 27
As far as is known there were two forms of ceremony. One would be held by
.the clan-head himself whenever there was a full moon, whilst the other form
was for the pilgrims who came to the centre for their own purposes.
In the first form of ceremony use was made of certain vessels set out in a
special order inside the centre but in front of the drums and spears. These were
six earthenware pots and receptacles which were placed in pairs. All had their
special uses. The smallest receptacles (13) (Fig. 2, A) were tended by either a
youth or maiden who kept a smoky fire burning in them throughout the time
the clan-head remained in the hut. The use of the largest receptacles (Fig. 2, B),
which are identical to the fumigators used in the preparation of bark cloth, has
not been fully clarified; they are considered to have been designed as fumigators
in the first place but are believed never to have been used for any purpose other
than ritual. Slices of flesh from the clan-head's animal-offering were thrown on
the glowing embers inside the remaining pair of pots (Fig. 2, C); there the meat
was burnt to cinders. Other pieces of flesh were thrown around the grove by
attendants. What remained of the offering was later eaten by the keepers.
Local pastoral people as well as Bahima from farther afield and pilgrims
from Ankole, Koki and Bunyoro, were frequent visitors to the centre. They
came to make offerings in many causes, such as for fertility or rain, for matters
connected with birth, marriage and death, for cattle, or for the cure of disease.
A gift of some white animal would be essential, usually a sheep but, in certain
cases, depending on the importance of the person, one or more cows. The
animals would be brought first into the enclosure by the main gate and later be
taken outside again to special pens. On certain occasions the clan-head himself
would officiate instead of the head keeper-who dealt only with the less impor-
The ceremonies would take place at dawn. The suppliant, walking at the right
hand of the officiating priest, would be taken into the hut of Rusama. On his
entrance a curtain, cutting off the centre of the interior, would be raised by
attendants to reveal the drum Rusama, other drums and spears. Clapping his
hands, the priest would kneel, the suppliant behind him following suit. The
priest, on both knees and facing the drum, would then intone the suppliant's
wants. Rising, he would then cross over to the drum and, with it on his right,
would face the suppliant and his people. Next, attendants would come forward
carrying spears of which the priest would select one, and then cross over to the
side of the hut to seat himself on a stool. There he would remain in silence for
some time. Later he would rise, bow to the drum and then approach the still
kneeling suppliant whom he would touch on the forehead with the fist in which
he clenched the spear, thus bringing the ceremony to a close. Gifts of animals
would be slaughtered for the feasting, drumming and singing that would follow,
except that cows would not be killed but kept alive for their milk.
Masaka Hill is still revered but the sacrifices and invocations of the past have
long given way to the observation of a single rite-in respect of fertility. Women
still wend their way to the silent grove to leave offerings of coffee beans wrapped
in banana leaves close to sherds said to be fragments of some of the clan-head's
.28. UGANDA JOURNAL
-ceremonial vessels. Others, apparently in some extension of the rite and under
the guidance of the head-keeper, bury their offerings among the roots of a
particular fig tree. It appears that this tree has been singled out because the iron
haft of a spear is buried at its foot. When dug up for my inspection this rusted
haft had the appearance of great age. Its purpose either was not divulged to me
or has been forgotten in the mists of the past. Suffice it to draw attention to
certain objects, apparently grave goods, excavated at Ntusi settlement in 1952;
amongst the contents of an inverted pot were found portions of the decomposed
haft of a spear (14).
The head-keeper, invariably clad in clean barkcloth, is the only one who still
,observes the tradition of never entering the enclosure unless in fresh or new
apparel. He is said to be the only person who knows the hiding place of the
remaining elephant tusk which, he and his assistants assert, is a relic of great
antiquity having been in the possession of the Bahima of this area long before
the arrival of the Bachwezi who were either superseded or absorbed by the
In 1888, when that part of the country was the scene of fighting between
Muhammadan and Christian, the local rulers and their people had perforce to
withdraw. Nanzigombe died. On their return to the hill his sons found the
,enclosure looted and the drum, Rusama, the other drums, the spears, and one
,of the tusks, gone. Very little had been saved (16).
It was eventually discovered that one of the Muiru attendants (called Bakitte)
at the enclosure had stolen a tusk and traded it somewhere to the north in the
vicinity of Mubende Hill. He was also suspected of having disposed of the drum
Rusama. Although held prisoner by Nanzigombe's sons and frequently beaten
he refused to talk. Today nothing is known of the fate of this drum though
some do believe that it found its way in mysterious circumstances to the sacred
centre at Mubende Hill. In this connection surviving servants of the last
Nakaima (the priestess at Mubende Hill) definitely speak of the existence at
Mubende of a sacred drum which they call Rusama. It is believed by them to
have been destroyed when the priestess was absent in Toro during the time of
the religious upheavals. It has not yet been possible to ascertain how the drum
came into the keeping of the priestess.
OTHER SITES NEARBY
As Masaka Hill provides considerable evidence of ancient occupation it is
tempting to seek elsewhere in the neighbourhood for other signs; these are not
Three miles north of Masaka Hill, on the top of Kinonye Hill (17), there is a
circular pit some 45 feet in diameter. It is about 10 feet deep and is heavily
-overgrown. The origin of the pit is unknown but the name suggests that chalk
(enoni) has been dug from it. Amongst people familiar with the region, it is
passed off as being the handiwork of the Bachwezi.
Some distance east of Masaka Hill, in the vicinity of the River Mukeberya, a
A B C
Copyright reserved Photo by Department of Public Relations, Uganda
The ceremonial vessels from Masaka Hill.
Photo by E. C. Lanntne
The elephant tusk from Masaka Hill.
[face p. 28.
MASAKA HILL-AN ANCIENT CENTRE OF WORSHIP 29
tributary of the Katonga, lie six circular, apparently man-made, water holes (18).
Five of them, varying in diameter from 15 to 25 feet, lie in a rough east towest
line at distances of 150 to 250 feet apart. The sixth lies at the extreme east,
slightly north of the line. When I visited the area some of the holes held water
and showed signs of being as much used by game-including hippopotamus
from the Katonga-as by the Bahima and their cattle. These water holes call to
mind the series of built-up pools-bwogero--reported in Masaka District (19).
Though these water holes have been used by countless generations of cattle
herds, their origin is unknown and whether or not for lack of a better idea, they
too are credited to the Bachwezi.
(1) Lat. 0* 13' 30" N.; long. 31 4' 20" E. From the main road on the Mubende side of
the causeway it is a 9-mile trek west across river and swamp.
(2) See Lanning (1953), p. 51.
(3) See Lanning (1953), p. 56.
(4) For an interesting summary of the legends and theories about these people see
Trowell and Wachsmann (1953), p. 24.
(5) Evidence of this is provided by the existence of 'tombs' in this area. As opposed
to the Hima custom of burying the common dead in dung heaps in the centre of their
kraals, the rulers are buried separately in graves some 25 feet deep which are then sheltered
by huts. This is, of course, very similar to the Bunyoro form of royal burial. With the
passage of time many of these tombs have been abandoned by their keepers, the huts have
fallen to pieces and the places have been forgotten. A few such tombs do survive, however,
particularly at Kaboma, in Masaka District, a few miles south of the Katonga, where there
is a group of six huts cared for by a Muiru keeper. Another is to be found at Makole near
to the Masaka-Mubende road, and is the burial place of a once powerful Muhima princess
Nyakaishika. In Nabubale, in Mubende District, is the tomb of a Muhima prince,
Busweeta, who is said to have lived in the reign of Kabaka Kamanya (who died about 1831).
All these tombs are of personages of the bushbuck clan.
(6) Sir Apolo Kagwa (1901) says that Ndahura had two residences, which he refers to
merely as having been 'in Bwera' and 'at Masaka'. Gorju (1920) also mentions Bwera
and Masaka in this connection but, according to his map, his 'Masaka' would be north-
east of Mubende Hill, some 40 miles north-east of Masaka Hill. Investigation on the spot
has, so far, produced no evidence that Gorju's position for 'Masaka' is correct.
(7) Masaka District Minute Paper H/8 and Gorju (1920), p. 75.
(8) An almost identical story is told of Ruhinda, the gate-guard, to whom was given
what now comprises Ankole District; cf. also the story of Bukuku, gate-guard to King
Isaza, left in charge of Kitara (K.W. (1935), pp. 156-8).
(9) Bwera remained unsettled until 1900 when it was incorporated in Buganda by the
Uganda Agreement. The events leading up to its absorption are worthy of note.
It was in 1898 that Kabaka Mwanga, fleeing from Buddu, entered Bwera. With Captain
Sitwell on his heels he encamped on Bukongote Hill, south of the Bigo earthworks. Having
already engaged the enemy once Sitwell finally came upon Mwanga's rear guard on 4 March
1898 as it was crossing the Katonga and engaged it. The elusive Mwanga had, however,
already fled north, and Sitwell had to be content with entering in his diary for that day
"Mwanga reported to have crossed by Nanzigombe's Crossing" (C. H. Sitwell, Uganda
Diary, 1895-1899. MS. Secretariat Library, Entebbe). Following this action Sitwell's
Baganda levies overran Bwera which, from that time, ceased to be an independent state.
(Masaka District Minute Paper H/8).
(10) See Gorju (1920), p. 75.
(11) See Lanning (1953), p. 59.
(12) Omwami J. L. Kagugube (latterly of the Uganda Medical Service), grandson of
Nanzigombe, hereditary Omulangira of the area Nabubale (Mubende District) and a
portion of present-day Mawogola (Masaka District), as far as Lwentale. Muntu, a brother
of Nanzigombe, was recognized as ruler of the portion of Bwera south of Lwentale includ-
ing a small part of Buddu. Muntu, said by some members of his family to have been the
30 UGANDA JOURNAL
'11th ruler of Bwera' was succeeded by his son Petero Mbugano. As a result of the
absorption of Bwera by Buganda, the latter's successor, Yosefu Njojo of Nambirizi (died
1953), held no official status but has been considered by his branch of the family as being
a ruler of Bwera de jure if not de facto.
(13) One of these has survived (Uganda Museum specimen: 52. 199). Of baked pottery,
it resembles the igicumbi used for ritual purposes in Ruanda (see Bequaert, 1949, p. 25).
(14) See Lanning (1953), p. 59.
(15) See Crazzolara (1950), p. 101.
(16) Those known to have been saved are:
(a) Three ceremonial vessels, said to have been in use when Nanzigombe assumed
responsibility for Masaka Hill (Fig. 2. A, B, C) (Uganda Museum specimens:
52. 199; 52. 200; 52. 201).
(b) A pot, used by the enclosure attendants for drinking beer. (Uganda Museum
specimen: 52. 198.)
(c) 1 elephant tusk (Fig. 3); retained at the site.
(d) Haft of spear; retained at the site.
(e) A large and unusual type of clay bead or amulet. (Uganda Museum specimen:
(f) A spear said to be in the possession of Omwami M. Lubega of Mawogola,
(17) See map Africa 1:250,000, sheet North A-36/T Mubendi (1911).
(18) Lat. 0 13' 0" N.; Long. 31" 6' 0" E.
(19) See Williams (1946), p. 67.
Bequaert, M. (1949). The Masaka cylinder: an interpretation of its use. Uganda J., 13,
Crazzolara, J. P. (1950). The Lwoo; Part I Lwoo migrations. Verona: Missioni
Gorju, J. (1920). Entre le Victoria l'Albert et I'Edouard. Rennes.
Kagwa, Sir A. (1901). Basekabaka be Buganda na be Bunyoro na be Koki na be Toro
na be Nkole. Kampala.
'K.W.' (1935). The kings of Bunyoro-Kitara. Uganda J., 3, 155-160.
Lanning, E. C. (1953). Ancient earthworks in western Uganda. Uganda 1., 17, 51-62.
Trowell, M., and Wachsmann, K.P. (1953). Tribal crafts of Uganda. Oxford Uni-
Williams, F. L. (1946). Myth, legend and lore in Uganda. Uganda J., 10, 64-75.
THE BAGANDA AND THE BAKONJO
By ROLAND OLIVER, PH.D.
A MONG the historical traditions of Buganda there are two rather strikingly
similar stories of regicide and its consequences. The first of these concerns
Kabaka Kayemba, the brother and successor of Juko, whose reign may be dated
with fair certainty to the closing years of the 17th century. It tells how Juko,
wishing to liquidate Kayemba, ordered a group of boat-builders to make a
canoe of clay for the use of his brother, who was due to lead a raid against the
island of Buvuma. While Kayemba was marching down to the port of embarka-
tion, however, he passed through Bweya, at Mile 11 on the Entebbe road, and
there he overheard some young goatherds singing a ditty the words of which
aroused his suspicions. He examined the canoe which had been provided for him
and wisely elected to travel in another. Later, when Juko died and Kayemba
become Kabaka, the boat-builders and all their kin fled for their lives westwards,
it is said, to the shores of Lake Edward. The second story is the celebrated one
of Kabaka Semakokiro, nearly a century later, who, having employed some
members of a kitongole known as the Bakunta to murder his brother Junju, after
his own accession made the ungrateful decision that the assassins must neverthe-
less suffer the penalty for regicide and arranged for them to be massacred at a
regimental feast. Fortunately, Semakokiro's ugly secret leaked out, and the
Bakunta, like the boat-builders of Juko, fled westwards through the then
independent principalities of southern Toro and north-western Ankole to Lake
Edward, where their main settlement in Bunyaruguru is now well known. Bishop
Gorju, writing in 1920, reported that the defendants of the Bakunta, while in
many respects indistinguishable from their Banyampaka neighbours, spoke of
the Creator by the Luganda name Katonda, and claimed to have introduced the
banana and the bark-cloth tree to the district.
In December 1949, while on sabbatical leave in Uganda, I made the aquain-
tance of Mr. M. E. Lind of the Seventh Day Adventist Mission at Mitandi, Fort
Portal, who first drew my attention to the existence of traditions among the
Bakonjo which seemed to indicate a connexion with Buganda. In January 1950
my interest was further aroused when I visited the Bakunta of Bunyaruguru.
There Abudara Baraba, a Mukunta of the Mpologoma clan who had worked for
some years in Buganda, gave me the following information. The Bakunta fled
first to Bwera, whence they were driven out by an army sent by Kabaka
Semakokiro. They then moved westwards into Kitakwenda, where some of them
settled. The main body, however, moved on to the neighbourhood of Kashenyi
on the Kazinga Channel, where some of their descendants are still to be found,
being known as Abayabindi, apparently from a type of loin-cloth called
omubindo which was worn in Buganda in the time of Semakokiro. From
Kashenyi a part of the group, consisting mainly of members of the Mamba,
Nsenene, Kobe and Nkata clans, went to Ruwenzori and became Bakonjo.
Others crossed the Kazinga Channel southwards. Of these some, drawn mainly
from the Mpologoma and Nkima clans, settled in Bunyaruguru; some, from the
32 UGANDA JOURNAL
Mbwa, Lugave and Nkata clans, settled in Igara; some went on to Kigezi. Of the
Kigezi group some, said Baraba, went to Urundi, while others returned to Lake
Edward, near the Rutshuru river, where they become Abagabu.
A week or two later I met in Kigezi a Mukunta of the Nkima clan who, though
himself a native of Bunyaruguru, had travelled among the Rutshuru Bakunta,
whose movements he described as follows. From Katwe they went round the
north and west of Lake Edward, through a place called Changwe in the Beni
district, then past Lubero, through the Buhimba hills, over a mountain called
Kabasa, and so to Rutshuru. Here, he said, they encountered the Bahororo, who
enslaved many of them, so they kept close to the Lake and went up to Bunyaru-
guru, leaving behind them small groups at several points on the lakeside, whose
descendants are now known as Abagabu.
Meantime, I had already received from Mr. Lind yet another account of the
flight of the Bakunta, this time from an aged Mukonjo, living on Ruwenzori, ten
miles from Mitandi. The Bakunta, this said, used to live near the king's palace
in Buganda. But one day they made trouble and killed a king. They were forced
to flee the country, and did so, passing through Ankole. Afterwards they reached
a place called Kirimanamoto. After staying there for a while they reached another
country called Rutshuru. In Rutshuru they greatly increased their numbers,
women bearing eight to ten children each. They therefore sent two men, Rubango
Rugwanasolo of the Babito or Basukali clan and Kisenge of the Basita clan, to
look for another country. These men found Isale and Bukobi, which the Batoro
call Bukenya. There Rubango became their first king, and it was he who brought
with him from Buganda the royal trumpets-Amakondere. When the Bakunta
lived in Buganda they used to wear loin-cloths called omubindo. A short time
after their arrival in Bukobi this land also proved too cramped for all the Bakunta.
