(navigation image)
Home American Libraries | Canadian Libraries | Universal Library | Community Texts | Project Gutenberg | Children's Library | Biodiversity Heritage Library | Additional Collections
Search: Advanced Search
Anonymous User (login or join us)
Upload
See other formats

Full text of "The carnivores of West Africa"

THE CARNIVORES 



OF 



WEST AFRICA 



BY 



D. R. ROSEVEAR 



\ 




.— cSt-T 



BRITISH MUSEUM (NATURAL HISTORY) 



z 



o-]^ 



/^ 



A 



a. 



O 



Price £18. so 



THE CARNIVORES 

OF 

WEST AFRICA 



«AT. HISf. '\ 

-4FEB1975J 

ft ^*^ 



Publication No. 723 



THE CARNIVORES 

OF 

WEST AFRICA 



BY 

D. R ROSEVEAR 




with II plates in colour 

by 

Rita Parsons 

and 172 line drawings 

by 

Patricia Wolseley, Monika Shaffer, 

Rita Parsons and the author 



TRUSTEES OF 
THE BRITISH MUSEUM (NATURAL HISTORY) 

LONDON: 1974 



-'•^'.^c 



First published 1974 

© Trustees ot the British Museum (Natural History) 1974 

Pubhcatioii number 723 

ISBN 0565 00723 X 



Primed in drcit Brilain by Sraplts Printers Limned at 'Hie George Press, Kerterinir Nortiidmptimsliire 



NOTE 

Definition of the area taken herein as representing West Africa and of other 
topographical terms will be found on page 518. 



"... I should have much stronger expectations than I dare yet 
entertain, to see philosophy solidly established, if men would 
more carefully distinguish those things that they know from 
those that they ignore or do but think, and then explicate clearly 
the things they conceive they understand, acknowledge ingen- 
uously what it is they ignore, and profess so candidly their 
doubts, that the industry of intelligent persons might be set on 
work to make further enquiries and the easiness of less discerning 
men might not be imposed on". 

ROBERT BOYLE 

The Sceptical Chymisi 

1661 

"Ce que nous connaissons est peu de choses, ce que nous ignorons 
est immense". 

'iltc Inst words of 
PIERRE SIMON, MARQUIS DE LAPLACE 

1S27 



PREFACE 

The research for this, the third volume on the West African mammals, was made 
possible by the generous support for three years of the Wellcome Trust, to whose 
Trustees, Director and staff I am most sincerely grateful. I must also once more express 
my thanks to the Trustees of the British Museum (Natural History) for the facilities 
again accorded to me. This present work to all intents and purposes brings to a close 
an association with the Mammal Room extending over almost forty years, and I cannot 
too strongly express my deep appreciation of the ever-friendly help and encouragement 
1 have received throughout from the members of this section and most particularly in 
recent times from Dr. Gordon Corbet and all his subordinate staff in their various 
capacities. I also gladly acknowledge my indebtedness to the Museum's Librarian, 
Mr. M. J. Rowlands, and his assistants who have repeatedly and readily gone out of 
their way to give me their help. 

I have been much encouraged by various correspondents who have found the 
previous volumes useful, and by the help I have received from collectors and field and 
museum workers who have kindly furnished me with specimens or items of informa- 
tion. These are more fully acknowledged in the appropriate places; but I would here 
mention A.J. Hopson, G. S. Child, R. H. Kemp, Sonia Jeffreys and Gerald Durrell for 
notes concerning distribution or behaviour ; Dr. P. H. J. van Bree of Amsterdam and 
Dr. Renate Angcrmann of Berlin for the loan of specimens; and J. J. C. Mallinson for 
photographs. I owe a very special debt to Theo S. Jones for his unflagging interest in 
this entire project, for his continual encouragement and for his practical help, despite 
an over-busy existence, in reading through the typescript and supplying numerous 
valuable notes from his wide experience in the field. The typescript has been read, too, 
by Robert Hayman who, although now in well-earned retirement, has once again as 
so often before kindly offered me helpful suggestions, thus adding yet more to my long- 
standing indebtedness to his wide knowledge of African mammalogy. 

No one who studies the morphology and taxonomy of the Carnivora can fail to be 
deeply impressed with the encyclopaedic knowledge of the Order possessed by the late 
R. I. Pocock, the outcome of a lifetime's acquaintance as much with living or recently 
dead specimens in zoos as with mere skulls or dried skins. I should be truly wanting in 
gratitude if I failed to acknowledge how very much tliis present work owes to the 
innumerable clearly illustrated papers of this remarkable and indefatigably inquisitive 
student. 

All who use this book will agree that it owes a great deal of its value to the painstaking 
work of the artists who have assisted in its production, Rita Parsons, Patricia Wolseley 
and Monika Shaffer. I have been fortunate in fmding such willing helpers and I am 
indeed grateful for their cheerful co-operation in the exacting task of achievingaccuracy 
and clarity. I must add m)' thanks also to Sylvia Oliver who has almost faultlessly typed 
the whole of my lengthy manuscript, skilfully and imcomplainingly taking in her stride 
a variety of unfamiliar languages, technical terms, columns of figures, niceties of spacing 
and abbreviations, often bedevilled with maddening complexities of punctuation. 

Though less immediately concerned with the day-to-day production of the 

vii 



work, I cannot overlook how much I owe initially to Lord Grey of Naunton whose 
foresight and faith originally got the whole of this project off the ground; and to Sir 
Terence Morrison-Scott whose friendly help and persistence in the face of financial 
difficulties kept it airborne over a period of years. All, however, would have been 
useless without the never-failing support and encouragement of my wife and her 
forbearance through several decades of collecting, skull cleaning and, ultimately, 
printer's litter. Lastly I make a wholeheartedly sincere acknowledgment of my indebted- 
ness to Mr. Stanley Raw, F.R.C.S., but for whose surgical skill this work would never 
have reached its conclusion. 

To all of these, as well as to many other unnamed friends and helpers, I am more 
deeply grateful than words can properly convey. It is only to be hoped that this account 
of a truly fascinating group of animals goes someway to justifying their faith, and that, 
in a region not ver\^ rich in local literature, the work will be found to fill a gap and 
continue to serve a useful purpose for some years to come. The subject is a complex 
one, far more so than is commonly imagined even by those who justifiably claim a 
considerable knowledge of the animal world. Those who study the often involved 
taxonomic sections will gather something of this author's disquiet regarding mammal- 
ogical taxonomy in general, with its insistence on reference to anciently named types 
often chiefly notable for their ridiculously sketchy diagnoses, and with its placid 
acceptance into an already encumbered nomenclature of new names — without question, 
until it is too late and synonymy is, under the present system, burdened with them for 
eternity. Turning from the museum to the field it is interesting to note the remarkable 
change which the last two decades have witnessed, in that the habits and behaviour of 
African mammals, once the recognized speciality of the professional hunter, have now 
become a fashionable field of intensive scientific study supported by large fimds and 
expensive equipment. This, as the exact niches in nature tilled by each species are more 
clearly appreciated, must result in benefit to the animals themselves — as, indeed, it has 
already done for some at least of those dealt with in this volume, which not so very 
long ago v/ere almost universally regarded as harmful and expendable predators. 

D. R. ROSEVEAR 

Hartley Wintney 
Hampshire 
April, 1972 



CONTENTS 



Page 



Preface ... 

List of Illustrations 

Checklist 

General Introduction 

Canoidea 

Canidae. 

Caninae 

Simocyoninae 
mustelidae 

mustblinae . 

Mellivorinae 

lutrinae 
Feloidea. 

ViVERRIDAE 

Viverrinae . 

Paradoxurinae . 

Herpestinae. 
Hyaenidae 
Felidae: Felinae 
Glossary of terms . 
Note on the area taken 
Note on vegetation 
References 
Index 



AS West Africa 



Vll 
X 

I 

5 
29 

30 
34 
75 
92 

93 

no 

128 

160 
161 
164 
229 
239 
341 
373 
513 
518 

519 
5^4 

538 



ILLUSTRATIONS 



Fig. 
I. 



3- 
4- 
5- 
6. 

7- 
8. 

9- 

10. 

II. 



13- 

14- 

15- 
i6. 

17- 

i8. 
19- 

20. 

ai. 



.11.26, (J, X I 

: i; 



TypRal carnivorous teeth, showing the position of closure: a. right upper 
jaw, palatal view above, lateral view below; b. right lower jaw, lateral view 
above, dorsal view below. Feiiiicais zcida, B.M. No. 25.5.12.22, $, x 2 . 
Carnassial teeth, lateral and surface views; a. right upper; b. right lower. 
Fclis iihir^driici aircmis, B.M. No. 67.1429, ,^, x 3 . 
Golden lackal [Cnnis aureus) .... 

Ciiiiis aureus: skull, B.M. No. 21. 2. 11. 28, 5, x § . 
Side-striped Jackal {Catiis adustus) 
Civus athistus: skull, B.M. No. 28.6.3.8, sex ?, x | 
Fenneais zerda: skull, B.M. No. 1939. 1746, ?, x i 
Vulpi's rueppelli: skull, Type ot caesia. B.M. No. 21 
Hunting Dog [Lycaoii pictus) .... 

Lycaoii pictus: skull. Type o( sharicus, B.M. No. 7.7.8.74, 
view ......... 

Lycaoii pictus: skull. Type of sharicus, B.M. No. 7.7.8.74, 
& dorsal views. ....... 

African Striped Weasel {Pcecilictis lihyca) and Striped Polecat {Ictony: 
striatus) .......... 

Ictcnyx striatus seiicgakusis: skull, Type, B.M. No. 50.7.8.38, $, 
lateral view ......... 

Ictonyx striatus seiiegalensis: skull. Type, B.M. No. 50.7.8.38, ?, 
palatal & dorsal views .... 

Poecilictus lihyca: skull. Type of rotlischiUi, B.M. No. 25.5.12.26, cJ, 
Mellivora capctisis: skull, B.M. No. 26.8.7.1, J, x |; lateral view 
Mellivora capciisis: skull, B.M. No. 26.8.7.1, cj, ^ |; palatal & 
views ....... 

Ratel or Honey Badger [Mellivora capeusis) . 
Mellivora capensis: dorsal coat patterns, showing the extent of white in the 
various forms: a. leucouota, b. buchauaui, c. concisa, d. signata, e. cottoni 
Lutra uiaculicollis: skull, B.M. No. 23.1.22.44, sex ?, x i ; lateral view 
Lutra iiwculicollis: skull, B.M. No. 23.1.22.44, sex ?, .• i; palatal & dorsal 
views ............ 

Lutrinae: posterior cheekteeth; top row, right upper p'^ and m^; bottom 

, ■ 2: a. Lii/cd omnf/ico///.';, B.M. No. 23.1.22.44; 

No. 10. II. 25. 2; c. Acnyx [Paraonyx) cougica. 



lateral 



palatal 



X 2; 



> 2; 



^3- 
24. 



row, left lower hii and m; 

b. Aoiiyx capeusis, B.M. 

B.M. No. 1938.9.29.8 

Aoiiyx [Paraoiiyx) cougica: 

view 

.-iouyx [Paraonyx) cougica: 

d< dorsal views. 



skull, B.M. No. 1938.9.29.4, sex ?, x |; lateral 
skull, B.M. No, 1938.9.29.4, sex ?, x |; palatal 



Page 



16 

17 
37 
45 
SI 
53 
59 
68 

77 
80 
81 

97 

98 

99 

107 
114 

115 
119 

1^5 
144 

145 



148 
156 

157 



ILLUSTRATIONS 



African Civet {Civettictis civetta) ...... 

Civettictis civetta: skull, B.M. No. 59.941, $, x f ; lateral view 

Civettictis civetta: skull, B.M. No. 59.941, ?, x f ; palatal & dorsal views 

Genetta: left bullae, x 3: a. G. genetta (? ajra), B.M. No. 9.11.2.34, S 

b. G. geneltoides, B.M. No. 46.397, (J; c. G. cristata, B.M. No. 39.323, J 

Geueffa iCHc^a/eo^-w: skull, B.M. No. 1939. 1766, c?, X i. 

Pdirtiia nVZ/rtrJic'/ii; skull, B.M. No. 98.3. 19.11, cJ, X I . 

Naiidinia binotata: skull, B.M. No, 48.814, sex ?, X I ; lateral view . 

Naiidinia hinotata: skull, B.M. No. 48.814, sex ?, x i ; palatal x dorsal views 

Herpestiuae: illustrating the two contrasting positions of the cheekteeth 

a. p^ well anterior to the root of the zygoma {Ichneumia albicauda, B.M 

No. 25. 5. 12. 13); b. posterior corner oi p* about level with the zygoma 

[Xeiwgale tiaso, B.M. No. 10.6. 1. 14) 

Miingos gainbiamis: skull, B.M. No. 36.10.30.12, cJ, x | 

Miingos gambianus: bulla, x 2 . 

Herpestes ichneumon: skull, B.M. No. 9.1 1.2. 10, sex ?, x i 

Crossarchus obsciirus: skull, B.M. No. 48.841, sex ?, x f. 

Atilax palndinosHs: skull, B.M. No. 13.4.1 i, (J, x i 

Iclmetwiia albicauda: skull, B.M. No. 12.4.3.3, (Ji X I 

Galerella sanguinea melanura: skull, B.M. No. 35.1.30.53, cJ, 

Galerella sanguinea: bulla, x 2 . 

Galeriscus nigripes: skull, B.M. No. 20.10.26.2, (J, X i; lateral view 

Galerisctis nigripes: skull, B.M. No. 20.10.26.2, <J, x i; palatal & dorsal 

views ........... 

Xenogale naso: skull. Type of nigerianus, B.M. No. 10. 6.1. 14, ?, X I 

Liberiictis kuhni: skull, Type, B.M. No. 58.507, sex ?, x i 

Striped Hyaena {Hyaena hyaena) and Spotted Hyaena {Crocuta crocuta) 

Hyaena hyaena: skull, B.M. No. 23.1. 1. 78, J, x |; lateral view 

Hyaena hyaena: skull, B.M. No. 23.1. 1.78, (J, X i; palatal & dorsal views 

Crocuta crocuta: skull, B.M. No. 59.272, $, (missing posterior detail filled 

in from B.M. No. 61.41, sex ?), x ^; lateral view .... 

Crocuta crocuta: skull, B.M. No. 59.272, $, (missing posterior detail filled 

in from B.M. No. 61.41, sex ?), x J; palatal & dorsal views. 

Ft'/i5 /i'fc)'M/o.vi; skull, B.M. No. 13.2.5.3, (J, X i ; lateral view 

Felis libyca foxi: skull, B.M. No. 13.2.5.3, cJ, x i; palatal & dorsal views 

Felis margarita airensis: skull. Type, B.M. No. 1939. 1673, ?. x I . 

Felis caracal: skull. Type of poecilotis, B.M. No. 21. 2. 11. 19, ?, X I ; lateral 

view ............ 

Felis caracal: skull. Type oi poecilotis, B.M. No. 21. 2. 11. 19, $, x i ; palatal 
& dorsal views . .......... 

Felis serval: skull, B.M. No. 99.10.23.3, (J, X 1; lateral view . 

Felis serval: skull, B.M. No. 99.10.23.3, cJ, x i; palatal & dorsal views 

Felis aurata: skull, B.M. No. 25. 10. 7.13, $, x ^o; lateral view 

Felis aurata: skull, B.M. No. 25. 10. 7.13, 3, x ^; palatal & dorsal views 



169 

172 
173 

180 

195 

223 

^34 
235 



241 
250 
264 
270 
284 
296 
304 
313 
314 
326 

3^7 
334 
338 
347 
350 
351 

358 

359 

388 
389 
398 

404 

405 
418 
419 
43- 
433 



XI 



Tin. (AUNIVOKIS C)l WlSl AIUKA 

60. Panthcra ihudm: pcLigc patterns, ■ I . . . . . . . 443 

61. Piii;f/i('rii /icirJib; skull, B.M. No. 35.10.22.71, o. ■ .1 ; lateral view . . 446 

62. Panihcra p,iidu<: skull, 15. M. Nci. 35.10.22.71, j, • !,; palatal &: dorsal 
views ............ 447 

63. Pii/if/itTd ic'i'.- skull, B.M. No. 21.7.23.1. o- ■ s ; lat'-T'i' view . . . 468 

64. Pii»r/u'ni /I'l'.- skull, B.M. No. 21.7.23.1, o, •, j^ ; palatal & dorsal views . 469 

65. .^(•/(;(l//)■.v /i//iiifi(.v; skull, B.M. No. 32.12.27.1, $, ;■; |^; lateral view . . 498 

66. Acitwnyx julhitiis: skull, B.M. No. 32.12.27.1, ?, •' |-; palatal 6c dorsal 
views ............ 499 



COLOUR PLATES Riciug 

Plate 1. Sand Fox [I'ulpcs pallitht); Fennec (Fciiiwciis zciila); Riippell's Fox 

{Viilpcs nicppclli) ......... 66 

2. Speckle-throated Otter [Lutra iitaatlicoUis); Cape Clawless Otter 
[Aoiiyx capensis) . . . . . . .142 

3. Senegal Genet (Gciictta scncgalciisis); F^ausa Genet (Gciidiii tliiciryi); 
Crested Genet [Gcnctta cristata) .190 

4. Pel's Genet {Genctta \ifiicttoii1cs); Small-spotted Genet [Gciuiia 
poaisis); Bcnm Gene: (Gcficttii hiiii) . .210 

5. Richardson's Linsang {Poiaiui licliaidsoin); Two-spotted Palm Civet 
[Natidinia hinolata) ......... 226 

6. Kusimanse {Civssiucliiis ohininis) ; Ganibian Mongoose {Miiii(;o.^ 
gamblaims); Talbot's Banded Mongoose {Miiiii^os niuni;;o lalboii) . 252 

7. Forest Ichneumon [Hcrpcstcs icliiieiiiuoii aitlios) ; Western Ichneumon 
(Hcrpestes ichneumon occukiitaUi) ; Marsh Mongoose [AtHax pahuhno- 

<ns) ............ 276 

M. Whitc-tailed Mongoose [Ichncuiwia alhicdtida); Black-tooted Mon- 
goose [Gali'iisciis iiigripes) ........ 302 

9. Forest Slender Mongoose {Galerclla siuif^iiiuca mcLimiia); Western 
Red-tailed Mongoose {GitleiclLi SiViiiuinca phoeiiicnni); Guinea Slender 
Mongoose [Galcrclla san^iiincii cana) . . 3 1 X 

10. African Wild Cat (Fc/i'i /i/iyci;); Caracal (fV/i'i (cJiiicii/) . . . 386 

11. Serval (Fi'/ii .<itim/) ; African Golden Cat (Fi7i'.\ ii/iM^i) . . . 414 



CHECKLIST 

Cohort FERUNGULATA Simpson, 1945 Carnivore — Herbivore Group 

Supcrorder FERAE Linnaeus, 1758 Carnivorous Mammals 

Order CARNIVORA Bowdicli, 1821 Land and Sea Carnivores 

Suborder FISSIPEDA Blumenbach, 1791 Land Carnivores 

Superfamily CANOIDEA Simpson, 193 1 Dog-like Superfamily 

Family CANIDAE Gray, 1821 Dogs, Foxes, etc. 

Subfamih' CANINAE Gill, 1872 Typical Dogs 

Genus CANIS. Linnaeus, 1758 Dogs 

C. rt/(/ri/i" Linnaeus, 1758 Golden Jackal 

C. adiistus Sundcvall, 1846 Side-striped Jackal 

Genus F EN N EC U S Dcsmarest, 1804 Feimecs 
F. zcrda (Zimmermann, 1780) Fcnnec 

Genus VULPES Fleming, 1822 True Foxes 

V. rncppdli (Scliinz, 1825) Riippell's Fox 
I '. pallidi (Crctzschmar, 1826) Sand Fox 

Subfamily SIMOCYONINAE Zittcl, 1893 Himting Dogs 

Genus LyCAO AT Brookes, 1827 Himting Dogs 

L. /)iVfi/_^ (Tcmminck, 1820) Hunting Dog 

Family MUSTELIDAE Swainson, 1835 Polecats, Weasels, Otters, etc. 

Subfamily MUSTELINAE Gill ,1872 Polecats, Weasels, etc. 

Genus ICTONYX Kaup, 1835 African Polecats 

/. striatiis (Pcrrv, 18 10) African Striped Polecat; Zorilla 
/. i. sciicgalcnsi!: {]. B. Fischer, 1829) Senegal Striped Pole- 
cat 

Genus POECILICTIS Thomas & Hiiuoii, 1920 Striped Weasels 
P. lihycn (Hemprich & Ehrenberg, 1832) Striped Weasel 
P. I. iiniltii'ittata (Wagner, 1841) Southerly Striped Weasel 

Subfamily MELLIVORINAE Gill, 1872 Ratels 

Genus MELLIVORA Storr, 1780 Ratels 

M. capciisis (Schrcber, 1776) Ratel; Honey Badger 
M. c. leticoiiotci P. L. Sclatcr, 1867 White-backed Ratel 
M. c. cottoiii Lydekker, 1906 Black Ratel 
M. c. coiicisa Thomas & Wroughton, 1907 Lake Chad 

Ratel 
A/, c. sigiiata Pocock, 1909 Speckled Ratel 
M. c. huchanani Thomas, 1925 Air Ratel 



THE CARNIVORES OF WI.ST A E R H A 

Subtamily LUTRINAE Baird, 1S57 Cotters 

Genus LUTRA lirisson, 1762 Typical Otters 

Subgenus HY'D RICTIS Pocock, 1921 African Long-clawed 

Otters 
L. iiuiailicollif Liclitenstein, 1835 Speckle-throated 

Otter 

Genus A ON FA' Lesson, 1S27 African Clawless Otters 

Subgenus /I O/Vy.Y Lesson, 1827 Typical African Clawless 

Otters 
A. Ciipciisis (Schinz, 182 f) Cape Clawless Otter 

Subgenus PARy40iVyA' Hinton, 1921 Small-toothed Clawless 

Otters 
A. coiigiCii (Louiiberg, 1910) Small-toothed Clawless 

Otter 

Supcrfamily FELOIDEA Smipson, 193 1 Cat-like Superfaniily 

Family VIVERRIDAE Gray, 1821 Civets, Genets, Mongooses, etc. 

