REVUE cJ6

D’ETHNOGRAPHIE

SOCIOLOGIE
I^UJBLIEE PAH

LISTITUT ETHPGRAPBKIDE iTERMTIOMl DE PARIS

PARIS
ERNEST LEROUX, EDITEER
28 , RUE BONAPARTE, VI®
 Digitized by the Internet Archive
in 2016

https ://arch ve org/detai Is/b2241 9032

i .
THE NAMES OF BANTU LANGUAGES
(Londres).

In an interesting paper contributed to the Reuue dns Idees for January 15,1907,
M. Van Gennep lays stress on the importance of studying the noun-classes of the
Bantu languages from a new point of view. The ideas underlying this arran-
gement have long been a puzzle to philologists. Beyond the facts that one class
consisted of nouns denoting the names of persons, another of verbs used as nouns
and a third (in some languages) of diminutives, while others contained, though
they did not exclusively consist of, the names of trees and of abstract qualities
respectively, it seemed impossible to discover the principle on which words were
included in one class rather than another. Most attempts in this direction have
been more or less fantastic in character, and were, M. Van Gennep thinks, fore-
doomed to failure, because they approached the question from a purely European
point of view. The solution, he suggests, may lie along the lines indicated by
Mr. Dennett in At the Back of the Black Mans Mind (now supplemented by Nige-
:

rian Studies], — —
viz., in discovering the « logical system » of the Bantu the prin-
ciple on which they classify the facts of the visible world, so far as these are
known to them :

« Ce systeme de classification des choses de I'univers, phenomene de I’ordre

social, entraine une classification correspondante des mots designant ces choses. »
We may remark, in passing, that considerable light is likely to be thrown on
this subject by M. Torday’s researches among the Bushongo, whose system of
sacred animals, intimately connected with their social organization, seems to
complete and explain the information obtained by Mr. Dennett from the Bavili,
and the hints as to the Warundi contained in P. Van der Burgt’s work, to which
M. Van Gennep refers at the end of his essay.
The present can hardly be called an attempt at solving the problem in ques^
lion — that would, for many reasons, be a task beyond my powers, — but an
endeavour to examine two points connected with it the way in which the names
:

of animals are distributed through the various noun -classes, and the distribution
of root-words denoting animals though the whole area of the Bantu language
-field. One of the most generally accepted opinions with regard to the Bantu lan-

guages is, that, though they have no grammatical gender as we understand it (i.
e. no distinction of sex expressed in Ihe forms of the language) they make a very

clear distinction between living things and things without life. This is borne out
by the known (or perhaps we shoudd say the least
fact that, in Swahili, the best
unfamiliar) Bantu language to most Europeans, all names of living beings are
included in the first, or person-class. But a little further study shows that this
is not entirely accurate. Not only do other languages place most names of animals

in entirely dilferent classes, but internal evidence shows that, even in Swahili, the

arrangement is an afterthought, a late grammatical development, evidently the

result of logic and reflection. For instance, simba « a lion », chni, « a leopard »,
which, as they stand, have no prefix and do not vary in the plural, belong to
 IIEVIE li’ETIl.NUGli.UnilE ET DE SOCIOLOGIE

IJleck's niiilh class (called llie third in mosl Swahili grammars) which originally

liad Ihe prefix ini. These words are treated as belonging to the first class, i. e. we

sav siniha mlaihtcfi afika., « the great lion comes », not iimba (n)kubwn ifika as it

would be according to the concords of Class 3. The plural, however, is not

irasiinba, but simba.

As a matter of fact, most names of animals (so far as I know, in all the Bantu
languages) belong to Class 3. In some cases there is a process of transference to
Class 1, eilher still going on or already completed. Professor Meinhof is probably
right in attributing this to a habit of personification due to the influence of folk-
tales :

« Die Anwendung der Klasse !) anf die Tiere ist bekannt. Ini Suaheli werden
die Tiere zwar nach dieser Klasse gebildet, aber die abhiingigen Worte gehen
nacb der Menschenklasse (nnter dem Kintluss der Tierfabel). Im Kele ist sogar die
Priiligierung von ba Kl. "2 regelmiissiger Plural zu Kl. h
But, apart from Ibis transference to tbe person-class, of which we find several
stages, animal-names are found with greater or less frequency in several other
classes, which propose to examine separately.
1

Perhaps the most convenient way of beginning our inquiry will be to give a list
of animal-names belonging to this class in Zulu, which will serve our purpose as
well as any other, and belter than some, having preserved its grammatical forms
much more fully than Swahili, and, in some respects than Nyanja.

imbube, ingonyama, lion

ingwe, leoiiard.

indhlovu, elephant
imvubu, hi[»[»opolamus.

imfene, baboon
inkau, mon key
imbodhla, wild cat.

inja, dog.
impisi, hyena.
impofu, eland.

imbabala, bush-buck.
innnngu, porcupine.
imbila, byrax.
impuku, mouse.
inkomo, ox.
imbuzi, goat.

imvu, sheep.
ingulube, pig (wild or tamej.
inyoni, bird.

inqe, vulture

insingizi, ground bornbill.

impangele, guinea-fowl
iidcuku, domestic fowl,
in taka, finch,

ingwenya, crocodile,

inyoka, snake.

1. Grundziige einer vergleichenden Graininalik der

Banlusprachen, p. 14. Kele is spoken in the

Claboon countrj’, French Congo.

 MISS WERNER : THE NAMES OE ANIMALS IN THE BANTU LANGUAGES 21

imbulu. monitor lizard

inhlatu. python.
inhlanzi. fish.

inkwetu. mussel.
impukane. fly.

inkala. crab.
impetu, maggot.
intutwane. kind of ant.
inyosi, bee.
imbambela. octopus.

The might easily be made much longer; hut it is sufTicient to show that
list

there no apparent distinction between large and small animals, or between the
is

different orders, all being represented in this class. The deficiency in generic
terms, which the Bantu languages share with other primitive tongues, has often
been remarked on. Here, it will be seen that there is a general term for « bird »
and one for « fish », but none for « reptile »; inyoka is generic for all snakes
(various kinds being distinguished as inhlali, imarnba, imfezi, etc.), but there is

no common designation for lizards as such, nor for insects. The different kinds
of beetles have each its own name, but there is no word corresponding to
« beetle »; on the other hand, there are two words for « butterfly » {ijnbajubane
and uvemvane), but none for the various species ;
or perhaps no difference of
species is recognised. Two
kinds of moth [inyundu, umounyn], have names of their
own, no doubt because their destructive propensities have brought them into
notice. Animals in general are called inyama (which also means « meat ») or
inyamazane (diminutive of the former word). These properly denote « game », or
at least any eatable animals, but are sometimes applied to others, hilo is any
wild or ferocious beast, but is now (as we shall see later on) usually applied to
the leopard. In Chinyanja, the distinction is made between nyama and chi-rombo
(same root as isi-lo, but with the suffix-?;i6o). The above remarks apply, on the
whole, to all the Bantu languages. Swahili has a word for « insect » [dudu) but
it is borrowed from the Arabic.
It would be easy to compile similar
lists to the above from the vocabulary of

any other Bantu language, but this will be sufficient for our purpose, as, with
some exceptions, even when the name is not identical, it is found in the same
grammatical class. Thus, had we selected Luganda, instead of Zulu, our list would
have run empologoma [or eniale), cnyo, etijovu, emviibu, enkobc, etc.
:

This class contains, besides names of animals, a variety of other substantives.