There was a great famine, during which many died. So they set out for other
places, some going to Busongora, and others up the great mountain. Another of
Mr. Lind's informants said that when they lived in the Congo they had a king
called Kengere, whose father was Tengetenge. I read this account to a gathering
of Bakunta in Bunyaruguru, and the place-names mentioned were readily identi-
fied by them. Kirimanamoto, they said, referred to the Mfumbiro mountains.
Isale and Bukobi were two neighboring hills on the western side of Lake
Edward, near the exit of the Semliki. They knew of Kengere as a chiefly name
in that district.
Having one day to spare before my next assignment, I telegraphed to Mr.
Lind, who very kindly invited a gathering of Bakonjo to meet me at his house at
Mutandi. There, in the very few hours between the double journey from and
back to Kichwamba, I recorded on a Wirek dictating machine three accounts in
Lukonjo from very old men, which are now available for study at the School of
Oriental and African Studies in London. The first tells, in very similar terms to
those of Abudara Baraba, the story of the flight of the Bakunta from Kyagwe.
The second is a typical aetiological legend of two eponymous ancestors, Kaganda
and Kakonjo, who quarrelled and went to live in different countries. The third
is a garbled version of the story of the clay canoe. Of the existence of a body of
Konjo tradition, distinct from that of the Bakunta and the earlier refugees from
Buganda, there was, at least in this one gathering of Bakonjo, no trace at all.
THE BAGANDA AND THE BAKONJO 33
Now what do all these strands of evidence amount to? Obviously, the 20,000
Bakonjo of Ruwenzori, the very much larger body of Banande whom they left
behind in the Beni and Lubero districts to the west of Lake Edward, the Abagabu
of the lakeshore and the 4,000 Bakunta of Bunyaruguru (not to mention those
who settled in Igara and Urundi) cannot all be the descendants of perhaps 500
Baganda who fled at different times from the wrath of Kayemba and Semakokiro.
Besides, it is well known that both cultural and linguistic classifications trace the
affiliations of the Bakonjo and the Banande rather more towards the forest
peoples of the eastern Congo than towards the lacustrine Bantu of Uganda. In
fact it would seem to be a reasonable conjecture that the masses of the present
Bakonjo and Banande descend from one of the most ancient immigrations of
Bantu-speaking peoples into the Lake Edward region. And yet in the field of
oral tradition, it is the descendants of a handful of Baganda refugees, the latest
and numerically the least important immigrants into the Konjo-Nande society
who, at least on a superficial view, dominate the scene. One can imagine how it
happened. The Baganda refugees, in the one case skilled boat-builders and
fishermen, in the other members of a well-drilled and well-armed kitongole,
arrived among the comparatively primitive people around Lake Edward, perhaps
with their banana seedlings and their tools for making bark-cloth. Naturally,
with all these superior skills, they became important people among the Nande-
Konjo of 200 and 250 years ago. Moreover, they brought with them the Kiganda
habit of teaching their children about the past, of handing down a corpus of
tradition from one generation to the next. They became the historians of the
Banande and the Bakonjo, the Banyampaka and the Bagabu, and, for the distant
past, they imposed their own traditions on the less historically-minded nations
of their adoption-with the result that today even someone who, like Mr. Lind,
lives among the Bakonjo, can learn only the traditions of the minute Kiganda
element in the population. Here in fact, as in so many other parts of Africa, it
seems that tradition is not the tradition of the mass of the people, but rather that
of a comparatively recent and culturally superior minority. It is none the less
interesting for that.
AFRICAN PERFORMANCE ON AN INTELLIGENCE
By JOHN McFIE, M.D.
(Uganda Medical Service)
Congenital ignorance is the condition of intelligence. Aldous Huxley.
THIS paper describes the performance on a modern intelligence test of a small
group of Uganda student nurses. I present it in this manner, rather than as a
more formal scientific paper, because I think that in the first place the results may
have a particular local significance; and in the second, because any conclusions
based on so restricted an investigation must necessarily be tentative, and their
discussion should perhaps be confined more to those who are well acquainted
with the people concerned, than to specialists in psychological techniques.
Intelligence testing is a well-established psychological technique, and one
in the perfection of which we have reached a landmark-have touched, some
may think, a nadir-in our materialistic evolution. In fact, in England, this type
of test enables us to assess an individual's innate ability almost as we might
compare calculating machines or wireless sets; to measure his "problem solving
power" more or less in isolation from influences of schooling and social position.
To many, this may seem a rejection of moral attributes, a deplorable denial of all
that is meant by "personality"; but the fact remains that these tests prove so
valuable when all sorts of decisions about individuals have to be made-about a
child's schooling, in vocational guidance, and even in medical diagnosis-that
they do appear to be valid instruments of measurement.
So much, as I have said, in England. In Africa, however, the situation is more
complicated; and, to introduce the various difficulties which arise, I cannot do
better than quote from Dr. S. Biesheuvel's book on "African Intelligence", in
which are discussed many of the problems which must be dealt with in a work
of this kind. He observes that a score achieved on intelligence tests must be
regarded "not as a direct measure of innate ability, but as a measure of hereditary
potentiality as it happens to have been realized by specific environmental circum-
stances". Clearly then, it is only when the environments of the subjects are fairly
uniform that one can be sure that differences in scores represent differences in
inherited ability alone. In other words, using an English test in England, where
we can be sure that acquired incentives, schooling, and spare-time activities are
not widely different, we can also be fairly sure that test scores will depend upon
the individual's innate potentialities. In Africa, using a test devised for applica-
tion in England, performance will obviously depend to an enormous extent on
cultural factors like language, motivation, attitude towards the test, and so forth.
It is for this reason that some anthropologists have used these tests, in the
expectation of distinguishing differences of upbringing and attitude rather than
differences of individual ability; and some points of interest in this connexion
have, I think, come out of the present investigation.
AFRICAN PERFORMANCE ON AN INTELLIGENCE TEST 35
On the other hand, the steady advance in Africa of European organization and
technology makes it relevant to consider whether or not Africans can be assessed
in terms of their ability to undertake a job of this or that degree of complexity
within these new spheres.1 I should have made it clear in the foregoing remarks
that no direct comparison can be expected, from these techniques, between the
innate abilities of African and European individuals; the immense cultural
separation precludes this; but it might prove possible to make valid distinctions
between the abilities of individual Africans if suitable tests were devised.
Furthermore, the present tests are not intended to be in any way a final selec-
tion. They were merely intended to explore the possible range of useful tests;
and after describing them, I propose to discuss first the significance of the scores
which were achieved, and then to describe some of the differences in types of
performance which I noticed.
THE TESTS USED
Before considering the tests in detail, I should say that intelligence tests fall
into two broad categories: the "group" tests, generally in the form of a special
sheet or booklet of problems, on which the subject records his answers in pencil,
and which may be given to whole classes of children or platoons of soldiers at
the same time; and "individual" tests, which are designed to be given to one
subject at a time, and generally involve asking him questions and setting him
certain specified performances. Group tests involve considerable reliance upon
the subjects' understanding of the instructions, understanding the test material,
and having an appropriate attitude toward the test. These would be well-nigh
impossible to ensure in an African group unless it had received a considerable
amount of specific instruction, and furthermore the use of a group test makes it
difficult to study individual idiosyncrasies or different types of performance.
Clearly, the maximum amount of qualitative information is to be gained from
the individual type of test-an opinion in which perhaps I am biased by my own
experience, which has been entirely with this latter type of test.
An intelligence test which has proved very useful in this connexion is an
American one, known as the Bellevue Scale. It is composed of a number of short
tests, each consisting of a series of questions or problems of a uniform type; thus
there is an arithmetic test, an information test, a constructional test, a pictorial
test, and so on. The particular point about these tests is that the majority of them
represent a particular type of intellectual operation, rather than involve learnt
material, and therefore it seemed to me that they might easily be rendered in a
vernacular language if necessary. This will be apparent when we consider the
tests themselves, which follow:
1. Comprehension. This test was intended more to prepare the subject for
subsequent tests than as a measure of a specific ability. In the original, many
questions were obviously dependent upon Western cultural experience, but I
omitted some and changed others, leaving six questions of the type: "Why
would most people rather have a car than a bicycle?" With a maximum of 2
points for each answer, the maximum total was 12.
1 A subject also discussed by Biesheuvel (1952).
36 UGANDA JOURNAL
2. Similarities, involving answering: "What is the same?" about pairs of
words, like dog and lion, air and water. Excluding some unlikely pairs and
altering others, left seven pairs. With a maximum of 2 points for each answer,
the maximum total was 14.
3. Analogies. This test involved an opposite logical process to the preceding
test. The subject was required to complete a series like "Brother, boy: sister,
......" There were some obvious difficulties in translating these into Luganda,
but nine questions were given, 1 point for each correct answer.
4. Arithmetic. The questions nearly all involved small purchases, like "If a
pound of sugar cost 50 cents, how many pounds could you buy for 1/50?" There
were eight questions, scoring 1 each for the first six, and 2 for the last two, giving
a maximum of 10 points.
5. Picture-description. This was inserted, largely in preparation for the suc-
ceeding test, and to see what the subjects' actual interpretation of pictures might
be. Four photographs of local African life were made available by the Depart-
ment of Information; and the subjects' descriptions were marked according to
their comprehensiveness. With a possible 3 points for each picture, the maximum
total was 12.
6. Picture-arrangement. This involved arranging sets of pictures, each to tell
a story. In the original, many of the pictures were of Western events, and for a
while I thought I should have to provide a series of pictures of African events.
Fortunately, however, the Department of Information came to my rescue again,
and enabled me to select some "Kapere" stories from Mawulire. I chose ones
which did not seem to depend for their comprehension on the captions, and cut
these off before presenting each group of four pictures, disarranged, to the sub-
ject, with the request to arrange them to make up a story. I showed how to do
one, as an example; but even then the first story was not as easy as is the first
story in the Bellevue test. There were four stories, with 3 points each, giving a
possible total of 12.
7. Block-designs. The subject was required to copy with coloured wooden
blocks a series of coloured designs presented on cards. No changes were
necessary from the original, and with six designs scoring 3 points each (1 point
was given if the first attempt required correction) the maximum total was 18.
In the original method of scoring, additional points were given for speed in
doing items of the last two tests. For reasons which will be discussed later, this
was not done in this version: the maximum total for the whole group of tests was
therefore 87 points.
My general aim throughout the tests was to avoid, as far as possible, the
creation of an artificial atmosphere, and to try to adapt the manner of testing to
the attitude of the subject. If a question was not understood in one form, I
rephrased it-without altering its content-in the endeavour not only to make
it intelligible, but also to make the proceedings more like a conversation and less
like an examination; and, in the performance tests, I explained, with demonstra-
tion, the nature of the performance required. When interpretation into Luganda
was necessary, the interpreter sat at the table also-and soon showed herself
capable of "giving the tests" in just the way I aimed at.
AFRICAN PERFORMANCE ON AN INTELLIGENCE TEST 37
The choice of subjects is of course all-important in a study like this. Entirely
uneducated ones might be found not to understand the tests at all, and might
give answers difficult to interpret, let alone to score, whereas ones too sophisti-
cated might produce performances in no way characteristic of the "average
African". For this sort of "pilot" study a "betwixt and between" group seemed
most desirable, and I was fortunate in the help given to me by the Sister Tutor
at Mulago Hospital in arranging for 30 student nurses to be tested; and also in,
the assistance of the nurse who acted as interpreter. An additional advantage
of these girls as subjects was that they had already been exposed to Europeans
as instructors and examiners, so that the present situation was not one likely
to inspire fear or suspicion. In fact, my impression was that the girls took quite
an interest in the proceedings, and did their best with the strange material put
As the study was not intended to result in "an intelligence test for Africans" it
is not necessary to go into details about the social and economic levels from
which the girls come. The majority of them were Baganda and Batoro; but the
Bakyiga, Banyankole, Acholi and Bagishu were also represented. The girls' ages
ranged from 16 to 20: all had completed their primary schooling, and some had
also been to secondary school. The general impression was that they formed a
pretty "average" group of subjects, neither exceeding in brilliance nor in dullness;
and therefore of course not at all representing a "cross-section" of the population
as a whole.
THE RESULTS: SCORES
In commencing the discussion of results with some arithmetical considera-
tions, I hope not to give the impression that these are the most important ones.
They are, however, an integral part of the evaluation of any testing procedure;
and, did they not tend to confirm the validity of the present tests as estimates of
intelligence, the subsequent review of individual performances would have no-
more than trivial significance. Nevertheless, those in whom the mere suggestion
of "figures" induces a sensation of despair may take it that the scores recorded
are entirely consistent with the assumption that this is in fact an "intelligence
test", and may pass on to the section on performance.
As stated above, the total possible score on the whole group of tests was 87: of
the 30 subjects tested, the highest score was 77, the lowest was 29, and the
average (mean) score was 54. The fact that no subject achieved either full marks,
or a zero score, means that the group of tests was neither too easy for the
brightest nor too difficult for the dullest of the subjects-in other words, that its
overall level of difficulty was reasonably well suited to the group tested- while
the range of scores, nearly 50 points for 30 individuals, suggests a fair chance
that the scores would be "spread out" according to the differences in subjects'
The actual way in which they were "spread out" is shown in Table 1, which
records the number of subjects within each group of scores. This arrangement,
or "distribution", of scores, with a few high and a few low and most clustered
38 UGANDA JOURNAL
round the average, is just the distribution which is usually found in intelligence
test scores. The fact that all those tested through an interpreter are represented
at the lower end of the distribution may indicate that the test is more difficult if
given under those conditions; but it must be mentioned that these were also the
girls who had had least secondary schooling, and further, that they were the
youngest ones of the group. Their scores might therefore be expected to be among
the lower ones, even if the tests were as easy in Luganda as they were in English.
Distribution of total scores.
Each letter represents one subject: E, those tested in English and L, those in Luganda.
E E E E
E E E E
L L E L E E
L L L L L L L E E E
Below 30 30-34 35-39 40-44 45-49 50-54 55-59 60-64 65-69 Above 70
But perhaps the most important question which may be asked about this
group of tests is "How can you tell that they are in fact testing (that is, measuring)
intelligence?" In England, there is ample circumstantial evidence that this sort
of test does in fact discriminate validly between persons of different innate
intellectual ability, but at present there is no such confirmation available for the
validity of this group of tests.'
I should like to suggest, however, that part of the answer may reside in the fact
that in England these are in fact recognized as tests of intelligence rather than of
learning; and, provided that the problems posed do not involve too much learnt
material, there is no reason why the same should not apply to Africans. Another
piece of evidence we have seen already, in the similarity of the distribution of
the total scores to what would be expected of such a distribution in England
(the total score is assumed to be the best estimate of general ability, composed
as it is of measures of performance on problems of very varied types). And yet
another point of similarity, and I think an even more striking one, arises if we
consider the relationship between the score of each subject on all the tests.
This relationship, known as the correlation of the scores, may be expressed as
the answer to the question, "How frequently do those who get high scores on one
test also get high scores on another test?-and on the whole group of tests?"
The answer to this question may in fact be expressed by a measure known as the
coefficient of correlation, which is zero when there is no relationship, and 1-00
when the relationship or correspondence between the subjects' scores is perfect.
In other words, when the coefficient between tests A and B is 1-00, then the top
subject on test A also gets top marks on B, the second on A is also second on B,
and so on down to the bottom subject, who is bottom on both tests; and when this
is the state of affairs, it is very likely that both tests are measuring the same thing,
I I learn that, of the girls who took this test, those who achieved the lowest marks have
since failed their examinations.
AFRICAN PERFORMANCE ON AN INTELLIGENCE TEST 39
or at least that their scores are both completely determined by the same factor.
When the coefficient of correlation is 0, no such relationship can be assumed.
The coefficient of correlation between scores on every pair of tests, and
between each test and the total scores, are set out in Table 2. The first point of
interest is that all the test scores are in fact interrelated to a considerable degree
-even tests so apparently different in content as Arithmetic and Analogies, and
Picture-arrangement and Block-designs (although the Similarities test has rather
low correlations)-a feature characteristic of normal English intelligence test
scores, and a feature which suggests that the common factor accounting for all
this intercorrelation is in fact general intelligence. With the present group of
subjects this factor was clearly not something external to the test, like influence
of schooling, similarity of tribe, or similarity of language used; for all these
were varied among the subjects. The only point in common was the relative
unsophistication of the girls, the fact that their experience had probably not led
them to diverge along lines of special ability (like "being good at arithmetic" or
"clever with one's hands") with the conclusion that this common factor was none
other than their general intelligence, none other than the undifferentiated ability
which they called into use to deal with every type of problem.
Coefficients of correlation between test scores.