Subfamily VIVERRINAE Gill, 1872 Civets, Genets, Linsangs 

Tribe VIVERRINI Wmge, 1895 Civet Tribe 

Genus CIVETTICTIS Pocock, 1915 African Civets 
C. civctta (Schrcber, 1778) African Civet 

Genus GEN ETTA G. Cuvier, 18 16 Genets 

Subgenus GENETTA G. Cuvier, 1816 Typical Genets 
G. qciictta Linnaeus, 1758 European Genet 
e.g. (ifra F. Cuvier, 1825 North African Genet 
G. sciicgiilcnsis {y B. Fischer, 1829) Senegal Genet 
G. pardiiiti L Gcoffroy, 1832 Pardme Genet 
G. ^e/)e/?('/Jt'5 Temminck, 1856 Pel's Genet 
G. pocnsis Waterhouse, 1838 Small-spotted Genet 
G. cristata Hay man, 1940 Crested Genet 
G. hini sp. iiou. Benin Genet 
G. tliicnyi Matschie, 1902 Hausa Genet 

Subgenus PARAGENETTA Kuhn, i960 Kuhn's Genets 
G. jolnistoiii Pocock, 1908 fohnston's Genet 

Genus POIANA Gray, 1864 African Linsangs 

P. I /(:/)ii)(/iii;;( (Thomson, 1842) Richardson's Linsang 
/'. lfi(^liioiii Pocock, 1908 Leiglitoii's Linsang 



CHECKLIST 3 

Subfamily PARADOXURINAE Gill, 1872 African and Asian Palm 

Civets, Binturong, etc. 

Tribe NANDINIINI Simpson, 1945 African Palm Civet Tribe 

Genus N AND INI A Gray, 1843 African Palm Civets 

N. binotata (Gray, 1830) Two-spotted Palm Civet 
N. h. binotata (Gray, 1830) Typical Two-spotted Palm 

Civet 

Subfamily HERPESTINAE Gill, 1872 Mongooses 

Tribe HERPESTINI Winge, 1895 Typical Mongoose Tribe 

Genus MUNGOS E. GeortVoy & G. Cuvier, 1795 African Mungos 
M. mungo (Gmelin, 1788) Banded Mongoose 
M. in. q^othnch (Heuglin & Fitzinger, 1866) Kordofan 

Banded Mongoose 
M. m. talboti (Thomas & Wroughton, 1907) Talbot's 

Banded 
Mongoose 
M. in. inandjamin (Scliwarz, 1915) Schwarz's Banded Mongoose 
Af. )/). caiiriims Thomas, 1926 North-west Banded Mongoose 
M. g^ainhianns [OgAhy, 1835) Gambian Mongoose 

Genus HERPESTES Illiger, 181 1 Typical Mongooses 

H. ichneumon (Linnaeus, 1758) Iclmeumon; Egyptian 

Mongoose 
H. i. occidentalis Monard, 1940 Western Iclmeumon 
H. i. aitlws siibsp. nov. Forest Iclmeumon 
H. i. intermedins subsp. nov. Nigerian Ichneumon 

Genus CROSSARCHUS F. Cuvier, 1825 Lesser Long-nosed 

Mongooses 
C. obscunis F. Cuvier, 1825 Kusimanse 

Genus ATILAX F. Cuvier, 1826 Marsh Mongooses 

..4. ^(jWi/i05!/i (G. Cuvier, 1829) Marsh Mongoose 

Genus ICHNEUMIA I. Geotfroy, 1837 White-tailed Mongooses 
/. albicauda (G. Cuvier, 1829) White-tailed Mongoose 

Genus GALERELLA Gray, 1865 Slender Mongooses 

G. sangiiinea (Riippell, 1836) Slender Mongoose 

G. s. melanura (Martin, 1S36) Forest Slender Mongoose 

G. 5. aina (Wroughton, 1907) Guinea Slender Mongoose 

G. i. phoenicnra (Thomas, 191 2) Western Red-tailed Mongoose 

G. s. saharae (Thomas, 1925) Air Rcd-tailcd Mongoose 



Tun CARNIVORES OF WFST AFRICA 

Genus GALERISCUS Thomas, 1894 Foiir-tocd Mongooses 
G. iiiiiiipcs (Puclicran, 1855) Black-footed Mongoose 
G. ;;. ;;((;) //>i'.'; (Pucheran, 1855) Typical Black-footed 

Mongoose 

Genus XENOGALEj. A. Allen, 1919 Greater Long-nosed Mon- 
gooses 
X. luisc (de Winton, 1901) Greater Long-nosed Mon- 
goose 
A'. II. iiaso (de Winton, 1901) Typical Greater Long- 
nosed Mongoose 
Genus LIBERIICTIS Hayman, 1958 -Liljcriau Mongoose 
L. Liiliiii Haynian, 1958 Kuhn's Mongoose 

Family HYAENIDAE Gray, 1869 Hyaenas 

Subfamily HYAENINAE Mivart, 1882 Hyaenas 

Genus HYAENA Brisson, 1762 Striped and Brown Hyaenas 

H. hychiui (Linnaeus, 1758) Striped Hyaena 
Genus CROCVTA Kaup, 1828 Spotted Hyaenas 

C. ciocHta (Erxleben, 1777) Spotted Hyaena 
Fanul)' FELIDAE Gray, 1821 Recent and Fossil Cats 
Subfamily FELINAE Trouessart, 1885 Cats 
Genus FEL/S' Linnaeus, 1758 Cats 

Subgenus FELIS Linnaeus, 1758 True Cats 

F. Uhyca Forster, 1780 African Wild Cat 
F. I. haussiJ Thomas & Hiiiton, 1921 Hausa Wild Cat 
F. /. foxi Pocock, 1944 Mid-belt Wild Cat 
F. iiiargarita Loche, 1858 Scind Cat 
F. //;. iiiiriisis Pocock, 1938 Air Sand Cat 
Subgenus CARACAL Gn^y, 1843 Caracals 

F. ctiiciail Schreber, 1776 Caracal; Desert Lynx 
Subgenus LEPTAILURUS Severtzov, 1858 Serval Cats 

F. scnnil Schreber, 1776 Serval Cat 
Subgenus PROFELIS Severtzov, 1858 Golden Cats 

F. iiiiiiitii Temminck, 1824 African Golden Cat 
Genus PANTHERA Oken, 1816 Great Cats 
P. /»iii(///.< (Linnaeus, 1758) Leopard 
P. p. paiihis (Linnaeus, 1758) W. African Open-country 

Leopard 
P. p. Icopdidiis (Schreber, 1775) W. African Forest Leopard 
P. Ico (Linnaeus, 1758) Lion 
Genus ACINONYX Brookes, 1828 Cheetahs 
.4. yi(/iii/ir.< (Schreber, 1775) Cheetah 



GENERAL INTRODUCTION 

The system of classification adopted throughout this work is in the main that of 
Simpson (1945) with relatively minor amendments in the lower taxonomic ranks. That 
author divides the Euthcrian (placental) mammals into four cohorts, the Carnivora 
forming part of the cohort Ferungulata, the other three cohorts being the Unguiculata, 
comprising the insectivores, chiroptera and primates; the Glircs, consisting chiefly of 
rodents; and the Mutica, the whales and dolphins. 

Cohort FERUNGULATA Simpson, 1945 

As far as recent, living, mammals arc concerned the Cohort Ferungulata covers the 
following Orders: Carnivora (Flesh-eating mammals), Tubulidcntata (Aardvarks), 
Proboscidca (Elephants), Hyracoidca (Hyraxcs, known also as Dassies or Conies), 
Sircnia (Manatees, Dugongs), Perissodactyla (Odd-toed Ungulates) and Artiodactyla 
(Even-toed Ungulates). 

From the point of view only of existing mammals some of these seem to make 
strange bed-fellows; especially in that the carnivores, that is to sav the wild cats, dogs, 
genets, mongooses and so forth, should be regarded as having any kind of near relation- 
ship with the artiodactyls or antelopes and their close kin — in relation to which, both 
superficially and in general mode of life, they appear not only to have nothing in 
common but to stand, rather, at opposite poles. The explanation of this seeming paradox 
lies, of course, in the fossil record, with the complications of which this present account 
cannot involve itself beyond the brief statement that differentiation from a common 
ancestor into what are now broadly the hunter and the hunted took place in early 
Tertiary times. It must be further added that Simpson's views on this matter are not 
those of all palaeontologists. 

The Ferungulata are regarded as comprising five Superorders, of which we are here 
concerned only with the Ferae. These latter may be broadly regarded as covering 
beasts of prey, the remaining Superorders consisting, by and large as far as living 
mammals are concerned, of herbivorous, or occasionally insectivorous, species. 

Superorder FERAE Linnaeus, 1758 

Order CARNFVORA Bowdich, 1821 

On the other hand, the Ferae comprise a single Order, named by reason ot the 
predominantly predacious habit of its constituent members the Carnivora, from the 
Latin auo, cainis flesh, and vow to devour. The animals included under this heading are 
divided into two Suborders, the Fissipeda or land-based carnivores, and the Pmnipedia, 
which spend the greater part ot their active lives in the sea, upon which they are wholly 
dependent for their food. The land carnivores are to be found everywhere with the 
exception of the soutliem polar regions and a few islands. The pinnipedcs, that is to say 
the seals and their kin, are of wide distribution in the oceans of the world but onlv one 



Tlir CAUNIVdlllA (H WIST M-IilrA 



genus, Moiuwhns, is partly tropical, and tliat is not known to occur withiji the Innits 
chosen for this book, the monk seal [M. iiioiiachiis) being recorded no further south 
along the western coast ot Africa than Cape Blanco (Rio dc Oro). This present account, 
therefore, is in effect concerned only with the Fissipeda. It must be added that there is 
an increasing tendency amongst systematists to accord to these two divisions full 
ordinal, rather than subordinal, rank; but though there are plausible arguments to 
support this such a classification does, in fact, tend to obscure the close relationship that 
exists between the two sections. 

Suborder FISSIPEDA Blumcnbach, 1 791 

The two Suborders of Carnivora are named from, and to some extent founded on, 
the general build of the feet. The Fissipeda are characterized by having the digits more 
or less clearly independent of each other, even though in some cases connected by 
interdigital webs. The name is, tor this reason, derived from the Latin words fissuiii 
cleft and pes, pedes foot. In the Pinnipcdia, on the other hand, the digits, notably those 
of the forclimbs, in conjunction with the contiguous bones fiuiction as a single unit 
adapted as a paddle for swimming, encased for this purpose in a common integument — 
though there are, of course, exceptions of greater or lesser degree to this broad plan. 
The name, thus, is built from the Latin word piiiiui a feather because the structure, in 
general shape, and in the unitary way it functions as a means of propulsion, recalls a 
bird's wing rather than the normal run of mammalian foot. 

The fissipede carnivores consist, in West Africa, of wild dogs (i.e. jackals and foxes), 
the hunting-dog, striped weasel, striped polecat, honey badger, otters, civet, palm- 
civet, genets, mongooses, hyaenas and wild cats of various kinds including the lion, 
leopard, cheetah and others less well-known. These are disposed in j families, 27 genera 
and 45 species embracing about 59 different forms. Of the two families which are 
lacking entirely from the African famia the more notable is the Ursidae or bears. The 
other, the Procyonidae, though of considerable importance in both northern and 
southern America and in eastern Asia, comprises less widelv known animals such as the 
raccoons, coatis, kinkajou and pandas. 

General description. The carnivores, especially the fissipede carnivores, are an 
extremely interesting group. They are certainly more intimately associated with man, 
in the capacity of close friend or bitter enemy, than any other Order of mammals. They 
have fairly big brains and are more often than not highly intelligent; they live by their 
wits; and under domestication, therefore, they are quick to grasp a situation and to 
learn. Without question, both the domestic cat and the domestic dog have, each in its 
own special way and ni return for assured food and snug shelter, turned themselves 
into man's most intimate and seemingly luidcrstanding associates. Further, many ot the 
normally wild species take iniexpectcdly kindly to captivity and become amazingly 
responsive to close association with human beings. Yet in the wild state there are no 
more unremitting foes to man's cattle, sheep and poultry, as well as to wild game, 
than the majority of fissipedes. 

The suborder exhibits a wide range of size, varying in West Africa from a 



GENERAL INTRODUCTION 



little weasel with a head & body length of only a few centimetres and weighing 
about 200 grammes (i lb) to an animal of the vast bulk of a lion which may weigh 
200 kg (4 cwt) or more. Between the different families there is a good deal of variety of 
shape; but in a general sense the bodies throughout are mostly relatively slender, long, 
of subcylindrical form, frequently without any very marked constriction at the neck, 
highly muscular and often, especially in the cats, of extreme suppleness and agility. 
Variety of external appearance is brought about by differences in the shapes and relative 
proportions of the head, the tail, the legs, and the feet, apart from any question of 
texture and pattern in the pelage, hi much of the suborder the common form of head 
is long, tapering m a greater or lesser degree to a pointed muzzle; but the cats differ 
markedly from this general shape in having a much rounder head with a notably 
short muzzle. 

Ears. Ears often form a very conspicuous feature. They are long and upstanding in 
the dogs and hyaenas, sometimes sharply pointed; but in the scrval cat they are very 
rounded besides being tall and broad. In other cats they may be rounded but of much 
less size. They reach their greatest reduction in the ratel, otters and mongooses. The 
backs are clad with short hairs, the inside of the pinna with long hairs, often densely; 
and in the cats the backs mostly bear a black or black and white patch, quite charac- 
teristic of the species. An apical tuft exists in the caracal. A curious aural character 
occurring in large sections of the Carnivora is a doubling of the rim of the pinna near 
the base of the outer margin, forming a little pocket, known as the bursa, with an 
anterior and a posterior flap. The latter is usually semi-lunar; the former emarginate. 
A bursa is always present in the Canidae, Felidac, Viverrinac and Paradoxurinae ; it is 
completely absent from the Hyacnidae, the West African Mustclidae and the Hcrpest- 
inac. Near the base of the inner margin of the pinna is a small process known as the 
tragus; and opposite this on the outer margin is another process, the antitragus, having 
a notch into which, in some cases, the tragus can fit tightly and help to cficct closure of 
the entrance to the ear passage. On the inner face of the pinna are three or four shallow 
folds or ridges crossing it in various directions. These also, in those species which fmd 
it necessary, aid in closing the ear when the piima is collapsed and folded. The termin- 
ology for these is not well fixed. 

Rhinarium. The rhinarium in this suborder, that is to say the region at the extreme 
anterior end of the muzzle surroiuiding the nostrils — the "nose" in common parlance — 
is naked over a larger or smaller area and, in the Canidae at least, has a slow secretion of 
mucus. The nostrils themselves may be narrow and slit-like or rather more open and 
comma-like; and the whole form of the rhinarium is gcncrically, or even specifically, 
characteristic. This has been studied and illustrated by Pocock in a number of published 
papers; and there are figures also in J. A. Allen (1924). 

Eyes. The eyes of the carnivores are mostly of a moderate, or sometimes relatively 
small, size; and they are situated well to the fore on the head, vision being thus 
effectively binocular, permitting that accuracy in the judgment of distance necessary 
to the successful striking of prey. Acuity of vision is in general good but in fact varies 
quite appreciably not only between species but amongst individuals too. It is a matter 
of common observation that some breeds of dog, or different members of a single 



8 TUP, CARNMVOKrS OF WHST AIIUCA 

breed, can see much better than others; and tliis apphes as much to wild species ot 
carnivores as it does to domestic animals. There is Httlc doubt that cheetahs liave con- 
siderable sharpness ot sight over long distances, whereas otters are in a different category 
and, indeed, have eyes probably far better adapted to seeing under water than on dry 
land. Yet however acute the power of vision may be, it is in many, if not all, cases aided 
to a great extent by the faculties of smell and hearing, both of which are better developed 
than m human beings, sometimes to an nicomparably greater extent. Equipped with a 
highly sensitive sense of smell, a carnivore can build up a mental picture of its surromid- 
ings almost beyond the comprehension of man. Sound frequently has no very great 
range, may be easily obstructed or blurred, and is often transitory; but scent sometimes 
lasts for long periods and may carry fir, and it enables a suitably equipped animal to 
"see round a corner" where sight is of no avail. Its chief drawback lies in the degree to 
which it is affected by air currents; for whereas it may lie for long in sheltered under- 
growth it will in open cotmtry certainly be carried considerable distances by the wind. 
Carnivores, therefore, if circumstances permit choice, approach their prey from down- 
wind, securing the double advantage of having the scent of their proposed victim 
brought to thcni whilst their own is borne away. The canidae are notable possessors of 
a very highly developed olfictory sense which they generally rely on in preference to 
sight and, where possible, use to confirm what their eyes appear to have told them. 
Sensitivity of nasal perception and analysis amongst some of the Canidae has been 
shown to be of such delicacy that one second's holding in a man's hand of a billet of 
wood suffices for a trained dog to be able to select it from a pile of numerous similar 
pieces. Cats are more dependent upon hearing as an auxiliary to vision. 

Many carnivores are able to sec unusually well in poor crepuscular or nocturnal light 
conditions, a power very necessary to the securing of prey in these circumstances. Some 
night-active species have irises that arc highly and rapidly responsive to variation in 
light intensity. Accommodation in this respect is achieved in the most typical cats by a 
pupil which in response to a lessening intensity can change from a narrow vertical slit, 
customary for a bright light, to a wide circular opening. By no means all felines have 
this power; and in some the pupil remains, within narrower limits, broadly elliptical. 
It is interesting to note that eyes of a similar type with a vertical slit are found amongst 
the owls, also nocturnal hunters; but the horizontal pupils of ungulates and kangaroos 
do not serve the same function. 

The effect of this enlargement of the pupillary aperture in dim conditions is aug- 
mented in the carnivores by the possession of a second means of increasing the efficacy 
of weak light. Anyone who has observed cats or dogs at night is acquainted with the 
remarkable way in which the eyes assume a green or yellow phosphorescent luminosity. 
Since mediaeval times this somewhat ghostly glowing ot the eyes has been popularly 
held to be something to do with "seeing in the dark" by the mysterious projection of 
beams as from a lamp to illuminate the object looked at. This last, of coiu'se, is not so. 
The effect is due to a subcircular area of tissue at the back of the eye situated immediately 
behind the retina around the optic nerve, having no common name but known in 
anatomy as the tapctum hicidiiiu. This possesses the property of virtually doubling the 
effective intensity of such poor light as may exist by reflecting it back again through the 



GENERAL INTRODUCTION 



receptive cells ot the retina instead of its being at once absorbed in the posterior layers 
of the eye as it is in organs unprovided with this structure. It does thus enable an animal 
to see virtually twice as well in the near-dark as it might otherwise do, but not by the 
generation of light in its own eyes. It follows that it is quite inoperative in absolute dark- 
ness — a condition which rarely exists in nature — and the eyes would emit no glow in 
such circumstances. This dual use of feeble illumination is, of course, of considerable 
advantage but nevertheless has the drawback ot some loss of sharpness of vision 
(Tansley, 1965). Eyes of this kind arc found also amongst whales and ungulates. A 
nictitating membrane, or so-called third eyelid, is fairly well developed in some of the 
carnivores, especially in the cats, though it is never wholly and constantly functional 
as it is in reptiles and birds. 

Vision varies markedly in its sensitivity to colour. Diicker (1957, 1964) found Genella 
and Fclis to be totally colourblind, but the mongoose Hcrpcstcs ichneumon to be relatively 
well equipped, having a positive colour sense though exliibiting difficulty in distinguish- 
ing between nearly related hues such as red and orange or green and blue. 

Legs and feet. Legs are possibly responsible for a greater variety of fissipede form 
than heads. They are in some cases long or very long, adapted almost wholly to 
running, as in the dogs or that in some ways rather dog-like feline the cheetah. In these 
cases the feet are digitigrade and carry moderately straight, non-retractile claws. In 
the typical cats the feet are also digitigrade but the legs are relatively shorter and a 
little stouter with greater flexibility of movement. The feet are armed, with strongly 
curved, very sharp claws which are retractile within sheaths to preserve them from 
abrasion in rmuiing and walking but extensible when a furm anchorage is required, 
either in climbing, taking off for a spring, or in clinging to prey. Such limbs are adapted 
rather to a bounding than to a rurunng gait, to slinking with lowered body, to climbing 
trees and other rough surfaces, and to performing powerful leaps. Between these two 
extremes lie a number of forms, mostly with short legs suited to a trotting gait and 
sometimes to climbing, either with fixed or, in the genets, partly retractile claws, 
digitigrade, semi-plantigrade or, in the ratel, nearly fully plantigrade. Bears, which do 
not occur as wild species anywhere in Africa, provide the best example of carnivores 
that are fully plantigrade, that is adapted to walking with the whole of the foot from 
toe to heel in contact with the ground. There may be four digits on each foot, or five, 
or a combination of the two. Where the ist digit is present it is mostly widely separated 
from the rest, more particularly on the hindfoot, where it is far removed from any 
possible contact with the ground. In the Canidae this is popularly known as the "dew 
claw", possibly because of its ephemeral nature since it often disappears in adolescence. 
Although the digits, unlike those of the pinnipcdes, remain virtually independent they 
are nevertheless often joined by webs at least basally though in semi-aquatic species 
they may be extensive. 