1 give a few examples (still in Zulu) without attempting to arrive at any principle
of classification.

Moon, inyanga (but not sun),

Star, inkanyezi.
Mountain, intaba (but not river or rock).
House, indhlu.
Blanket, ingubo.
Basket, imbenge.
Stick, induku
Earthen pot. imbiza.
Chattels, furniture, etc.. impahla.
Story, discussion, etc., indaba.
grass pad (for carrying loads on head), inkata.
22 REVUE d'etenograpiiie et de sociologie

Also many names of plants, and some nouns (formed from verbs) denoting
the agent, as imOongi, praiser, imOazi, « carver », etc.

Then we have a number of animal-names belonging to the li-ma class (Bleek’s

5th and Glib, e g. .

Ibubesi, lion. Plural ama-bubesi.

ibizi,
)
zebra.
idube, ) .

ikati, (domestic) cat.

ihashi, horse.
ihlosi, another name for leopard,
ijele, male leopard,
iqina, steinbuck.
iula, oribi.

Several kinds of rats, as : iOiizi, igicence, ivoudiue, etc.

Birds : igwahaba, « crow », ihlungulu, « raven », ihobe, « dove », ljuba « pigeon »,

itiCthogn, « plover », iseme, « bustard », idnda, « duck », igude « cock ».

Reptiles ; ibtilulu, ivuzamanzi, kinds of snake, ixoxo, « toad », isdesele, « frog ».

Insects, ipelu, « cockroach », ibiuigane, « boring beetle »; ikazane, « tick «

;

igiigca, « kind of termite )> (usual kind, inhitva, innuhliva), itulu, « cattle-tick «,

izenze « Ilea ». It might almost seem as if tlie insects belonging to this class were
all more or less noxious ones; but this principle will not apply to the other

animals, and even among insects we have ijubajubane « butterlly » and ilshon-
gololo, « millepede », which last is at all events harmless. In considering the
other names, we may perhaps note two points, though I should hesitate to gene-
ralise; -some names arc those of imported animals, as Ikall, ihashi, and (in the
Xosa dialect of the Cape) ibokwe, (imported) goat; igiishn, woolled sheep; ihagu
or ihangu, domestic pig (English « hog »). Again, some are of the nature of
synonyms or appellatives, used instead of, or alternatively with the real name;
ibubesi, idelabulongo (another name for « hyena ») ;
or denote a particular sex or
age of the animal in cpiestion : igude, ijele, icalura, « male dog ». This, however,
cannot be pressed.
The seventh or in and eigth or ku classes (Bleek's ITth and loth) contain no
names of animals, the latter consisting solely of infinitives of verbs and the former
of abstract nouns (e. g. ubu-hle « goodness with one or two exceptions, such
»),

as ubu-tshwala, « beer » u [bu)-fshane, « grass »). I have found only one name of a
living creature ubu-cubu, « a chick ».
The lu class (Bleek's llthy contains a few names of birds, reptiles and insects ;

Ukozi, uhefshe, hawks, icnivaba, chameleon, u[lu\fudu, « tortoise », u(lu]lembu,

« spider ». It is sometimes said that this class contains the names of « long or high
things » ;
but this does not seem to apply.
Some animals are found in the mu-mi class (Bleek’s third and fourth) of which
the most marked characteristic is that it contains the names of trees, or of things
(such as parts of the body), which, though living, have no independent life
[umkono, « hand » umlomo, « lip », umlenze, « leg », etc.) Thus we have umkorno,

« whale », « », umkolwane, « crowned hornbill », etc. (In Nyanja,

iimvundhla, hare
mkango, « lion », and mbulu, « wild dog » (Lycaon pictus) belong to this class). We

also have umnenke, « snail » umuhhra, « white ant », umvunya « fish moth »,
urnkonya, « kind of locust », etc.

Lastly, names of animals belonging to the first class are in a somewhat peculiar
position. Normally, the plural prefix for this class is aba-, but animals take the
 MISS AVERNER : THE NAMES OF ANIMALS IN THE BANTU LANGUAGES 23

plural prefix o. Thus, umeyane, a mosquilo » has for its plural orneyane upejnne, ;

(( rhinoceros » opejane. This is the same form of the plural as that taken by proper
names, as oMpande (used instead of uMpande as a plural of respect) or in the
idiom oZatshuke, meaning « Zutshuke and his people ». Probably this arises from
the fact that when the name of an animal is treated as a proper name, as in sto-
ries, the personal prefix u is substituted for that properly belonging to the w'ord.
thus i -caklde is
>;, but he figures in fairy-tales as a quasi human
a <( weasel
being with the name of uCakijana.
Again, many names of animals belonging to this class are clearly compounded
with t(nina, u mother » (contracted to no, as unogivaja, rabbit », unohemii, « crow- «.

ned crane »). It is difficult to trace the etymology of these words; possibly unog-
waja may mean mother of numbers » iikidi gtvaja
c( be thronged together ») <(
‘

but usually the word to which no- is prefixed does not seem to have a separate
meaning of its own. Comparison with other languages and considerations pre-
sently to be adduced suggest the possibility that the compound does not mean
« mother of » anything but is merely the animal’s ordinary name preceded by a

honorific particle, Mother Gwaja », Mother Hemu », and so on.

(( ((

We find an analogous development in Suto. The special prefix l/o- is used in an

exactly similar wmy to the Zulu o, among others, to form the plural of proper
nouns with a collective meaning. Some nouns of animals are treated in the
((

same way, but it will be noticed that they are all formed with the name 'ma
(mother of). Ex. 'mankhane bat, pi. bo-mankhana ; mampharoano lizard, pi. bo-
mamplia roane; mamnlangvane the secretary-bird, pi. bo-inamolanguane, etc. ^ ».
Cf. also w'hat Endemann says [Versuch einer Grammailk des Soiho, I, p. 34) :
—
« Manche Tiere haben —
was vielleicht mit urspriinglichem heidnischen Mytbus
zusammenhiingt —
mit Ma beginnende Personalnamen als Gattungsnamen und
werden daher sprachlich auch als Personen behandelt. Solche sind z. B. Makovele
(Habicht), PI. vo-makovele ; Maranlhane (Fledermaus) u. A. «.
Rolong dialect of Chwana (wliich is, of course, very similar to Suto), we
In the
find thesame rule, e. g. bo-mamathoane, bats », but it is not confined to nouns (c

compounded with ma-. « Bo-, is prefixed to form the plural of many w'ords
which refer to the smaller kinds of animals, e. g. samane ; a meercat, pi. bosamane;
lobolu, achameleon pi. boloboln; kgnntlapane, a lizard, pi. bokganllapane ® ».
In (the language of the tribe usually called « Mashukulumbwe « in the
Ila
Zamliezi Valley, closely relately to Subiya), we find, similarly, nouns compounded
with na- ((( mother ») and also with ska- (a father ») Thus :

Shankole, -wart-hog, pi. ba-shankole.