Analogies -21 -07
S Arithmetic -42 -30 -77
description -48 -03 -61 -44
F-' arrangement -17 -27 -27 -36 -34
Block-designs -31 -24 -38 -48 -56 -65
Total score -59 -43 -60 -71 -70 -71 -88
Following up this argument, I think the low correlations found with the
Similarities test can be explained on the grounds that it represents perhaps the
one "ability" (the ability involved is essentially the grasping of abstract relation-
ships) which the subjects may have had occasion to exercise during the course
of their normal development (after all, even ordinary conversation involves
frequent discussion of such relations as "same" and "opposite", and Africans
are not less addicted to conversation than are the English) and which may there-
fore show evidence of development independently of "general" intelligence. The
higher correlations observed with the Block-designs test can, by the same token,
be explained by the total novelty of the situation, and by the assumption that the
problems were solved solely by the exercise of the general intelligence available.
Moreover, as this test shows the highest correlation of all the tests with the total
40 UGANDA JOURNAL
scores, it follows that, if one were restricted to the use of a single test on this
group of subjects, the Block-designs test would provide the best estimate of their
relative general abilities.
THE RESULTS: PERFORMANCES
Having observed earlier that the subjects showed no reluctance or hesitation
in attempting the tests, and even showed a considerable amount of persistence
when faced with difficult problems, we can consider the performances noted
on the individual tests.
In the Comprehension test no particular difficulty was encountered, and the
subjects failed questions in much the same way as do English subjects-by
inadequate explanation and by frank guessing. Unexpected answers to the
question "Why would most people rather have a car than a bicycle?" included
"Because in a car you keep out of the sun". This was clearly appropriate to local
conditions, and was therefore marked correct.
The Similarities test likewise presented no difficulties, and failures were of
the same kind as are found in England.
The Analogies test proved hard to render into Luganda, as often (as in
day, light: night ....... ) nouns had to be translated by verbs, thus altering
the force of the question. Nevertheless, the type of problem did not appear to
present any difficulty to the subjects.
Arithmetic produced just the same expressions of surprise and dismay as
come from English subjects; but, as no girl scored less than 3 and one achieved
full marks, it was clear that arithmetical operations were not peculiarly diffi-
cult for the subjects. I noticed a tendency with some girls to do multiplication
by the subterfuge of repeated addition; but English subjects also resort to this
if they have not learnt their "tables" thoroughly.
In the Picture-description test, no subject held the pictures upside-down,
or failed to recognize at least individual features of the pictures. The pictures
were, however, photographs of a high quality, and represented scenes with
which the subjects were all more or less familiar.
The Picture-arrangement test, by contrast, was composed of sets of small
pictures which were essentially line drawings with little shading, and indi-
vidual scenes were frequently puzzling to the subjects; but the same difficulties
are found with English subjects on the Bellevue pictures. A more striking
irregularity was observed in two cases, in which the girls included in their
arrangement a picture which happened to be "lying on its side" (i.e., rotated
through 900) although this did not affect their correct interpretation of the
The main point of interest, however, is that the performance of these
subjects argues strongly against a recent analysis of "African thinking" which
has come from the neighboring territory of Kenya (Carothers, 1951): in that
article the author concluded, as a result of a number of anecdotal and sub-
jective impressions of African behaviour, that the African suffers from a
deficient function of the fore-part of his brain. This conclusion, with its
implication of mental defect, which the same author has amplified consider-
AFRICAN PERFORMANCE ON AN INTELLIGENCE TEST 41
ably in his latest publication (Carothers, 1953), is one which should clearly
not be advanced without the strongest evidence in its support; and we find in the
ability of these subjects to do the Picture-arrangement test that this support is
not forthcoming. For evidence has accumulated, on both sides of the Atlantic,
that ability to do this test is reduced by interference with the function of the
fore-part of the brain-as may result from operation or disease; yet, in spite of
the rather crude drawings, and the relative difficulty of the stories, none of the
subjects scored zero on this test, and three girls scored the maximum possible
marks. Moreover, the manner of failure shown in unsuccessful attempts was
quite unlike the type of failure characteristic of persons with damage to the fore-
part of the brain.
The Block-designs test, on the other hand, presented an entirely novel situation
to the subjects; it was a type of problem less likely to have been encompassed by
their previous experience than that presented by the other tests; and although
again no subject scored zero, and two subjects achieved full marks, the type of
performance showed two interesting differences from the English average.
The first of these concerns the speed of performance. As has been mentioned,
the original form of the tests gave additional marks for speed of completion of
each design, and also rather strict time limits for each design; but it very soon
became clear to me that these standards were not related to the ability of the
subjects to do the test. From a quantitative point of view, the scores including
"time-credits" gave a lower correlation with the total scores; and from the
qualitative point of view, the girls showed little tendency to "hurry" over the
problems, as is common with English subjects. They tended to examine the
situation for some time before commencing to build the appropriate pattern.
and often made various "exploratory" moves before what appeared to be a
"flash of insight" showed them the correct solution. It was evident therefore that
speed could not be regarded as an integral part of every performance, as it can
generally be in English subjects; and that the speed with which the girls achieved
the correct solution bore little relation to their ability actually to solve the
problems concerned. An easy explanation of this would be the observation, by
no means original, that speed or "hurry" is not generally an imperative in African
culture; but I think that another, perhaps more realistic, explanation may be
found in the second feature of the subjects' performance.
This second distinctive feature is more easily described than summarized. In
its most extreme form, it consisted of the subject copying the design correctly
but upside-down; this was particularly remarkable in the case of one design,
which consists of a red V on a white ground, and which was often copied as
A, -< or > Less severe forms of this error consisted of producing a "mirror-
image" of the original design; while the feature appeared in its slightest form
when the subjects showed considerable difficulty in fitting perhaps the last one
or two blocks into the pattern, although the greater part had been correctly done;
and finally in actually not recognizing the correct solution when they had already
completed it. Lest anyone should rashly conclude that this peculiarity of per-
formance indicates deficient function of the part of the brain concerned with
spatial problems, let me hasten to say that the type of performance described
above does not resemble in any way the deficiency observed in cases of damage
42 UGANDA JOURNAL
to this part of the brain-though it is occasionally seen in younger English
children. The fundamental difficulty in these African subjects seemed to be a
failure to achieve an accurate comparison between the pattern presented and
the design constructed. Whereas this often led to inability to complete the design
correctly, yet in many cases, although it led to inversion of the design (as with
the V), the essential problem of the construction was solved, but of course with
much more difficulty than if the subject had copied the design the right way up.
This difficulty in correct comparison, if it may be thus described, consequently
not only made each problem more difficult than it would have been for subjects
who had started off with the correct orientation, but made the time taken over
each design unusually long, and also made it dependent upon the almost random
chance of whether the pattern was copied right way up or inverted. (This does
not mean to say that the subjects were just left to flounder on with the designs
until they got them right; observation shows that there tends to be a limit beyond
which the subject never gets the design right, and reasonable time-limits for
these subjects were only about one minute longer for each design than were the
The cause or origin of this difficulty in accurate comparison is a matter which
I hardly feel competent to discuss. It is evident that there is no absolute deficiency
of ability to deal with this sort of problem, as the patterns were often correctly
copied though incorrectly orientated. It seems to me more likely that the usual
African unbringing, with its absence of toys and constructional games, and
relatively slight emphasis on accurate measurements and comparisons, simply
does not encourage a rigid standard of orientation and imitation. This impreci-
sion may also account for the errors noted in the Picture-arrangement test, in
which pictures were left "lying on one side", though correctly interpreted; and
it may in fact account for features of African performance in other walks of life,
and may even suggest certain points of emphasis needed in the education of
By contrast, it should be emphasized that there appeared to be none of this
imprecision or uncertainty of comparison in the "verbal" tests-whether Com-
prehension, Similarities, or Arithmetic. This accords with Biesheuvel's point,
that in general the African is skilled in verbal expression and in the exercise of
his auditory sense.
In conclusion, this study demonstrates that the subjects were capable of solv-
ing problems covering a wide range of mental abilities, but that their methods of
solving them depended to a considerable extent upon the material involved in
the problems. Thus verbal and arithmetical problems presented no particular
difficulty apart from the technical one of "translatability" into Luganda; while
one of the "performance" tests revealed a difficulty in achieving a direct compari-
son between the task set and the solution attempted. In spite of this difficulty,
however, the test did not prove impossible for any of the subjects, and in fact the
scores suggested that, of all the tests, this one provided the best single measure
This evidence has a particular bearing on the problem of devising intelligence
tests for this type of subject. It seems that suitable verbal questions will receive
reliable answers, which may be scored in the usual manner; but that the novelty
AFRICAN PERFORMANCE ON AN INTELLIGENCE TEST 43
of material in "non-verbal" or "performance" tests may call for adjustment of
the methods of scoring to suit the subjects-an adjustment which can only be
determined by the study of individual performances. The type of "group" test
currently in vogue in England, in which the problems are merely sets of incom-
plete patterns or groups of figures, and which are allegedly "culture-free", seem
to be in fact peculiarly "culture-dependent", and might be found to be almost
incomprehensible to the average African subject.
I am indebted to the Hon. the Director of Medical Services for permission to
submit this paper for publication.
Biesheuvel, S. (1943). African intelligence. Johannesburg: S. African Inst. Race
-- (1952). Occupational abilities of Africans. Colonial Development, No. 12, 12.
Carothers, J. C. (1951). Frontal lobe function and the African. J. ment. Sc., 97, 12-48.
- (1953). The African mind in health and disease. World Health Organization:
Monograph Series, No 17.
THE PREHISTORY OF THE ENTEBBE PENINSULA
By K. MARSHALL
T HE first part of this paper consists of notes on the methods employed during
an archaeological survey of the Entebbe peninsula; a description of the
discoveries made forms the second part.
Publications which deal with archaeology or prehistory are of a necessity of
two main types, firstly those which deal with field survey, and secondly, those
which are principally excavation reports. In some cases where publication has
been delayed it is possible to combine the two-to describe the discovery and
preliminary investigation of the sites as a group and their relations one to
another, and to continue with detailed descriptions of the complete excavation
-of some or all of them, and a dissertation on the antiquities which are found
during the process.
Field survey is a necessary introduction to archaeological work in any locality
and before a spade is put into the ground it is important that the whole area
should be thoroughly examined on the surface, by walking over it and from
aerial photographs if such aids are available. The latter are invaluable if properly
interpreted, and sites invisible on the ground are often clearly shown, particu-
larly if pairs of photographs with a fifty per cent. or greater overlap are obtained
and a study can be made using a stereoscope. In the case of the Entebbe peninsula
two short lines of photographs, crossing the area from east to west in the neigh-
bourhood of Nsamuzi Hill showed the ancient lake-beaches at several points.
It is not intended here to discuss the archaeological interpretation of such
photographs; Crawford (1925) and St. Joseph (1945) deal with the subject.
Adequate maps are important for a successful field survey and in the case of
Entebbe maps were available up to a scale of 1/5,000. They were used both for
the plotting of physical features, such as old lake-levels and cliffs, and as a basis
for small surveys carried out in connexion with work on individual sites. Where
maps cannot be obtained it is often necessary to carry out a rough survey of the
whole area under consideration.
Sites which may be found during a survey are varied. Cliffs on the modern
and on old lake-shores quite frequently contain 'sea caves' excavated by the
action of the lake at some period during its history, and caves or rock shelters
may be found in outcrops of massive quartzites or granites where the collapse
of blocks along joint planes or lines of weakness has occurred. Other caves which
owe their existence to the solution of the surrounding rock by percolating water
are sometimes found; these are only found in rock which is soluble in this way,
such as limestone, and are not common in Uganda. The value of a cave or rock
shelter as an archaeological site may usually be assessed by examining its floor.
A bare rock floor is unlikely to be productive although in a cave with a rocky
floor it is as well to penetrate to the back in order to be sure that there are no
THE PREHISTORY OF THE ENTEBBE PENINSULA 45
holes which contain pockets of cave earth. Any considerable depth of earthy
deposit may repay investigation and be found to contain one or more levels
which yield debris of the period of human occupation. This is true of Aerodrome
Cave at Entebbe, where two separate, implement-bearing levels were found,
divided by barren levels of natural accumulation. When a cave which has clearly
been occupied during the stone age period is discovered, it is also advisable to
search for rock paintings both on the walls and roof of the cave itself and on
neighboring rocks. It should, however, be noted that paintings are far more
likely to be found on the quartzites and granites which form rock shelters than
on the laterite cliffs into which the Entebbe 'sea caves' are cut.
Open sites of the later stone age culture are very likely to be found during a
survey of the lake-shores area of Uganda and occasionally finds from the more
recent cultures of the old stone age may be made (e.g. the site at Sango Bay in
Masaka District which has given its name to the Sangoan culture). An open
site of the middle or new stone age usually reveals its presence by a dense scatter
of chips or ddbitage of white quartz, with rare flakes of clear vein quartz, lying
on the surface of the ground just above the existing lake cliffs where the thin-
ness of the soil has precluded the formation of dense forest growth. Such sites
must have had a similar attraction for the stone age peoples of this part of
Africa as did the open chalk uplands of southern England, with their supply of
flint for the making of tools for the neolithic peoples of England videe Fox
(1947) and Fleure (1951)). The source of supply of quartz for the Entebbe site is
still a matter for speculation, but it seems likely that at some point in the
peninsula a vein of the material out-crops and large pieces were obtained by the
toolmakers who then carried them to their working site for preparation. Amongst
the debris of these sites a few well made artefacts are to be found, as will be
described below, in the section on the archaeology of the area, along with a few
very crude sherds of pottery. It is hoped at a later date to excavate a part of one
site where the level containing tools is not merely eroded downwash, but is
sealed below an earth fill at the foot of an old cliff face. On open sites the presence
of burial mounds or hut circles may be revealed by irregularities of contour at
the surface or by careful study of aerial photographs on which vegetational
differences may often be seen quite clearly, giving the complete outline of what-
ever site is buried beneath (Crawford; St. Joseph; loc. cit.).
A field survey as described, will, in all probability, lead to the discovery of a
number of sites which will obviously repay further investigation and several
may be selected for treatment as has been the case in the Entebbe area. Their
value may be further proved by preliminary excavation, which involves the
driving of test trenches at specified points. In the case of caves containing deep
deposits, these cuts are made through the cone of debris outside the entrance, to
the mouth of the cave itself. By this method an idea can be obtained of the nature
of the deposit which will be met with inside and, since cave dwellers did not
confine their activities to the interior of their shelters, it is quite likely that
occupation debris will be found in the cone, having been dropped there or
carried out from the cave by the forces of erosion. However, before even a small
excavation of this nature is begun a detailed survey must be made of the site and
its immediate surroundings, and if possible the plan obtained should be related
46 UGANDA JOURNAL
to known points both for height above datum and for position. The accuracy of
such a survey depends, of course, on the training and skill of the archaeologist,
and the instruments used will also depend upon the accuracy which it is con-
sidered will be required. The usual method for rapid survey is by plane-table,
chain or tape, and compass. In the case of a cave site with considerable changes
of slope and varying depths of deposit a surveyor's level is of great assistance,
and the management of a simple form of this instrument is not difficult to learn.
By its use the comparative depths of the various deposits can be assessed quickly
and with considerable accuracy (Atkinson, 1946).
Preliminary excavations of open sites consist of making narrow trenches at
fairly widely separated points in order to test the depth of the deposit and at
the same time learn something of the stratification. Surface collections are also
made of material found lying about, artefacts and debitage, pottery and any
other objects which appear to be of interest. Bones and shells should always be
collected in order to obtain information about the diet of the people who lived
there. In northern Europe the so-called 'kitchen middens' consist largely of
great piles of sea-shells and bones which accumulated as refuse heaps on sites
occupied during the middle stone age period. Beads of shell, stone or pottery
sometimes come to light and both the interior and exterior of pots or sherds
occasionally retain the impressions of grasses or grains with which they have
been in contact before baking, when the clay of which they are made was still
wet. Graves in which grave goods have been deposited at the time of burial are
of great value and are often the only source of complete pots; the skeletons them-
selves are important to both the archaeologist and the anthropologist. In Uganda
there is always the possibility that the remains of fossil man may come to light,
contained in some of the recent geological deposits. The preservation of early
man as a fossil is invariably due to some accident, since he was less likely to be
trapped by natural hazards than were the lower animals, and it is for this reason
that his remains are so rarely found. In any case burials, either natural or
deliberate, require extreme care in the excavation, in order that they may reveal
the maximum information.
It is this stage which has been reached in the study of the Entebbe area. The
field survey has been completed and preliminary excavations have been made
at three cave sites and two open sites. As can be seen from the foregoing, the
study of the prehistory and of the geology of the more recent periods, the
Pleistocene and to some extent the Pliocene. are closely related. This is
especially true of East Africa, where a very lengthy development of the stone
age, from the earliest pebble tools to the neolithic cultures can be observed. The
examination of man's occupation sites and working floors on the ancient lake-
beaches and river terraces, is inextricably mixed with a study of these features
themselves and with an estimation of the climatic conditions, the flora and the
fauna which obtained during their formation.