A notable feature of the fissipede foot is its sole, consisting largely of rather rubbery, 
mostly independent cusliions, sometimes collectively referred to, for the fore and 
hindfeet respectively, as the palmar and plantar pads, though these terms have also a 
more restricted application as explained shortly. These pedal pads, in general, fall into 
three categories: those situated below the distal portions of the digits, which may be 



THE CARNIVORES OF WEST AFRICA 



conveniently referred to as the digital or apical pads; those lying in the middle ot'thc 
foot, basically comprising tive pads surrounding a central depression but often reduced 
in number or indistinguishably merged, termed by J. A. Allen (1924) interdigital pads 
but more often called the palmar (forefoot) and plantar (hindfoot) pads. Finally, 
posterior to these, there is often a narrow, longer or shorter metacarpal or metatarsal 
pad, frequently compound. Pads follow set patterns in ditferent species or genera and 
are often wholly diagnostic; but family resemblance is more unitonn in the Felidae and 
Canidae than in the \'iverridae and Mustelidac (Pocock, I9r4b). 

Tails. The tail in the fissipedes is a structure of considerable variet\' and is, with few 
exceptions, usually a showy as well as a characteristically formed appendage. In the 
cats and the genets it is for the most part long and of considerable suppleness of move- 
ment; and it is clad with hairs of subequal, moderate length resulting in a more or less 
cylindrical structure. In the dogs it is far less flexible, capable of little more than stiff 
side to side or up and down motion from the root only. It is sometimes uniformly long 
haired and of striking bushiness. hi the mongooses the tail is tapering, in some species 
markedly so; in some it is rclativelv short-coated, in others of somewhat asymmetrical 
coutour owing to the drooping carriage of its long hairs. The lion's tail is tuiusual 
amongst the African fissipedes in not only being close-haired but in carrying also a 
promir.cnt. blackish, long-haired terminal tuft — hidden amongst which is the famous 
"thorn" of hard skin with which in ancient times it was reputed to lash itself into fury 
before attacking its prey. On a far smaller scale, a terminal tuft, sometimes black but 
much less bushy than in the lion, occurs also in some of the mongooses. Apart from 
variety of shape, tails in this suborder may be unicoloured. bicoloured, annulated, 
spotted or speckled. 

Scent glands. One of the most notable features of the carnivores is the possession by 
many members of the Order of scent glands. These arc mostly situated in the region of 
the sexual organs or the anal area, and their functions are recognition, sexual attraction 
and stimulation, demarcation of territory, warning or defence. As an example of the 
last in West Africa the striped polecat {Ictoiiyx) can, if alarmed, by the deliberate and 
abrupt contraction of the circumscribing muscles emit to a surprising distance a spray 
of fluid with an extremely disagreeable and long-lasting stink. Despite the existence of 
glands, in many cases, including that of this striped polecat, little or no offensive odour 
is under normal circumstances apparent to human beings; but in some animals, notably 
some of the dogs and h)-aenas, a fetid stench unremittingly pervades the whole animal 
bv a continual imcontroUed oozing of the offensive secretion. Involuntary emission of 
odour also in many species accompanies the acts of defaecation and micturition, both 
of which functions, but especially the latter, are in consequence used to indicate the 
boundaries of a territory and publish a warning against trespass by prospective intruders. 
In other cases, establishment of ownership, or other scent-borne message, is conveyed 
by the rubbing of circumanal glands against trees and rocks or by dragging the anus 
across the ground. Scent-marking in the Canidae, especiallv in respect of postures 
adopted, has been dealt with by Kleiman (1966). 

The precise siting of fissipede scent glands is, within a relatively limited bodily reginrt 
vcr)' varied. They may be associated with the anus, either within the rectum oronexe,- 



GENERAL INTRODUCTION II 

ally aroimd the orifice (circumanal); between the anus and the scrotum or vulva 
(perineal) ; in association with the prepuce (preputial) ; between the scrotum and the 
penis (prcscrotal) ; on the tail, above or below, (caudal); in hyaenas, between the root 
of the tail and the anus (supra-anal); or, in the palm civet, anterior to the vulva and the 
penis (pregenital). Usually, but not invariably, the position follows a family pattern. 
The secretion is sometimes liquid, sometimes of a waxy consistency as in the case of the 
civet. It is interesting to note that this latter secretion, while nauseating to humans 
when in concentration, is in dilution pleasing; and "civet", extracted with a spoon 
from the glandular sacs of captive animals kept for the purpose, has for many centuries 
commonly been one of the ingredients of a number of commercially successful per- 
fumes in Africa, Asia and Europe. 

In connexion with the wide occurrence of scent glands it must be remembered that 
many carnivores, except at breeding time, lead a more or less solitary existence and 
some more reliable means than chance encounter is needed to enable the sexes to find 
one another at considerable distances. The emission of powerful and significant odours 
that may cling for long to undergrowth or be carried far on the wind, in conjunction 
with a higlily sensitive and critically analytical sense ot smell, is an ideal way of sur- 
mounting difficulties of time, distance, intervening obstructions or darkness that niay 
render sight useless in bringing the sexes together. Sound, though sometimes similarly 
used as a remote inter-sexual signal, is often not so effective in overcoming difficulties 
of time and space as well as having a less emotive impact. 

Mammae. The mammary glands are, so far as known, mostly purely abdominal 
and number between one and five pairs except in the cheetah, where they are far more 
numerous. 

Pelage. The pelage, by its length, texture, colouring and pattern, accounts for very 
marked differences of appearance not only between families but often between genera 
and species. There is, unfortunately, no clearly defined and commonly accepted 
terminology in English for the various types of hair found in mammalian coats. At 
one extreme there are bristles, or vibrissae as they are more technically termed, which 
are of circular section and considerably stouter than the greater part of the pelage. 
These occur most noticeably as the "whiskers", or mystacial vibrissae, on the upper 
lip; but also, in the carnivores, singly, in pairs or limited groups in other set positions, 
of which there are five on the head besides the mystacial area. These are submental, on 
the anterior part of the chin; intcrramal, on the posterior part of the chin; genal, on the 
cheeks; superciliary, above the eyes; and subocular, below the eyes. Apart trom the 
mystacial vibrissae not all these are always present; in the fissipedes the interramal 
vibrissae are lacking in the Felidae alone. Vibrissae occur, mostly singly, on other parts 
of the body, in some cases on the legs, but more commonly at wide intervals over the 
back and flanks. The function of these stout and sensitive hairs, wherever they are 
situated, is to give tactile warning or information. Those on the face and on the body, 
for example, furnish vital information in the dark when creeping into a narrowing hole 
or through a restricted rock fissure; the whiskers and possibly other facial vibrissae 
clearly tell the direction of the wind; and those of the otters may play a part in the 
detection of prey in ill-lit or muddy waters by revealing currents set up by the powerful 



12 THE CARNIVORES OE WEST AERKA 

swisli ot .1 fish's tail. Tlic mystacial vibrissac arc generally stouter and longer than the 
others, and sometimes very stitf — as thev are in tlic leopard. Indeed, in some parts of 
West Africa in the past the whiskers of this animal were looked upon as a certain means 
of causing death if chopped up and mixed with anyone's food, being reputed to 
penetrate, or at least cause fatal inflammation of, the stomach wall. 

At the other extreme ot liair is the underfur, so called because in many mammals it 
torms the lowest, and often rather insignificant, constituent of the pelage. But this is 
not always so; and in some species it is the major element — as, for example, the wool 
ot sheep. For this reason it is not a good term but it is in general use since there is 
nothing else with which to replace it. Undertur is always of fmc diameter, sometimes 
extremely fuie; in the carnivores it may be short or moderately long and is otten 
slightly sinuous. Occasionally it is almost completely lacking, as in the Banded and 
Gambian mongooses of the genus Miin(;os; but in the majority of cases it is dense; 
sometimes exceedingly so, as tor example in the otters where it forms, on submergence, 
a waterproof coat aniongst which air is trapped, thus keeping the body both dry and 
warm. It is also remarkably dense and, amongst West African carnivores, at its longest 
in the striped polecat {Icloiiyx) in which it in etiect torms the entire pelage apart from a 
few widely dispersed lengthy bristles. In this case it probably serves to shield the body 
from the ettccts ot direct tropical siuishine and hot dry winds. 

Above the undertur lies the longer outer fur, which may be so abiuidaiit as to conceal 
the underfur completely or, as in the case just mentioned, it may be merely a widely 
scattered and comparatively unimportant element. The tormcr is by far the commoner. 
This part of the pelage is, in a greater or lesser degree, harsher than the undertur, being 
composed ot stouter hairs though by no means so stout and strong as the vibrissae — to 
which the unqualified term bristles is best confined. More often than not there arc two 
distinct elements in this outer coat. The more important and abundant consists of 
stoutish hairs of round or sometimes slightly flattened section, tapering distally to a 
slender point but not decreasing much in diameter towards the base. When the fur is 
turned back they therefore usually stand out against the background ot undertur by 
size and sometimes by colour. In the carnivores they never exhibit a concavo-convex 
section commonly found in the Rodentia (Roscvear, 1969). These hairs are in this 
present work, for want ot any better recognised term, referred to as bristle-hairs. They 
are often broadly aniiulated with different colours, the tips being almost invariably 
black, the basal portion very often pale or even white. In many species there is a second 
element in the outer fur, termed herein sub-bristlc-hairs, shortened to sub-bristles. 
These, like the bristle-hairs, are longer than the undertur and play their part, too, in 
overlying and concealing it; but they are of different form, consisting of a long, slender, 
pale stalk (the petiole) which passes into a broader, terete or slightly flattened coloured 
blade ending in a darker, pointed tip. This distal part, except under magnification, 
resembles the bnstle-hair, and the fundamental difference ot form has therefore often 
been overlooked; the slender proximal halt though ot somewhat greater diameter than 
the underfur is nevertheless not so disparate as to stand out clearly from it when t!ie 
pelage is turned back as the bristle-hairs do. 

This gross morphology of the pelage can be seen w ith the naked e\e or iiiuler low 



GENERAl. INTItODUCTION 13 

magiiificatioii. No reference is made here to the surface patterns occurring on hairs 
since this is a matter for the microscope or cicctroscan and beyond the scope of this 
present account. It need only be said of tlicse that, though little or no research has yet 
been carried out into African carnivores, it is possible that they may follow at least 
generic patterns and might be of use for identification, as between other mammalian 
groups. It is the surface scales giving rise to these patterns that underlie the property ot 
close adherence together known as "felting" which takes place when a mass of hairs 
is suitably beaten. 

The coat is nearly always plentifully developed to afford a more or less complete 
covering to the skin; but it may range between very short, as in the lion, to very long, 
as in some hyaenas and the striped weasel. However, coat length and quality and 
sometimes the relative proportions of the constituent elements are affected very con- 
siderably by season, moult and condition of health; and owing to these factors skins 
in museum collections are often misleading and sometimes not susceptible to valid 
comparison. Wrong taxonomic deductions have sometimes been based upon them. 
Little is at present known of these factors in comiexion with the majority of West 
African carnivores. In some species, apart from the normal coat there may be a well- 
developed collar or "mane", or a nuchal or spinal crest ot exceptionally lengthy 
hairs. These growths are affected by age, sex and possibly season. 

Pelage patterns. Pelage texture follows a general family plan. In the typical dogs 
it is frequently long and rather harsh; in the cats mostly much shorter and somewhat 
softer; the otters have fairly short but extremely dense, rather silky fur. Pelage pattern 
in this Order varies very widely but often similarly follows some sort of family, or at 
least subfamily, plan, to which, however, there are sometimes notable exceptions. 
The larger dogs in West Africa have coats characterized by broad, irregular blotches 
but the foxes are more self-coloured; a speckled, "pepper and salt" appearance occurs 
in many mongooses, though, again, some are predominantly unicolorous. It is amongst 
the cats that the widest diversity is found. Spots are common — the bold markings ot 
the leopard, cheetah and scrval are familiar; and though the pelage of the adult lion 
is plain its spotted heritage is clearly displayed m the cubs. A more curious incidence 
of the loss of spots in the Felidac is shown in the golden cat, which within a single 
species exhibits in the adult both spotted and unspotted forms, a difference which 
appears to be related to distribution and is thus racial. The West African Viverridac, 
that is the civet, genets and their km, also have spotted coats. 

Though spots may in some individuals coalesce into lines, pure transverse striping 
as the sole pattern as in the tiger does not occur in West Africa; but through the regular 
disposition of the light and dark annulations of the hairs ot its dorsal fur one of the 
mongooses is cross-banded. Bold, broad, longitudinal bands are exhibited by the 
African polecat and weasel. The ratel is unusual amongst the African carnivores in its 
coloration. Most mammals have the belly white or at least paler than the back; the 
ratel exactly reverses this; and so to a lesser extent do the striped polecat and striped 
weasel. 

Pattern is produced in three quite different \\a\s. The hairs may be unicolorous 
throughout their lengths as in the polecat and weasel, where they are in separate bands. 



14 THt < ARNIVORF.S OI \^ EST APRU A 

wholK' black or wholly wiiitc. More ohen, bold spots or lines arc produced, as in the 
cheetah and other cats, by groups ot hairs that are indistinguishable from the rest of the 
tur in their pale proximal portions but difter in having black ends. Much more rarely, 
and in West Africa onlv in the single case ot the banded mons^oose referred to in the 
previous paragraph, the pattern is brought about by the regularity with which each of 
the dirterent coloured annulatic>ns on the hairs tall together along the back instead ot 
Knig hcterogencously mixed as in the ordinar^• "pepper and salt" coat. The backs ot 
the ears, notablv in the Felidae, otten contrast with the rest ot the coat. In this tamily 
thev may be wholly grey, whollv black, or have bold black markings with or without 
an additional white spot. 

Melanism. Melanism is not uncommon, but albinism appears to be very rare, lied 
as an outstanding feature ot the pelage crops up in a number ot dirterent families ot the 
carnivores. It occurs in the dogs, where in the foxes it sometimes constitutes the main 
colour. It is found also in the cats; in the genets, where it is in some cases at least a 
component of the spots; and in the mongooses. The striking appearance of a leopard 
skin when it is seen as a trophy m a room makes it seem that such an animal would 
be very conspicuous and easily picked out in the field; but the reverse is mostly the 
case; for w'hen one of these animals is at rest in the undergrowth or amongst the foliage 
of a tree the pattern merges completely into the dappling ot light and shade produced 
m the vegetation by sunlight. Only when the animal stands fully exposed in the open 
or is on the move can it be clearly seen. The alarmed surprise of suddenly discovering 
oneself in unexpectedly close proximity to a leopard in the bush must possibly first be 
experienced before the real truth of this can be fully appreciated. At birth and soon 
after, the coat is mostly short and rather silky, the adult markings either absent or very 
taint; but lions are born with a pattern that is lost at maturity; and in the cheetah the 
juvenile coat is very long and greyish-white, c]uite unlike that ot the adult. Sometimes, 
as in the mongooses, the entire pelage is erectile in anger or alarm; sometimes only 
the hair along the spine. 

Skull. A great part of the tissipedes have a rather narrow skull m which the total 
length IS markedly more than the zygomatic breadth, the bramcase otten not a great 
deal wider than the posterior end of the rostrum. This last is itself relatively narrow 
and often long and tapering. In the Felidae, however, the skull is clearly rounder, 
its breadth across the zygomata not a great deal less than its length, the braincasc 
broad and subglobular, the rostrum very short. Throughout the suborder there are 
nearly always pronounced, sometimes ilangc-hke, supraoccipital crests, and often, 
in mature skulls, especially males, a sharp sagittal crest, sometimes short, but often 
extending the whole length of the braincase. The nasals are always narrow, often 
relatively short, their upper front margins always situated well posterior of the forward 
limit of the premaxillae and incisors, thus leaving a very open, anterior nasal aperture 
lacking a bony roof The slender coiled bony laminae within the nasal cavities, known 
as the turbinals, which plav a major part in the sense of smell support a total surface 
area of receptive membrane incomparably greater than that available to man. 

The zygomata are alvyays strong, sometimes very strong, and this, in conjunction 
with the cranial crests, is tn give the nccessar\- anchorage to the extremely powerful 



GENERAL INTRODUCTION I5 

biting muscles. Thcjugal boiic plays an important role since it torms the greater part 
of the anterior half of the arch, the maxillary zygomatic process being relatively short. 
With few exceptions as in the otters, the infraorbital canal is of comparatively small 
size. The orbit is incompletely ringed with bone, there being a longer or shorter gap 
between the postorbital process from the frontal and, in the majority of genera, an 
upwardly curving process arising from the jugal. In the ratel and some otters one or 
both of these orbital processes may be poorly developed or virtually absent. 

The bullae may occasionally be small but are more often large or very large and 
betoken a highly developed sense of hearing. They furnish a point of considerable 
taxonomic interest, providing a distinction between the two superfamilies into which 
the suborder is separated. In the Canoidca the bulla is simple, consisting of a single 
cavity. In the Fcloidca, on the other hand, it is partially or almost completely divided 
into two chambers, of varying relative sizes, by a septum which can usually be clearly 
seen in a prepared skull through the auditory meatus. The division of the bulla into 
two parts can generally be detected externally as a shallow fiurrow curving around the 
body of the structure. It must be added, however, that this classic conception is ques- 
tioned by Hough (1948) by whom the canid auditory region is regarded as more like 
the fclid hi its essential structure than has been commonly supposed. This paper furnishes 
details of this region in the Canoidea in comparison with those of some other present 
and past Carnivora and draws conclusions regarding their significance in the phylogeny 
of the Order. 

The bony palate sometimes terminates about the level of the back of the molar 
row but very often extends far posterior to this. The anterior palatal foramina are of 
no great size, round or oval, and lie more or less between the canines; the posterior 
foramina, often two pairs, arc situated at very different points of the palate according 
to genus. 

A highly important feature of: the skull is the form ot the mandibular condyle, that 
IS the hinge between the lower and the upper jaws. In many mammals, including 
human beings, this is a fairly flexible affair permitting movement of the lower member 
in various directions, up and down, from side to side, or back and forth, enabling the 
occlusal surfaces of the upper and lower teeth to alter their relative positions and to 
slide across each other, and thus produce a grinding effect. This is possibly best seen 
in the ungulates in the action known as "chewing the cud". In the carnivores, on the 
other hand, the hinge consists of a long subcyltndrical condyle on the mandible firmly 
embedded in a complementary-shaped receptacle, the glenoid fossa, in the upper jaw, 
completely inhibiting any side to side or back and forth action, ensuring that the jaws 
close, like a comparable liinged door, in one firmly fixed position. Without this, 
the cutting action of the carnassial teeth, as described below, would be as ineffectual 
as a pair of scissors with play at their axis of coupling. 

Dentition. The dentition is always, even in the smallest carnivores, patently ot a 
powerful nature (fig. i). All the roots are closed, that is to say that once any tooth has 
reached its appointed size there is no further growth to replace wear as, for example, 
in the open-rooted gnawing-teeth of the rodents. The incisors arc invariably f . They 
arc relativelv small but the outer one, especially above, is often larger than the others. 



16 



THE (AUMVIMIIS (II U 1,M M 11 1 ( A 




"^cg-Q-O o 



^ 




^--crrcs-o _Q^^ 



Fii;. I. Typical carnivorous teeth, showing the position ot closure: a. right upper |a\v, 

palatal view above, lateral view below; b. right lower jaw, lateral view above, 

dorsal view below, rcitiurns zcnia, B.M. No. 25.5.12.22, V, -■ 2 

sometimes appreciably .so, even to taking on a subcaniiiitonii appearance. Tlic jaw is 
nearly always strikingly dominated by the exceptionally tall, strong, curved canines — 
the "dog teeth" in fact — one above and one below on each side; but the cheekteeth 
are also ot a remarkable and unique torm. "Cheekteeth" is a convenient term used 
tor all the teeth which lie posterior to the canitie when it is not desired to ditierentiate 
between premolars and molars. The symbols /, c, p and in are used to denote 
incisors, canines, premolars and molars respectively, a tigure being added above or 
below as index or sutiix to indicate whether the toodi is m the upper or lower jaw and 
its position in each category reckoning trom trout to back. Thus, p^ is the third pre- 
molar from the tront in the upper jaw; 1112 tlic second molar in the lower jaw. 

The total number of cheekteeth ni the Carnivora is very variable according to 
family or genus, ranging in West Africa between 14 and 26 tor the entire mouth. 
The premolars may be 3 or 4 above on each side ot the jaw, and 2, 3 or 4 below; 
the molars i or 2 above and i, 2 or 3 below. Complications exist in that some teeth 
are occasionally deciduous, being shed with advancing age, the jaw in such cases 
appearing not to correspond to the accepted dental formula. Moreover, in a few species 
certain teetli seem to be in tlic process of evolutionary loss, forming components of 
the dentition in some specimens but not in others, without being actuallv deciduous. 
'I'lie dental formula is thus variable. 



GENFRAr, INTRODUCTION 



17 




Fig. 2. Camassial tcctli, lateral and surface views: a. right upper; b. right lower. 
Fclis marfiarita ainiisis, B.M. No. 67.1429, o. - 3 



The most characteristic feature of typical carnivorous cheekteeth is the modification 
of some of them into a highly efficient, sharp, cutting ("sectorial") form, adapted to 
dealing with masses of flesh and bone. This chiefly concerns the last premolar of the 
upper jaw (p*) and the first molar of the lower jaw (nn). The crowns of these are 
either almost wholly compressed from side to side or have at least some of their lobes 
thus flattened, the occlusal surfaces being in this way reduced to keen, knife-like edges 
which on closure slide over the face of the opposing tooth in the manner of a pair of 
scissors. The upper teeth always close over the outside of the lower. They are maintained 
in the essentia! firm close contact necessarv' to effective shearing by the rigid condyles, 
as explained earlier. Because of their flesh-cutting function these teeth arc kno\\ n as 
the "carnassials". In this work, as in most others, the upper carnassial is alwavs reckoned 
to be the 4th premolar (p*), it being assumed when only 3 premolars actually occur 
in this toothrow that it is the ancestral ist premolar (p^) that has disappeared and that 
the anterior premolar as it today exists is therefore p-. The lower carnassial is always mi. 