Shasubila, the pallah antelope, pi. ba-shasubila.
Shapidio, a hawk, pi. ba-shapidio.

Nachindwe, the Oriln, pi. ba-nachindwe.

Nakasha, the duiker, pi. ba-nakasha.
Nam u wane. the crested crane, pi. ba-namuwane,
etc., etc.

It isperhaps somewhat curious that names compounded with « father » should

be lesscommon than those with « mother but in most languages this seems to >
,

be tbe case. In Zulu, I have only been able to discover u-somhelxhe [so is flic usual

1. Colenso’s Diciionary
2. Jacottet, Elementui’y Sketch of Sesido Grammar, Moiijn, 1893, p. I.?.

3. Crisp, Notes towards a Secoana Grammar, London (S. P. C. K.), 1900, p. C.

 REVUE d'etiinograpuie et de sociouogie

ponti'alion for uijse, as in proper names sucli as Somkeli, Sotemha, etc.) a kind of
hawk, also called hi)-lielslic. In Ganda, bolli se « father » and na, « mother » arc
ii{

used lo form llie names of animals (and oilier words as well). E. g. Sekanijoltja
« a crane » .vc’ kesa, « a kind of centipede », semiiliuulii, « a large kind of lish »,
na/mlii, « spider » ;
lunmdimi, «. ant-hear »; naivolovii « chameleon » (cf. Solo
lohoJu).
Nyanja a large number of animal-names begin with na, the explanation
In
(which does not seem to have occurred to D. C. Scotl, see Dicllonary p. 418] being
obviously the same. ^Ye may instance nadzikamhe, « chameleon », nankahai,
<(hawk », nakodzire, « waterhuck », nmnf'uko, « mole », nanudindl the « lily wa-
der » [Parra africaiia\ etc., etc. Some of these names are also used without
the jirelix, as mfuko. Na-, in these compounds, is often followed by -ka- or -aka,
for which no salisfactory explanation suggests itself: aakaftimlnve « a weevil »,

naakahai, « a hawk », nankalizi, « a scorpion ». Ml these nouns are reckoned as

having no prolix in the singular — that is, they do not lake mu, rn, omu, or unm,
and simply ])relix the sign of the jilural lo the unaltered singular form. In many
cases, there is really examples given above, samanc
no jirelix, thus, among the
and kyaatlapaan, hut loholu clearly belongs lo the sixth class, which in Chwana
has Ihe singular prelix In pi. //, in Zulu u[lu] and iz\a This shows that />o-is addrd
in the plural lo Ihe oriylaal prefix and [halsamane and similar words originally

belonged, like most names of animals, to the third class, which in Chwana has
entirely lost its prefix in Ihe singular.
seems Iherefore as if we were here dealing with names of animals transferred
It

from their original class lo the person-class, as a result of personification. We find

the same tendency at work in Siihiya, where Ihe lion is called indanu, hut, when
liguring as a personage in the folk-lales, undavu. In Lala and henje, Ihe plural of
courtesy is used nod,'alula, hare :
—
a'ankadniiu, lion, and so on. ;

In Sumhwa folk-lales, animals have the [irelix AV/ ; nakaini, the hare, is A'ana-
kinni; in Ganda, W a : U'u-njoou.
In Yao stories, the animals are distinguished as Che Snngiila, Che Ndernho, etc.,
-Che being the ordinary prelix of courtesy used to men and women, and Nyana
is similarly used in .\ngola. These two prefixes are less inlimalely connected with

the words than those iirevioiisly mentioned, and remind us of « Hrer Uahhit » and
« Michie Dinde », in the folk-lore ofGeorgia and Louisiana.

fhe rule that names of animals, of whatever class, take the plural of the first is

found in Lenje, where we have njovu, wanjovu, nsefu, iransefii, etc. Yyanja seems
in this respect lo he in a transition stage : one hears « anjohvu a li ankanve and

njohvu zi ri za nkarwe ».

In 11a we find that several words which elsewhere (a) belong to the third class

in both numbers, or (h)- belong lo the third class in the singular anti Ihe first in
Ihe plural, as we have just seen in Lenje, have taken the first class prelix in
Ihe singular also, e. g.

mu-nyati ~ buffalo pi. ha-nyati

mu-sefu = eland pi. ha-sefu
mu-7.ovLi = elephant pi. ha-/,ovu.

But, besides these, we have the compounds (N sha and im already referred lo,
and names of animals beginning with m, (or mu) shi- mi-, ka- and chi -all of which
prefix ha- in the plural, as hashaluzuke (kind of fish) hashiluwe (leopards), hamishikn
(hawks), hakahwenria (hyenas) hachiwena (crocodiles] bachihizi (zebras) baniwaha
(jackals) ;
the author adds, « Nouns of this class have really U as their singular
 MISS WERNER : THE NAMES OF ANIMALS IN THE BANTU LANGUAGES 2o

classifier, but with most nouns heard only when special empliasis is put on
it is

the word. Thus at the may hear U-chibizi, U-shiluive »

beginning of sentences one
It seems clear, therefore, on the whole, that names of animals have only been

placed in the person-class at a late stage of development, and as a result of per-

sonification, probably, in the first instance, through the influence of the animal-
story which is so marked a feature in Bantu folk-lore.