Further excavations are planned in the Entebbe district, both at the occupa-
tion site at Bugonga Point (where the actual dwelling places, if such were indeed
constructed, may be revealed by the presence of post holes in the lower deposits,
and where a good deal more pottery may be expected in close proximity to
microlithic tools of the new stone age) and at Aerodrome Cave. At this latter
THE PREHISTORY OF THE ENTEBBE PENINSULA 47
place a typical cave deposit will be exposed at an easily accessible point, where
it can be inspected by those interested. All the sites mentioned will be included
in due course in a schedule of sites of archaeological interest and annexed to a
revised form of the Antiquities Ordinance in order that they may be completely
protected. After the completion of the excavations a full report will be prepared.
The following description is concerned with details of the work which has been
done to date.
Although this account is entitled "The Prehistory of the Entebbe Peninsula"
it is necessary to go back into recent geological time in order to give the
geomorphological background. It is in the setting described below that early
man lived, making use of the materials which were found locally in the construc-
tion of his tools and weapons. Only stone, and pottery from the latter periods,
have survived. Early man's other materials, wood, skin and bone, were unable
to resist decay in the soil. Thus it is from these few artefacts and broken frag-
ments of pottery, along with knowledge of the ancient lake-beaches and cliffs,
whose caves must have provided shelter for some at least of the population of
the time, that prehistory must be pieced together. At best the picture revealed
is vague in outline, its colours merge, and the men and women illustrated are
shadowy creatures, since in prehistory there can be no personal identification
and the prehistorian must deal in types rather than individuals. The first section
of the following work is, therefore, principally concerned with the old lake-
levels of Entebbe, and is continued by a series of short notes on several sites of
archaeological interest which have been discovered during the study of the raised
beaches for the purpose of a Geological Survey Memoir to be published at -a
The presence of a series of high-level beaches of Lake Victoria on the Entebbe
Peninsula was first noted by the then Director of the Geological Survey, E. J.
Wayland and his assistant, W. C. Simmons, in a paper (1921) which appears
in the first Annual Report of the Survey. During his later career Wayland
measured the height of a number of these beaches by aneroid and tabulated his
results alongside levels taken at other points around the lake-shore, in all the
three East African territories. Along with this he excavated partially two associ-
ated 'sea-caves', a number of which are to be found in the old laterite cliffs which
are a feature of the ancient lake-levels in the area. Even by such a simple method
as altitude measurement by aneroid, Wayland was able to arrive at a useful
correlation of levels; and his figures have, to a considerable degree, been sub-
stantiated by the use of an accurate level. The excavation of a cave near to one
examined by Wayland has produced concrete evidence for an old high-level
beach in the form of the water-washed sand and gravel of the beach itself, still
surviving in its original beach bedding.
The old beach and cliff lines are exceptionally well illustrated in the peninsula
and have, therefore, been examined with some care in order that they may be
used as a yard-stick, against which the numerous other ancient lake-levels to be
observed in the Protectorate can be measured. Artefacts of the prehistoric period
are also found in all the levels, and pottery and other remains, apparently of the
protohistoric age of Uganda, about which little is known, are to be found in
considerable quantity on and near the lake-shore, and on Nsamuzi Hill which
rises behind Government House.
It has been observed from the table of ancient lake-levels in the Entebbe area
prepared by Wayland, that variations of up to 10 feet occur in what appears to
be the same beach, measured at different points in the peninsula but this may
easily be assigned to the comparative inaccuracy of an aneroid and the difficulty
of deciding on an identical level on each beach for measurement. For the purpose
of the present search an attempt was made in each case to determine a final
station for each line of levels at a point on the backshore, if no nick point at the
foot of an old cliff, or wave-cut platform was visible (Fig. 1).
Briefly, three lower levels may be seen with clarity, above these are two ill-
defined higher levels above which is the laterite plateau surface, only occurring
on Nsamuzi Hill. The first three old beaches are 25, 50 and 80 feet, respectively,
above present lake level and on all three there are remnants of cliffs cut in the
lateirite. Below these is a beach terrace, only visible where headlands give rise to
laterite cliffs, marked by a number of water-worn caves and one natural arch.
The lower levels and the beach terrace are dealt with first.
At 10 feet above present lake-level a number of water-worn caves are to be
found in the laterite, at the following places: on a headland at the southern end of
the Botanical Gardens; near Bugonga Point; below the Entebbe Swimming Club;
and in headlands to the south of the rifle range. This level has not been observed
in the form of a beach between these places nor is it to be seen on the westward
or leeward side of the peninsula. It therefore seems probable that the 10-foot
notch on the headlands noted above represents either a bench cut during a com-
paratively short period, when the lake was at a slightly higher level than at
present (the force of the prevailing winds on the exposed eastern side of the
peninsula accounting for its being found only on that side and not on the sheltered
western coast) or it may represent a notch cut at the head of a beach terrace which
had been built up on a beach to which the lake has once more regressed. If the
latter conclusion is accepted the water-level must have receded, at least to the
stage reached at the present day, only to transgress once more, building up a storm
beach, above which the caves were excavated presumably by pneumatic action
in the remains of cross-jointed quartzite ribs which can be seen to correspond
to the lines taken by the caves. In support of this theory is the fact that in all
cases, below the 10-foot level caves and below the natural arch on the 10-foot
level at the extreme south-eastern point near the rifle-range, a storm beach
reaching to the level of the cave floors is still in existence although it is badly
eroded in places. The surface of the beach terrace is in all cases covered by a
downwash deposit of earth carried away from the areas above the cliffs. In the
bays between the several headlands noted above, remains of a 10-foot level are
apparently not in evidence, since any erosion from the gentle slope reaching to
the top of the 25-foot beach has been replaced by storm beach deposits thrown
THE PREHISTORY OF THE ENTEBBE PENINSULA
up on to that slope at the same time as beach terraces were under construction
around the headlands. This process is still going on and it may be observed as
follows. Between June and December 1952, the lake fell several feet; at and
around the laterite points of the peninsula the area uncovered can be seen to
consist of masses of detritus from the cliffs above whilst the change has left no
mark on the gravelly beaches and is only represented at the time of writing (June
1953) on sandy beaches by heaped up areas of sand which are rapidly being
removed by wind erosion. This 'level' was reported by Wayland from the
Botanical Gardens area and from Bendegere, and has recently been measured
at the points near the rifle range and near the Botanical Gardens. The natural
arch noted above is presumably an old cave which has been cut back until the
headland, which extends across the line of the prevailing wind, has been com-
pletely penetrated. Its position is shown on the general map of the peninsula
which accompanies this paper (Fig. 1). Other caves at this level nearby contain
a deposit of earth several feet in depth, but erosion appears to have removed all
but a few inches from the arch, the floor of which is littered with large laterite
blocks apparently fallen from the roof and back.
Above the 10-foot beach comes the 25-foot level which can be recognized at
almost any point around the shores of the peninsula and is well developed both
as laterite cliffs above the existing water level or as raised beaches in the embay-
ments between the headlands and on the sheltered western side of this stretch
of coast. It has been noted as being virtually continuous and has been levelled
with complete success at a number of points. On the exposed headlands it can
be seen as a notch in the laterite cliffs and is marked by the presence of a number
of well developed sea caves at Bugonga Point to the east of the White Fathers'
Mission. Between the cliffs the old lake-level can be traced with ease in the form
of a raised storm beach; this is particularly illustrated in the bay which lies to
the south of the point on which the Geological Survey Department is situated
(where the ancient storm beach rises steeply behind the modern beach and. forms
a distinctive feature) and also at the eastern end of the 3,300 yard runway of the
Entebbe Airport, although at this point much of the line of the old beach has
disappeared under the aerodrome runway.
At almost every point on this level, although the actual high water mark may
be difficult to recognize, the beach deposits of water-sorted sands and gravels,
left when the lake receded, can be seen on the old foreshore and have been
revealed in situ at one point by excavation. It is also possible that some of the
caves in the 25-foot level mentioned above, may have been overdeepened and
their deposits lost by undermining wave action, leaving high and shallow caves
at the 10-foot 'level'-such as one at the point near the Botanical Gardens. The
lake's standing at the 25-foot level made little basic difference to the geography
of the peninsula, except at the extreme south-eastern end, where the cliffs to the
south of the rifle-range must have been left as a small island; elsewhere the lake
merely stood above the present shoreline, filling several minor valleys on the
southern shore. Archaeological evidence on the ground immediately above the
25-foot beach is considerable and shows little of an intermixture of cultural
types. The sites are certain well-marked areas, usually close beside or actually
on the headlands which are a feature of the eastern side of the peninsula; they are
50 UGANDA JOURNAL
dealt with in detail in the section devoted to the archaeology of the Entebbe
The 50-foot level is also, for the most part, clearly defined, and its course may
be traced with a high degree of certainty, although laterite cliffs rising from 50
feet above the lake are rarer than those of the 25-foot level, since the higher
beach and cliff series have been exposed to the forces of erosion for much longer
than those of the lower. In many places the foot of the 50-foot cliff is hidden by a
slope composed of detritus carried down from above and in others the old cliff
line can only be located by observing a break of slope or sharp vegetation change,
the flats above the old cliff being covered by thin grass whilst much denser growth
covers the area below the cliff line which has a thicker soil cover. There are,
however, localities where the 25-foot level runs up to form the base of the 50-foot
level and at such points two-tier cliffs are often well developed, for example, at
Bugonga Point and at Rifle Range Point (cf. Cotton (1945); pp. 421, 469; figs.
The location of laterite cliffs in the 50-foot is very much as in the 25-foot level;
they are without exception on the eastern and southern sides of the peninsula.
On the western side, the shore of Waiya Bay, the 50-foot level takes the form of
a short slope up from the top of the 25-foot shore and then an abrupt and steep
rise above the 50-foot high water mark. A high lake at the 50-foot mark had far
more effect on the geography of Entebbe than did the 25-foot lake, since with
'the lake at this height the southern end of the peninsula was no longer con-
tinuous, but consisted of a series of islands.The southernmost of these is now
Kigungu Hill, which stands out clearly above the 50-foot level when viewed
from the lake to the south, or from the western end of the main runway at
Entebbe airport. Two distinct breaks of slope may then be seen on it, the upper
of which will be discussed in dealing with the 80-foot level; low cliffs, badly
eroded, may be seen at its low south-eastern end. A larger island above the
50-foot level was formed by the hill now known as Kilima, which rises behind
the Veterinary Department buildings, and Buku Hill, now destroyed by the
extension of the main airport runway. The only cliffs to be seen on this old island
are again on its south-eastern side overlooking the airport. These are slightly
higher than those on Kigungu and, since they were seemingly more exposed,
possess a feature which the Kigungu cliffs do not, a series of well developed
water-worn caves one of which is very large and has been partially excavated by
Wayland videe page 52 below). The course of the northern side of the Kilima
island has now been lost beneath the large runway extension, but an assessment
can be made of its probable line.
Other small islands on the 50-foot level can be seen at Campbell's Plot on the
southern end of the peninsula and in the form of an arc to the south of the present
rifle range. The contours also show a small area more than 50 feet above lake
level inside the boundary of Musoli Forest, but this has not been examined. The
area covered by Nsamuzi Hill and modern Entebbe is entirely above the 50-foot
level, and, although the isthmus connecting Entebbe to the remainder of the
peninsula was narrower when the lake stood at that level, no waterway was
established there to separate the Nsamuzi section from the mainland. On the
Nsamuzi-Entebbe promontory the most notable sites for examination of the
THE PREHISTORY OF THE ENTEBBE PENINSULA 51
50-foot beach are at the foot of the bluff below the Chief Secretary's Lodge and
across the golf-course, on the eastern side of the secondary runway at Entebbe
Airport, and on the western slope to Waiya Bay from the hill known as Nakiwogo.
At none of these points, except possibly below the Chief Secretary's Lodge, can
the 50-foot be described as a cliff but in all cases it is easily recognizable in the
form of a steep slope rising from the old beach or foreshore. At many places,
notably on the aerodrome flat and on the Nakiwogo slope the water-sorted beach-
sand and gravel of the ancient beach are to be found mixed with later downwash.
It should be noted here that the average level of the aerodrome flat above modern
lake-level is of the order of 30 to 35 feet rising at its edges to 50 feet where the
bluff of the second lake-level rises abruptly from it for a height of twenty to
The 80-foot beach and cliff line is less well represented than the two already
described, and depends for its identification almost entirely upon changes in
degree of slope, in places only detectable by reference to sections obtained by
the use of a levelling instrument. Considering the peninsula from south to north,
as in the last example, it is not observable on Rifle Range Point which must have
been covered by a shallow depth of water in 80-foot level times. On Kigungu
Hill it is visible, but it is thought that the cliff which reaches almost to the summit
of that hill on its south-eastern side is an erosion feature which cannot be
ascribed principally to lake action, since it is discontinuous and does not take the
form of a waterwQrn cliff; it extends through the 80-foot level to the old 50-foot
beach and yet shows no sign of the double tiered cliff effect as referred to above
and is due to the existence at this point of a band of quartzite in the granite which
covers the southern end of the peninsula. True sea cliffs in laterite may only be
seen at one point in this level, that is on the south-eastern slope of Kilima where
there is a section of low cliff which extends for several hundred yards and whose
base is at the 80-foot beach. The lake at 80 feet above present level was high
enough to separate Nsamuzi Hill from the rest of the peninsula to the north and
in those times a further island was added to those already delineated in describ-
ing the 50-foot series, the separation being from the narrow inlet to the north of
Nsamuzi across to the northern end of the Botanical Gardens on the opposite
side of the peninsula. The 80-foot beach is poorly marked on this section and
for a good deal of its length it has been hidden by the buildings of modern
Entebbe, although it is clearly visible in the Botanical Gardens themselves and
in the neighbourhood of the Secretariat.
The two hills at the southern end of the peninsula are not of sufficient height
to show either of the upper levels which are only to be seen on Nsamuzi Hill and
on the Gowers Park flat. In the times of the 175-foot level the summit of Kilima
was probably only just exposed and awash as a reef, similar in fact to the exist-
ing Alice Reef, off Bugonga Point. Kigungu Hill must have been completely
covered by the lake water. The 175-foot level shows as the flat below Govern-
ment House and extends roughly along the line taken by Circular Road and
Gowers Road. On the northern and western slopes of Nsamuzi the level is far
from clear except where it may be seen on the road to the Nakiwogo ferry as it
descends the hill from the Virus Research Institute.
The highest possible lake-level is also shown on Nsamuzi and takes the form
52 UGANDA JOURNAL
of an abrupt change of slope on the hill proper above the Virus Research Insti-
tute flat; the summit of Nsamuzi was the only part of the Entebbe section of the
peninsula not covered by water at the time. The height of this extreme level above
the modem lake is 250 feet, which height, despite its magnitude, may be matched
by river terraces (with which it may be associated) on the Kagera River. No
ancient beach deposits have so far been detected at the foot of either of the two
high levels just mentioned; this may be due either to their removal over the long
period which must have elapsed since the formation and growth of these levels
or their covering by downwash from the summit of the hill, or possibly by a
combination and interplay of both factors.
The above account of the ancient lake-levels and cliffs in Entebbe is purely
descriptive and it is not intended at this juncture to present detailed arguments
which will shortly be published as part of a Geological Survey Memoir on the
Pleistocene Geology of the Uganda Protectorate. It is sufficient, for the moment,
to suggest that the levels and cliffs owe their existence to a much greater Lake
Victoria which developed in Pliocene or early Pleistocene times and which fell
by a series of steps, with at least one possible transgression, to its present level.
The lake levels represent halting points during this gradual decline, and are
probably due to a combination of geological, geomorphological and meteoro-
logical phenomena, involving rift valley formation, river capture, and periods
of increased and decreased rainfall.
From the point of view of archaeology, Entebbe is of undoubted interest and
it is a simple matter for the most unskilled layman to discover the remains of
prehistoric cultures which are scattered, often on the surface, throughout the
peninsula. At this point it should be stressed that the sites mentioned below are
protected by the Uganda Antiquities Ordinance and any chance finds which
appear to be of especial interest should be reported to the Geological Survey
Department. The caves too should be approached with some caution, both from
the point of view of the reptilian inhabitants, and, in the case of Aerodrome
Cave, the dangerous nature of its entrance which is at once gloomy and booby-
trapped by a deep excavation trench dug across its mouth.
Palaeolithic finds have been rare in Entebbe, but a careful search may yet
reveal stations of this period. However, Wayland discovered Oldowan and
Chellean tools at Buku Hill and at the northern end of the township, both points
being above the 75-foot level. But traces of later cultures are numerous.