The crowns of the cheekteeth lying immediately anterior to the carnassials are tor 
the most part als(5 laterally compressed and pointed and thus play their part too in a 
shearing action though in a relatively minor way. Those teeth, if any, posterior to the 
carnassials are of a less specialized kind suited rather to crushing than cutting, though 
they are often of much reduced size and then almost functionless. The detailed form 
of the carnassials varies somewhat from genus to genus ; but, speaking in broad terms, 
the upper one consists of a laterally compressed outer blade divided into cither two or 



IS Tin; (AUNivouis oi wcsr atuk a 

thicc j>oiiiti.'(.l Liisps, toi;(.tlicr uitli .1 Initial, miali lower, cusp whicii plays no role 
in the sectorial attion (tig. 2). Tlicrc are three roots; two below the blade and the third 
beneath the inner cusp. The lower carnassial is onl\- two-rooted but has, in general, 
a broader and more complex crown tliaii the upper one, comprising, basically, six 
cusps, though most are indistinct and several may be lacking. There is a good deal ot 
variety ot sliapc and disposition of these cusps; but most commonly it is the two anterior 
buccal ones wlijcli arc enlarged and laterally compressed to form tlie sectorial blade 
that exercises a shearing action against the upper. 

The virtual absence of grinding surfaces in the t\pical carnivore dentition means 
that cluinks of flesh are cut off and swallowed whole. But not all tissipede teeth conform 
to this predoniinautly sectorial shape, which is best exemplified in the cats and d<igs. 
In the otters, for example, there occur molars and premolars with broad crushing 
occlusal surfaces adequate to the disposal of the relatively soft flesh offish and amplubians 
or to crushing of the shells ot crustaceans (fig. zi). Teeth oi a less purely sectorial form 
arc found also in the mongooses, where they vary a good deal but in the smaller 
species are ot a highly cusped, insectivorous type (fig. 37). The evolution away from 
cutting to crushing in the Viverridae has been dealt with b\ G. Petter (1969). 

In some carnivores there is a wide interval between the canines and the anterior 
premolars — the post-canine gap. This permits the canines to be siuik into flesh to their 
full depth with a considerable band ot flesh held firmly in the jaw, thus bringing about 
an almost imshakeable grip into a struggling pre^■. It is best developed in the Fehdac; 
in other groups it is small or lacking. 

Way of life. Mammals arc tor the most part secretive creatures. They are ottcn 
nocturnal, or live concealed m dense bush or the toliage ot trees, or in holes in the 
ground, or in crevices amongst rocks, and are theretore not easy to see let alone observe 
with much continuity. Most ot them are, understandably, of a nervous, wary dis- 
position reacting, by immediate retreat, to warnings conveyed by slight sounds or 
smell as well as by sight. Many are dangerous to approach or disturb, especially when 
breeding or feeding. It is not surprising, theretore, tliac their lives are ottcn not so well 
known as those of some other, less difficult creatures. It is a tact that alter many centuries 
ot close acquaintance with and study ot mammals in Europe the habits of some ot them 
arc still not understood in detail. In Africa with its much more plentiful mammalian 
tauna and shorter history ot study this ignorance is much greater. Without question, 
there has been a ver\' large number of Atricans deeply conversant with the habits of 
mammals and other creatures. Their livelihood as professional hunters armed only 
with crude weapons and devoting a great part of every day or night to the pursuit of 
elusive quarry tostercd keen and accurate observation. But tliough a little ot this, 
especially as regards the more sought-after "big game" species, has trom time to time 
been passed on to European travellers and sportsmen, the majority ot it concerning 
the more insignificant creatures has found no place in the corpus ot recorded natural 
historical know^lcdgc and is now seemingly lost. 

Such biographical data as we have are to a great extent the outcome ot chance 
observations bv European travellers or part-time naturalists. The study of mammalian 
habits calls tor both patience and time. But though the former mav have existed in 



CENERAL INTRODUCTION It) 

plenty amongst keen held naturalists, the latter, owing to the daily need to pursue 
some more certain and lucrative occupation, has regrettably usually been lacking. 
Prolonged and continuous study has, to the majority, been impossible. Our knowledge 
of most animals is, in fact, made up of disjointed observations, often of extremely 
brief duration, of different individuals of a species in widely separated areas of the 
continent and which conceivably may not have been acting in a typical manner. 
Moreover, notes of behaviour have been made by observers ot diverse competence 
and not always correctly interpreted. Habit data are thus often far sketchier and less 
reliable than might be supposed b)- those who have not examined the subject closely. 

All this is now undergoing great change. Interest in wild animals has increased 
enormously. In consequence, part-time, intermittent amateur observation is slowly 
giving place to full-time continuous professional research carried out by the permanent, 
trained personnel of nature reserves, or specific studies by selected individuals or teams 
subsidized by scientific fomidations. Further tlian this, the apparatus at the disposal of 
observers today is vastly more varied and sophisticated. Li the past a naturalist had as 
his equipment at most a gun and a pair ot binoculars, and could thus study an animal 
dead or for a brief moment alive, at some remove. It was virtually impossible to follow 
for any distance a large mammal in rapid progress, except occasionally with the aid 
of a horse — which itself introduced complications. Nowadays the observer commands 
the services of a Land-Rover or similar conveyance which carries not only himself 
and possibly a co-observer but all his complex apparatus as well at high speed across 
rough terrain. It is a strange thing that many animals have clearly shown that, provided 
reasonable caution is exercised, they are not disturbed by or, indeed, much interested 
in motor vehicles, which can thus follow them in the chase or approach them fairly 
closely at rest. Such a truck, therefore, fulfdls not only the function of rapid transport 
but acts as a protection and a hide, enabling the observer to approach closer, to sit for 
long periods in relative comfort while taking notes or working his scientific apparatus 
without revealing his presence by movement, sound or smell. In Africa this, of course, 
applies mainly to the open woodlands; for while some of these advantages remain, 
that of free range is lost in the tangle and resistant imdcrgrowth of the West African 
forest and the tall, dense grass of the Guinea woodland in the r.iins. 

Apart from this invaluable mobility the modern held observer has at his disposal a 
wide range ot scientific aids which cither just did not exist till recent years or existed 
only in a relatively crude form. The improvements that have taken place in photo- 
graphic material alone may seem unbelievable to those who did not know West 
Africa half a century, or even less, ago. Basic to much of the improved facilities is 
also the simple modern fact of reliable portable electricity. More esoteric methods and 
apparatus apart, the main aids to animal study m the tield today are lightweight pre- 
cision cameras witli a wide selection ot easily changed lenses ot far greater etliciency 
than formerly, together with unrecognizably improved films; there are also reliable 
cinematography, infra-red photography, powerful flashlights and spotlights. In addition 
there are now sound recorders and videotapes, together with immobilizing drugs 
which enable even the most dangerous animals to be carefully examined while yet 
alive and to be distinctively marked for future recognition. 



20 I 111 (MIMVOIJIS (11 \\ I M Al UK A 

All this has brought about a change in the observation ot anunal behaviour and way 
oi' life that would be almost incredible to former generations of naturalists. But the 
new research has up to now been directed towards relatively tew species — as far as 
this present work is concerned, towards not much more than halt-a-dozen of the 
larger African carnivores. Sucli studies as have been carried out have revealed a good 
deal that is new besides showing tliat some ot the old beliefs are inaccurate. For the 
many species which have not been investigated reliance must still be placed on old 
field observations, though the picture of probable habits in nature has been in some 
cases augmented by behaviour studies made under laboratory conditions and breeding 
data derived from zoos. Nevertheless, it must be remembered that animals in their 
natural environment may not necessarily behave exactly as they do in captivity. Nor 
nmst it be overlooked that thcv are individuals, each with a recognizable character 
of its own which may difter in some degree from a common specific pattern. Such 
idiosyncrasies, however, cannot of course diverge so tar from the evolved norm as to 
jeopardize survival. 

Finally in this general survey of the state ot our present knowledge it must be pointed 
out that since what we know of the habits of most of the carnivores dealt with in this 
present work is still extremeh- sketchy and open to some doubt — as will become very 
apparent in the accounts given later ot the various species — it is desirable that even 
seeminglv insignificant items relating to the life ot animals they may come across 
should be recorded by collectors or field observers. It mav often appear to the naturalist 
that the scrap ot information at his disposal must sureK' iiave been recorded before; 
but every note he contributes, if not original, either adds valuable confirmation to 
what has been asserted by others or leads to its being brought into question. Examination 
of thousands of collectors' labels on museum specimens reveals how astonishingly 
little has been recorded in the field about sucli commonplace matters as vegetation, 
time of dav, season, food, stomach contents, lairs, litter size, and other similar bio- 
graphical data. There is still a vast, new and interesting field open to the keen mamma- 
logist in West Africa. 

The carnivores, as their name implies, are as a group predominaiuK- Hesh-eatiiig 
mammals. Yet though flesh may constitute the staple diet of the vast majority, a certain 
degree of vegetarianism takes place, hideed, extralimitally some, the pandas, live 
almost exclusively on this type of food. Those whose experience, in the north temperate 
zone, is limited to the poultry-stealing red fox and a virtual absence of vineyards very 
possibly regard the reference in the Song of Solomon to "the little foxes that spoil 
the vines" as nothing other than rather puzzling figurative poetic prose. But it is 
almost certainlv a perfectly accurate field observation; for while little is blown in 
detail of the diet of the small foxes of West Africa it will be seen later in this work 
that golden jackals and other normally flesh-eating carnivores do sometimes deliberately 
make a meal of berries. Species inhabiting arid country may find these at certain seasons 
,1 convenient and valuable source of essential fluid. 

Possibly all fissipedes on occasion take a certain amount of vegetable food. The 
well-known habit of domestic dogs of from time to time eating grass is shared by wild 
species also. This ma\' possiblv be to meet some vitamin deficiency; by reason of its 



GENERAL INTRODUCTION 21 

intermittent nature and the small quantities involved this habit is generally regarded 
as being not so much a matter of diet as concerned with some regulator}^ action on 
the bowels. Yet the larger wild fissipcdes do take in much more green food than appears 
at first sight because many of them start on their prey by ripping open the belly and 
consuming the entrails, the stomachs being well filled with grass. Fungi are also some- 
times eaten; and later in this work a genet is recorded with its stomach full of rotten 
wood. In West Africa a regular alternative to an otherwise prevailingly carnivorous 
regime is adopted by the ratel, which frequently makes a meal of honey and insect 
grubs, and is known to consume also eggs, bulbs and fruits as well as small mammals. 
It is often erroneously assumed that the flesh taken by carnivores is always that of other 
mammals; yet, besides a vast number of rodents, the smaller species depend also for 
their food in large measure upon birds, snakes, lizards, frogs, fish, crustaceans, insects 
and their larvae. And though the larger species rely mainly upon antelopes for their 
sustenance tliey are not above taking some of these humbler foods too when necessity 
dictates. Some carnivores habitually eat carrion; others take it only rarely or when 
hard-pressed for fresh meat — a situation sometimes brought about by their own 
inability, through injury or sickness, to hunt for themselves. Smaller species in captivity 
have been found readily to accept prepared starchy foods, and domestic dogs are often 
made to depend very largel)- upon them; but there seems to be no evidence that farina- 
ceous foods are ever taken by any carnivores in the wild. 

Since the greater part of the diet is, in general, living and capable of swift escape, 
both speed and cunning are commonly called for in its procurement, as well as, in the 
case of larger prey, considerable strength, determination and, often, endurance. The 
exercise of these attributes in the chase is one of the most characteristic features of the 
carnivores. However, there are widely different methods of hunting. A broad distinction 
between those of the dog tribe and those of the cats is that the former track, chase and 
pull down, whereas the latter stalk, patiently await opportunity, and poimce. There 
are of course exceptions. At times the excursion after food has the appearance of 
being nothing more than a rather haphazard affair, as when a ratel or a civet proceeds 
solitarily at a leisurely jog-trot along a route seemingly trusting that chance will 
present some suitable prey. In the sociable mongooses equally random foraging 
becomes, however, a rather fussily noisy proceeding, individual members of the party 
each gathering what morsel may fall his way, unless some larger and more difficult 
prey, such as a snake, calls for more concerted attack, bi other cases the hunt is without 
question a coldly planned assault on a selected victim, either on the part of a single 
animal or as a co-operative measure by a family or pack during which there is every 
semblance of messages being in some way conveyed from one member to another, 
serving to concert the strategy and bring the attack to a successful outcome. Such 
methods as these last are employed by lions and hunting-dogs. In contrast, little beyond 
patience is called for in the case of jackals or other scavengers which conunonly live 
on the left-overs of large predators, and need only exercise their acute sense of hearing 
to pick up news of a kill to become assured of the eventual effortless acquisition of a 
satisfi'ing meal. 

The search for and killing of prey may take place either by day or by night, though 



22 THL CARNIVORES OF WEST AnilCA 

in the Utter case tlie victim ni.ty have been picked out earlier wliile there was still 
some fair degree ot light. Hunting in relative darkness is aided, as mentioned above, 
by the peculiar structure of the eye by which the reflecting layer known as the tapctmii 
lucidiim doubles the effective strength of weak light though, probablv, scent and hearing 
at the same time play at least an equal part. Yet despite the usual donnnance of these 
two latter, sight is sometimes obviously the key sense, as u ith cheetahs initially distantlv 
observing their intended prey and subsequently pursuing it at liigh speed; or w^th 
hunting-dogs in full chase leaping at intervals over intervening grass cover m order 
to keep on the line ot their quarrv. 

Though food is of first importance iii anv animal's lite the search tor a convement 
meal is not the sole purpose ot a carnivore's daily or nighth excursion trom the resting 
place. Protection of hunting territory or nursery groimds from intruders and, at least 
at seasons, the discovery of breeding partners are both of considerable moment hi the 
activities of the majority ot adults. In these vital matters sight, sound and smell all 
plav their parts, probablv in the reverse ot this order in importance. As tar as repulsion 
ot trespassers is concerned sight may be emploved in two ways. It may furnish the tirst 
alert to danger at a distance; but, further, at closer quarters warning postures often 
become a major factor in defence. These involve not only an overall threatening attitude 
but in some species the presentation as well of intimidating, generally black, marks on 
the body conspicuously brought to the opponent's view by erection ot the tail or 
twist of the cars (Hingston, 1933). The role of sight in wooing is less; for though 
courtship displays are known to take place in a few species then- importance in carnivores 
in no way approaches their standing amongst birds. 

Sound enters into both personal and territorial detence in the tonn ot the caunonar\' 
growl or hiss, the more purposeful snarl and bark, or in social species communal 
yapping. Its part in breeding is limited to the initial attraction at a distance of the 
attention of the opposite sex by a recognised mating call. But little is known ot this 
together with the parts that touch and taste also sometimes play in the mating ntual. 
It IS smell, however, that in a great number ot carnivores overshadows all else in 
importance. Scent can, and does, convev clear messages ot species, numbers, sex, 
warning or of ripeness tor coupling that are fir more enduring and frequently more 
distantlv carried than those of sight or hearing. Its importance in this Order is deducible 
from the widespread existence and variety ot odoriferous glands; and is evinced 111 
practice bv the ceaseless inquisitive snitfuig ot most species on the move, except possibK' 
the cats, together with the repeated deposition ot their own odour. 

To judge trom the development ot the bullae, acutencss ot hearing is common 
throughiHit the suborder in West Africa. At least in the Canidae, the upper limits ot 
pitch received are well beyond that which is effective to human ears. Voice is extremely 
varied, embracing the roar ot the lion, the snarl ot the leopard, the purring of the cat, 
the ululation of the hyaena, the mournful hoo-jioo ot the jackal, the musical call of the 
hunting-dog and the chirruping ot social mongooses. These, and other characteristic 
notes, obvi(Hisly express contentment, satisfaction, warning, tear, detiance or excite- 
ment, though it is by no means alw.ivs easy to be sure which of these emotions is, 
in fact, the source of a particular utterance. Most carnivores can produce a variety ot 



GENERAL INTRODUCTION 23 

sounds according to mood; and there is not uncommonly a difference between the 
calls of the two sexes. In the case of social animals, the voice, or a change in its tone, 
may convey to young animals, or in the chase to other adults as well, an actual meaning, 
sometimes tantamount to an instruction. 

The degree of sociability amongst the carnivores is widely diverse. The marine 
pinnipedes, not dealt with in this book, exhibit grcgariousness to an extent not in any 
way approached by the land-based fissipedes ot West Africa or, for that matter, any- 
where else. A high proportion of the carnivores are to all intents and purposes solitarv' 
except at mating time or in the earl\- stages of parenthood. Lions and cheetahs may 
occur as a small family group or as a combination of groups up to a dozen or so indivi- 
duals, though this rarely occurs in the territory now under consideration. Hunting- 
dogs collect in packs, sometimes, elsewhere in the continent, of large size. The kusi- 
manse and banded mongooses also forage in companies. These latter animals are 
consistently social ; but grcgariousness may in other species be of a transitory — though 
possibly recurrent — nature, the outcome of passing circumstances rather than of inbuilt 
sociability, as when jackals, for the most part rather solitary in habit, gather round the 
common attraction of a lion's kill to feed. Of other West African carnivores, foxes, 
most of the cat tribe, manv mongooses, the genets, weasel, polecat and others are 
basicallv solitary though at post-breeding times loosely and temporarily biit into 
small family imits. 

Few carnivora actually burrow, as some other mammals do; but several of them take 
advantage, for breeding or shelter, of holes made b)- other animals such as aardvarks 
and porcupines. These and other holes they may enlarge and adapt to their ov^ai pur- 
poses; but probably the only real burro wers amongst the West African carnivores 
are to be tound amongst the dogs, that is to say the fennec, the foxes and possibly, 
to some slight degree, the jackals. There is also the ratcl, which is the most persistent 
and expert digger of all since it regularly unearths terrestrial bees' nests in search of 
food and has been known to make tumiels to effect entrance beneath stone walls 
protecting fowl houses. It breeds subtcrraneously in a hole that is almost always self- 
excavated. Otters make breeding chambers in river banks, usually with an uiider-water 
entrance. No nest, as such, is ever actually constructed bv the carnivores but these 
subterranean shelters, either selt-dug or adapted, arc smooth-walled, warm and com- 
fortable. Whether they are lined in any way is not recorded as far as most African 
species are concerned, but it is known that the fennec's is, and that grass, ferns, moss 
and leaves are used bv other species extralimitalK' as bedding. Certain arboreal species 
shelter and breed in holes in trees; and it is possible that in West Africa, as elsewhere, 
squirrel "'drevs" may be taken over and adapted, and perhaps some of the larger 
birds' nests, as tor example those of the Hammer-headed Stork {Scopus mnbretta). 

Skill in climbing is developed in widely different degrees in different sections of the 
Suborder. This is to some extent a question not only of inborn inclmation but of leg 
and claw shape as well. In the Canidae and Hvaenidae, apart from the ability to scramble 
over relatively low obstacles, there is practically no capacity for climbing in the true 
sense, that is to say the ascent of almost vertical surfaces to any considerable height. 
Leg and claw shape, together with the relative inflexibility of their articulations, are 



24 Tiir CAKNivours or wrsr airiua 

against it. The adult clicctah is not a great deal better oti; but the rest ot the Felidae, 
with their needle-sharp, abruptly curved claws and the supple movement of their 
limbs, are amongst the tmest exponents ot climbing m the mammalian world. In this 
they are closely followed by the genets, palm civet, polecat and, to a lesser degree, 
the ratel. The larger mongooses arc poor and unwilling climbers; some of the smaller 
oncs, with light bodies, quite expert. 

Although nrost ot the West African carnivores have at one time or another been 
kept in zoos the amount of published information regarding their breeding habits 
under these conditions is relatively limited. Extremely little is known of what actualK- 
takes place in the natural state though our knowledge has in recent years been welcomcly 
augmented in this respect as regards a few of the larger and commoner species by the 
patient observations of specialized researchers, as recorded later in the appropriate 
places. The pattern of courtship, mating, gestation, number ot voung, and their care 
and upbringuig vanes considerably, as is to be expected in a Suborder which ranges 
so widely in size from the lion to the striped weasel and lesser mongooses, and in mode 
of life from tree-dwelling genets to aquatic otters. However, large or small, tlie young 
al^^■avs come into the world in an incompletely developed state, born secluded in a 
den, with their eyes closed, their pelage rudimentary, their limbs too weak to 
enable them to walk for several days. This is in sharp contrast to the antelopes, which 
must face the harshness of nature from birth and can walk or even run after their 
mothers within an hour or two of parturition. 

The long, carefully supervised childhood is accompanied by interesting develop- 
mental ploys in connexion witli the upbringing ot a new generation. These are play 
and more deliberate training, both of which arc exhibited by carnivores to a greater 
degree than in any other group with the possible exception of the primates. This does 
not hold true of the whole Order; it possibly reaches its maximum development in 
the Felidae. Plav that appears to be sheer exuberance ot yoiuig spirits has none the 
less a more important design underlying it in that it brings into practice attitudes and 
movements which later in life become essential factors in seeking and killing prey. The 
playful pounce, the sidewavs strike of the kitten's paw are examples cxactK- parallelled 
in earnest by the adult in action after food. 