II

Even a superficial acquaintance with the Bantu languages discloses the fact that
some names of animals have practically a universal dill'usion, while others are
curiously limited in range, two or three adjacent languages sometimes using words
which have no etymological connection whatever. Sir Harry Johnston’s early
theory, which he has since seen reason to modify, will hardly fit the facts. He sup-
posed that the widely spread names belonged to creatures known to the undivided
Bantu in their original habitat, and thence deduced the conclusion that they came
from « a well-forested and well-watered region in West Central Africa. »
« The fact that similar words are used in the remotest and most widely separated

members of the group to e.xpress « leopard », elephant », « hippopotamus », « buf-

falo », « pig », « ape », « monkey » « grey parrot », « bee », etc., leads us to
believe that these creatures were familiar to the primitive Bantu race, while we
may infer that because the rhinoceros, giraffe, lion, ostrich and zebra are known
by many different terms in the dialects of those tribes acquainted with them, they
Iwere unknown to the forefathers of the people home
in their ancestral The
oopard, elephant, apes and monkeys mostly affect densely wooded regions, the
hippopotamus only inhabits big rivers or lakes. The buffalo is rarely found
westward of the Niger. On the other hand the lion, giraffe, rhinoceros, zebra and
ostrich decidedly prefer the sparsely wooded steppes and savannahs of Eastern
and Southern Africa »

It had occurred to me, before 1 knew that this theory had been suggested by

Professor Meinhof, and possibly others, that this difference in names might, in
some cases at least, be the result of hlonipa, knowing that the ordinary names of
certain animals are forbidden to be uttered, or at least, not used willingly, accor-
ding a world-wide usage of which it would be superfluous to give examples.
to
It iswith the object of obtaining a little more light in this point that I shall now
proceed to examine the names of the animals enumerated by Sir Harry Johns-
ton, or at any rate some of them, in those Bantu languages in which I have found
adequate material for comparison.

(.4 suivre).

(1) Smith, E.\V., A handbook of the ila language (Oxfonl, 1007), pp. IS-10,

(2) The Kilimanjaro Expedilion, p. 48G.

 REMARQUES SUR L’lMAGERIE POPULAIRE
Par A. VAN Gennep.

Le problernc qui se pose ci propos cle Pimagerie dile populaire est exactemenL
le meine que celni qui se pose a propos des lilteralurcs populaires, des croyances
populaires, des costumes populaires, el il se formule dans les memes termes :

etant donne que populaire se rapporle d’une pari a I’acte de creation, d’autre part
ailphenomtMie dc la transmission, el cnlin a celui de Pulilisalion d’une forme pre-
lendue collective d'aclivile, peut-on cepcndant, sous ce caractere a premifme Mie
colleclif, discerner des tendances determinees, el sont elles reellemenl collectives,
oil bien individuelles?
Ce jiroblerne, j’ai tenle de le resoiidre a diverses reprises deja, el pour d’autres
activilesque I'aclivile esllieli(|ue, soil, dans mes Mi/ilies ei iJfjoules d'Auslralie, a
propos dll fail parliculier de I'invenlion des coutumes religieuses el sociales, soil
dans ma Formation des Lhjendes A propos des recils de toute calegorie qui out
cours dans le people.
En matiere d’art, il esl beaucoup plus difficile d’alleindre A des generalisations
acceplaliles qu’en maliere d’instilulions ou de legendes, presfpie aussi difficile que,
par exemple, en maliere de conies. C'est que lA aussi I’anonymat A pen pres regii-
lier oppose A la recherche approfondie un obstacle presque invincible. Je dis

« presipie », parce que A bien eludier les techniques cl les manieres, un criliipie
d’art pent arriver A dislinguer des « mains », c’est-A-dire des caracterisliques indi-
viduelles, indiscernables aii jiremicr venu. Quelle que soil Eequalion personnelle
propre dii critique, el qui jiourrait vicier son Jiigeinenl, les resultats acquis par les
archeologues ou les criliipies d'art s’inleressanl A la Renaissance par exemple, cl
mieux encore A des ecoles delerminees, ue sont point niables. II se pent aussi
(jue la melhode des mensurations, lelles les recherclies de S. Reinach sur I’indice
mammaire, ou cedes de Laran sur I’ecole liourguignonne, conduironl plus lard A
des acquisitions plus solides que la critique A demi subjective encore ordinaire-
menl en usage. Jacques Mesnil cependant a fait, dans la Ilevue des Idees, aux sys-
lemes pretendus objectifs de critique d’art des objections bien fortes; mais elles
prouvent plulot que I'instrumenl est encore mal conslruit, non pas que le principe
qui guide ses invenleurs soil A rejeter d’emblee et des avanl son perfeclionnement.
Le mieux, d’ailleurs, en ces maliAres, c’est toujours de choisir un terrain de faits
et de n’y avancer que pas A pas, Et Ton comprendra combien doit etre bienvenue
la publication, par M.U. E. van Ileurck, d'Anvers, et J. Roekenoogen, de Leyde, de
leur admirable Histoire de flamande (7d0 pages, in-i'’,
I'lrnar/erie populaire
Bruxelles, van QEst, 30 richement illuslree de cliches en noir et d’originaux
fr.i,

en couleurs —
et ceci surtout si j’ajoute qu’un long Appendice donne un expose

compare de I’imagerie populaire dans les aulres pays d’Europe il). Comme les

(1) Me trouvant a Anvers, il y a quelques semaines, j’ai eu la

ctiance de pouvoir visiter en la com-
pagnie de M. van Ileurck le musee de Folk-Lore rScemment organis6 dans cette ville grace aux
soins de Max Elskanip et de M. van Ileurck lui-meme. It m’a montre aussi plusicurs series de sa
belle collection d’images populaires et a pouss6 I’obligeance jusqu’a me preter, pour illustrcr le
present article, non seulemcnt des cliches, mais aussi quelques bois originaux auciens.
THE NAMES OF ANIMALS IN THE BANTU LANGUAGES
By Miss A. Werner (Londres).

(SaUe et fin).

1. The Lion.found throughout Africa south of the Sahara, with the exception
Is

of parts of The Congo forest country. Sir Harry .lohnston says « Over the greater ‘

part of West - Central and Western Africa the lion is either non - existent
or has all the appearance of a recently-arrived animal ». Nevertheless he seems
to be known, if only by report or tradition, to nearly all Bantu tribes, and I am
inclined to doubt whether my failure to find a name for him m some languages,
such as the Fang, is not merely due to the scantiness of the materials avai-
lable.
The names by which the lion is known differ, as already slated, very widely
from each other. Mkango, slmha, tan, nyalupala, nkaJamu, niale, nkosl, iamho,
are some of them.
In considering the distribution of these and other names, I shall begin with the
Nyanja mkangn, which, so far as 1 am aware, stands alone. Entirely different words
are used by the Yaos [lisimhn) and Makua [karnmu) who are the neighbours of the
Anyanja on the east. The Sena people on the Zambe/.i (who indeed speak a dialect
of chi-Nyanja) say nkalainti, as do the Basenga, the Balala and the Babemba on
the west, and, in a slightly varied form the Konde people in the north. In no
other language have been able to found the word mkango, while it is also remar-
I

kable that this word belongs to the mu-mi class rnkango u-lumpa, « the lion
:

leaps », (not a-lumpa), ri po rnkango

i « there are lions », (not a ii po icakango, as

it would be if rnkango belonged to the person-class). Native usage is decisive on

this point.
It was only recently discovered that the Bushman word for « lion » is Ka. This
I

seems quite a possible origin for the word, when we take into consideration that
there seems reason to think that a large Bushman preceded and to a certain extent
became amalgamated with the Bantu A-Nyanja, at any rate west of the Shire.
(See R. E. E. S. March. 1908).
Leaving aside this suggestion, which further inquiry may or may not invalidate,
Iproceed to take next in order the name which occurs most frequently in the
table of languages I have drawn up, viz. simba. This seems to be confined to the
eastern side of the continent. In Pokomo, (Ihe most northerly Bantu lan-
guage of the coast region), it is dzimba, in Nika ’. Swahili, Chaga, Kami,
Nguru, Bondei, Gogo, Zigiila and Pogoro, simba, which is also the word in
Sukuma, the northern dialect of Nyamwezi. Variants of this form are Nyamwezi
isimba (or ishimba), Riindi, ishimba, Shambala, skimba, Sumbwa. nshimba, Fipa
(South Tanganyika) nsimba, Yao lisiinba, Matengo (Eastern side of North Nyasa),