Dealing first with the caves, that in the 50-foot level cliff (Fig. 1) overlooking
the aerodrome flat and referred to by Wayland and by van Riet Lowe (1952;
pp. 100-1) as Aerodrome Cave No. 1, contains 10 feet of deposit and a descrip-
tion of its stratification and implementiferous levels is to be found in the work
mentioned below. Van Riet Lowe came to the conclusion that the artefacts dis-
covered represented an industry which was homogeneous throughout the
deposit and represented an advanced phase of the Wilton (Late Stone Age)
Culture as developed in this part of Africa.
Another cave in Entebbe has also been partially excavated, this time in the
25-foot level; it is referred to as Bugonga Point Cave No. 4 (Fig. 2). This cave
N ANCIENT LAKE LEVELS
/ 1-t woo s 9 logopO 2900 3000 4000 5000
/ .-Musoli.II. :. ,<:'-.. .
Forest .., '
/. / .... / II i i ;
SWAMP SWAMP S AMP
,. ,, / <, ,'/
point (rWS (" ... ,
P'.. ... No 9 Site ..~~ a.... 175 -.
LAKE VICTORIA PugoyaPt stPont cale"
Geological Survey of Uiganda N't343
FIG. 1 (face p. 52.
CAVES AT BUGONGA POINT
Down wabh with Earth down wash
Implte n.t aond pottery 50'cliff
Clay with Pabbl as. C'0-,
Wetrsr ,ITd Sand s Craol d Nd &ool
Earthy do.t wash ask A Section
Recent ccuulatlon with L
Scle 2 3 fast
GENTLE GRASS SLOPE
" iO .. . u.
rs as w, o a o n o t o a SOaOeOOV
54 UGANDA JOURNAL
was unremarkable for purely archaeological discoveries, revealing no well
stratified artificial deposits, although a section 7 feet in thickness was cut before
the cave floor was reached. The upper part of the deposit appears to be down-
wash carried through fissures behind the cave and contains only a few Uganda
Wilton artefacts and bone fragments probably transported from an open site
which has been discovered on the flat above the cave and which is discussed later.
Below the lateritic downwash a level of considerable interest was discovered
-consisting of clean water-sorted sand and gravel which is almost certainly an old
lake-beach preserved in the cave. The average height of this beach surface at the
cave mouth is 24-6 feet above the present average lake-level.
In addition to the cave sites in Entebbe, there are a number of open stations,
particularly on or near the headlands which are well developed on the south-
eastern side of the peninsula. The first of such sites which was examined was
that at Pumping Station Point (Fig. 1). The site covers an area some 100 by 800
feet and is characterized by a dense scattering of quartz ddbitage amongst which
a number of microliths are to be found. The tools are principally crescents and
awls, with pigmy cores and end scrapers, the assemblage being fairly typical of
the Uganda expression of the Wilton Culture and may be compared with collec-
tions from Aerodrome Cave No. 1, Entebbe, and from the Nsongezi rock shelter.
Artefacts and potsherds are found on this site to a depth of one foot below the
surface, at which point the lateritic soil gives way to virtually solid laterite. A
scatter of artefacts below the 25-foot level is almost certainly due to downwash.
On the southern side of this small headland, a nick point representing the head
of the 25-foot level is clearly visible but is lost on its northern side in a cliff of
almost vertical laterite, more than 30 feet in height.
The pottery from Pumping Station Point is of especial interest and does not
resemble either in paste, form or decoration the sherds which are frequently
found on the modern shore line of the Lake. It does, however, compare with
discoveries made during 1944 in Central Kavirondo, Kenya, by Leakey and
others (1948), but unfortunately the Entebbe discoveries are all fragmentary and
no pieces of base have come to light. It is therefore impossible to say whether or
not the dimple base form found in Kavirondo was produced. The largest frag-
ment from Pumping Station Point (Fig. 3) consists of some one-third of a medium
sized bowl with incurved rim, and the assembled pieces reach to the curve of the
base. The paste is a coarse, dark brown material containing a large number of
quartz grains which may be several millimetres in diameter; there are, however,
no quartz grains or pebbles which pass through the thickness and protrude on
either side. Five other segments of rim were recovered from this site, all
apparently parts of incurved rim bowls, but no two were from the same pot, as
may be seen from the differing decoration motifs and from the several diameters
calculated from these sherds (Fig. 4b, c). The colour of the Entebbe specimens
varies from yellowish brown to dark brown and whilst all have suffered from
Weathering, some are in much better condition than the others despite the fact
that they were all discovered at approximately the same level. There is no trace
on any of the specimens of burnish, paint, or the use of slip. Slight irregularities
on the surface of the large sherd referred to above suggest the use of the coil or
rings technique in its construction. The decoration in all cases consists of incised
THE PREHISTORY OF THE ENTEBBE PENINSULA 55
grooves or punched dots usually arranged in a chevron pattern at the rim with
horizontals below, the whole pattern rarely extending for more than two inches
below the rim; occasionally there are traces of vertical incised grooves below the
rim on the interior. All these pots are reminiscent of the Peterborough (Neolithic
B) culture of the south of England (British Museum, 1921).
1 0 1 2 3 4 _
A pottery sherd from Pumping Station Point
A second open site has been investigated in the area to the west of Bugonga
Point between that headland and the White Fathers' Mission (Fig. 1), where
conditions are very similar to those obtaining at Pumping Station Point. The site
is, however, on the old beach of the 50-foot level, and immediately below the low
cliff of that level, this fact being explained by the two-tier cliffs around Bugonga
Point preventing the development of a wide flat immediately above the 25-foot
level; this situation is clarified in the section describing this part of the coastline.
Archaeological material is in even greater abundance at this site than at Pump-
ing Station Point, very dense scattering of quartz chips, including many artefacts,
occurring at several points. The laterite surface is covered by one to two feet of
downwash which consists of dark earth, pebbly gravel and decomposed laterite,
1 0 1 1 3 4 $
Sherds from Nsamuzi Hill (a) and from Pumping Station Point (b & c).
throughout which artefacts and potsherds are to be found. Irregularities in the
surface suggest hut circles but although several of these low mounds have been
sectioned it is impossible to prove this and no post holes are visible. It is true,
however, that such a mound usually contains a greater proportion of archaeo-
logical debris than similar excavated area outside its circumference. Neither
bones, fish or animal, nor shells have been discovered in any of the excavations
to date and the mound cannot, therefore be referred to as a midden. Little can be
suggested of the mode of life of the people who inhabited this settlement,
although the use of pottery, some sherds of which have been blackened by
contact with fire, tends towards the view that the culture was, to some degree at
least, settled. Many of the artefacts, microliths finely worked in quartz which is
difficult medium, show a comparatively high degree of skill. It is unfortunate
that both sites should be so exposed to erosion, causing a jumbling of the evidence
and probably the total destruction of a great part.
At the Bugonga Point site it is possible that a sealed deposit may exist at the
foot of the 50-foot cliff where a small embayment has been filled by black earth
downwash to a depth of five feet above implement and pottery bearing levels.
THE PREHISTORY OF THE ENTEBBE PENINSULA 57
This section is still under excavation and it is impossible to discuss its stratifica-
tion until the work has been completed.
The sinking of foundations for new buildings on the summit level of Nsamuzi
Hill, Entebbe (Fig. 1) has revealed a quantity of potsherds but no stone tools,
suggesting a protohistoric or iron age date. One sherd, which is illustrated at
Fig. 4 (a) is worthy of special note. This is a rim sherd of an incurved bowl,
remarkable in an assemblage which reveals a concentration on the everted rim
type; its paste is dense black and evenly fired. The exterior shows signs of a red
burnish, except where the surface is hatched in diagonals to form a simple
pattern, and the whole would appear to have been a fairly well-made bowl, its
depth being rather more than half its diameter at the mouth. It seems unlikely
that Nsamuzi was a settlement site but may well have been a ritual centre similar
to, but much smaller than, that which survived at the summit of Mubende Hill
into the early years of the present century.
From this brief description, it can be seen that Entebbe, along with the rest
of Uganda, is rich in traces of the past, some of comparatively recent date and
others of considerable antiquity. Legislation exists to protect archaeological
specimens and areas which are rich in such specimens, to check any possible flow
of antiquities from the country and to prevent vandalism at well-known sites,
rather than to put a rein on the interested amateur. In Great Britain many of the
greatest discoveries during the last half-century have been made, in the first
instance, by amateur archaeologists. Indeed much field reconnaissance must be
left to the amateur; but when excavation with its special skills, or preservation
requiring laboratory techniques, are necessary, the professional should be called
upon in order that the maximum information can be gained from a site before it
is destroyed, as it always is by the very process of excavation. It is hoped that
any discoveries will be reported to the co-ordinating authority, the Geological
Survey Department. If doubt exists it is better to send in what may prove to be
merely rubbish than to allow what may be a valuable find to go unrecorded.
Atkinson, R. J. C. (1946). Field Archaeology. London: Methuen and Co.
British Museum (1921). Guide to the Stone Age Antiquities. London.
Cotton, C. A. (1945). Elements of Geomorphology. Wellington, New Zealand.
Crawford, 0. G. S. (1923). Air Survey and Archaeology. Geographical J., 61, 342.
Fleure, H. J. (1951). A Natural History of Man in Britain. London: Collins.
Fox, Sir C. (1947). The Personality of Britain. Cardiff: Nat. Mus. Wales.
Leakey, M. D., Owen, the late W. E., and Leakey, L. S. B. (1948). Dimple-based
pottery from Central Kavirondo, Kenya Colony. Coryndon Museum
Occasional Paper, No. 2.
Lowe, C. van R. (1952). The Pleistocene geology and prehistory of Uganda. Part II-
Prehistory. Geological Survey of Uganda Memoir, No. 6. Entebbe: Geological
St. Joseph, J. K. (1945). Air photography and archaeology. Geographical J., 105, 47.
Wayland, E. J., and Simmons, W. C. (1921). Notes on the Geology of Entebbe.
Annual Report of the Geological Survey Department, Uganda Protectorate,
for the year 1920, pp. 40-3. Entebbe: Govt. Printer, Uganda.
THE NORTHERN PROVINCE MOUNTAINS:
SPECULATIONS ON CLIMATE AND VEGETATION
By H. C. DAWKINS, M.A. (OxoN.)
AMONG the most striking features of the Acholi and Karamoja Districts are
the solitary mountains or inselbergs separated by wide stretches of seem-
ingly level plain. These inselbergs with their steeply soaring slopes and broken
rocky profiles are far from unique in Africa, but they provide some of the finest
scenery in Uganda, and afford a broad panorama of wild mountains from Elgon
to West Nile, into Turkana, Equatoria and beyond.
Most of the mountains are granite with various gneissose and schistose
degrees of foliation, but a few syenitic and more basic types occur, and in the
south-east is the well-known group of volcanoes of which Napak and Elgon are
the best-formed examples. Some of the granitic masses are indistinguishable
from surrounding rocks and may simply represent harder knots of basement
complex gneisses left prominent by denudation of their surroundings. Others
are evidently more recent bosses of less metamorphosed material.
At first sight the inselbergs appear to rise directly from the peneplain itself,
the line of demarcation being definable to within a few yards. But more careful
examination, particularly of a profile from a few miles distance, shows that
below the steep hillside there is usually a low angled and very slightly concave
slope stretching away from the hill foot and merging into the peneplain from one
to five miles away from the mountain. This low-angled slope is particularly
clear in the case of the more isolated and larger inselbergs, whatever their origin,
such as Toror, Madi Opei, Rom, Ogili, Kadam, and on the south side of the
Imatong Mountains and the west side of Moroto Mountain. Each inselberg is
surrounded by an astonishingly regular 'apron' of detritus, which is distinguished
from the peneplain by drainage pattern and nature of watercourses as well as
by its even slope.
The detritus apron usually contains highly varied material, masses of sand,
stones and boulders mixed in confusion. The larger fragments tend to be more
frequent nearer the hill foot, but no systematic sorting or widespread layering
seems to occur in these deposits. Complications such as islands or half-concealed
domes of rock occur in the prevailing sea of deposits and occasionally an
unusually precipitous crag gives rise to a scree slope which confuses the other-
wise clear profile. But whereas scree slopes are usually in the 150 to 35 range,
the slope of the main detritus apron seldom exceeds 4 in its steepest part and
is mostly less. Excellent examples of the sharp divisions which exists between
eroding hillside, relatively stable scree and old detritus apron may be seen at
Kalonga Mission in Acholi, near Iriri camp in Karamoja, and under the cliffs
of Langia on the south-east edge of the Imatong Mountains.
It has long been recognized that the characteristic 'apron' formation
described, which lacks both the homogeneity resulting from pure wind erosion
THE NORTHERN PROVINCE MOUNTAINS 59
and the systematic sorting characteristic of riverain deposits, is a product of
ephemeral 'sheet wash' in an arid climate. The process has been described as a
current phenomenon in many areas especially in South-west Africa, but a
particularly good example which one used to see in the old daylight flying-boat
era between Khartoum and Wadi Halfa exists in the Jebel Kuror in the Nubian
Desert. On this remarkable jagged inselberg one observes deeply ravined hill-
sides surrounded by a gently sloping zone of deposits on which the lines of
sheet wash and resulting deposition are clearly shown. The deposits lie at a
low angle to the desert peneplain and are sharply defined from the mountain
slopes, just as in the Northern Province of Uganda. But the lack of vegetation
in the desert stage shows up clearly the extent of the ephemeral torrents, and
the manner in which, after leaving the hillsides, they spread out as sheet-wash
on the mountain pediment and leave fans of detritus at the limits of their rapidly
absorbed flow. In this way each desert inselberg becomes surrounded by a
zone of coalesced detritus fans forming a continuous apron. Continued accre-
tion of the apron is inevitable provided that the rainfall continues to come in
rare and short but heavy storms, insufficient to support a protective vegetation
or prolonged stream flow. In Fig. 1 the relative position of hillside, detritus
apron and pediment are shown by means of a block diagram. A very similar
process may be seen on a heap of sand dumped in the open by a builder's lorry
and exposed to heavy rains. The flanks of the heap soon become scored by
miniature ravines while the base developed a low-angled smooth-surfaced, not
ravined, detritus apron on the desert inselberg pattern. It therefore seems that
our Northern Province mountains stood in recent geological time in desert
country with a low and poorly distributed rainfall and sparse or negligible
Just how long ago the postulated desert conditions prevailed makes interest-
ing speculation, much of which must be left to the geologist. But there are some
striking features of stream erosion in the detritus apron which allow some
deduction about climatic history. These features are most clearly shown along
the Imatong Mountains, which country is partly used for illustration of the
argument to follow. The block diagrams of Figs. 1 to 4 show the type of country
met with in that area.
The important and striking fact is that most streams derived from small
catchments pass through the detritus in deep, steep-sided gulleys, while those
draining larger areas flow through wider open-V-shaped valleys with gently
shelving sides. A good example of the latter type is the River Aringa which
leaves the steep granite slopes of Lututuru at a level very little above that of
the peneplain, having long since penetrated and removed the hundred feet or so
of apron deposits that may have existed there. The stream then continues down
a relatively gentle slope through its wide-V-section valley, whose sides diminish
to disappear where the apron ends on the peneplain. The valley sides are rocky
near the mountain where boulders, left prominent by removal of the finer
elements of detritus, strew the surface. Further from the mountain the valley
sides are relatively free of rocks and the apron deposits become more uniform.
(Figs. 2-4, left-hand valley.)
A good example of the gulley type of valley is provided by the Lakure, a
60 UGANDA JOURNAL
semi-permanent stream nearby, east of Lututuru. This stream leaves the moun-
tain slopes by a fall of a few hundred feet over exfoliating gneiss, but carries
less water than the Aringa as it taps but two square miles of catchment to the
Aringa's thirty or forty. As in the case of the Aringa the Lakure continues down
the steep mountain profile well into the thick detritus apron, where it enters
a gulley nearly a hundred feet in depth and with vertical walls. Having there
reached bedrock-believed to be that of the mountain pediment if not of the
peneplain itself-the stream flows gently for about two miles in the gulley whose
walls diminish in height as the apron deposits wane towards the peneplain
surface (Figs. 3-4, right-hand valley). The erosion profile (side elevation of the
stream bed) of the Lakure is therefore similar if not identical to that of the
Aringa and both give a picture of the appearance of mountain, pediment and
peneplain profile before the formation of the detritus apron.
Gulleys of the Lakure type are clearly young; their sides, but for the protect-
ing blanket of shrubs and climbing herbage, would be far from stable and they
meet the apron surface at sharp or barely-bevelled angles. Valleys of the Aringa
type are older; no sharp angles exist and there is even a system of tributary
runnels draining the neighboring area of the apron and giving rise to sub-
sidiary valleys. Assuming these age differences, there is a striking correlation
between the maturity of valleys traversing the apron and size of their mountain
drainage areas. The streams from large catchments have evidently been eroding
through the apron for very much longer than those from small catchments.