Play is most evident and constant among the \-oung, but in some ot it the parents 
take part, demonstrating movements that their otlspring may copy. I'lay is, indeed, 
sometimes though much less frequently, indulged in by adults without the instigation 
of the young; and some of it would seem to have no other purpose than pure fiui. 
Otters, even as adults, are renowned for their apparently purposeless acrobatics and 
gambolling both in and out ot water. Some ot the later training tor adult life has 
nothing of juvenile high spirits about it at all, being given directly and purposefully 
by the parents, more especially the female. Young cheetahs are made to sit and observe 
the mother in action after game before they themselves are allowed to hunt. In the 
Canidae the smaller packs of hunting dogs are essentially family groups under the 
direct tutelage of the mother, though eventually two or more ot such groups may 
become associated under tlie leadership of a dominant male. Co-operative luuiting, 
though it mav to some extent be instinctive, nevertheless calls tor the acquirement. 



GENERAL INTRODUCTION 25 

by observation, of clan methods and the understanding of the implications of different 
calls. Hunting in packs does not necessarily always imply planned co-operation, 
or certainly not at the high level of intelligence found amongst, say, the larger Canidae 
and Felidae; for, as already pointed out, some of the smaller mongooses forage in 
company while seemingly acting each for itself alone. 

Economics etc. The carnivores fill a key role in maintaining the balance of nature 
siiicc without them many other kinds of animals, especially other mammals, would 
multiply to harmful proportions. One practical aspect of this as far as man's activities 
are concerned has been demonstrated in recent years, when the ruthless destruction of 
leopards led in many places to a vast increase in the number of baboons and, as a 
consequence, a grave destruction of farm crops by these marauders. The carnivores 
do in fact, alive or dead, play a fairly constant part in man's economy, sometimes, 
as in other Orders, through domestication. Dead, the pelts of several kinds, especially 
those with dense or strikingly patterned pelage, are valuable in the fur trade. Of these, 
the leopard is in Africa the chief; and it is the widely increased demand to meet this 
market together with the general fear and hatred in which this animal is held, combined 
with more effective methods of killing it that have led to its excessive destruction — 
with incidental repercussions such as that mentioned earlier in this paragraph. Outside 
the limits of this book, or indeed of the African continent, other carnivores furnish 
pelts of the highest value, some of the foxes and the mink ranking high in this respect; 
ijut there arc other African skins which find a market as ornamental furs, such as the 
cheetah and to a lesser extent the scrval and the otters. At a lower level, in local trade, 
genet and mongoose pelts fill a constant demand for bags and matchet-scabbards ; 
and throughout a good deal of Africa the skin of the caracal finds a ready sale as an 
anti-rheumatic garment. 

The most important economic function of the carnivores in the living state has 
already been glanced at — that of keeping in reasonable check all manner ot animals, 
from rodents to antelopes, that would otherwise certainly assume, for man, pest 
proportions. But there are, too, certain minor economic roles. The undoubted com- 
mercial value of living carnivores as highly attractive zoo exhibits cannot be ignored 
in a review of the Order's economic position, though the spectacle of a lion impatiently 
pacing the narrow confines of a cage may not in itself be to everyone ethically justi- 
fiable — an objection which does not apply to the possibly even greater economic 
success of making such animals part of the attraction of a national park. The importance 
of the dog in many spheres, not only as the friend and sometimes protector of man but 
most notably as a vital aid in the discovery, seizure or retrieval of food must not be 
overlooked, especially in countries and communities where such use is an urgent 
matter of business not a "sport". Yet, as a whole, wild carnivores as living entities play 
httlc beneficial part in man's economy. The civet furnishes the perfume to which it 
gives its name; but apart from that it is difficult to think of any other direct service of 
value rendered by undomesticated species. 

The adverse side of the picture is different. Wild carnivores from the largest to the 
smallest impinge detrimentallv upon human activities, stealing and killing cattle, 
sheep and poultry, and sometimes even man himself, so that his hand is raised in 



26 Tlir. CAUNIVOFrS OF Wl:ST AFRICA 

relentless battle against all sueh predators. There is another, more covertK" harnihil 
role ot the carnivores, their function as reservoirs and vectors ot' disease. Everyone is 
aware of the key part plaved by domestic dogs in the transmission of rabies. Wild 
species are equally guilty though their impact is, of cour^^e. less direct. There is little 
doubt that the carnivores as a whole are capable of spreading, dircctK' or indirectlv, 
a number of other diseases and intestinal parasites; and the\' are well-known hosts of 
ticks and fleas which themselves are recognized as having high medical and vcterinarv 
significance. But so far in West Africa very httle investigation lias been carried out 
into these matters. 

Taxonomy. Simpson (194s) treats this at considerable length, and those to whom 
this aspect of the carnivores is important should consult his account. There is little 
purpose in a book of the present nature in entering into a wealth of not verv relevant 
historical detail. Fundamentally there has for long been broad agreement with the 
general classification, if not the naming, adopted herein — which is that ot Simpson 
(1945), itself basicalK- that of Winge (1S95). There is much about the animals mckided 
in the Carnivora, their structural morphology, dentition, external form and habits 
which make their apparent relationship both as a group and \\ ithin that group prettv 
clear-cut, and for this reason their taxonomic arrangement has in essence remained 
much as it was in the 10th edition of Linnaeus' Sysu-uui Natuiac. 1758. What has been 
more in question than the C(Mistitution of the Order as a whole or of its mam sub- 
divisions is the precise degree ot relationship to one amulicr of these subdivisionai 
groups within the Order. 

The separation of the Order Carnivora into two main categories, Pinmpcdia and 
Fissipeda, is of very long standing, the name for the latter dating from Blumenbach 
in 1791, and for the former from Illigcr twenty years later. The fissipedes, alone ot 
concern to this book, are likew ise divided hito two mam sections, herein given Simp- 
son's Superfamilv titles Canoidea and Feloidea. Precise morphological defhiitinn 
of the major subdivisions of the fissipedes in the modern sense dates from just a centur\- 
ago when Flower came to the conclusion that three prlmar^■ groups could be recognisetl 
from features of the skulls: the Acluroidea (broadly the cats and their relatives), the 
Cynoidea (the dogs), and the Arctoidea (the bears). The two last, previously separated 
partl\- as being on the one hand digitigradc and on the other plantigrade, were later 
regarded by Winge (1895) as sufficiently close to be included in a common supertamily; 
and this grouping, using a variety of names, has to all intents and purposes been recog- 
nized ever since. The name Aeluroidea for the cat group is still often come across and 
was employed in a major work as recently as I'ocock, 1951 ; but Simpson holds that 
it is hivalid and otherwise objectionable and has therefore substituted the term Feloidea. 
used herein. The degree of relationship ot the families within these superfamilics lias 
also been subject to differences of opinion; but these matters will be dealt with in the 
appropriate places later. 

With the cx'ception ot the H\aenidae, wliich in some measure occupy .ni niterniediate 
position, on skull characters the ditierences between the two supertamilies are clear 
and in most cases easily ob':erved. The question of drawing up a brief key from external 
features is altogether another matter. For wliile all the tannlies are individually recog- 



KEYS TO THE SUPERFAMILIES 27 

nizable easily enough by those who already have some experience of the animals with 
which wc are here concerned, the characters upon which distinctions might be based 
are either partly shared by the two superfamilies or differ in ways that are incapable 
of such defuiition as is succinct and at the same time not productive of doubt in the 
minds of those who are dependent upon an artificial key. The two keys to the major 
scientific groups which immediately follow, and which attempt to preserve cohesion 
within each subfamily, illustrate these points. A third key, which entirely ignores 
the classic major grouping, leads to the families irrespective of their larger relationship, 
therefore, in a more direct fashion. 

KEYS TO THE SUPERFAMILIES OF FISSIPEDA 

A. Cranial characters 

I . The bulla (except in the Hyaenidae) almost completely divided internally by a 
more or less erect septum which can generally be clearly seen (except in the 
otters and some mongooses) through the auditory meatus, which is in 
nearly every case extremely short, being devoid of any marked external 
lip or lips; the position ot the septum is often (but not always) indicated 
externally by a shallow furrow dividing the bulla into two sections. In the 
Hyaenidae the auditory meatus is long with a well-developed externa! 
anterior lip, and the floor of this meatus is continued inwards to form a 
partial horizontal septum, which thus camiot be detected through the 
orifice. The paroccipital process spreads over the posterior part of the 
bulla, being sometimes very closely applied . Feloidea (page 1 60) 

1. The bulla with no, or only a very small, internal septum; the auditory meatus 
long with a well-developed external lip or lips; no trace of any furrow 
across the bulla wall, except sometimes round the lips of the meatus. The 
paroccipital process rarely spreading much over the hinder part of the 
bulla ........ Canoidea {page 29) 

B. External characters 

I . The pelage displaying one of the following characters: 

(a) a regular pattern of spots, at least on the underparts, sometimes on the 
spine coalesced into a stripe or stripes; 

(b) transverse stripes ; 

(c) speckled or ticked ("pepper and salt"), mostly plentifully but sometimes 
obscurely and only on the head, neck and shoulders; 

(d) more or less self-coloured sandy-brown, and either the ears black and 
pencilled or the tail with a blackish terminal tuft; 

(c) the longish fur indctlnitcly patterned except for a black transverse stripe 
on the inside of the foreleg, sometimes two or more on the outside of the 
hindleg; the flexible tail always with a black tip and often one or more 
rings .,,,,... Feloidea {page 160) 



28 THE CARNIVORES OF WEST AFRICA 

2. The pelage with one ot the following characters or patterns: 

(a) sleek and sheeny, the throat white with or without verv irregular spots, 
or the tips ot the hairs "frosted" white; in all cases the siniilarlv-haired tail 
markedly tapers troni a broad base to a slender tip; 

(b) blotched in large, irregular, varicoloured patches; 

(c) with longitudinal stripes on the flanks or dorsum; 

(d) the back wholly, or more rcstrictedlv, grev, the underside dark; 

(e) the longish fur ot indefuiite pattern, the tail sMnmetricallv bu^hv, the 
muzzle long and dog-like. .... Canoidea Q'iJft' 29) 

KEYS TO THE FAMILIES OF FISSIPEDA 
A. Cranial characters 

1. Cheekteeth 14 or less ...... . Felidae [p^igc iji) 

Cheekteeth more than 14 . . .... . . .2 

2. Cheekteeth 26 ....... . Canidae [pa^c 30) 

Cheekteeth less than 26 . . . . . . . . . .3 

3. Cheekteeth 22 or 24 ...... Vivcrridae {page 161) 

Cheekteeth 20 or less .......... 4 

4. Cheekteeth 3-1=18 (sometimes in Crocuta apparently 16 through the loss ot 

the minute upper molar) ..... Hyaenidae (page ■^i) 

Cheekteeth 16. 20 or if 18 then ^ .... Miistelidae [page O-) 

B. External characters 

1. Head roiuided, muzzle short, backs ot ears (exxept F. lyhica) wholly or partly 

black; coat often spotted at least on the underside; claws [except Aciiionyx) 
very curved, ver)' sharp and wholv retractile. Cats. Felidae [page 373) 
Not like this 2 

2. Of fairly large or large size, standing higher at the shoulders (30 inches, more 

or less) than at the hindquarters; coat spotted or transversely striped; 
jaw very powerfully built; both front and hindfeet with only 4 digits 

Hyaenidae [page 341) 
Not like this 3 

3. Of small or fairly small size; the coat spotted and the tail ringed; or the coat 

speckled (hi dark torms often obscurely) and the tail generallv long-haired 
but not symmetrically bushy, and the legs short . Viverridae [page 161) 
Not like this 4 

4. Coat with pronounced longitudinal dorsal stripes; or wholly (occasionally 

restrictedK') light grey and the underside dark ; or close-lymg, glistening, 
dark sepia with a more or less white throat . Mustclidae [page 92) 

Coat with varicoloured blotches, or a single side-stripe, or of an indefuiite 
mi.\ture with no set pattern: tail symmetricallv very bushv. Dogs 

Canidae [page 30) 



CANOIDEA 29 

Superfamily CANOIDEA Simpson, 1931 

The first major division of the fissipedes to be considered comprises four extant 
fainiHes, the Canidac (dogs), the Ursidae (bears, entirely absent from Africa), the 
Procyonidae (raccoons, coatis and other similar New World and East Asian animals), 
and the Mustelidae (a large and heterogeneous group of stoats, weasels, polecats, skunks, 
badgers, ratels, otters and several others, all characterized by powerful and sometimes 
highly offensive scent glands). From this it can be gathered that the whole superfamiK- 
is a pretty mixed assemblage. Even as far as the only two families occurring in West 
Africa are concerned, the Canidae and the Mustelidae, it covers such diverse creatures 
as the dog-like, carrion-eating jackals, the bulky, short-legged, honey-hiuiting ratel, 
the small and slender, wamingly-coloured weasel, and the aquatic fish- and crab- 
eating otters. The group was once known as the Arctoidea. This was, in fact, a com- 
bination of two major groups proposed by Flower in 1 869, the bears and the dogs, 
Arctoidea and Cynoidea; but since there were objections to the use of these terms 
Simpson (1945) coined the name now most commonly employed. 

It is almost impossible to fmd characters that firmly and clearly separate the Canoidca 
and Feloidca. Those given in textbooks are mostly found in practice to be at least 
partially untrue; and, indeed, Diicker (1957) on other than morphological grounds 
has questioned some of the existing classification (see page 163). What follows, though 
for the most part of general application, takes into consideration only those species 
which are of concern to West Africa. There are certain obvious external differences 
between individual famihes or subfamilies; but these are often hard to define succinctly 
and impossible of application to the superfamily as a whole. The characters chiefly 
used lie in the skull and especially in and around the auditory region. In the Canoidea 
the tympanic bulla is to all intents and purposes unobstructed internally and the meatus 
is of some length, having often a lengthened lower lip. Li the Fcloidea, on the other 
hand, the bulla is almost completely divided into inner and outer chambers by a 
septum, and the meatus is little more than a ring — although the septum is differently 
placed and of possibly different origin in the Hyaenidae (Pocock, 1916c); and both 
in this family and in the Herpestinae the meatus is long. The paroccipital process, 
where present, is in the Feloidea closely applied to the posterior face of the tympanic 
bulla, over which it tends to spread and fuse; but in the Canoidea the process is often 
manifestly independent, though in some cases, as in Ictonyx, Poecilictis and Mellivora 
it simulates the feloid form. It is also said that the carotid canal is long in the Canoidea 
and short in the Feloidea; but tliis, again, is not invariably so. The dentition is widely 
variable throughout the two superfamilies. 



30 III! (AHM\illlls HI W ) s 1 AIKU.A 

Fainilv CANIDAE Cir.iv, iSii 
DCXIS, ICWES. ETC. 

Distribution and general. The C.inid.K.-, or i-logs and their cldse km, are to be tound 
111 a wild state over most ot tile world with the exception of New Zealand, Antarctica, 
and a number ot islands m various parts ot the globe trom the West Indies to Oceania. 
.As tar as Australia is eoiieeriied it is believed that the now native wild d<ig, the dingo, 
must have been introduced into this essentialK' marsupial island-contiiK'nt bv early 
settlers or visitors man\' centuries ago. 

The animals included m this taimlv torm a very important section ot the carnivora 
and, one \\a\- and another, have a great impact upon both the natural tauiia and 
domestic stocks. This, however, has considerably lessened m recent times, due in part, 
b\ reason ot human population expansion, to apprcciablv' lesser numbers ot wild 
c.mids than tormerK' ; and in part to more etfectivc methods ot control <">r extermination, 
brought into use to cc^mbat predation upon valuable tarm stock. 

Canidae are to be tound in verv diverse ecological settings, trom the arctic snows 
to hot and dry deserts. In West Africa wild species, with whicJr tliis account is alone 
concerned, do n<it occur in the closed torest or, iiulced, nearer the coast than the 
middle ot the Cluinea wo<idlands. On the other hand some ot tlte foxes range well into 
the Sahara. Wild dogs mav be to all intents and purposes solitary, except at mating 
and breeding times; or they may habitualK- occur in packs, sometimes ot large size, 
as in the case of the wolves (extrahmital) and the hunting-dogs ot Atrica. These packs 
otten have a familv-group toundation. 

Description. Though the members ot this tamiK vary very considerably in size 
the\ are all built on a tairly well-detined plan that makes tlicm readily recognizable 
as canines, hi West Atrica there is ,it one extreme the little tennec weighing about 
i-S to 2 kg and at the other the jackals which attain s to X kg. Despite such disparities 
ot size, general overall tanuly resemblance is seen in v.irious parts eit the body: a liead, 
tuniishcci with conspicuous erect e.irs and narrowing to a long tapering muzzle that 
ends m a naked black rhmariuni; a deep-chested body; a long and abundantiv bushy 
tail; and ratlier slender but muscular legs terminating in well-padded, digitigrade teet 
armed with straightisli claws, well-designed tor running. On the outer margin of the 
ear a bursa is alw,i\s present, the iip|ier part of its posterior flap not being continuous 
with the mam rim ot the pinna but arising behind it. 

The pelage is generalK pretty long and is alwa)s dense. Thotigh, according to 
genus or species, it exjiibits widely diverse colouring trom blackish-brown to bright 
rcil, It is r.ircK entircK selt-coloured but most often ticked or speckled with lighter or 
darker tips ti> the individual hairs, which ma\' .iKo in luher ways var\' amongst tliem- 
selves in eoloiu or .it least tone. Areas ot black or white mav torm a well-marked 
feature in i ert.nii species; but there is scarceK' ever anything that could be regarded 
as ,1 formal, fixed and regular pattern ot stripes or spots. Even in the so-called side- 
slriped laik.d of West Atrii.i, tliL- bl.ick b.nul on the tl.ink, which gives rise to the 
n.niie. is shiucw ii.it iiu iviisi.nit .nid i>lteii obsi iiii . 'ilie extreme bushiness ot the t.nl 



C'AN'IDAE 31 

(Plate i) gives It a striking appearance and and a characteristic shape that furnishes an 
immediate point ot recognition. 

Skull. The canid skull (tig. 4) is long and narrow, tire elongated rostrum reflecting 
the wcli-kiiown sharp face of all but certain highly specialized domestic tonus of the 
family. The braincase is rounded ; but, considering the reputed intelligence ot the 
Canidae, seems disproportionately small and to constitute a relatively minor part of the 
total skull volume. As tar as West Atrican species are concerned there is always at 
least some development ot sagittal and supraoccipital crests except in Fciiiicciis (fig. S) 
in v\hich the former is almost completely lacking, and the latter relatively insignificant, 
hi Lyciioii (fig. 10) the sagittal crest is tall and knite-like; in Caiiis (fig. 4) it is low; and 
in Viilpcs (fig. 8) it IS restricted to the extreme posterior part of the cranium. The 
frontal region is marked bv fairly well-developed, subtriangular postorbital processes, 
otten torming a slight flange over the orbit itself but lacking any finger-like extension; 
and since the jugal process is also short there is thus a wide gap in the circumorbital ring. 
The nasals are long and narrow and always reach back at least as tar as the front of the 
orbit. 

The up-curved zygomatic arch is of moderate strength, the jugal bone pla\ing a 
major role in its constitution. The palate broadens posteriori)-, the cheekteeth, from 
front to back, curving gradually out and then, more sharply, in again, the dental row 
from the canine to the posterior molar thus forming a flattened S. The mesopterygoid 
fossa is mostly broad and deep. The bullae in the West African forms are large or, 
in Fciiueais, extremely large. In view of the manifest importance of the teeth in this 
family and the relatively strong construction of the upper jaw, the mandible appears 
unduly slender and weak, the slightly up-curving rami being, except in Lycaoii, 
shallow; but it must be remembered that there is very little chewing carried out in 
this family, the teeth being mostly used for severing chunks ot flesh which are swallowed 
whole. The coronoid process rises steeply, high above the t\pical carnivore sub- 
cylindrical condyle. The angular process is small and sharply divided from the main 
body of the ramus. 

Dentirion. The dental formula in the Canidae is basically jYjt 4- throughout 
the family with a few (extralimital) exceptions, of which Otocyoii, the bat-eared fox, 
occurring from Ethiopia to Cape Province, is the onl\- African example, hi this genus 
the cheekteeth may be ^ or |. 

In the West African genera, with which this work is concerned, the incisors, though 
well-developed, are relatively small, their sharp cutting edges sometimes trilobed. 
The canines — a name, of general use throughout mammalian dental nomenclature, 
deriving from their prominence in this family — are always very tall, recurved, strong 
and tapering to a sharp point, ideally suited, to sinking deep into flesh and maintaining 
a secure anchorage upon a struggling prew Their build prevents their plaving any 
further role in mastication. The post-canine gap is at most of verv moderate size, 
and in Lycaoii non-existent. 

All the cheekteeth (fig. 6) have cingula, most promiiK-nt!\- developed in the posterior 
part of the toothrow. The four premolars ot the upper jaw nicrease progressively in 



THE CARNIVORES OF WEST AFRICA 



oe a 



size troin the tirst to tiic last, /)', tlic upper caniassial. beiiii; always the tallest check' 
tootii. All, when iiinvorii, are sharp and ot triangular profile, though there may b 
secondary, hir lower, cusp anterior or posterior, or both, to the main one. p^ has a 
single root, p'~ and /!■' two each longitudinally sited; p^, which is of somewhat more 
complex construction, has a third root situated transversely to the first, surmounted 
by a small cusp; and there is a larger secondary narrow cusp in line with and posterior 
to the main one, forming an important component of the sectorial blade. The lower 
premolars arc ot the same torin as the upper ones except that /m is simple precisely 
similar to y)2 and ;)3. 