1. George Grenfell and Lhe Congo, II, 918.

2. Some dialects have dzirnba.
94 REVUE d’eTUNOGRAPUIE ET DE SOCIOLOGIE

lihimba, and Zwina (the language of the Mashona, which, according to Torrend, is
a dialect of Karanga) sliumha.
In Ila ^Upper Zambezi), shumbwa is used, as well as indcwu.
Generally this word belongs to the lliird class, and is invariable in the plural.
In Yao andMatengo, however, it belongs to the class, and the plural would de
masimba, mahhnba. It is possible Ihal originally it belonged everywhere to this
class and the loss of the prefix has caused it to be placed erroneously in the third.
But Fipa and Sumbwa have the n- prefix, and it is perhaps more probable that the
converse process has taken place, and that it acquired the li- prefix in Yao and
Matengo from an idea that the ai>ocopated stem belonged to this class. In Swahili
there are many apocopated nouns of which it is impossible to say, till their con-
cords are known, whether they belong to the third or the tilth \U-mn) class.
liere is a wonl slmbn (as well as Uhnnha] in Matengo
1 but it is applied to a '

dillerent animal. In Nyanja, by the bye, shnba means a small kind of spotteil
bush-cat (Fells caracal?)
Krapf and Hebmann’s
In Ficllorarfi (Imndon, S. 1’. C. K., 1887), we find
a very interesting note on the subject nf hloaipa, s. v., inaniambo [muniinmho).
which is the word for lion in the Kigiryama dialect of Nika. « The word is only
used in superstitious dread, while the jirojier word both in Kigiryama and Kira-
hiii [the dialect of the Uabai iieople close to Mombasa] is D/.imba. » KIsewhere, as
has been said, the same authority gives the Nika word as shnba.
.\nothcr animal wdiose name the Wanika avoid [ironouncing is tlie loma pro- <(

bably the badger » (loc. cil., p. ^14), which « burrow's in old ant-hills. « When one
is killed, a lamentation is held over it, hoping that by so doing the meat will prove
very fat. » When digging for they do not,
to mention it. call it
it, if they have
loma but Godzo from the siqierstilions dread Ihal, if they called it by its real
»

name, it would go further from them in its hole and, if caught, would not prove
as fat as it ought to be. » (Id., p. 107.)
In Sukuma, shnba is used along with nijanurcre, which may be the same word

as niialapala, to be considered later on. In Gogo, besides shnba^ nijalapala and

nijaiuni are used, the latter a Masai word.
Ill Zwina, Mondoro (hnonduro) is used, as well as slnnnba. Mondoro^ according
to Father Hiehler, « means especially a man jiossessed by a spirit as valiant as a
lion’s >, and seems also to be used for « a spirit ». What its etymology is, I am
unable to say; it is of the third class, the plural being dzhnondoro. There appears
to be a belief [irevalent in tliis part of Africa that the spirits of deceased chiefs
assume the form of lions.
Mpondoro is also used by the Tete and Sena people on the Middle and Low'er
Zambezi.
The Ila synonyms for sluoalnva are, lapurnpu, indaon, Shamanc/ana, Shanza,
Shanza-Mnlala (« Great Shanza »), Shclwi, and M a'anda-llamjama (« Hunter of
Animals »). It is probable that some of these are titles, used to avoid mentioning
the real name. 1 can find no indication of this custom jirevailing witli regard to
animals among the Ba-lla, but certain names of persons are always avoided, e. g.
a person may never utter his own name or that of a namesake, or the names of
his father, mother, sister-in-law, and wife i. e. the latter’s name before lief*
marriage. Kvery married woiniin has to receive a new name from her husband.
Coming back to the north, we tind that the nearest Bantu neighbours, on the

1. I*. Hiitliger, in Afrikanisclte ^luflien '.Mitteilungen <les Seminars (iii' orient. S[U'achen), 190'.i_

[1. ISO.
 MISS WERNER : THE NAMES OF ANIMALS IN THE BANTU LANGUAGES 95

west, of the Wapokomo and Wanika are the Akikuyu and Akamba. The former
use two words for lion, Murothi, which I can find nowhere else, except that
Mololhi, given as the Kamba for >< leopard », appears be the same word, and
to
Mupaka, which is also found in Kamba. This last, in other languages means « wild
cat » ;
in Swahili, paka is also used for a domestic cat, as is mpaka in Nyanja.
(Nyanja has a great many names for different species of wild cats, as mjuzi, vuinbwe,
simba and others; but, unless 1 am mistaken mpaka is not included among them). In
Zulu, impaka is seldom mentioned, having an ill omened repu-
thewild cat, but it is

tation as a wizard’s familiar. noticed that where the word is used for
It will be
a lion it has the mu prefix. I have not been able to ascertain whether the plural
takes mi or ba. Kamba bas, besides mupaka, munyambo (which we tind also in Nika'
and munabwe (see Last, Polyglotla Africana Orienlali.i, p. 277), which may be an
erroneous rendering of the same word.
In Ganda the word in general use is empologoma *. Which, as pointed out to me
by the Rev. W. A. Crabtree, seems to mean « The Roarer », being derived from
the verb ku icologoma. It is therefore probably not the original name, which I infer
was Utah; But, as learn from the same authority, that, as it is used in folk-
I

stories with the personifying prefix {ivampologoma], « presumably nlab^ is not even
recognized as old Lu-Ganda ». (/z]nlale is the regular word in N'yoro, Ziba and (in
the form intarp.), Riindi. Cf. also the Kon jo ndali
I feel uncertain wheter we can venture to connect ntalp w'ilh the 11a and Subiya
indavu, wliich, by all the laws of sound-shifting is clearly identical with the
Chwana iau. It is not impossible that they might spring from a common Proto-
Bantu root. In Ronga (Delagoa Bay) we tind it in the form ndjnu. Possibly ;but here,
again, I hesitate), the Luyi (Rotse) nde might belong to this group.
It has already been pointed out that in Snbiya folk-tales, indavu becomes
undaou. (Mr. Smith suggests that indavu, in this language is borrow'ed from the
neighbouring Totela dialect; but w'hy? — —
unless to avoid the use of shumbwa].
In like manner the Basuto, in folk-tales, frequently call the lion aiob/u; — e. g.