Similarly it is evident that the latter type of stream, when it did get started, was
quick to make its mark, and vertical erosion of the stream bed greatly outpaced
lateral erosion of the banks. Little or no valley can have existed before the era
of gulley formation, as even a small valley of mature type would be shown up
by a dipping of the ground surface towards the lip of the present gulley. In the
Lakure and in many other cases of small catchments no such dip exists, so that
any previous channel cannot have been wider than forty or fifty feet; the Aringa
affects the apron surface for several hundred yards on both sides.
There seems to be only one explanation for this correlation, namely that the
arid climate under which no marked stream-valleys were formed gradually
changed to one with a well-distributed rainfall, while the mountains became
clothed in an effective blanket of vegetation (Fig. 2). Had the rainfall increased
in mere quantity with too poor a distribution for perennial vegetation cover,
gulleys would naturally have formed from the smaller as well as the larger
catchments, though they might not have been so deep. And these in turn would
have developed the open-V section in course of time. But a vegetation cover of
sufficient density would have restricted the run-off so that only the accumula-
tions of larger catchments would be sufficient for significant valley erosion
across the detritus apron.
It therefore seems that the desert climate was followed by one of well-
distributed rainfall and a vegetation cover approaching, and possibly achieving,
the state of closed forest. During this period streams derived from catchments
of around and over ten square miles produced sufficient water to wear down
valleys for themselves in the apron. The smaller catchments and streams
evidently exerted little or no erosive influence.
THE NORTHERN PROVINCE MOUNTAINS
Desert erosion stage.
Stream erosion stage.
THE NORTHERN PROVINCE MOUNTAINS 63
The only conditions under which the Lakure type of gulley can have been
formed are those of torrential and intermittent run-off (Fig. 3). Permanent
streams usually carry riparian vegetation which has a most definite protective
influence on the banks. The Lakure is not now eroding fast-for at the present
time it is well-clothed with vegetation and has large mahogany (Khaya grandi-
foliola) trees growing from the bottom of the gulley. At the time the gulley was
formed, both the catchment and stream-bed itself must have been devoid, or
seriously denuded, of forest vegetation.
The same argument applies to all the larger but now stabilized gulleys, and
leads to the supposition that the forested stage was followed by one of severe
deforestation of the mountain catchments. Now since the gulleys are so recent
that even those whose vegetation is now stabilized can exhibit only juvenile
types of forest, and since the present climate is capable of supporting closed
forest on the mountain, it is assumed that the deforestation was relatively recent
and the work of modern man (Fig. 3).
In the Imatong Mountains, as elsewhere in Acholi and in parts of Karamoja,
old karega-grinding places-occur abundantly over the now deserted moun-
tains. These consist of elliptical hollows worn in the solid granite by the grinding
of grain. They may often be found in groups of five or six, situated along the
radii of a half-circle whose centre was the kneeling position of the woman
flourgrinder. Each grinding hollow is used only until worn down to an incon-
venient depth, so that after a time it is abandoned and another spot is chosen a
few inches away. One can even, at the focus of the larger karega groups, detect
a twin smoothness at the kneeling place worn by the horny and thorn-hardened
knees of generations of mountain womenfolk. The southern Imatong Mountains
were evidently thickly peopled for a prolonged period, up to a very recent time,
the beginning of the present century. Whether it was the karega-making popula-
tion which cleared the forest or their predecessors, it is believed that the
deforestation which caused the gulley-forming run-off might have occurred a
mere two or three hundred years ago. In some parts, relics of Podocarpus,
Juniperus, Olea and other mountain forest types exist where they were protected
from fire and cultivation by precipitous rock slopes. These usually contain a
few old dying trees and testify to the larger area of closed and valuable forest
which recently covered all suitable sites on the mountain.
As has been noted, gulleys of the Lakure type, however active they may have
been one or two hundred years ago, are now stabilized by forest vegetation
wherever the country is still uninhabited. In the Lakure gulley itself, one can
touch the crowns of seventy-foot mahogany trees whose boles arise from the
sunken stream bed and whose foliage brushes the lip of the gulley at apron
surface level (Fig. 4). The stabilization process is believed to be between forty
and eighty years old, and to be the result of an exodus from the hills culminating
in administrative removal of the people in the last forty years. In the upper
Lakure valley, the traces of recent cultivation and grazing are most clearly shown
by stands of Cassia singueana and Acacia seyal var. Multijuga, now, with other
regrowth, so dense, as to provide first-class cover for game and, apparently, to
slow up the potentially considerable run-off.
It can be observed throughout the Northern Province mountains that,
64 UGANDA JOURNAL
wherever population and game are absent, colonization by vegetation of
previously serious gulleys is dominant and outpaces further denudation. This
is a most convincing pointer to the fact that the present climate is, if given a
chance, favourable and far removed from the aridity the area once knew. But
it is equally obvious that, wherever excessive game, man or stock operate, or
where they have returned to stabilized areas, active erosion of the detritus apron
is taking place and that the reopening of gulleys has begun and is much on the
While this is most evident in Karamoja, excellent examples on a smaller scale
may be seen among the Labatilwonga cultivations west of Lututuru. Here both
the Jom (Nimur) and the Aringa valleys of the wide-V type are developing
minor tributaries of gulley form (Fig. 3); and formerly healthy riparian trees
are becoming 'stag-headed' as their roots are exposed.
The detritus aprons of the Northern Province mountains are a most important
economic feature of Uganda. They provide some of the deepest soils that we
have (being almost universally free of ironstone or murram), and they catch, or
could catch, practically all the mountain run-off in their capacious interstices.
Yet for both these reasons they are liable to irrevocable erosion if the uncon-
trolled processes of a few decades or centuries ago are allowed to regain
dominance (Fig. 4). The flanks of our mountains, as well as the heights, require
most careful management if they are to continue to be as useful to the populace
as they have evidently been in the past.
I am grateful to Dr. R. B. McConnell and Mr. K. Marshall of the Geological
Survey Department for their help in reading this manuscript and for their
correction of some misconceptions.
By P. H. GULLIVER, PH.D.
THE staple crop of the Jie farmers is sorghum (mumwa in the local dialect)
which forms about 95 per cent. of all cultivation. Small 'luxury' crops of
groundnuts (epuli), maize (ngaboret), eleusine (ngakima) and tobacco (etaba)
are grown on old homestead sites and next to the occupied homesteads.
Garden lands are usually scattered amongst and around the owners' settle-
ments, and it is general, though not in any way prohibited, that plots of adjacent
settlements do not intermingle. Plots are owned individually by women; each
wife's holding of cultivated and fallow land is scattered throughout the settle-
ment area. Older men may also own plots but this is not general, although with
the introduction of the ox-plough it is becoming more common. Women own
and, with their daughters, cultivate at least 9/10 of the garden land. The
appendix gives a summary of a small garden survey made in 1951 in three types
of hamlets. The average holding per wife in the survey (64 wives) was 1-63 acres,
and this appears to be a general optimum size. Again, the introduction of the
plough may well increase the area cultivated. Ploughs were first used in 1950,
being introduced by Indian traders. I have no records of actual production and
this must vary a good deal since the average rainfall is barely enough for agri-
culture, and perhaps in two years out of five the rainfall is either too small or
badly distributed, so that crops are severely damaged. Soils are generally poor
and are grossly overcultivated in the settlement area, where erosion is serious.
On the whole agriculture is a rather precarious occupation; yet on the average
enough crops are produced to make cereal food a staple part of Jie diet.
Garden land is the property of the woman who first tilled it, or her proper
successor. Untouched bushland is free to all women, irrespective of clan or
settlement, if they wish to clear it and establish rights of occupation. Even
though a plot is left untouched for many years and seems to have returned to
bushland, it still remains the property of the woman who last used it and may
not be used without her permission. Nowadays her rights will be upheld in the
native courts. There is no notion of a clan or settlement having jurisdiction over
any land, nor of clan or settlement elders having authority in the matter. When
a woman dies her daughters may continue to use her plots if they wish; if she
was an old woman they would probably already have taken over most of her
cultivation. Usually, after marriage when the daughters go to live in their
husbands' settlements, they give up their land in their natal settlements and
obtain plots more conveniently near their new homes. A new wife expects to
be provided with adequate garden land near her husband's homestead, or
rather she expects her husband to support her for she herself must take the
initiative as cultivation is a woman's affair. Land which is not taken over by
daughters or sons' wives does not revert to commonage but remains under the
control of the dead woman's house i.e., the group formed by her husband and
his full-brothers, or, if they are all dead, that formed by her own sons with their
66 UGANDA JOURNAL
wives and children. The senior wife of this group assumes control over vacant,
but formerly tilled land. Other wives of the group have of course the right to
any of this land, either immediately or later on when their own is to be left
fallow. Once a particular wife begins to cultivate a plot, however, it comes
thereafter under her sole control, whether it is cultivated or fallow. A house
which has a surplus of land due, say, to a diminution of the number of wives,
can easily be persuaded to give it to a neighboring wife or kinswoman and
such a gift is non-returnable. A friendly request of this sort is rarely refused
unless the land is obviously required at the time. Land is still plentiful enough
for its value to keep low, far below the values of kinship and good neighbourli-
ness. There is, in general, no land problem, but one might well arise if the people
could be persuaded to rest much of the land immediately surrounding the
homesteads, which has been overcultivated for many years. Under the present
system, however, even in the more central settlements near which new bushland
is not very plentiful, the size of population is not so great as to make a real
shortage. Nevertheless, population is rapidly increasing and is likely to be an
important factor in the next decade or so under the present, traditional, mode
of settlement. Already women who live near the periphery of the settled area are
beginning to take in virgin land outside it. As the area of settlement in Jieland
is small this outward spread is not difficult. It may be added that there is no
cultivation in the pasture lands to the east and west, away from the. homesteads
of the centre of the country.
The crude agricultural system is that a plot is used until by its crop it shows
marked signs of falling fertility-the Jie reckon 3 to 6 years-then it is left
fallow for several years; fallow land is rapidly covered with new low bush.
Usually each wife has about three lots of garden land, one in use and two
fallow; but there is no wholesale changing over from one lot to another at any
given time, merely a gradual relinquishing of some plots and a taking up of
others. After the hot, dry season of at least six months, the Jie do not like to
begin cultivation before the first rains in late March or early April have softened
the baked earth. By this technique the people frequently miss the valuable
impetus afforded to their crops by these early rains, and only too often their
gardens suffer from later spells of drought. Harvesting begins about early
Agriculture is women's work; some men assist them, but never in my exper-
ience are the younger men to be seen at work in the gardens. They are common
sayings of the Jie that "Men own cattle, women own gardens", and, "Sorghum
is the cattle of women".1 This division of labour has been given a certain social
rigidity by long tradition. As the Jie see it, only the men can adequately protect
the herds (and this applied especially in the old formative days of tribal warfare),
and domestic animals are clearly tied up with the solidarity and continuity of
the family, with marriage, with cognatic and affinal kinship relations, with
ritual and law, and formerly with warmaking. All these in this country are
primarily masculine concerns, and therefore it is in'keeping that men should
1 Just as men have an ox-name derived from a favourite bell-ox, so women have their
'ox-names' based on gardening activities. Cf. my 'Bell-oxen and ox-names among the Jie'
(Uganda J., 16 (1952), 72-5).
also be responsible for the practical side of pastoralism. It may be added, how-
ever, that in the homesteads women and girls are the dairymaids. To the women,
the stay-at-homes, the food providers, go the duties of cultivation. To the
women also go the privileges of their role. They can plant as and when they
like, and when harvested the crops are entirely their own to use and dispose of
as they wish. Husbands have a right to cereal food and beer, but they cannot
legitimately complain if their wives sell or give away part of their stores. Each
wife has her own granary baskets in her own yard in which she often also keeps
her other private possessions. Conversely, the women have a right to milk from
the herds but have no voice in the use and disposal of the animals. This descrip-
tion is, of course, in the most formal terms; actually these matters are arranged
within the normal give-and-take of marital and family life.
It was at one time generally supposed that the Jie, along with the Karamo-
jong and the Dodoth, had little time, opportunity or inclination for agriculture.
It is now realized how facile this assumption was. The Jie have a mixed economy
wherein cereal foods are really no less important than animal foods; and this
has been the case for very many generations. Unfortunately however the Jie do
not understand the values of mixed farming, the use of manure, etc. The
expressed ideal of these people is to produce enough grain to suffice the
inhabitants of the homesteads (some 4/5 of the population) throughout the dry
season as a staple item of diet. Their produce, eked out with additional supplies
of grain bartered from their western neighbours (Labwor, Acholi, Ngiangeya),
and with milk and some meat and blood, affords the people a continuous
although exiguous cereal diet. Indeed, since most homesteads are without dairy
herds in the latter half of the dry season, the people are almost entirely dependent
on agricultural produce, self-grown or bartered. My wife kept records of the
daily diet in two homesteads in the early wet season, 1951 (i.e., before a new
harvest). Scarcely a day passed when cereal food was not eaten; it frequently
formed almost the only food until the dairy herds returned in May and the milk
supply was resumed. Even then and until harvest time, porridge and especially
beer continued to be a main food. Much of this food was local produce, but
the new money income (from cattle sales) and the increased desire for cereals
are resulting in increased demands for cereals at the local trading stores.
So often has it been remarked that the pastoral peoples of eastern Africa look
down on agriculture that it is worthwhile pointing out that this attitude is not so
universal as might be thought. At any rate the Jie and their cousins of the
Karamojong Cluster do not have this attitude. In this matter we must first
distinguish between the purely economic, and the social and ritual, aspects of
animals and agriculture. Secondly we should be careful to distinguish between
disliking or despising agriculture as such, and despising people who have few
or no stock and are purely agricultural, or hunters and gatherers.
From the purely economic point of view-that is in terms of the relative
importance of animals and agriculture as providers of foods and other necessi-
ties, and the time spent on each-such well known pastoralists as the Nuer
(Evans-Pritchard, 1940, Ch. 2 and passim) and the Nandi (Huntingford, 1951,
Ch. 4) as well as the Jie and the Karamojong have a mixed economy wherein
the practice of agriculture is quite essential for the maintenance of traditional
68 UGANDA JOURNAL
and modern standards of living. Indeed we may add that these peoples would
starve without their agricultural produce. On the other hand, the Masai are
reported to have no agriculture at all. The nomadic Turkana of Kenya cultivate
where they can, but production is pitifully small and of little importance in
their total economy; much the same can be said of some of the camel nomads of
northern Africa. Apart from the strongly pastoral peoples, there are of course
many African tribes which have a predominantly agricultural economy, and
derive little or no economic benefit from their small herds. On the other hand,
it is a truism that many tribes-perhaps the majority, even in the modern age
of money-of eastern Africa give a high status to domestic animals from a
non-economic point of view. In ritual and religion, in kinship and community
affairs and in law and war, domestic animals, chiefly but not only cattle, are of
the utmost importance. A mixed economy or one heavily weighted in favour of
agriculture can exist in a society where livestock are of great significance,
perhaps overwhelmingly so, in other spheres of social activity.
On the second point I must emphasize that the Jie and their companion tribes
of the region do not in any way despise agriculture and its products. Everyone
wants as much cereal food as possible; porridge and beer are both as much
elements of everyday and ceremonial diet as blood and milk. These peoples do
tend to feel superior to agriculturalists, but this must not be overemphasized.
The Jie's western neighbours, the Acholi, have nowadays few stock as compared
with the Jie themselves, but they have what the latter feel to be a good and
regular supply of cereal food, and for this they are envied and even respected.
The nomadic Turkana openly envy the agricultural fortunes of their Karamo-
jong neighbours, which are denied them in their semi-desert homelands. On the
other hand the Nuer of the Anglo-Egyption Sudan regard agriculture as an
unfortunate necessity (Evans-Pritchard, 1940; p. 80) whilst the Masai, who
practise no agriculture at all, and some of the Saharan camel herders with a
more truly mixed economy (Cline, 1950) both openly despise agriculture and its
practitioners. It is not easy to understand these divergent attitudes, and the
range of social variations and the shortcomings of the data should preclude
facile generalizations. The Saharan herders' attitude, as that of the Arabs, seems
probably to be associated with the institution of slave labour. Agriculture
became one of the principal tasks of slaves and has therefore become associated
with social inferiority, poverty and loss of freedom. Conversely the Turkana can
grow very few crops and would like to have more; this attitude comes from the
culture which they brought with them from the rather better lands to the west.
On the other hand the Masai could easily undertake agriculture if they wished,
with no more difficulty than the Jie already experience, for parts of their country
are reasonably well watered and fertile nor are the Masai so truly nomadic as,
for instance, the Turkana and Bedouin. Perhaps like many other social attitudes,
e.g. the wearing of clothes and nudity, or the abstinence from certain foods, they
are ossified cultural conventions whose explanation lies beyond historical reach.
JIE AGRICULTURE 69
THE EXTENT OF CULTIVATION IN THREE JIE HAMLETS
This is a summary of the results of surveys carried out at the end of May 1951,
when all the garden land had been cleared and sown for that year's crops.