The molars are far more complex besides being ot markedly diHerent torms in the 
upper and lower series. They also ditter somewhat in the two subfamilies. ;»i, by reason 
ot Its great breadth, is by tar the bulkiest tooth in either jaw. Its cingulum is well- 
developed and anteriorly torms a small subsidiary cusp. Apart from this, in the Caninae 
the crown comprises two outer cusps, two much lower inner ones, and an internal 
heel. Ill- is of similar construction but lesser size. The mandibular molars arc not broad. 
/!/i, the lower caniassial, consists of a large anterior narrow blade, divided into 2 cusps; 
posterior to and in line with this is a third much lower external cusp; and there are 
two similarly small internal cusps, one opposite this last, and one opposite the rear 
section ot the blade. 1112 is a much smaller tooth, with four cusps; and 1113, the smallest 
tooth in the mouth, little more than a peg, has two cusps, hi the second subfamily, 
the Simocyoninae (Lycdon), the molars, both above and below, though of the same 
general form are r.ither simpler in their cuspidation. 

Habits. Some of the wild dogs, the foxes, have always been recognized as habitual 
predators, seeking out and killing their own meals; others, the jackals, have for long 
been commonly regarded as, next to the hyaenas, the great scavengers, living almost 
entirely on the remains of the lion's or the cheetah's kill. Recent intensive observation 
has shown this latter notion to be less true than was thought and that the jackals do, 
in tact, hunt more on their own account than was supposed. This is dealt withmorefulh' 
later. However, though all the C.-uiidaeare preponderantly flesh caters they do, never- 
theless, consume an unexpectedly high proportion ot insects and fallen fruits. Most of 
them are to a very large extent nocturnal, or at least crepuscular, avoiding, except in 
necessity, direct exposure to the sun. Daylight shelter tor the purpose ot rest, or more 
especially for breeding and the earlv protection of the ^■oung, is most commonly found 
in holes in the groimd, "earths" as they are popularly termed. But sometimes, more 
particularly in the case of young as yet unmated, and hence solitary, adults, temporary 
concealment is sought in naturally occurring craiuiics amongst rocks, or even in 
dense grass opened sufficiently for the purpose by a rotatory movement of the animal 
before lying down. Earths mav be selt-e.xcavated but arc probably most often basically 
holes made bv other animals, hares, aardvarks, pangolins or termitc■^. improved and 
adapted. There arc often two or more exits. 

The Cauidae all appear to be monogamous, the pairs remaining together tor some 
tune. They share in feeding and bringing up the family, though in the early stages 
the male is often kept at some distance bv his mate. Litter size amongst wild dogs 
seems to range between 2 and as many as ly; and there may be one or two litters a 



CANIDAE 33 

year, die age of the mother possibly being a determining factor in both. No set breeding 
seasons have been cstabhshed for Africa. The average period of gestation is in the nature 
of 9 weeks but, though not well-investigated, probably varies a good deal amongst 
different species and may lie anywhere between 7 and 11 weeks. The young, variously 
known as "cubs" (foxes) or "pups" (jackals and hunting-dog) are breast fed for 6 to 
10 weeks, being gradually weaned to solid food, mostly regurgitated by the parents. 
An interesting ceremony in this connexion is described later imder the jackals. 

The gait in the Canidac varies a good deal according to circumstances but follows a 
common pattern throughout the family. A slow walk is rarely adopted except within 
a limited range of a few yards extent. Over longer distances the normal means of 
progression is either, at slow speeds, a four-legged run, sometimes varied by bouncing 
the hindquarters as a unit; or, at higher speeds, a canter; or, in full piu'suit, a gallop. 
In this last, the two hindfeet touch the ground in rapid succession, impcllmg the animal 
onwards and slightly upwards while the forelegs arc stretched forwards until the two 
pairs of limbs arc fully extended and the whole body is in flight out of any contact 
with the groiuid. The two forefeet then consecutively touch the ground, and give the 
animal a second onward thrust while the hindlegs are brought forward until they cross 
the now backwardly directed forelimbs, which leave the ground, the whole body 
becoming a second time suspended in flight, but this time with the legs tucked under it, 
not outstretched as previously. 

The Canidae have no great powers of climbing; but they can overcome low obstruc- 
tions in their paths both by leaping and by scrambling over those that offer adequate 
footholds. The hunting-dog when in full cry intermittently performs leaps to obtain 
a view of the prc\' which may be hidden by the tall grass. 

Voices in the African Canidae, so far as they have been recorded, are pretty varied, 
not only between species but according to circumstances as well. There is, without 
doubt, an extensive vocabulary of signal notes, for the attraction of a mate, the control 
of the yomig or the co-ordination of the pack, that has not as yet been investigated or 
recorded. Nothing truly resembling the familiar bark of the domestic dog seems to be 
uttered by West African wild species, most of tire sounds being characterised as harsh 
yaps, reiterated melancholy whoops or long-drawn-out notes. 

Several factors serve to hold the number of Canidae in check. When yoimg and 
relatively defenceless they are, unless actively protected by the parents, subject to the 
same attacks as other small mammals from pythons, eagles, other carnivores or driver 
ants, hi the adult stage they may be killed by hyaenas, angry lions and the like ; but the 
risks of destruction they run from these or similar enemies seem to be slight. Their 
numbers arc kept in control more by diseases, and by loss of efficiency from luider- 
mincd health or accident which prevent them from maintaining their place in the 
pack or against stronger, more active and thrusting members of their own kind. The 
availability of food has a considerable influence on the numerical size of the litter or 
the proportion of it that can be successfully reared. The relatively low density of antelope 
population in West Africa, for example, is directly responsible for the general rarity 
of the hunting-dog there and the small size of the packs in comparison with East 
Africa. 



? I 1 1 1 1 ( \ I! N I \ c ) iM s o 1 w I s r M in I A 

Tiixoiioni) . XltliiuiiJlli tin C .inid.ic might .ip|H'ar to lie .1 i.lc,u-tut, casiU icnii;- 
in/.ihlc gnnip. tlKir t.woiioim has, in tact, inorc c(iiiipK\itv tliaii at first sl-chis hkclv. 
I his Is a matter tor more speeiahzed works and it is pointless to Jo more tlian glance 
.It the position here. Simpson (i<)4_s) gives a long siimiiiar\ ot the Kinsiderations and 
opinions involved and tnrnishes reterences to the very extensive hteratnre on the snbject. 

The tossil record is unusually ricli and the intorination which it others leads to a 
diversit\ ot possibilities in tlie matter ot plivlogem- and ol consequent views. One ot 
the chiet questions at issue is the limits which should be placed upon the fanuh' and the 
closeness ot its assocuilion with, or degree ot separation troin. other group.s, and in 
particular the Ursidae, the bears. The latter, though with their heavy build and lum- 
bering plantigrade gait appareiuK' so dissimilar trom the dogs, are, at least in Simpson's 
opinion, \-er\' closcK' related and, in point ot tact, a tairK- late otisho(it trom them. 
Bur this IS ot no great concern to this present work. 

On ,1 narrower issue, the extent ot, or even the propriet\' ot am , subdivision within 
the ver\ uiiitorm tamil\ Canidae is a ttirther point 111 some dispute. In practical terms, 
three subtamihes ot living caiiids are in tact tairlv gcneralK' recognized, all occurring 
111 Atrica though oiil\- two in the region with which this work deals. The third, the 
C")toc\ oninae comprising soIeK Liiocyoii, the b.it-eared tox, is contined to southern 
Atrica and the eastern side ot the continent as tar north as Ethiopia. The question as 
tar as West Atrica is concernei.1, theretore, reduces to the validit\' or otherwise ot 
recognizing two subtamihes, and the distinction which may be drawn between them. 

No one could mistake Lydioii tor anything but a dog, tjiough it is true that it dithers 
shghtlv 111 minor matters ot general appearance trom the more typical members ot 
the tamiK', the |ack.ds, toxes. wolves and so torth. 1 lowever, on the score ot its slightK' 
simplitied molars and its possession ot onl\- 4 (hgits on the toretoot it is retained herein 
as representing the subtamilv Simocvoninae. though it is admittedly doubtful whether 
these and other minor distmcticsus should lo^icalK- be accorded .iin greater than 
generic significance. 

KEY TO THE SUBFAMILIES OY CANinAI' 
(Previous ke\' page 2iS) 

( 'o.it with varicoloured blotches; forefoot with oiil\ 4 iligits: .idult skull lengtli 
about 200 mm; /;/' with little or ]io sign <if lingual cusps; 111- only about one- 
third or less ot the size ot ///' , . , . Siitwcyonitidc (jhi[;( 75) 

Co, It somerimes gri/zled or speckled, ,uid often w itli splashes of black, but not a 
varicoloured patchwork; forefoot with 5 digits; adult skull length not more 
than about 1 ;!0 mm; 1//' with two small internal cusps and a heel; ;//'-' about 
li.ilt .IS big .IS i;i' or more ..... Cdiiiiiae (/ii;i,'c 34) 

Subfamily CANINAE C;ill, iSyj 
TYPICAL noes 

The C.uiin.ie .ire in distribution co-extensive with the t.iiniK, ot whuh the\- form 



CANIS , 35 

by tar the greater and more varied part. Eight genera arc now recognized covering 
the wolves, jackals, coyotes, foxes and various other less familiar animals as well as 
the domestic dogs, the precise ancestry of which is in some dispute and is doubtless of 
a diverse nature. The general characters of the subfamily are those detailed above for 
the family with the exception mainl\- of the points brought out in the foregoing key. 

KEY TO THE GENERA OF CANINAE 
(Previous key page 34) 

r. Size small, liead & body length 350 to 400 mm; ears extremely large tor the 

size of the body (about 90 mm) ; hindfoot about equal in length to the ear 

or not much longer; bullae very large, their length almost equal to the 

breadth across the back margins of;;;- — //)-. Fenneciis [pdgc 54) 

Size larger, usually at least 400 mm ; ears moderate and the hindfoot appreciably 

more than their length; bullae not so large. ..... 2 

2. Back and/or flanks with at least some fairly defuiite and obvious splashes ot 
black; of fairly large size, head & body measuring 590 mm or more, 
the tail less than half their length; maximum skull length over 130 mm; 
postorbital processes without any depression on their upper side 

Canis {page 35) 
Back and flanks without any defmite black markuigs, though sometimes 
greyish in tone; size moderate, head & body generalK' measuring 400 
mm or more, the tail well over half their length; skull under 120 mm; 
postorbital processes with a slight dorsal depression Vtilpes {page 62) 

Genus CANIS Linnaeus, 175S 

Cam's Linnaeus, iy$S, Sys/eiiia Naturae, lothcd. I: 3 S. Type species Cams /rt/Hi/iVins Liimaciis. the dumcstic 

dog. Canis was the Latin word tor a dog. 
rlios Oken, iSifi, LchrbKcli titr Natur<;cschichh- 3, 2: 1037. Type species Thos riil^aris Okcii (-- Caiiis 

(iiircMs Liruiaeiis). Thos was tlie Greek name for (probably) a jackal. Oken's work lias been ruled to 

be taxonomically unavailable. 
Lupus Okcii, 1X16, Lvlirhiicli dcr Nalur<<vsiliiclili' 3, 1: tojg. Type species Canis hipns Linnaeus. Lii/jm'; 

was the Latin for a wolf. Unavailable. 
]'nlpiianis IJlainville, 1837, Annis Sci. nal., Z.00I. 8, 2: 279. Type species Canis anrcm Linnaeus. This 

name was compounded trom the Latin nulpcs fox and canis dog. 
Dicha Gray, 1869, Catalofim: <>/ the Carnivorous, Padtydorniatous and Edeulalc Mammalia in the British Museum : 

180. Type species C<im/s anthus F. Cuvicr. The name is a Latinized form of rfiV/i, a North Atrican vern- 
acular name for a jackal. 
Sehae'fia Hilzheimer, 1906, Zool. Beob. 47: 364. Type species Canis adustus Sundevall. This was called 

after Dr. Ernst Schaff in recognition of his help in the investigation of jackals. 

There is little doubt but that this genus has had a greater impact upon man than 
any other group of mammals with the possible exception of the ungulates; tor apart 
trom wild species which, despite greatly reduced numbers, in most parts ot the world 
even today impinge considerably upon human activities, it has given rise to the millions 



36 111! (MINl\'nl!lS OI WI.SI AIKK A 

ot domestic dogs upon whose varied special abilities man is dependent in many spheres — • 
without taking into any account a more intangible emotive intimacy, widespread 
though this last may be. The limits ot the genus, taxonomicalK' speaking, have altered 
a good deal through the years, Caiiis at one time being held to cover almost any 
species ot dog-like appearance (exemplified in Mivart, 1890), and at others to be 
reasonably separable into a number ot independent genera, as partly indicated by the 
above synonymic list, which includes only names of concern to West Africa. It is not a 
function ot this present work to deal with domestic varieties, though it may be observed 
ni passmg that there are several different strains in West Africa whose appearances, 
attributes, origins and possible relationships merit some study and record. Ot these, 
the small, black Ogbomosho hunting dog is, or was, amongst the most highly trained 
and skilled. 

Wild species ot Civiis occur in Europe (wolf), Asia (wolf, golden jackal). North 
America (wolf, coyote), Africa (golden, side-striped and black-backed jackals), and 
Australia (duigo), this last probably a domestic term anciently introduced by man and 
subsequently gone feral. The foxes were at one time also considered appropriate to 
this genus but there is miw fairly general agreement that they are best separated as 
I 'iilpcs. Certainly the two genera, m the forms ot the European tox and the domestic 
dog, though not infrequently in popular belief held to hybridize, have never, in tact, 
been shown to do so, even with artificial insemination (Gray, 1954). This leaves, 
therefore, only three true wdld dogs of the genus Caiiis in Africa, of which two, C 
iViivHi and C. (i<hisii:s, the golden and side-striped jackals respectively, are known to 
occur within the limits licrein dealt with. It will be appreciated that, with a range 
extending over much of the globe from Australia to the Arctic, there must be a good 
deal of variety ot appearance and form in the genus. No useful purpose is served b\' 
attempting to provide a general diagnosis ot it, the main characters of which can be 
sutficieiuK- gathered truni the descriptions <it the two West African jackals which 
follow. 

These two ma\- be ditlereiitiated thus: 

KEY TO THE GENUS CANIS IN WEST AFRICA 
(previous key page 35) 

Backs of the ears gre\ish; Hank generally with a blackish l<ingitudinal stripe 
beneath a white one; skull profile shallow, almost flat from the nasals to 
the frontals; the zygomatic arch .shallowly curved in the vertical plane 

adustus [pai^c 49) 

Backs of the e.us reddish; dorsal pelage often with irregular black patches; skull 
profile fairK- arched and with a marked change of level about the middle of 
the nasals; the z\goinatie arch stronglv up-curved . aureus {pn(;c 36) 

CANIS AUREUS Linnaeus Golden Jackal 

(.iim5 rtiirciii Lmn.icus, I7S'S, Sytfiiia Nntiirae lotli cJ. 1 : 40. Bciin.a Mts., L.inst.in. S. Pcrsi.i ( /?(/(■ Thoina?, 
i<;ii). The specific ii.unc is Litiii for goUfeii. 



CANIS 



37 














o 



THr CAKNIVOIIFS Oi: WRST AIUICA 



(Cciiiii) antJius F. Ciivicr, 1S20. in CJcotiroy & Ciivicr, Histoiic NatnrcUc ilcs Mainniifeics, pt. 17, pi. 173 
and text. Senegal. Amhiis was, in .ancient Greece, the n.nne of" an Arcadian family of winch, according 
to the Roman author Phny (Book S) a member was Lx-licved to he periodically chosen by lot to be 
metamorphosed into a wolt tor a period of nine ye.irs. 

Catiis varii<^at\[s Cretzschm.ar, iS^rt, m Riippell, Alias zu dir Rfis<- im Nt'nilkhcn Afrik,!, Stiiii^fthicn-: 
31, pi. 10. Nubia and Upper Egypt. (Not Caiiis fomdmis varicgaiin Gmelin, 17.S8). The name varie<;atus 
is Latin meaning composed ot various colours; given by reason of the pelage. 

Caiiis ripariiis Hemprich .X Ehrenberg, 1S32, Synthcltie Pliysicae sen Icoiics cr Dcsiripliom-s AUiiiiiiuiliuiii . . . , 
deciis sccunda. Coast of Abyssinia, near Arkiko. 

Caiiis vulpcs tiorsalis Gray, 183S, Proc. zoo]. Soc. Lofid. for 1837: 132. Senegal. {Vido Ellerman & Morrison- 
Scott, 195 1 : 224). This name is the mediaeval Latin adjective meaning pertaining to the back, probably 
given in reference to a distinctive dark s.iddle. 

Thos sciic^alcnsis Hamilton Smith, i S39, Jardine's Nnliualist's Libniiy. 25 (of the scries), 9 (of Mannnals) : 
210, pi. 13. Senegal. 

Cmiis anthus soudanicus Thomas, 1903, Proc. zool. Soc. Lomi, I. 295. El Obeid, Kordofm, Sudan. 

Canis liocdcrleini Hilzheimer, 1906, Zool Anz. 30: 16. Upper Egypt. Named after Professor Dr. Doderlein. 

Canis lliooidos Hilzheimer, 1906, Zool. Bcob. 47: 364. Senaar. This name is compounded of the Greek 
ihos, wolf, and the termination -ocdidcs implying resemblance. 

Thos mirctis imbiciiius Cibiera. 1921, Bol. Soc. csp. Hist. nat. 21: 264. This name was to replace Catiis 
t'iiric\;atiis Cretzschmar, preoccupied. 



Distribution and general. The golden jackal (hg. 3) is sometimes called the 
common jackal, but this mav be a niisleading term, certainly m West Africa. It is also 
frecjuently referred to as the Asiatic jackal since it is spread across much of that con- 
thient and was more commonly and better known from there than from Africa. 
Its range is, indeed, wide. In Africa it extends from Kenya and north-eastern Congo 
northwards and westwards to Senegal, Morocco and the countries bordering the 
Mediterranean. Thence the range continues across south-eastern Europe, tlie Arabian 
peninsula, and a good deal ot southern Asia, including tlie whole of India and Ce\lon, 
to as far east as Burma and Thailand. 

From West Africa specimens exist in the British Museum from Takoukout (Damer- 
gou). Lake Chad, Tchsiderak, Manakaoki (Air) and (just extralimital, in Ahaggar) 
Tazerruk. The species has also been recorded from near Timbuctu, from Portuguese 
Guinea (Madina de Boc and Gabu), fr<im I^ahomey and from Cameroun; but it 
almost certainly occurs throughout the Sahel vegetation zone. This together with the 
Subdesert are its main habitat. It is replaced in the Sudan and 13oka by the side-striped 
jackal, C a Justus. 

Taxonomy. With this wide distribution it will be readily understood that a good 
deal of variation of colour and size occur and that, in consequence, a large number of 
local races has been described. Several attempts have been made to syiionymizc many 
of the proposed names; but the taxonomic position with regard to the golden jackal 
in Africa remains so confused and obscure that it is impossible to conic to any precise 
conclusion. The reasons arc twofold. Although jackals are, in suitable localities, com- 
mon enough 111 the field, the collected study material is for the most part meagre and 
of scattered provenance. Yet more fundamentally frustrating than this is the plain 
fact that the early type descriptions on which alleged races are based arc so inadequate 



CANIS 39 

as to be without any real meaning or ability to afford any proper criteria by which 
clearly to differentiate from others the animals to which they refer. 

A glance at the above synonymic list, compiled from names which have at one 
time or another been associated, directly or indirectly, with West Africa, shows that 
of the 10 forms listed 6 were described before 1840, rehance being almost entirely on 
colour; one is simply a replacement name; leaving merely 3 dating from relatively 
recent times, the early days of this century. One only of these, soudaiiicus, has ever in a 
direct manner been held to occur within the area covered by this present account; 
and it alone furnished in its diagnosis cranial and dental data as well as external measure- 
ments. 

Workers have thus for over a century been in the position of attempting to equate 
collectors' specimens with ill-dcfuied taxonomically classical forms, partly by external 
appearance and partly according to locality. It is possible, and indeed probable, that 
certain African forms oi aureus differ from Asiatic forms and from each other in colour 
and size sufficiently markedly and constantly as to merit trinominal distinction; but 
attempts to do this tor West Africa on the basis of existing classical names and in the 
face of the lamentable deficiency of material is little more than an exercise in plausible 
ingenuity. Dependence upon pelage pattern is vitiated by the almost certainty of a 
high degree or individual variation, especially of the chief character taken into account, 
the amount of black in the coat, both dorsally and on the tail. There can be very little 
doubt indeed that within any one population or family this cannot help but be a 
matter ot considerable variability and one upon which little sound argument can be 
based. This, to judge from British Museum specimens, would seem to apply to the 
amount ot rufous colouring in the pelage make-up as well. Colour and pattern are 
also affected by age and moult. This is well-evidenced by two skins from Rio de Oro 
(extralimital), B.M. Nos. 5.9.1.1 and 5.9.1.2, one a fairly old animal with scarcely a 
trace of black, the other a youngish one, with a relatively large amount. Had they 
been collected at ditferent times at slightly different places they might well have been 
given separate distinctive names. 

It seems, indeed, very possible that the existence or form of black markings, often 
deduced from single specimens, or even from illustrations demonstrably wrong in 
other respects, have been credited with a constancy of occurrence and a taxonomic 
significance which they do not possess. 