« Joale hosasane mutau moliolo a tsoha, a ea pele letseng la liphoofolo — Joale

motau moholo a bolsa, a re », etc.
Inkalamu seems to belong chiefly to the country north of the Middle Zambezi.
We find it in Bemba, Wisa, Lala, Lenje, Senga, Sena, Konde {ingalamu) and (sepa-
rated by a considerable interval), Makua {karamu).
Mnyalupala {nyalopala, nyarupala), seems to be confined to the upper course ot
the Rufiji and the depression north of it, called by the Germans « das abflusslose
Gebiel ». It is found in Kaguru, Gogo, Hebe and Sango.
Turning to the Congo region, we tind the following variants of one form.
Fiote (Lower Congo, also called « Kongo »,) -nkoshi (spelt by Bentley nkoxi).
Bang! (above Stanley Pool) -nkosi.
Lolo nkosi.
Mbundu (Angola) hoji [j probably pronounced as in French).
Kwanyama (south of Angola) ongushi.
Ndonga (Ovambo country) onkoshi.
One cannot help wondering whether the Zulu word for « Chief », which
occurs also in Suto [k-fosi] and Thonga [hosi], hot not, Itink, elsewhere connected
with the same root as these, f do not recollect any instance of a chief being called

1. I have not I'ound this word elsewhere, except curiously enough — — in the Chwabo dialect

f Kilimane (Portuguese East Africa) where it has the form podogoma.

2. Jacottet, Treasury of Basuto, p. 13.
•J6 REVIIE d’ETHNOGHAPHIE ET 1)E SOCIOLOGIE

« Lionwith the exception of Siinha Mwene in East Africa, Uiough « leopard » is

»,
frequently a royaltitle. But the suggestion might be worth considering. Professor

Meinhof, howewer, derives the word frow koka, « draw, lead », cf. as dux frow
duco.
The Zulu name for lion which seems to be most commonly used is Ihuhesl, evi-
dently connected with huba = « perish », causative hub'isa — « destroy ». Inibube,
which not does so frequently occur, comes from the same root. I do not know what
word has been replaced by these. Ingomiuma is given in the dictionaries, but I do
not remember hearing any word but ibubesi used in Natal. In Nyasaland, however,
an old « Angoni » woman who spoke Zulu, gave me the word ingwenyama, which
pu7.7.1ed me, as being evidently a compound of iugwe « leopard » and nyamu

« animal » but perhajis it is an instance of corrui)tion through « folk-etymology ».

I had therefore always looked on this word as a derivative and perhaps employed
equally with ibubesi to replace some tabooed original. But the existence of the form
nngeama in llerero suggests that it may be a modification (in accordance, as already
said, wilh folk-etymology) of some pi'itnitive root. I can find no other examples of
this or any cognate form.
This avoitlance of the lion’s primitive name in Zulu might have originated (a)
in a superstitions fear of the animal itself, as we know to be the case with the names

of the leopard and the wild cat [iin/uika), or (b)if it had heeii the personal name of
a chief, in the rule of hloui/xi which forbids the use of such a name. 1 do not know
whether it is possible that —
sup|»osing the tabued name to be iiikosi, it might, —
when the tabu had lapsed and its original meaning had been forgolten, ac(|uire lh(‘
meaning of chief ». <i

in II crero we have also ougezu « maned lion » and oudumbi, stated by Brinckei'
to be a « Hciwune », a surname or nickname. The initial u, by the bye, do(!S not

here denote the person- class in llerero the class corres|)onding to the one we
;

have been calling the third has the prefixes o in the singular and ozo in Ihe plural :

as, oudu « sheep », pi. ozondu.

find no parallel to the .Nilonga o/o/ac, which is used alongside
I oukoshi. It may,
like oudumbi, be an appellative.
The Bushongo of the Kasai do not themselves know the lion, which is not found
in their name for
country, but they have a it — lainbo. This is borrowed from

Iheir neighbours, the Balunda, whose word for lion is lambu (Senhor Bias Car-
valho writes it Idbui. In Rundi we have
‘
(besides the two other a[)pellations
already mentioned) iulnmbwe and Rega (s])oken west of Rundi and Ruanda, in
in

the forest region between the great lakes and the Congoi ndambive, which seem to
be the same word.
Ihalu is the name for the lion in the very interesting .Masaba language spoken
near .Mount Elgon, wdiich ajtpears to be one of the most archaic among the Bantu
tongues. In Chopi (« Xilenge) » spoken in the Inhambane district 'Portuguese East
Africa), nearly at the other end of the Bantu field, we have nghala. Duala, again, at the
northwestern extremity of the area, has ngila, ujila, which (nay be connected,
though in Bantu vowels are slower to change than consonants, and the transition
from A to is so violent as to seem improbable.
1

The only name remaining on our list belongs to Kinga, a language spoken in the
mountains bordering the western shore of Lake Nyasa at the northern end. This is
elibonzu, a word to which have failed to discover an analogue in any other lan-
I

guage, nor can I ascertain its etymology; —

it has every appearance of being a

1. Metodo pratico para fallar a lingua de I.unda. (Lisbon, 1890).

 A. WER.VER : THE AAMES OF ANlJIALS I.\ THE BAiNTU LANGUAGES 97’

Idonipa-wovd recently coined, and one miylit luizard a conjecture that donza (which
in not in WoIfiTs vocabnlary) is onoinatopoetic for growling or roaring. It will
be noticed that it belongs to the U-ma class. A curious point mentioned by Wolff is

that ania (the Kinga plural prefix of this class) sometimes used with nouns of
is

other classes, to express I” size and contempt, as amaheki aga (for emlbeki egi],
« these huge trees », amanu aga, « these follows », for aoa>iu uva, « these men ».
Ifwe suppose the idea to be that contempt or detestation of an enemy can be safely
expressed when a pseudonym is employed, this might throw some light on the
reason for the class oi ibubexi. Can any examples be found to support this view?
We may add a few notes on the position of the lion inBantu folklore. He is by no
means so prominent as some other animals, e. g. the Elephant, the Hare and the Tor-
toise indeed, we find that he does not always take the highest place. In Angola, for
;

instance - « the elephant is the supreme king of all animal creation, and the special
chief of the edible tribe of wdld animals. Next to him in rank, the lion is special
chief of the tribe of ferocious beasts, and highest vassal of the elephant. The lion >>

as chief of the animals figures in Suto tah;s, e. g. « The Nyamatsanes » in Jacottet,