HAMLET A. Lothorgut old settlement containing the clan-hamlets of Lothorgut and
This is an old established settlement in Kotido district, in the centre of the settled
area. Most or all of the land has been cultivated on and off for many generations,
and suffers from heavy erosion. Yields are frequently poor and always uncertain.
HAMLET B. Lodera, a 'colony' of Lothorgut, on the edge of the Kotido district.
This is inhabited by a group of people who moved here some three years pre-
viously. Much of the land had been under cultivation for less than that time. Yields
are both higher and more reliable than at Lothorgut, three miles away.
HAMLET C. Etagwara clan-hamlet, Lokuruk settlement.
This is an old established hamlet in the eastern part of the settled area (Kapelimuru)
where rainfall is normally more favourable than at Lothorgut though no more than
six miles from it. Some of the gardens had only just been cut out of the peripheral
bush, others had been in use periodically for generations. Yields are almost always
rather higher than in central settlements.
The number of homesteads, population, area of cultivation, etc., at 3 Jie hamlets.
Number Number Acreage Number Acreage Acreage persons in
Hamlet of of oard of per owned 'average
s people p end wives wife by men family
steads per head unit'
A 6 168 0-21 29 1-74 5-3 5-8
B 4 63 0-35 12 1-86 0-1 5-3
C 6 99 0-3 23 1-3 0 4-3
1 These figures only apply to the area actually under cultivation and not to production.
As compared with more favoured parts of Africa, the yield per acre is doubtless low, due
to inadequate, badly distributed rains and poor, overcultivated soils.
2 The 'average family unit' is the number of people per wife in a given area; e.g., in
Lothorgut the unit is 5'8 people, indicating an average of one wife with 4-8 dependants
(husband, children and others). The wife is taken as the basis of the calculation of the
number of people dependent upon the food products of a group of plots, because women
own at least 90 per cent of all garden land.
3 The higher amount of land owned by men at Lothorgut (Hamlet A) is due to the use
of a plough. One of the first ploughs in Jieland is owned by a man here, and he lent it to
friends and also hired it out. One plough is in use at Lodera (Hamlet B), but there the
owner and his friends ploughed up land for their wives and not for themselves. There were
no ploughs at Etagwara (Hamlet C).
70 UGANDA JOURNAL
4 Comparable figures are appended for (a) a tribe with a mixed economy and (b) two
tribes with basically agricultural economies:
(a) Nandi, Kenya Man, wife and 2 or 3 children: circa 0-75 acres (Huntingford,
1950; p. 62).
(b) Lamba, N. Rhodesia ... Family of 5-6 people: 5-9 acres (Allan, 1949; p. 33).
Ngoni, Tanganyika. .. Family of 3-8 people: 1-95 acres (Gulliver, unpublished).
Allan, W. (1949). Studies in African land usage in Northern Rhodesia. Rhodes-
Livingstone Papers, No. 15, p. 33.
Cline, W. (1950). The Teda of Tibesti, Borku and Kawar in the Eastern Sahara.
Banta General Studies in Anthropology (U.S.A.), No. 12.
Evans-Pritchard, E. E. (1940). The Nuer. Oxford: Clarendon Press.
Huntingford, G. W. B. (1951). Nandi work and culture. London: H.M.S.O.
WESTERN UGANDA RAILWAY EXTENSION
M10taO 5 0 w a so 0 o hO Sn Bo 00 aO*aies
THE RAILWAY EXTENSION TO WESTERN UGANDA
By P. H. HICKS, B.Sc., A.C.G.I., A.M.INsT.C.E.
THE extension of the Uganda Railway to the western boundary of the
Protectorate has received more or less consistent support for many years.
Only a series of military and political impediments-such as have occurred
during the past 50 years-could have deferred a project having such potenti-
ality for the development of the territory.
The history of the proposed extension is singularly marked by deferments.
As a first practical move to study the possibilities, Captain (later Major-General)
A. G. Stevenson, R.E., carried out in 1907-8 surveys of two alternative routes
to Lake Albert, one trace leading to Butiaba and the other to the southern
end of the lake. Neither project was carried through then for financial reasons,
but in 1914 the Governor of Uganda (Sir Frederick Jackson) and the General
Manager of the Uganda Railway (Major H. B. Taylor) gave renewed considera-
tion to the construction of a railway link with the southern end of Lake Albert.
In 1914 the Great War intervened, prohibiting further progress, and it was not
until six years later, in 1920, that the Uganda Development Commission pub-
lished a report urging the construction of a railway from Kampala to Mityana,
with subsequent extension to Toro. This plan, however, had once again to be
abandoned in favour of the proposed Jinja-Kampala extension of the main line
from the coast.
In 1929, Mr. H. F. Varian submitted a report on a possible railway route to
Western Uganda, and in the same year the Colonial Development Act gave
encouragement to further investigations. A detailed engineering survey was
therefore carried out in 1930 and 1931 by Mr. H. F. Birchal, O.B.E., of the
Kenya and Uganda Railways and Harbours, and an economic survey of the
route was completed at the same time. This route followed the main road from
Kampala to a point beyond Mubende and then proceeded to Kasese along a
trace skirting the northern end of the Lake George swamp, finally ending at the
border with the Belgian Congo. For various reasons, connected mainly with the
financial depression that followed, the project had yet again to be abandoned;
and when interest might have re-awakened yet another world war intervened.
Finally, in December 1950, a new engineering survey was urgently called for
and a Committee was set up to carry out an up-to-date economic survey. In
order to make it possible for decisions regarding the construction of the line to
be made concurrently with those regarding the development of Kilembe Mine-
each project largely affecting the other-the report and estimates of the survey
were called for urgently for completion by the end of April 1951. The author
was appointed Resident Engineer.
The ruling gradients specified for the railway are those of main line standard,
i.e., 1-50 per cent. compensated against UP traffic and 1-18 per cent. com-
pensated against DOWN traffic, and a maximum curvature of 8' was fixed. The
country to be traversed consists largely of bush, forest, elephant grass and papyrus
swamp. The trace follows flat swampy valleys for the first 150 miles, and then
enters broken hilly country in the faulted region on the east side of the escarp-
ment overlooking Lake George and Ruwenzori. Here it is necessary to find a
route to descend 1,200 feet down the Rift Valley escarpment and debouch on to
the large swampy plain lying north of Lake George. The choice of a long
divergence of the railway around the northern edge of the swamp, or of a direct
crossing of the plain, posed a problem which required to be solved under great
pressure for time.
With less than five months in which to prepare, survey, estimate and report
on about 250 miles of railway, careful planning was more than ever necessary.
During December 1950, efforts were made to find staff to form four field survey
parties, and in addition-due to shortage of personnel-to arrange for two field
parties on contract to carry out the final staking of the first 45 miles to Mityana,
the construction of which had by then been authorized under the financial
guarantee of the Uganda Government.
The general location of the route as far as the border with the Belgian Congo
was defined by the route to Mityana, the valley of the Katonga River and the
Lake George Escarpment. Controlling points were examined by aerial recon-
naissance and by stereoscopic examination of the aerial photographs taken in
1950 for the mapping of Uganda by the Colonial Survey Directorate. The
'controlling points along the new route having been viewed in general from the
air, and followed up by ground study, it was then necessary to examine the
practicability and economy of the lines of the route between these points by
:survey in further detail.
While such reconnaissance was in progress, field equipment and staff for the
survey were got together-with some difficulty-during December 1950. By the
beginning of January 1951, however, staff, equipment, and the essential African
porter corps began to come together. As much of the country to be traversed
was then devoid of motorable tracks, porterage by head loads had to be
arranged despite the feelings now held against this practice in modern Uganda.
Much is therefore due to the cheerful co-operation of the volunteers who were
recruited to the task.
Field work built up to its full output by the end of January and was main-
tained at high pressure during February and March; all parties were recalled
from the field at the end of March. In about ten weeks a total of 247 miles of
railway location was mapped, partly by the revision of some of the 1930
-survey plans by H. F. Birchal and the Katonga Canal Project by E. P. Nicolay,
and by aerial mapping of certain sections. This result still left 145 miles of
reconnaissance or detailed survey to be carried out on the ground. The field
,conditions of isolation and the abundance of big game of all kinds- elephant
and buffalo in large herds-gave a background interest to the technical side of
,the work. The final withdrawal of all field parties by the end of March on
'completion of the bulk of the task was fortunately achieved, as the long rains
then broke with unusual violence.
In the meantime, the final staking of the line westwards from Kampala was
proceeding through the thick forest fringes of the Mayanja Kato valley. By the
'end of April about half the distance to Mityana had been completed, and the
final plans were under preparation. The field works at this stage were continued
,under difficulties due to heavy rains and flooding.
When at the end of March 1951, all parties had returned from the field, a
THE RAILWAY EXTENSION TO WESTERN UGANDA 73
month remained to reduce the field work to terms of construction plans and
costs. The computations and plotting were distributed in sections for comple-
tion, and the plans, sections, abstracts and earthwork and bridging quantities
were then produced by all the field staff working together, in a steady flow
calculated to be complete before the end of April 1951. Meanwhile the report
and estimates were written section by section around the plans and figures as
these were completed. Consequently the final production of the official report
and estimates was in hand while the survey work was under process of statistical
In May 1951, the report and estimates were submitted in due time, five months
from the start of the work, the field staff were mainly dispersed once again to
various spheres throughout East Africa, and by June 1951, the direction of the
work was turned to the completion of staking out the first section of railway to
Mityana, prior to its construction.
The preceding notes outline only briefly the background history, the pre-
liminary work and the technical results. The multiplicity of detail behind the
final abstracts of costs and figures remains-possibly to all but the engineers-
submerged in innumerable plans and field books, quantity sheets, roughed-out
and final drawings, and in appendices and various other arid data in the tech-
nical reports. The original staff and planning dispositions and the day to day
events are now buried in the welter of files inevitably accumulated in the
development of an organization responsible for the expenditure of several
In order to try to enlarge the detail slightly, it may be of interest to describe
more fully a few selected points. One of the commonest misconceptions of the
organization of the railway extension is that the work progresses steadily from
Kampala; mile by mile it is assumed to make a laboured but continuous advance
to the border with the Belgian Congo. If this were so, nothing could be simpler
for the Resident Engineer. Progress, planning and technical problems could all
be dealt with one by one in an orderly and unhurried fashion. The time required
to do this would, however, necessarily extend to rather longer than is acceptable
to the planners who specify the time limit. This puts the matter of organization
in an entirely different light; it is necessary to plan not only the first stages of the
project, but sections of work along the whole route, simultaneously from the
outset. In one instance of this, to survey the whole extension in the detail and
limits of accuracy required within the time allowed, the basic consideration was
the amount of work an 'average' survey party of fixed composition could cover
in that time. Here as usual, all stands or falls by the value of the word 'average'.
The route must be divided into sections which can be covered by the 'average'
party in that period, not forgetting to take into account the 'average' conditions
of the ground. The general alignment of the route having been determined and
taking account of unknown gaps in the route, the straightforward sections of
country could be apportioned to survey, while detailed attention was concen-
trated on controlling points such as heights of watersheds, grading and
curvature problems of the line in hill sections, and the restrictions and economics
of various earthworks. The two major problems of the route were the descent
74 UGANDA JOURNAL
of the Lake George Escarpment on easy grade and curvature, and the crossing
of the wide swamps north of Lake George.
Each survey party operated independently from selected starting points, and
the survey traverses were based on bearings and levels tied to the main land
survey system. Over much of the area traversed, the triangulation system is,
however, sparse and scattered, and some of the original stations in flat bush-
country were most difficult to locate and determine. As many of the original
beacons were stones buried several feet underground some forty or fifty years
before, and as the once-cut bush and trees have regenerated thickly, the difficul-
ties were by no means negligible, and in some cases the triangulation work had
necessarily to be completed after the close of the main survey.
The field survey work provided the drawings and basic plans. The next step
was to plot a railway route on the ruling standards of grade and curvature,
adjusted, amended and readjusted to economize. Finally earthwork and other
quantities were extracted mile by mile, classified and totalled and the whole
summarized in a handful of intelligible figures.
The calculation of earthwork quantities is a relatively simple task, although
laborious. But at the preliminary stage the difficulties of estimating types of
subsoil-whether hard or soft rock, waterlogged or dry-pose a special problem
whose solution is not assisted by the lack of general knowledge of the whole
area. The problems of bridging and the provision of waterways can also be of
considerable technical difficulty and complication, particularly in country where
hydrological data-or even local traditions-are not available.
Although beginning and ending in the wetter zones, the major part of the
railway route lies in the drier regions of Uganda. As regards waterways the
general first impression is of sluggish swampy rivers and little run-off throughout
the year during normal seasons. Conditions in early 1951 were, however, far
from being normal; the rains broke very early in March after unusual sporadic
outbursts earlier still, with the result that all the field work was much handi-
capped. In March, the weather became so adverse that by the middle of the
month large areas of country were flooded. Once-dry grassy valleys were swept
by freshets which developed in a few hours into torrents several feet deep, then
dwindled, leaving behind wide swathes of flat-beaten grass and bush. This
occurred even in the drier regions of the territory, and the difficulties of moving
stores and equipment-particularly the essential rations-in and out of the
waterlogged areas were a constant problem. Two field parties were for a short
time isolated, except by runner over distances of 20 or 30 miles, and another
party could only be rationed by a lorry travelling 400 miles via Fort Portal,
Katwe and Mbarara, although the location of their camp was only 175 miles in
a straight line from Kampala. The difficulties of field work, of recruiting labour, of
administration and, lastly of obtaining the survey data and plotting it under field
conditions, were appreciable at the time, and the ultimate achievement reflects
credit on the men-European, Asian and African-who carried through the work.
From the location of the line on paper to calculating earthworks and bridges,
the work must take a multitude of details into account, e.g.: the spacing of
stations to suit the traffic operating and to fulfil public necessities; the watering
of locomotives at required intervals; the arguments for tunnelling at one site or
Copyright reserved Photo by East African Railways and Harbours
Survey under relatively easy conditions.
Copyright reserved Photo by East African Railways and Harbours
Tractors and scrapers build the line across a swamp.
[ ace p. 74.
Photo by East African Railways and Harbours
Bridge construction in a typical swamp.
Copyright reserved Photo by East African Railways and Harbours
The bridge completed: note standing water level.
~ .t'*. *. .* -
Copyright reserved Photo by East African Railways and Harbours
Kasese landing-strip near the terminus of the railway. Ruwenzori in the
Copyright reserved Photo by East African Railways and Harbours
A construction train at work ballasting the line.
Photo by East African Railways and Harbours
Gangmen packing the newly laid line.
Copyright reserved Photo by East African Railways and Harbours
The finished railway running across a swamp.
THE RAILWAY EXTENSION TO WESTERN UGANDA 75
of constructing a viaduct at another, as against other alternatives. The details
summary, on which the estimate of costs can be based. Much of the detailed
work is unfortunately complicated by the fact that the possibilities and the
solutions are infinite; queries may arise in turn, about compensation for land
and crops, the platelaying organization, the housing of the staff, the provision of
water supplies, and so on. The problems posed do not fail in variety and interest.
A general picture of the scope of the work can be obtained from the quantities
and figures. The total amount of earthworks in cuttings and embankments will
probably exceed 200,000,000 cubic feet of soil, soft rock and hard rock. Nearly
300 miles of drainage of various sorts must be constructed, and many miles of
.concrete pipes, 2 feet and 3 feet in diameter, will be installed in culverts. Con-
crete for bridges and arch culverts, while amounting in some individual items to
considerable totals, is for the whole extension a relatively small item; neverthe-
less, the total will amount probably to about 800,000 cubic feet. Some 4,000
track signs, mileage posts and suchlike will require to be made and distributed
over the 209 mile route. Over 5,000,000 cubic feet of selected ballast will be
excavated, loaded into construction trains, and distributed. About 800 buildings
of all sorts (station offices, staff quarters, etc.) will be built; telegraph lines and
instruments to serve the whole route to Kasese will be installed. Many miles of
pipe lines to provide water supplies for locomotives will be laid and pumping
Over and above the permanent works to be built, housing and administrative
arrangements for the care and welfare of up to 100 Europeans and Asians and
5,000 African staff have to be arranged and provided. Finally, the progress of
these must be arranged so that platelaying and building construction can
advance without delay--camps and stores being ready in advance-as each
section of earthworks is completed.
The building of a railway thus involves many problems besides those of
plain engineering, and the field to be covered may be surprisingly wide. Behind
all, economics, inevitably, influence every technical decision, and political con-
siderations may suddenly become important. Science and engineering alone do
not solve all the problems. Lastly, it is likely to be forgotten that some of the
most difficult work may have been done, not in the works actually constructed
but in preparing for others not carried out in the final event.