Skull character and size would appear to provide a soimdcr line of reasoning; but 
here, the material available is often so limited and of such scattered origin that, certainly 
as far as West Africa goes, it is scarcely possible to base any convincing argument upon 
the threadbare data it yields. C. lupastcr skulls from Egypt arc indisputably far 
larger than any others; but when it comes to the consideration of the remainder of 
alleged forms the position is not so clear-cut. That size is not always dependable, 
especially when only two or tlirec skulls are taken into account, is demonstrated by 
a pair of apparently similar age from the Plain of Tokar (extralimital), one being 15 to 
20 per cent larger all roiuid than the other. It is true that one is male, the other female, 
but no such marked sexual disparity is evidenced elsewhere as sex-linked. The size 
and proportions of the teeth appear to be more reliable than the skull itself, but in 



40 Tin; CAUN'IVOKF.S Ol WEST MRICA 

these, as in otlier charaeters, it must be repeated that tlie material available trdiii West 
Africa is so limited that it is scarcely possible to come to any precise conclusion; and 
in anv case it is difticult, it not impossible, to relate cranial and dental measurements to 
corporeally diagnosed classic forms. 

Before leaving the general aspect ot this matter one factor contributor\ to racial 
obscurity must be mentioned. There can be little doubt about the wide-ranging 
capabilities ot these wild diigs. Like their more domestic relatives they can cover 
considerable distances, untiringly and often, by human standards, at a tair speed. They 
may do this, without any marked impulse to return upon their tracks, in the chase, 
in response to sexual urge, seasonally as the result ot the movement ot game and 
other food supplies, or just by reason of a roving nature. There is, in the part of Africa 
under consideration, no vegetational or physical barrier to wide transcontinental 
movement; and dogs by their verv nature, the wiry strength ot their build, their 
stamina, their readiness and eager persistence in pursuit, and tlieir ability to be self- 
supporting, are better able to take full advantage of this frectlom tlian most. Thus, 
while species must always remain separate, the maintenance ot discrete races under 
such circumstances would seem to be altogether anotlier matter. 

We must turn now from these general considerations to a more particular investiga- 
tion of the reputed West African forms. The position can only be summarized; enough 
has been said to show that there is little profit in too searching an examination. The 
above synonymy shows that, basic species aureus apart, three forms have been speciticallv 
attributed to western Africa, antluis. donalis and saicgalcmis, all typically from Senegal. 
Considering these first, it may be said at once that donalis, cieseribcd by Gray as a fox, 
is unquestionably a jackal, so juvenile as to be indeterminable but m all probabilit)' 
aureus. The other two forms at once plunge us into doubt. Cuvier described aiiihiis 
as a species in its own right, dirtercntiating it from aureus not very exactly in words 
and onlv slightly more clearly in pictures — whicli, however, tend to belie the meagre 
verbal diagnoses. There are other ditticultics. Cuvier stressed the fact that the two 
jackals had been found in captivity to interbreed fertilely. Tjiis, it nothing else, seemed 
to make it improbable that two separate species were involved; and the modern view- 
is that the golden jackal exists as a single species, aureus, spread across Asia and the 
more northerly part of Africa as described earlier on. It this is so, and there is little 
reason to doubt the correctness of the view, aiitluis can be regarded as nothing more 
than a race of aureus; and such is, in tact, the status universally accorded it today. 
Yet a complication remains. Cuvier indicated that the two reputed forms existed 
svmpatrically; and as this could hardly be so with two races of one species — and 
especially since they had been demonstrated to interbreed — one is left wondering 
whether they do, in fact, occupy the same country or what his so-called aureus, with 
which he contrasted authus, might be. 

What, indeed, is anthnsl No one seems very clear from Cuvier himself onwards. 
The original type specimen had been a female kept in captivity in Paris. Ten years 
later, Cuvier who, as indicated in the previous paragraph, had already laid the founda- 
tions of future doubts, further confusLxl the issue by describing, and illustrating, as the 
male o? authus an animal which has since been pretty generally regarded as something 



CANIS 41 

quite different. Hilzheimcr (1908) in a long monograph attempting to sort out the by 
then even more obscure position remarked on the confusion which, consequent upon 
the erection of antlnis, the tendency to refer all North African jackals to tliis species 
had given rise to — a situation not least due to Cuvier himself Oldfield Thomas, 
certainly, was never very clear regarding West African golden jackals. C. antlnis 
was, by diagnosis, expressly related to Senegal ; but no specimen from this area existed 
in the British Museum, and there was thus nothing which could with confidence and 
any justitication be regarded as truly representing the form. Li 1903 Thomas especially 
remarked on the value of the two specimens, mentioned earlier, newly received from 
Rio de Oro, "as being more nearly typical of the Senegal jackal described by F. Cuvier 
than the North African examples which have usually had to do duty as such". Rio de 
Oro is well north of Senegal, outside the Tropics and of a different vegetation. It may 
therefore be wondered how much more typical of the Senegal jackal they really were. 
This is the more so since, as mentioned in an earlier paragraph, they differ from one 
another totally in appearance and very much in age and size. The yoiuig one with its 
blackish back has more likeness to Cuvier's illustration of anthus than the older one, 
which bears no resemblance to this whatsoever; but it may here be remarked that no 
specimen of any kind in the British Museum shows any sign of the rufous colouring 
on the fore part of the belly which constitutes one of the clearest pelage distinctions, 
in Cuvier's illustrations, o( anthus from aureus. 

Thomas (1921) made only one other reference m literature to anthus when he attri- 
buted to this form a very much more heavily blackened specimen, B.M. No. 21. 2. 11.28, 
collected by Angus Buclianan at Takoukout (Damergou). This is, in fact, for all 
practical purposes indistinguishable from Buchanan's five later specimens from Air, 
which Thomas (1925) then named riparius, as recorded below. From all this it will be 
appreciated that there is very considerable uncertainty regarding the identity of anthus. 

This brings us to the consideration of the third reputed West African form, scnc- 
galcusis. Hamilton Smith devised this name for the second, male, jackal which had been 
described and figured bv Cuvier as anthus, as mentioned above; because, as Smith 
wrote of Cuvier's two illustrations, "an artist seeing both would hardly admit more 
than the approximation of the two species". Smith's comment applies with incom- 
parably greater force to his own picture of scucgaknsis, for this bears not the slightest 
resemblance to Cuvier's plate. His description is also largely at variance with Cuvier's; 
and, indeed, one is left in some doubt as to whether it really was the male rather than 
the female that he had in mind. 

Part of the argument put forward in support of the various alleged forms lies in 
coat colouring, part in different size or build. Verbal description of all these three 
categories, in the t)'pe diagnoses now under review, is regrettably inadequate. In the 
matter of size, no account is taken of the possible influence of age; and the few measure- 
ments turnished are sometimes given in different forms or are in other cases not com- 
parable. Cuvier, for example, gives his male as standing 17 inches [pcuccs) at the shoulder 
and 16 uiches at the hindquarters, but his female as 15 inches at the mid-back. Hamilton 
Smith gives no actual measurements other than what is presumably an estimate from 
these figures, that one jackal is at least an inch higher at the shoulder than the other. 



THE CARNIVORES OF WEST AI1U<:A 



A good deal ot the argument must necessarily depend upon the illustrations. It is 
extremely improbable that any dnect comparison ot the appearances of living animals 
can have been made, as Cuvier's female was described and illustrated in 1820, his 
male in 1830, and 1 laniilton Smiths sciicgdknsis in 1839. This brings into prominence 
the considerable importance attaching to the circumstances of the making of the 
three illustrations involved: that is to say, the competence in draughtmanship and 
colour matching of the artists concerned; and whether the paintings were in fact 
executed from the living animals or from pelts and measurements made from them ; 
or were reconstructions ot build made in response to verbal prompting, hi addition 
to this, the possibility of inaccuracy of reproduction of colour in printed books during 
the early days of the 19th century cannot be overlooked. How poor, in fact, Hamilton 
Smith's artist was can be gathered from a glance at his version of Lycdo/! pictiis. The 
whole matter of description, both verbal and pictorial, of the animals with which we 
are here concerned is fraught with doubt. 

Over the years many systcmatists have spent a good deal ot effort and ink in attempts 
to establish exactly what Cuvier's Caiiis aiitluis, male and female, and Hamilton Smith's 
scuc(ialeiisls might be. In view of the facts outlined above, the present writer regards 
such attempts as little more than a futile waste ot time; and in this work it is proposed 
to draw 110 nomenclatural distinction beyond the specific name aiirais. This is not to 
hold that West African golden jackals may not prove to be validly distinguishable 
from those of Asia or elsewhere in Africa. It is simply an expression ot the opinion 
that to trv to et]uate any existing inadequate material to these insufficiently and con- 
fusingly diagnosed forms, merely because they are classical and repeatedly and lui- 
questiouingly appear in literature as occurring in this area, is both misleading and 
unrelated to the reality of the situation. Nothing short of a wide-reacliing review of 
the species based on first-hand comparison of very considerably broader study material 
than today exists can lay sotuid foiuidations of the subspeciation of aureus and justify 
categorical assertions regarding it. Anything else is frankly unhelpful. 

Because of the transcontinental nature of the major ecological zones, certain forms 
named chiefly for Egypt and Sudan have been said to occur also in the west — as, 
in view of the uninterrupted nature of these biotopes and the absence of impassable 
physical barriers, they well might. A brief examination of the remaining names in the 
synonymic list must therefore be made. Cretzschmar's varkgaUn, though frequently 
mentioned in literature, is excluded by its prior use by Gmelin elsewhere in the genus. 
Cabrera therefore intended that it should be replaced by iiiihiamis; but Schwarz (1936) 
thought that there was much to be said tor identifying Thomas's so\idauicus with 
varicgiUus; and if this is so — and there is good reason to support the view — vaiiegatus 
must be replaced by soudankus rather than by iiuhiauus which it antedates. 

Sctzer (iQSfi), following G. M. Allen (1939), further synonymizes doedcrlciiii and 
thooidcs with soudaiiiais; and though this may possibly not be justified, the two animals 
thus named by Hilzheimer appear to have little connexion with West Africa and are, 
in any case, antedated by soudaiiicus. There remains only yiparius. Under the influence 
of Hilzheimer (1906), Thomas (19^5) came to regard his soudaiiiais as identical with 
riparius, described much earlier by Hemprich & Ehrenberg from the "coast of 



CANIS 43 

Abyssinia". Setzer (1956) by implication rejected this. The position is not easy to 
evaluate. When Thomas accepted Hilzheimer's view he had no skin or skull from the 
type locality of riparius on the Red Sea coast with which to make any kind of direct 
comparison of his soudanicus. On the face of it, the distribution of a single race between 
such an area and West Africa, with the Abyssinian mountains in between, seems 
unlikely; but the vegetation map (Keay et al., 1939) shows that there are continuous 
vegetation zones passing around these from the cast to the west coasts. The possibility 
of such a range is thus not so improbable as at first appears; and some of the available 
Ethiopian skins in fact agree well with the Air series. Of all the north-eastern specimens 
in the British Museum those which would seem to correspond in provenance most 
nearly to riparius are two from the Plain of Tokar, near the Red Sea. These correspond 
in appearance to West African skins, and also fairly closely in cranial and dental 
measurements, as the table on page 5 5 shows. 

The table referred to at the end of the previous paragraph is chiefly concerned with 
mean cranial and dental measurements; for though some external data are provided 
they are not wholly reliable. It demonstrates a succession of specimens in the 
British Museum across the Sahel vegetation zone from Takoukout in Damergou to 
the Red Sea. Unfortunately nothing is available from further west, that is to say 
from Senegal ; but the area generally implied by this geographical term, and especially 
in the early days of last century when anthus and senegalensis were described, lies in 
this same vegetation belt. To these Sahel specimens are added for comparison three 
skulls from the Saharan highlands of Air and Ahaggar, and also one from Rio de 
Oro, all these localities Iving nominally m Subdesert, the rather more arid contiguous 
zone. 

It will be seen that though there is some variation of size between the largest and 
the smallest there is a broad general agreement. It must be remembered that the means 
are derived from very limited numbers which fail to take into account the range of 
size normally occurring in the different populations. Only in one area are as many as 
five specimens involved. The size variation becomes very apparent when the skulls 
are laid out in order; but there is no regularity in this from which clinal or racial 
trends might be deduced. Given single specimens only, it would be difficult for any 
taxonomist to equate the relatively small Takoukout skull, a female, with its next, 
larger, neighbour from Lake Chad, a male; but two from Tokar, of equal age yet 
vastly different size, indicate that it would be unwise to lay too great a stress on apparent 
size difference argued from scant material. The smaller of these two, a female, matches 
the Takoukout animal; the larger, male, that from Chad. All but one of the skins 
embraced by the table bear a pretty close resemblance to each other; and it is probably 
justifiable to hold that, with one possible exception, all the animals in this transcon- 
tinental series could be assigned to a single race. 

The possible exception is that from Rio de Oro. In this animal the upper camassial, 
p*, is appreciably larger than in the others and occupies nearly 26 per cent of the 
toothrow from c-iu~ as against about 23 per cent in the rest. This fact combined 
with the different appearance of the pelage, as recorded earlier, as well as its extra- 
tropical provenance, could indicate that ^\■c are in this dealing with a different. North 



44 THH CARNIVORES OF WEST AFRICA 

African, race. But, once again, it must not be overlooked that argument here is based 
on a single specimen alone; and also that the second, very much \'ounger, ot the Rio 
de Oro skins has the dorsal pelage sprinkled with black in the manner of the Sahel 
zone animals. 

To sum up: trom what has been said it seems likely that, in so far as present purely 
West African material is concerned, we arc dealing with a single form; and such 
departures from a mean as occur are in all likelihood individual rather than racial, 
and tall within the limits ot normal idiosyncratic variation. This applies to reputed 
ditlerences ot muzzle shape and length ot leg as well as to marking and size. Further 
th.iii this, without very greatly increased data, both morphological and biological, 
it docs not seem justified to go. 

Finally, in connexion with the question ot races m general it is ot interest to note 
that Dobroruka (1959 and 1963) has recorded a change ot reputed race between young 
jackals born m captivity and their parents. 

General description. A fully grown golden jackal weighs some 6-5 kg. All the 
West African skins available tor study in the British Museum are sufficiently similar 
to merit a single description. The dorsal pelage is dense, long and rather harsh. It 
consists of a mixture of very long, wiry, terete bristle-hairs and shorter, finer, butl- 
coloured undcrfur, the general appearance being a mixture of buff and black. The 
bristles, about 65 mm long, arc four-zoned alternate black and white, the base being 
white, the terminal portion black. Tliis latter is of variable length. Wliere the black 
occupies only the tip the general buff colour is scarcely affected; where it is long it 
forms, in combination with contiguous hairs, a conspicuous element of the coat, 
giving rise to larger or smaller splashes of colour. Owing to the more or less regular 
length of the bristles these marks sometimes tend, in a greater or lesser degree, towards 
forming slight transverse patterns. But in general the black markings are scattered 
irregularly and vary in amount and pattern individually, with no taxonomic signifi- 
cance. Sometimes instead of the intensely pigmented terminal zone being sharply 
boiuided it merges through a longer or shorter weakly pigmented zone into the white, 
the area of mild pigmentation being reddish, the luidcrtur being also palely reddish. 
This gives rise to rufous areas in the pelage, often on the nape and mid-back, generally 
present in a minor degree, though sometimes very pronounced. There is insufficient 
study material to draw any firm taxonomic conclusions from tliis; but it is very 
possible that this, too, is a matter of simple individual variation. 

The black speckhng is carried on to the very long-haired bushy tail; but the hairs 
are only black-tipped, not banded. These terminal zones aggregate into a pronounced 
black mark, usually some 50 to 75 mm beyond the root of the tail; and at the end of 
the tail they become lengthy, jointly forming a conspicuous black tip. 

The nose is rufous; the crown speckled bufiy-grey; the cars golden-brown on their 
backs, with long white hairs on their iimcr face and a marginal bursa. The upper lips 
are white. The flanks are progressively less speckled than the back, the w hole underside 
from chill to anus is generally buffy-white, though there is sometimes a little obscure 
speckling and colour on the throat. There is no sign in British Museum specimens of 
the rufous zone on the fore-part of the belly indicated in Cuvier's illustration ot atiilins. 



CANIS 



45 




Fig. 4. ^'fl"/.> i7i/fc/i.<: >kiill, li.M. No. 21,2.11.28 : . 



i6 rill, CARNIVORES OI WEST AFRICA 



The legs, tdic and liiiid, are palely red on their outer aspect; and the forelegs carry a 
longitudinal black streak to the wrist, sometimes clear, sometimes very much reduced. 

Skull (tig. 4). In all West African skulls the profde show s a marked descent from the 
trontals to the nasals as opposed to the relatively flat outline at adiistiis. The zygomatic 
arch upcurves strongly, forming in some cases an almost semicircular outline. The 
sagittal crest is low over most of the braincase but becomes sharp and keel-like poster- 
iorly and joins equally pronounced lambdoidal crests to form an acute pyramidal 
pronnnence. The rostrum is shorter, less tapering and slender tlian in adtistus; the 
nasals are shorter. The mandible is more curved and rather more powerfully built 
than m aJiistiii, the depth of the ramus being greater. The coronoid process, too, 
is broader, generally with an incurved hind-margin and a slight backward hook at 
the top, the front margin descending in a broad convex curve from this hook to the 
base. In iuhisiiis there is no hook, and both margins are almost perfectly straight. 

The chief feature of the dentition is the relatively greater length of the carnassials. 
It will be seen from the table of measurements, page 55, that both /)■* and iiii are 
longer than they are ni luhisliis; and iu addition to this the length from front to back of 
p* is at least 82 per cent of the length oi ui^ + iifi. and that 0(1111 is well over 130 per 
cent of /;i2 4- 1113. 

Habits. Close studies of these animals in the field have been made in recent years 
by Wyinan (1967) and Goodall (1970), and much of the following accoimt is derived 
from these sources. Golden jackals arc to be seen on the move b)' both day and night, 
especially it the latter is brightly moonlit. Like most dogs they can sleep or be active 
at a moment's notice as occasion demands; and though t\\c\ do most of their feeding 
by day it is sometimes necessary to follow up kills made at night by lions or hyaenas 
even though they may probably, apart from a few hastily snatched mouthfuls, have 
to wait at a safe distance till well after siuirise before feeding can commence in earnest. 

Often golden jackals are solitary animals; but at mating times and during the raising 
of a family they are always in pairs; and sometimes these or small family units hang 
together for extended periods. A. J. Hopson noted that they were frequentl\- to be 
seen up to tour in a party in the Sahcl woodland and Sulradora pcrsiai (Salt Bush) 
thickets near Lake Chad (private communication). Their popular reputation is solely 
that of scavengers, cleaning up, in company of the vultures, when other larger carnivores 
have killed and eaten their fill. That thc\- do this is true; but it is only part of the story. 
It is possibly their easiest way of procuring a full meal since it calls tor little more than 
patience or the skill to see an opening and the agility to dash m, seize a few mouthfuls 
and nip quickly awav before the heavier and slower-moving feeders can prevent 
them. 

But, in fact, their dietary is much wider than this, and they do a good deal of killing 
n their own accoiuit. They mostly confine themselves to small prey, poiuicing upon 
hares, rats, ground squirrels, cutting-grass (Tliryoiioinys) and the like; they arc known 
to take lizards and not infrequently to kill and eat snakes. Ground-haimting birds 
such as francohns and bustards tall prey to them. They also consume a surprisingly 
large amount of insects: dung beetles, larvae, termites, or grasshoppers, the last ot 
which thev may either pounce <->n or catch in flight. The\' also eat a good deal of 



<i 



CANI5 47 

fruit at the right season, fallen figs and the berries oi Zisyplnis jiijiiba (Hausa iiingarlya) 
being said to be particularly welcome. Another sort of vegetable food they have been 
observed to eat is certain kinds of fungi. Like other medium- to large-size carnivores, 
golden jackals steal the more or less helpless yoimg of antelopes unless the mother is 
by to fend off the attack ; but even if she docs make a determined defence it is unlikely 
to succeed if more than one jackal is concerned, since while she is dealing with one 
aggressor another leaps in and carries off her tawn. Adult gazelles, duikers, warthogs 
or anything else that would make a suitable meal and is incapacitated by wounding 
or sickness also fall prey to these predators. Further, it is not generally realized that 
golden jackals quite frequently themselves hunt and kill fully grown antelopes of the 
smaller kinds, duikers or gazelles; but this they do, for the most part, only if three or 
more jackals arc working in concert. However, Goodall (1970) once saw a lone jackal 
chase a female gazelle for some 3 km, unsuccessfully, both animals being by then too 
exhausted to do more than trot. They are often held responsible for the theft of fowls, 
lambs, kids or even calves from domestic stock; and they are, hence, themselves 
ruthlesslv hunted. Besides the varied articles of diet enumerated above golden jackals 
will mark down and follow the females of antelopes that have just given birth in 
order to be able to pick up the after-birth. They also readily cat fresh droppings from 
rhinos. Droppings of all kinds are, of course, a source of dung beetles. 

When golden jackals himt antelopes for themselves and succeed in making a capture 
they, like others of the dog tribe, make no deliberate attempt to kill their prey but 
rip open the belly and start eating the entrails. Smaller creatures, such as rats or snakes, 
they kill bv shaking; but Goodall once saw a jackal eat a snake alive beginning from 
the tail end. These animals often carry away more food than they can consume and 
this surplus is then buried, usually in several different places. It is generally recovered 
from these within 34 hours; but the caches are sometimes robbed by hyaenas. It would 
appear that golden jackals can survive for long periods without drinking. 