Treasurg of Basulo Lore, vol. I, n“ I :
(Conies pop. des Bassoutos, « Le petit lievre »),

and possibly in Ilonga, (Junod, Chants et Conies, pp. 108, 127, where Masingue wa
Njao seems to mean the lion). In the Sesuto story he is outwitted by the Hare, who
induces liim to help in thatching a hut, fastens his tail into the thatch and leaves
him to die. In a Nyanja story (Ratiray, p. 130], the Hare, by a trick catches a suc-
cession of animals in a game-pit and kills them. The Lion, inhis turn, also falls in,
bill, the Hare falling in at the same time, he catches him, and only spares his life

on condition that he will follow him home and nurse his children. He then plays
on the Lion the trick which, in so many other tales he is made to play on the Leo-
pard, and, after eating the cubs, escapes, and contrives to put the blame on the
Baboons, who arc killed by the Lion.
In another story collected by Mr. Rattray, a lion is overcome by a blind mau
who employs a series of subterfuges reminding one of the way in wdiich, in a
Chinamwanga story, the lion is « blull'ed » by the shrew’-mouse. There are Nyanja
and Chinamwanga tales relating how a girl married a lion who had temporarily
assumed the form of a man but it is more commonly a hyena who figures in this
;

kind of werewolf story. In Ha, too, it is the lion whose cubs the Hare pretends to
take care of and eats, and there are three other stories of the Lion and the Hare.
The above is not the only instance where the Lion in some tribes takes the
part assigned to the Leopard in the same story by others.
2. The Elephant. Out of seventy Bantu languages we find that lifty have virtually

the same name for this animal. Perhaps it may be as well to give the list, which
is by no means com plete.
Pokomo ndsofu. Swahili ndovu L
Kamba unzoo, nzou. Go go nzofu 'L

Kik uyu njogu. Gan da enjovu.

Nika ndzovu. Masaba injoji, injovu.

Tai ta Nyam wezi nzovii.

Chaga ndofii. Sumbwa nzovu.
Taveta nzobu. Nyoro enjoju.
Tikuu (northern ndofu. Ru ndi i nzovu
dialect of Swahili ) Pega njoLi.

1. Archiv I'ur das Studiuui deutscher Kolunialspi'achcii, liand 111.

i. Clialelain, Folk-Talcs of .Vngola, p. 22.

:|. Tliesc languages have other names as well. These will be discussed later.
 •JH RliVUE D'ETllNOGKAi'lliE ET 1)E SOCIOLOGlE

Kon jo ndsogu Senga nzovu.

(Ruwenzori) Bonga ndlopfu.
Kivamba (do) mlsogu Chopi nzo fu.
Kihoko (do) ohgu. Zwina nzou.
Fipa nzovu Zulu indhlovii.
San go nzovu ‘. Bushongo njoko '

Konde isofu. Kele Congo) njoku.

Chinamwanga inzovu. Colo — njokn .

Bemba imsofu. Bangi — nzoku

Wisa nzovu. Ngala — njoku.
lla muzovu Ben a Kanioka nzovu.
Snbiya unzovu. Fiole nzou
Ltila in so fu .
Fan ensok, nzox
Lenje nsofu. Duala njou.
Lunda u/avo. Mpongwe indyogu.
Boise ondopu .
llcrero ondjou.
Ayanja njobvu. Chwana and Snto Ilou.

Sena nzou.

All llieso forms ;irc derivoil, arcordin^- lo Moiidiof, I'roin a I'rolo-haiilii rool /.oyu.
Ill llu? majorilyof llie remaining languages, some form of llie word In.mho is used.
This, according lo Professor Mcinhof, originally meaiiL «alnsk » and, from llie

vast proiiorlions assumed by the ivory trade, came lo he used for llie whole
elciihaiit and gradually su|)|)lanted Ihe older designal ion. It is now used in Swahili
almost entirely to the exclusion (unless in some of Ihe norlhem dialeels) ol'

It is also found Mondei (n/cm/io), (logo [nhemho),

in {mlambn'i') Kami [nhemlxi]^
.Nguru, MaUua Zignla, Pogoro, Shamhala, Van '/ic/cmAn) and Matengo.
(c/c/io),

It is possible, though at preseni can bring no evidence in support of the

I

suggestion, that the Piole nzeunha (used side by side wilh Ihe .Mbundu and
-Ndoiiga ondjinnba and the Kwanyama ondjnha can be referred to the same rool.
The Kinga etsunfiira is an isolated form of which know not what to make. So I

are Ziba mjiuliiu (if correctly reporled), Sukunri inliuli (also used along wilh
Nvamwe/.i nzovu Matengo mpicfiijawn/o, delined by P. Ilofliger as « grosser Elefanl,
,

die zu zweien gehen «, Zignla lujofjomi, said lo mean « an old elephant (Kisbey) <>
;

Fipa nllnja and Sango yo/i^/cr/, which are given as synonyms oi nzovu.
The Polo (Middle Congo) word mhonpu may be connected wilb a non-Bantu root.
In Bushongo, besides .y/o/>o, i-nmOa and /Juloboare used. The former belongs to the
archaic and now obsolete language which the Bushongo peojile brought wilh them
from their original home in Ihe far north, and which seems nol lo be Bantu.
VossWAy nznm/ju and ondjamha should be connected with Ibis, and not with lemho,
and, if so, they are probably due lo Bushongo inlluence.
I have not met, so far, wilh any inslanco of Ihe elephant being snbjeci to lilonipa

as regards his name, though, of course where he is Ihe totem of a clan (as with the
.\i/anja dan iXjoOvn) he must nol be killed or eaten. Hollis menlions an elephant
clan among Ihe A'andi, whose members think they have sulliciently observed Ibe
rule if, after killing Iheir totem, they apologise and say they took him for a rhi-
noceros.
But there is a curious point inNyanja folk-lore, which may have some reference
in this connection. There is a creature figuring in many of the animal tales under

1. These languages have other names as well. These will be discussed later.
 A. WERNER : THE NAMES OF ANIMALS IN TIIE BANTU LANGUAGES 99

Ihe name of dzimive (Likoma dialect j/aiwe) which is by some explained to be the
elephant, and indeed in some cases the identical tales are elsewhere told of the
elephant under his usual name. But there is some vagueness on the point, as one
native authority says that the dzinuve is « either an elephant or a big spirit
[chimzimu) », and another « that it is only a creature in a story ». And it is clearly
the same word as Zulu i-zimu, Sato lelimo^ Duala edimo, Kikuyu irbnu, which
denotes a cannibal ogre, usally of more or less preternatural attributes.
It does not seem in accordance with the usual tenor of Bantu ideas that the name
of a preternatural being should be adopted in order to avoid naming the elephant,
who, though (as Chatelain points out) recognized as supreme chief of all animals,
is almost invariably represented as (for all his wisdom and dignity) rather stupid,
at any rate no match for the keen wdts of the Hare who usually makes a butt of
him. If the foregoing be correct, the absence of hlonipa in the case of the elephant,
coinciding w ith a great uniformity of nomenclature, and the diversity of the names
of the lion, while there is presumption,
if not proof, of hlonipa being applied to

him, might go someway support the idea that hlonipa has had considerable in-
to

tluence in the differentiation of language. I am obliged to reserve for a future

occasion the examination, from this point of view', of the names for leopard,
hippopotamus, hyena, buffalo, hare, crocodile, tortoise and some others, w'hich
I think might throw a little further light on the question.
 ANALYSES ET NOTICES