The author is grateful to W. Urquhart, Esq., O.B.E., B.Sc., M.Inst.C.E.,
authorities-too numerous to name-not concerned in the actual construction
which has made it possible to overcome many practical difficulties. In particular,
acknowledgment is due to S. S. Tindall, Esq., C.M.G., late Uganda Develop-
The author is grateful to W. Urquhart, Esq., O.B.E., B.Sc., M.I.Inst.C.E.,
Chief Engineer of the East African Railways and Harbours, for permission to
publish this paper, together with the illustrations which are taken from official
UGANDA PLACE NAMES
The following observations may serve to supplement H. B. Thomas and I. R.
Dale's study of Uganda Place Names (Uganda Journal, 17 (1953), p. 101).
BROOKES' CORNER. Before Mr. Brookes' arrival this corner was commonly
known in Soroti as 'Mile 17'. His house was in the form of a tower or the keep
of a castle, with a garage and workshop nearby, and the whole was often
referred to as Brookes' Castle. He was a 'character', with a touch of genius
about him. I recollect that he invented a clever gadget for towing cars: this could
be quickly fastened to the track-rods of a car which, when towed, automatically
steered itself and so obviated the need of a driver.
GREEK RIVER. At one time there was a police post on the south side of the
river crossing, and I think it was to this camp site that the name really applied-
This post was manned from Mbale, and was the last camp on the plains before
travelling north towards Karamoja or round Mount Elgon to Kitale. The police
were called on to check native traders and travellers into and out of Karamoja,.
and to prevent ivory-smuggling, stock-thefts and border raids by the Karamo-
jong. In the dry weather the Karamojong were in the habit of grazing up the
Siroko valley as far as Muyembe, and this sometimes led to fights. The lower
Siroko valley was mostly uninhabited and the Karamojong seemed to me to
have a strong claim to grazing rights there.
LAKE KIRKPATRICK. On earlier maps (e.g. the Survey Department's 1915
Sketch Map of the Eastern Province) this was shown to the north of Aketa: but
it was dropped from the 1928 map. About 1930 I spent two days at a place then
known to the Ateso as 'Kiriki' (no doubt the Greek River in Teso which is
referred to in the article); this was a day's safari north of Aketa. There were two
or more fair-sized watering-places or lakes obviously part of the bed of a
shallow river. I wondered then whether this was what was once referred to as
Lake Kirkpatrick. These were dry-weather watering-places for the Ateso and
even more for the Karamojong. R. C. U. FISHER.
Charmwood, Farnborough, Kent.
14 October 1953.
Sir William Gowers, K.C.M.G. writes:
I recently heard of another thing that might have seemed to Macdonald an
additional reason for commemorating Roddy Owen at the Owen Falls-that
Owen made the first enlistments of Sudanese under the British flag as from
1st April 1893 (Lugard's Sudanese having been enlisted in the I.B.E.A.
Company's service). The 'long-rolls' prepared by Owen showing their names
(450 of them) are still extant among the Nairobi archives. This gives the
4th (Uganda) Battalion of the King's African Rifles the right to claim seniority
as 'soldiers of the Queen' in East Africa. I hope some allusion will be made to
this when the 4th K.A.R. are furnishing the guard of honour when the Queen
opens the power-station at the Falls which are called after their first officer. It is
also appropriate to remember, knowing Her Majesty's love of horses, that the
works will be opened within a few days of the anniversary-the 62nd-of Father
O'Flynn's win (R. Owen up) in the Grand National of 1892.
Sir John Gray writes:
With regard to Mount Gordon Bennett it may be remarked that in 1888-9
Stanley at first insisted that the Ruwenzori which he then saw was not the
mountain he had caught sight of in 1876 on his way to what is now known as
In Stanley's In Darkest Africa ii, 293 (1890) Mount Gordon Bennett is an
"isolated outlying fort" of the Ruwenzori range, and in the accompanying bird's
eye view it is an isolated 'sugar-loaf' to the east of the range with a coating of
snow on top. It had already been depicted as an isolated 'sugar-loaf' in the
illustration facing p. 427 of Through the Dark Continent, vol. i, (1878); and,
becoming lyrical (chap. xv, ibid), Stanley refers to "that King of mountains,
Mount Gordon Bennett, which towers sheer up to the azure with a white veil
about his crown".
Writing on 8 September 1888 to Colonel J. A. Grant from Batundu on the
Ituri River, Stanley says, "My interest is greatly excited, as you may imagine,
by the discovery of Ruwenzori I conclude that it is not Mount Gordon
Bennett, seen in 1876 (though it may be so), which the natives said had only
snow occasionally. At the time I saw the latter there was no snow visible. It is a
little farther east, according to the position I give it, than Ruwenzori" (The Story
of Emin's Rescue as told in Stanley's letters by J. Scott-Keltie (1890), p. 152).
In a long letter to the Daily Telegraph dated 18 January 1876 from "Kawanga
frontier village between Unyoro and Uganda" Stanley gives details of informa-
tion received regarding Gambaragara, "Snow is frequently seen on it, though
not perpetual. Upon its summit dwell the chief medicine-men of Kabba Rega, a
people of European complexion". The letter is printed in the Journal of the
Royal Geographical Society, 46 (1876), p. 27, and is followed on p. 33 by a
letter from Stanley's assistant Frank Pocock to his parents dated 18 April 1876.
Describing the journey to Lake George, he says, "On the road we passed a fine
mountain covered with snow". Did Pocock see with his own eyes snow, which
Stanley had failed to detect? In the region of Lake George, Pocock would only
have been about 45 miles from the Portal Peaks of Ruwenzori. If he did actually
see snow, Stanley may have disbelieved him, as he professed to disbelieve
Surgeon Parke and Mounteney Jephson in 1888 (see Uganda Journal, 2 (1934-5),
249). Was history in 1888 repeating what had happened twelve years before?
Was Stanley anticipated on each occasion by one of his companions in catching
sight of snow on Ruwenzori: and was he, reluctant to accord precedence to his
subordinates, led to deny credit where credit was due?
Incidentally Stanley (In Darkest Africa, ii, 290) reproaches the American
Colonel Mason Bey for "failure to see what ought to have been seen". In
fact, in the course of his circumnavigation of Lake Albert in 1877, Mason saw
"a large isolated mountain" (Proceedings R.G.S., 22 (1877-8), 225) to the south
of the lake, i.e. about 45 miles from the place where the Semliki enters the lake.
If therefore Stanley is, at his own request, to be allowed to deprive himself of
78 UGANDA JOURNAL
the distinction of having discovered Ruwenzori in 1876 the honour should be
given to Mason.
I might add that when, in 1876, Stanley fastened the name of Beatrice Gulf on
the present Lake George, he was convinced that he had reached the southern
end of Baker's Albert Nyanza. This is clear from the letter of 18 January 1876
referred to above. But in Through the Dark Continent published on his return
to Europe in 1878, when the report of Mason's circumnavigation would be
available, he plays for safety and refers to it only as Muta Nzige. It is curious
that, fortified by firm knowledge obtained in 1888, Stanley was not above a
small gibe at Samuel Baker for having, in 1864, believed in an apparently
boundless extension southwards of Lake Albert (In Darkest Africa, i, 306).
But these problems are perhaps less geographical than psychological ones, to
be resolved only by a proper understanding of Stanley's mind, which, in the
words of Chief Justice Hale, "is not triable for the devil himself knoweth it not".
CORRIGENDUM. Uganda J. 17 (1953), p. 110 LAKE GEDGE-the date of the death of
Mr. Ernest Gedge was 1 August 1935 (not 1932).
ANCIENT EARTHWORKS IN WESTERN UGANDA
I should be grateful if the following correction and addition to my paper
'Ancient earthworks in western Uganda' (Uganda J., 17, (1953), 51-62) could
The correct name of the earthwork referred to as 'Kalisisi' (pp. 52, 55, 60, and
plan) is 'Kisalizi'.
Recent investigation has revealed that what has been thought to be an earth-
work at Kijwenge (pp. 52, 59, and plan) is, in all probability, a natural feature;
a similar feature lies near the earthwork at Madudu in Buwekula County. The
trench in question lies further west than the Kijwenge site shown on the plan. It
is known as Nongo and lies in lat. 1V 5' 30" N., long. 31 14' 00" E.
E. C. PLANNING.
C/o Royal Geographical Society,
19 August 1953.
In an Editorial Note to Mr. E. C. Lanning's article referred to above it was
implied that he had been stationed in the Masaka-Mubende area since 1950. In
fact his first connexion with the area was in January 1952. [EDS.]
AFRICAN HIGHWAY. The Battle for Health in Central Africa. By SIR MALCOLM
WATSON, M.D., C.M.. D.P.H., LL.D. London: John Murray, 1953. xxix+
294 pp. 7 maps, 79 plates, 17 text figures. 3 3s.
A book appearing over the signature of Sir Malcolm Watson is approached
with anticipation of good things to come. It is, however, difficult to understand
why this particular title was chosen and the author's exact objective. The book
is not an account of "the" battle for health in Central Africa-it deals with some
of the work done to control health conditions in the copper belt of Northern
Rhodesia, where a group of industrialists started to open up four copper mines
some twenty-five years ago. Realizing that the success of their venture might
well be jeopardized by sickness due to the many tropical and other diseases
abounding in the area and the necessity for establishing early control over these
and for decent sanitary conditions for their labour officers, the Directors sought
active help and advice from the Ross Institute of Tropical Hygiene, where, in
1929, Sir Malcolm was Director of the Malaria Department.
In the Part I of the book, we are given some account of the results of this
approach with particular reference to the attack on malaria, as planned by Sir
Malcolm and executed by the staff selected by him and working under his
guidance. They were faced with formidable problems, Anopheles funestus,
one of the principal vectors of malaria in the area, proving to be a difficult and
stubborn foe. As one reads this part, there emerges a picture of planned control
well up to the standard of work achieved in Malaya by this eminent malario-
logist, and he tells the day-to-day story in extracts from reports by the men
actually working on the problem. To these pioneers Sir Malcolm pays a well-
In Part II the author deals with some of the many other health problems that
demanded attention if the African and European staff were to profit fully from
being relieved of the former menaces of malaria and blackwater fever.
In each part an attempt is made to present the story in terms mid-way between
a scientific discussion of the problems and what is, quite frankly, a eulogy of the
administration for having had the intelligence to apply in the right quarter for
advice on these matters and for implementing that advice when it was given.
Thus to the reviewer the book falls to some degree between two stools and at a
price of three guineas it is not easy to decide to whom it is addressed. It contains
much valuable information for the medical reader but it is far from being a
medical text book; at the same time it is too little of an 'advertisement' for the
The book is well printed and presented with an adequate index and many
maps, charts and photographs. C. A. BozMAN.
NOTICE TO CONTRIBUTORS
Not all contributors to the Journal will have access to facilities to enable them to
follow strictly the accompanying notes. Inability to do so should never be allowed to
discourage the submission of a manuscript; but as close adherence as possible will
save both contributors and editors much unnecessary work in the preparation of the
material for the printer.
1. All correspondence should be addressed to the Editors, The Uganda Journal,
The Uganda Society, Private Bag, Kampala, Uganda, unless in answer to a letter
received direct from an Editor giving his own postal address.
2. Text. MSS should be typed double-spaced on one side of paper leaving top and
left hand margins at least an inch wide. Two copies should be sent. Only words to be
printed in italics should be underlined.
3. Latin names of plants and animals. The name should be underlined and, when
first mentioned, should be followed by its author, e.g., "Balearica regulorum
gibbericeps Reichenow". It is desirable to refer the nomenclature to some standard
work, if such is available.
4. References. Quotation of sources in the text should be clear, and full references
to them should be given. In all cases the author names and initials, the year of publi-
cation and the title are required; for a book, the place of publication and the
publisher should be added, for a paper in a periodical, the name of the periodical, the
volume number, and the pages of the beginning and end of the paper. Care should
be taken to verify references. Volume numbers, to avoid confusion, should be given
in arabic numbers even if originally given in roman. Titles of papers and books
should not be included in inverted commas. For simplicity and brevity the following
general arrangement is preferred: Sources are quoted in the text by the author's
name and the year of publication, e.g., "Carpenter (1920) states that .... ", or, "In
those days it was the custom .... (Speke, 1863, p. 345)". References are assembled
in alphabetical order at the end of the paper in the form:
Gumba, E., and Kafuho, E. (1946). Two Lusoga fables. Uganda J., 10, 17-24.
Thomas, H. B., and Scott, R. (1935). Uganda. London: Oxford University Press.
5. Tables should be on separate sheets and numbered in arabic numerals. Their
headings should'as far as possible make the tables self-explanatory.
6. Figures. Drawings should be in black Indian ink, on hard white paper, Bristol
board, or (white or blue) tracing paper or cloth. Shading should be by lines or stipple
(indicated by blue crayon), not by washes which require half-tone reproduction.
Intricate line work and lettering, as for instance on maps, are liable to blur in
reproduction; it is important therefore that such should be drawn on a large scale
to allow for reduction. Blue ink is particularly unsuitable for reproduction. The
scale of drawing is best shown by the inclusion of a measure divided appropriately
which will remain in the same proportion to the drawing whatever the reduction in
block making. Lettering, numerals, etc. may be shown in pencil to be inserted by
the blockmaker. The author's name and the number of the figure should be written
on the back of each figure. Legends for figures should be collected together at the
end of the MS.
7. Plates. Photographs should be submitted in the form of unmounted half-plate
glossy prints of good detail and medium contrast, or, if this is not possible, as nega-
tives. They should be numbered and each should bear on its back the author's name
and its number.
8. Reprints. Authors are supplied with 20 reprints gratis. Additional copies may be
purchased if ordered at least one month before the due date of publication of the
Journal issue in which the paper is contained. Costs vary between a maximum of
about Sh. 0/50 (6d.) per printed page, for a few copies of a short paper, to Sh. 0/10
or less for a large order of a long paper.
9. African Languages. Spelling should follow some standard orthography.
10. Place-names should be as they appear on current official maps.
11. Book Reviews. A full description of the book should be given, e.g.: THE
PLEISTOCENE GEOLOGY AND PREHISTORY OF UGANDA. Part II-Pre-
history. By Professor C. van Riet Lowe. ix+113 pp., map, diagram, 53 plates.
Geological Survey of Uganda, Memoir VI. Entebbe: Geological Survey Department,
1952. Shs. 30.
Tiik :Uii A6A s 6C Irw i n"I'l bi
R, A" --0- -d
Inmf sub iog- fv ,a b4p
Oifta ffisillfutiun'i S) 2 bsefit) rrO&As,4w
two j nm'bm, 4'a- -faimly to -all,rivilegci-of ifietybotship butjb, yr p
-o*, of 'the -,Iournal.
Tbf:AsslociteiMeiiibers theannOt'subscription-is S4s_2/5Q,- they m4y a'
-due, y on.-Istj4quary aTe, b1c,'; tO, the
OW h -P
p*, _44gjjsh cjpeAo, "a
-,.vho ig '4tdbojvIze4 ts pili, b16& of t
W -d, -Vi issue IwANO
p'," -,6;41it-drawAiafavo ofllle, nda bel Y
'itha, tryrilb& ot FaWjjgh n wsoa*rs -AN"jophAls. ,Tbp,,-
iW4- i4d 4reopepto rniii4iif. Nfu qay "to r' ay, 9
tQ 74'jpnj." be bi
-0,: m-"-,Rdokjs may anow
l "WiO set-out'in' rk'q' R.
T6 t I
jkh4r4y-' t%"S W
I is, 14ic of thill rn
11 bta"itr1se ku (tee Vvlmd-
Ome thi-SOO, w -Of-w S-of local
n EPIP ation
,G&W' o Me _k 'Id
a, Sn' es of bg& a-9,'
,4]tai 'L Z4
od kn RL.- 0.
..... . . .
,ooa by #T
WE JU, Nos.,l VoL J
a#:lAfrlcqa-,ond VTeanda'Niitur4l, HIS my $,Pdm. VQU., jt 14
40'ik So& 6,;U, 24, 29
'y drag ne,
a ir, oi Th so a6w?";
1 i Rho
o aAftd:-q4_ 4-*M'R vity afA&Mls'. A-T. r
M;WgJ94 :O, diid-,p
qme 4,,Tp,,v y.,_amiqh th' At'
4v u ffr
oVsherftm of '14 A*,gw, v
vn' brban'5tu Y
7 '6 3d
n 0)zp, t,4a PlaWu T
man; P6iers r'
'4`6itwkmjq A 949.', ZK pp,
'Pat torn Wer dW-4
I'M ljildiiiodMh lawAM CiW6tn,,, hdi loso,61, d.-
"Lod Uwagei'k b, UTO'lelm
*ctg` 4 n xh
A _iM UWt'S'L 4, '#j*ro& of"' lzlin'a oo' pidv
ppea "44rjy it S 6i,