CoM ICHK.NCKS l)K I M -I'l ITK.N-NK Ecolc [irathllU’ 1*. Alicl siir Mdiidire. C‘('dail le nom dc la

iii- 18 20 liicalili' (III .\hraliani inslaila son ram|io-

,ri-:iude>< liihtiiliics, l',)09-l'.M() , ,
:i p.,

IjCconVc. — .1. (iiilialila, I'.MO, mcnl :

il s'y Ironvail soil un idicne, soil un

:i IV. aO. .
Icreliinllie sacn-. 1,'aiilenr commence par
comparer criliquonieni Ics divers passages,
Kn 1S',)() .sc I'oiula a .Icriisaleiii , sur I'ini- d(' la liililc cl des commenlaleiirs rdalil's a
lialive ilu li. 1*. l.aLraiif^e, liililciuc. ce licu-dil, place pres d'llcliron. II inonlre
I'.llc pulilia la Itcvuo liililiquc cl unc sciic conimeni, au coiirs do I’liisloirc juivc, la
iriAtudcs I’alcsliiiieimc.s lUuil Ic |ucscnl loralilc cn inl qiic sacree, a eld lanlol un
l

volume c^l Ic iroisiemc. (icsl tin lclllcil olijcl dc veneralion, tanidl nii lieu d'lior-
(l(‘ si.v cuiirciciiccs pronoma'cs an coinciil renr, cl commcnl, par un proces de con-
lie Saiul-Kliciine a .li ru.salcm cn lOO'.t- vergence, divers arhres giganlcsipics silucs
I'.i I (I (Icvaiil nil pulili'' Irl'S di\ ei s : ccclc.sias- dans le voisinauc d llcliron avaienl cIc iden-
liqucs, iniiiiics dc lulls iirdi'cs, piadcsianis, lilii'S avee I'arlirc dWliraliam. Adrien ins-
laics, cic. lie I’cs conlV'ccnccs. II n'y cii a laCa du Tcn'diinllic unc I'oire, qui dura
pri'S

ipi ti lie, colic dll II. 1’ . I' c \ I'icr siir sai II I I'.lic, (d acipiil unc grande cclidirilc on saisil ;

pali'iari lie dc .Icnisalcin, ([ui sidl on didiors iiicnie la la genese d'nn renouveau cer;'-
dll doiiiaiiic do la llcviic. l.a premiere, cellc nionicl; des (dirclicns ed des juilssy ren-
do I’. Iiliuriiic l•\pllsc Ics (Irioiiics llaliylo- daienl on masse, iiiais « on y lionorail les
iiiciiiics: oil >
tromcra iiii lum I'.Nposi’ dc anges par des lilialions de vin, des sacrilices
la i; o^rajdiic di' la .M6su|iolamio, avoc dis- de volailles on dehclesa comes des lu ii- i.d,

riissioiis pliilidii;;i pu'S d apres Ics docii- Icries ircncens »: dans Ic puils d .Miraliam

mciils Ics pill'' ii'm ciiIs, el des di'dillls do on jelail des galcaux, des monnaics on dc
1 liisli'irc sumcro-aiaaidioiiiic cl assy ro-lialiy- rcnceiis, on versail du\in cl des [larlums,
loniciiiio. Colic dll I’. I.ai;rani;c csl iiililiilcc on rcnviroiinail dc liimicrcs » (So/omcnc,
A lianors papyrus orocs Icsc osl d apres : Socralc, (dc). I'.onslanlin lildidriiirc I’aiilel

les papyrus (ru.wrliympi , d llilicli, c|c. uii el (•oiislrnire line hasiliipie iiour « idirislia-

1 ildcaii aiiiiiic cl amii^aiil. do la \ic qiioli- niscr I) cel eni|ilaccnicnl. I. arlirc moiiriil

dieiiiic Ciiiilumos, m;ciii '., croyaiiccs, droil. ccniimairc sous Ic rijgne dc l.oiislance

rclii.'ion, sciilimeiil- dii I'ayoum, empire du (:j:i7-:il)l) mais Ic pi’derinage a Mamlm’

ilioii i rociidilc, sous les I’tolciimcs. I.'‘ 1’. continiia, coiiime on Idmoignc le rcjil du
licrmcr-Diiraiid oliidie, on iirciiaiil puiir |ialcslinicn So/,oni(''iie (v sii-cle). I’uis le

pidiil do dcparl des claloiis cii piorro 1>. Aliid monire conimeni, par siiilc du
n'rciiimenl di'-coiivcrls, les nicsiircs do synert'disme ordinaire en Oricnl, paiens el

capacilcs lies Iltdireu.x (pliolo). liii dr'crivaiil clindiciis cx(';ciilaienl en ce lieu, cliaciin a sa
scs excursions aiilour du lac do Tihdriade, niani(''re, les riles el ceiadnonics; Aliraliani
dmii Zcphyriii lliever note qiiolqiies ac(|uil nicrnc dans 1(‘ jiopiilairc le r(‘nom
li'^'oiidcs, eiilro aiilres la rorme acliicllo do d'liiic soldo de magicicn, vers(‘ dans l occnl-

la Icfjcndo do .loll, cas iiili'Tcssanl do lixilV lisnic (liald(''en. I, 'Islam survcnanl coni-
locale d’uii llii'iiic I'cril : p. CIO, ilcilcaussi pliqiia le syncriHisme anir'rienr: Ics noiiis

la legciidc qiii allriliuc a Ali odli lilialior, aralies ai duels des accidcnis dc terrain

cliol' dc liri^ands du xviii'i sicrio, des (coniine le vallon du I’alriarclic, le crciix

oiivraL'OS liydraiiliqucs dataiil dc la liii loul du T(ii((diinllic, Ic liain dc Sara, la collinc

au moiiis do roccupalioii nirnaine, cl e'est d’Aliraliam) soul cxplicilcs a cel ('gard. —

laun joli cas dc a lajeiiiiissemcnl liislo- Puis vienl unc discussion d’ordre aiclieolo-
rique ». Cf. ma I'onnalion des leijcndrs. gi([ue, sur ridenlificalioii des mines exis-

Enliu je sigiialc, dans le memc ordre lanl en cel endroil (au liamel el Klialil,
d'idccs, lo tri'S intoressani mi'moiro du soi-disant piscine ilWbralianV, doni on iie
.1