When at rest golden jackals curl up, like a domestic dog, and go to sleep. This 
they may do on the surface; but the lives of these smallish predators, lurlike the far 
more nomadic hunting dog, arc, at other times as well as at breeding periods, very much 
centred roiuid a home burrow. These shelters — "earths" — may be dug out by the jackals 
themselves but are probably more usually enlargements of already existing, promising- 
looking holes made by other, smaller creatures; or adaptations ot larger burrows 
previously the home of an aardvark or warthog. Not much has been done to explore 
and describe these earths; but there are reputedly one central chamber and two or 
three independent escape routes to the surface. These homes may be in a secluded spot 
or, sometimes, not far from those of other den-dwelliiig carnivores, hyaenas or hunting 
dogs. Golden jackals have rather loosely defmed hunting ranges the defence of which 
is not very scriouslv undertaken and, according to Goodall, appears somewhat arbitrary, 
some intruders being driven back, others peacefully accepted. The area concerned also 
varies considerably in extent according to the circumstances of the environment. It 
may be only about zi square kilometres or, where game is more scattered and the 
chance of picking up a meal correspondingly reduced, as much as 20 square kilometres 
or more. That there is no meticulous insistence upon exclusive ownership is demon- 



4S Tin; cAUNuiiurs oi west muk.a 

stnitcd by tlu' luit unusual presence at a lions" or hyaenas' kill ot more than one pair, 
or one taniilv. ot jackals. Goodall once observed as many as 14 t;olden jackals at 
one such teast; but as onl\' 6 ot these were hilly adult perhaps no more than three 
families were represented. The ditierent nature ot the receptions accorded to intruders 
upon the range may possibly be accounted tor bv those liostilcK- treated being com- 
plete strangers while those calmlv adnntted are erstwhile, though now independent, 
members of the tamiK'. The much more restricted territory around the burrow is 
i.|uite a different matter. It is well defined, the boundaries being clearly demarcated by 
constant urination or the deposition of faeces, and it is hotly defended against trespass 
by other jackals. When there are young to be pr(Hected dangerous visitors of other 
kinds, such as hyaenas, are made very unwelcome. 

Fights ot a serious nature between golden jackals break out sometimes at kills, 
sometimes 111 the defence of territory. In these, which are accompanied by a good deal 
ot snarling, the main object is to bite into the opponent's neck, or failing that the face. 
Hyaenas arc driven off b\ snapping at their hindleet and legs. The j,\ckals are aided 
in this by their extreme nimbleness, being always verv quick indeed in their move- 
ments which thus enable them to contend successfulK- with much stronger but more 
clumsy opponents. A swift movement emplo\'ed bv |ackals 111 common with other 
members ot the Canidae is a quick rotation ot the body sideways, striking the opponent 
with the whole flank, thus knocking it off its stance. This would not, of course, be tried 
against an obviouslv bulkier opponent but is used amongst the jackals themselves and 
is an effective method ot frightening vultures awav from kills. The numbers of jackals 
must somehow be constantly kept in check. A full list of lethal enemies is not known; 
but possibK' leopards and other large telids attack them. Certainly in their juvenile 
stages they are in danger from h\'aenas and possibk hunting dogs as well as the cats 
and large birds of prey. Undoubtedly disease pla\s its part; Goodall observed the total 
demise ot a litter ot five weakly cubs 111 a period ot .1 tew days. Golden jackals are 
themselves, like all dogs, potential reservoirs and spreaders ot rabies; an accurate 
knowledge ot the extent to which they may be expected to range in search of mates 
or better feeding grounds is therefore ot some importance. 

Since golden jackals associate, most of the time, oiiK in limited tamily groups there 
Is obviously not the same scope tor the development of a recognized social hierarchy 
as there is amongst animals which live 111 larger and more diverse packs. Nevertheless, 
Goodall (1970) observed clear indications of social order established amongst the cubs 
of a single litter. There is no question ot the parents entering into this; but between 
themselves the male seems always to be dominant, and W\ man indicates, in a photo- 
graph, th.it he always feeds first. Goodall, on the other hand, clearlv refers to the 
pair feeding together off their recently-made kill. 

One of Wvman's most interesting observations is the existence of a behaviour 
pattern relating to the solicitation ot food. This is a greeting ceremony primarily 
employed bv the cubs when their elders return from foraging. They dance around 
the parent waggini; their tails and laving their ears back but alwa\s keeping their 
mu/zles to the ciirnci ot the month of the adult, who in response regurgitates chunks 
(it meat iipoii wlikli the \(Uingsters then teed. Soinetmies the\' regurgit.Ue more tli.ui 



CANIS 49 

die cubs can cope with. The same ceremony, vvitli a similar result, is used also b\' fully 
adult jackals, possibly of a previous litter, which have for some reason remained at the 
burrow instead of joining the hunt. 

Courtship starts with periods of prolonged grooming of the other's tur by each 
prospective partner. As the time ot actual coupling draws nearer the male makes more 
meaningful advances to the female with his tail stretched out horizontally, his head 
lowered and his ears laid forward. The female repels these advances with a bite. Coupl- 
ing probably takes place in darkness. The mean period of gestation is 9 weeks but it 
may vary by as much as 5 days on either side of this. There may be from i to 5 cubs, 
2 or 3 being common numbers in a litter. Tlieir eyes open at about the loth day. Their 
first pelage is very dark, almost black, and they do not assume their lighter, golden- 
grey colour until they are nearly a month old. 

The cubs arc milk fed for about three weeks and arc thereafter gradually introduced 
to solid food For the first few days suckling takes place in the den; but when the cubs 
can move about they come to the mouth of the burrow in response to a whining call 
of the mother and are fed there or just outside the entrance. When they are strong 
enough suckling takes place in a standing position with the cubs on their hindlegs, 
their forefeet pressed against the mammae. They are fully weaned at about two months. 
Unlike the hunting dogs or li)'aenas, golden jackal cubs are often left alone for long 
periods while both the parents are out hunting. When the parents return the cubs 
greet them by nibbling at their muzzles and ears; and ultimately when they are old 
enough they carry out the begging ceremony, as related above, and are fed by re- 
gurgitation. 

The father as well as the mother takes part in the upbringing. When feeding is over 
the female carries out a good deal of grooming; and the pups themselves indulge in 
much boisterous play. Goodall observed a female to have a second litter within six 
months of having weaned the first. Another activity carried out as a family, fairly 
rcgularlv morning and evening and sometimes at night, is howling, the head and cars 
laid back with nose stretched upwards to the sky, the mouth wide open. The sound is a 
high-pitched, two-toned "0000". It is always answered by other families in the distance. 
A golden jackal has been known to live in captivity for about 16 years; but survival 
tor as long as this in the wild is probablv unusual. 

CANIS ADUSTUS Sundevall Side-striped Jackal 

Catiis adiisliis Sundevall, 1846, Ofvers. VctmskAkad. Forli., Siockh. 3: 121. According to Simdcvall, the 
interior of South Africa; given by Roberts (195 1) ^5 Magalicsbcrg, Transvaal. The name adiisnn is 
the Latin word for sunburnt or swarthy. 

Caiiis ajiistiis wiitralis Schwar?, 191 5, Jh. iiassaii. Vcr. NatiirL'. 68: 60-62. Bate, near the Uhain River, 
Cameroun. The Latin subspecific name was intended to indicate that the type came from the middle 
of Africa. Sonietinics regarded .as possibly covering all West African jackals of this species. 

Distribution. The side-striped jackal (fig. 5) occurs from the Transvaal northwards 
to Ethiopia on the east and to Cameroun, Northern Nigeria, i5ahomey, Guinea and 
upper Gambia on the west. It inhabits rather moister regions than the golden jackal. 



50 nil rMlMVOlUS (M Ul-.ST AimcA 

tli.it IS to say 111 West Atnca tlic Sudan and Doka zones. Only two spcciiiiciis exist 
troiii West Africa in the Britisli Museum, one from near Gombe, Northern Nigeria, 
the otlier from Gambia, the exact locahtv unrecorded, tlie skin, but no skull, having 
been received from a zoo. It has also been reliably reported trom the Shendam area ot 
Nigeria, somewhat north ot the Benuc at about X 55 N, 9 28 E, in Doka woodland; 
and bv G. S. Child who has found it and its dens in the Borgu Game Reserve in extreme 
western Nigeria. 

Description. One of the distinguishing features generally cited between this species 
and the last is that the tail tip is white as contrasted with the black of iiiirciis. While 
this is true of South African and many East African specimens it docs not appear to 
hold for West African animals and certainly does not occur in the two examples from 
that region in the British Museum or those from the same vegetation belt as far east as 
Bahr-cl-Ghazal. Another reputed distinction between the two species is that which 
gives the common name to ihhistiis, a black stripe along the flank of this latter, lacking 
in the golden jackal. Such a mark certainly occurs in the majority of specimens, but 
it is not always easily discernible and often bv no means as clear as it is made to appear 
in most illustrations. It is in some measure dependent upon individual idiosyncracy, 
111 some upon moult, and in some upon the actual lie ot the coat, which may emphasize 
or obscure it. It can, indeed, vary very much in distinctness on the two flanks ot one 
jackal. This black flank-stripe is emphasized by the existence of a white band above it 
which divides it from the darkish, speckled back. Questions of set coat patterns apart, 
side-striped jackals are altogether much darker animals than golden jackals, partly 
because the black in the coat is more evenly distributed, seldom, except on the flanks, 
aggregating into black patches, and then only ot small size. There is always one certain 
point of distinction between the t\\o; the ears in aihistiis are grey on their backs, not 
golden-brown as m aiiiciis. 

Aside from gross difterences ot appearance the pelages ot the two species arc appre- 
ciably discrepant in detail. That a[ Mhistiis is long but not quite so long and dense as in 
iUirciis; and it is possibK' a trifle less harsh. It similarly consists of abundant tuie undertur 
and very long, banded, terete bristle-hairs: the former being approximately 20 mm 
long as compared with about 30 mm in aniens; and the latter so to 60 mm compared 
with 70 to So mm m the other. When the coat is turned up backwards a sharp dis- 
tinction between the two species becomes evident: (uhntiis is seen to be largely pale 
reddish-brown in contrast to the buffish-white ot aureus; and also, the white subter- 
iiiinal bands of the bristle-hairs, although very much shorter than in the latter species, 
show up far more conspicuously because of their greater contrast to the deeper-coloured 
lower part of the tur. The red of the luiderfur often pervades the dark speckled saddle- 
patch of the back, the extent to which it does this depending upon the state of moult 
and the consequent density of the bristle-hairs at the time. 

The nose is red-brown; the crown of the head and the nape are speckled, the white 
subtcrminal bands to the short hairs covering these areas making a conspicuous "frost- 
ing". The upper lips are narrowly yyhitish but this pallidness scarcely spreads to 
the cheeks. The chin is grey, the throat mostly whitish or buflish but there is usually a 
speckled collar at its Io\yer end. The rest of the luiderside is clothed w ith long hair 



CANI5 



51 




•i2 Tin (AllNIVllHIS (i| W I s I MUKA 

b.iving long liuHish tips over a pak' cliocolatc base. Tlic limhs and tcct arc a deeper 
shade nt red than ni .iii/ei/.v; and tliere is often some sign ot a black longitndinal line 
on the torelegs. but this is not always clear and is sometniies lacking. The bush)', 
verv long-liaired tail is basically burtish, often with a tcuicli ot red, but it alwa\'s has a 
considerable ainonnt of black on it. tar more than m iUircii<. hi West Africa the tip is 
black. 

It is diHicult to give am- precise idea ot the relative sizes ot (uliisliis and iiiiiriii since 
111 the British Mnseum there are no external measiirenients relating to West Africa 
and verv tew from elsewhere. Tiie figures tor the usual four bodily measiu'ements given 
in the table on page ss are nierelv estimates compiled from a tew e>ctralimita! data 
in conjimction with Schwarz"s (1915) figures published for cciilnilis — which themselves 
were derived from a dry skin. From these it would seem that there is not a great deal 
of dirterence in size except that the ear is possibly shorter in ndiistiis. One animal from 
liahr-el-Ghazal was said to weigh yy kg, that is slightly more than iiiirciis. 

Skull (tig. 6). The first obvious ditterence between the skull ot <uhistiis and that of 
diiicus lies in the former's longer rostrum, sloping evenly from the supraorbital region 
instead of markedly dipping below the frontal. The nasals arc generally considerably 
longer in adiistii.<: and extend further forward over the anterior uares, which arc thus 
less open dorsallv. The skull is narrower all round; the zygomatic breadth, the brain- 
case, the interorbital and postorbital widths are tor the most part less. The supra- 
occipital crest viewed from the front is much less angular, almost semicircular in 
outline. The zygomatic arch is far less curved vertically than the subsemicircular 
arch ot JI/IV//.V The mandible appears to be less powerftilK' built than in lUiictis. The 
ramus is not so deep and is only sliglitly curved, so tliat tlic angular process is not much 
elevated above tlie level of its lower margin. Both anterior and posterior edges ot the 
coronoid process are practically straight, the process tluis being ot a truncated wedge- 
shape. 

The carnassials are smaller than in (iinciis both in absolute terms and in their ratio 
to the molars, hi tins species p^ is not much more than 70 per cent ot the length ot 
;/i' - ;//-; and iii] is well under 130 per cent ot iih -p /;/;j. 

Taxonomy. The taxonomy ot adiisliis, at least in so tar as West Africa is concerned, 
IS much less contused than that ot i/kiii/.v. As a species it was originally the subject of a 
perfectly clear diagnosis; and most ot the nine races subset]uently named have been 
fairly adequatcK- described during the present century. Only one ot these, cciitialis, 
probably came from within the limits set tor this book; but it is ditticult, or impossible 
now, to t'lx the tvpe locality, Bate, with any certainty, h was said b)' Schwarz (1920) 
to lie on the Uham lliver on the derman side ot the Cameroun frontier; but this river 
does not appear to cross that boiuidarv. The most likely position of Bate seems to be 
roughK' 7 N, IS E, that is in the Guinea wdodland. The next nearest recorded speci- 
mens to the tvpe oi tciilnilis seem to be those from Chak-Cliak, liahr-el-Cihazal (I5oka 
woodland), which Setzer (1956) assigned to hwclui Heller. 

It IS obvious that with the extremely limited material from West Africa available 
tor study — two specimens only, of widely separated provenance and possibly ditterent 
vee:etatioii zones, and differing somewhat in appearance from each cvther — it is almost 



CANIS 



53 




Fig. 6. Canis adusnis: skull, B.M. No. 28.6.3.8, sex ?, x 



54 THE CARNIVORES OF WEST APIIK A 

pomtlcss to attempt reasonable comparison witli descriptions ot tlicse races troni 
further cast in Africa. Neither West African specimen would seem to match Schwarz's 
diagnosis ot ccntnilis, described as noteworthy for its very light ground-colour. That 
from Gombe, on the other hand, is very similar to the British Museum speeunens 
held by Setzer to be hirclia. All, indeed, seem to be a rather closer match to this race 
than to cciiindls, though they tail to correspond wholly to Heller's description. There 
is very little to choose iji size between any of the specimens referred to in this section 
except that the hii'clui type seems to be a little larger. 

Much more material is needed before any usetul purpose can be served by attempting 
to attach a third name to West African animals. Doubtless the issue of geographical 
races is obscured in all these wild dogs by a good deal of individual variation, the extent 
ot which it is quite impossible to estimate from exiguous collections. 

Habits. These are not very well recorded but do not seem to differ in their essentials 
verv markedly from those of the golden jackal. Side-striped jackals are by preference 
nocturnal but, like aureus, are also to be seen on the move in the early morning or late 
evening. It must be remembered that this species is probably less often seen than the 
golden jackal because for most of the year in West Africa, except at the height of the 
dry-season when the grass has been bm'iit. it occupies vegetation, the Sudan and Doka 
zones, in which the groimd-cover is considerably denser. At night these jackals make 
their presence known by sitting and uttering at intervals their sad yapping howl, 
a relatively slow woo-it'oo-woo. Fecduig habits are m general very much those ot the 
golden jackal, but ndustus is reputed not to steal farm stock. It seems, indeed, to be a 
rather shyer, more timid and secretive animal, mostly solitary in its habits or accom- 
panied by a mate at the right season, or by pups in due course. 

This jackal, like the other, shelters and breeds in a hole in the ground. The period 
ot gestation is similar to that n[ aureus, aroiuid 9 weeks but varying by 4 or 5 days on 
cither side of tliis. As many as 7 pups have been recorded in a litter, but 3 or 4 is probably 
a common number. A side-striped jackal has been known to live in captivit\ for 10 
vcars. 



CANIS 55 



o 









t^ «-< 






jHrnr-*ccoc !?*c\t~^o pt*it-^'C 7^7^" 0*^00 












pisqo B 










a. 



.a — : 



a -s-t isi.e -£->--^-s-^-^ 






, , 'J CJJ— Q 'J _ ti W 



S != -^ = 






56 I HE CARNIVOlilS Ol WIST AFKICA 



Gciius FENNECUS Dcsmarcst. TS04 
Feiuiccs 

reiinccus Dcsm.irest, A. G., 1S04, l\'oui'i:im Diciionnaire d' Hisloire Niitiin-llc, 24, Tableau niL-tliodiqiiL- dcs 

Mamniifercs: iS. Type species FiHiiecus arabicus Desniarest(= Ctinis zcrda Ziniincniiann). The name 

IS a Latinized form of the Moorish word tor a fox, fainec. 
Mcgtilolis IlHgcr, 181 1, Proilroniiis systciiwlii Mammalium ct Avium . . . : 131. Type species Caiiii ccrdo 

Gmehn (= Caiiis zerda Zimmermann). This name was made up from the Greek words tncj^as (mc^al-) 

large and oiis (otoi) ear. 

Tills is a monospecific genus distributed over a small area of northern Africa from 
Morocco to Egypt, as far south only as Air and Sudan, and thence across to parts of 
Arabia. In other words, it is an animal of dry sandy deserts or subdcscrts. Since there is 
only one species there is no point in entering into a generic description. 



FENNECUS ZERDA (Zimmermann) Feniicc 

I'lilpcs minimus siiannsis Skjoldebrand, 1777, K. svciiska I'etenskAkad. Haudl. 38: 267, pi. 6. Algerian 

Sahara. This name is regarded as invalid because, as given to a species, not a subspecies, it was tri- 
nomial. The second name, minimus, is Latin for smallest, given because ot this annnal's dnninutive 

size for a fox; the third name is a Latinizarion of Sahara. 
Ciinis zcrda Zimniermami, 1780, Gcographische Gcschichle dcs Akiisclicn und da vicrluszi\;i-n Tltiere 2: 

247-24S. Sahara and North Africa behind the Atlas Mountains. The name was said by Zimmermann 

to be that used in "Barbary". 
Cdnis ardo Gmelin, 1788, Linnaeus' Syslvnui Natumc, 13th ed. 1 : 75. Sahara. This name is another spelling 

of zi-rdii. 
I'ii'ina anrita Meyer, 1793, Systeniatisch-summarische Uebersiclit dcr ncuesteii zoohgischcn Entdeckungcn in 

Naiholland und AJrika: 91. Bisk.a etc., Algeria. The Latin adjective aurila means having large ears. 
Fcnncais arabicus Desmarest, A. G., 1804, Nouvcau Diaionnaiic d'Hisioire i\'atiin-Uc, 2i, Tableau metho- 

dique dcs Mammiferes: 18. Barbary, Nubia, Abyssinia. 
Mt'^ahiis ccrda Illiger, 1811, Prodromus systcmatis Maiiimaliuni ti Avunn . . . ; 131. The name is a variant 

oi zcrda. 
h'cnuccus brucci Desmarest, A. G., 1820, Encyclopedic Mcthcdique, Mammalogie: 235. Libya, Tunis, Algeria, 

Sena.ar. James Bruce, after whom this was called, was a well-known explorer of northern Africa in 

the second half of the 1 8th century. 
GjMij^eimcaK Lesson, 1827, Manuel dc Maninialoi;ic : 168. 
Vnlpcs denhami Boitard, 1842, Le Jardin dcs Planlcs: 213. Interior of Africa. This was named in honour 

of Lt. Col. Dixon Denham, a famous traveller m the Siliara and explorer of Lake Chad in the early 

iQth century. 

Distribution and general. The range of this essentially Saharan animal has already 
been given above. In suitable localities it is not uncommon and because oi its small 
size and rather charming appearance it has tor long been a favourite with writers on 
natural history, tiguruig far more copiouslv in literature than mam- more widely 



FENNECUS 57 

distributed and rather more important animals. In general appearance it is a mimatiuc 
fox with a sandy coat, huge ears and a very bushy tail, and it is, indeed, often know^l 
as the fennec fox, though any really close relationship to the true foxes has been brought 
into question. It does not occur ever)'where iji the Desert and Subdesert zones bccaure 
it must have soft sand into which to burrow. Sand dmies are therefore ideal, but not 
in utterly barren situations since food must, of course, be available in fair quantities. 

Description. The femiec (Plate i) has a head & body length of about 300 to 370 
nun and a tail of 160 to 240 mm. It stands about 150 to 175 mm at the shoulder and 
weighs about 2 kg. The pelage is long and very soft, both above and below. Broadly 
speaking, dorsally it is sandy, but there is a certain amount of variation, some specimens 
beiug rather greyer, some rather redder. The same apphes to a darker, richer coloured, 
band along the spine, almost absent from some, clear m others, especially on the hinder 
part of the back. In this area, in some skins, the bristles are dark-tipped with a pure 
white subterminal band, forming a very prettily-patterned patch. In others the pattern 
is more diffuse, the majority of specimens having fme bristles with long black tips 
thinly dispersed over the entire dorsal region. The pelage is so soft to the touch because 
it consists, apart from these scattered bristles, entirely of dense, very fme, very long, 
luiderfur. Tliis, iji the North African examples, is pale chocolate-grey in the basal half 
the extreme base being narrowly white; but in the western specimen from A'ir there 
is no trace of this basal tijiting; while in the Dongola (Sudan) skin it is relatively pale. 
The West African examples, too, are much shorter-fiu-red; but they are all youngish 
animals. The tuidcrparts aiid the insides of the limbs are pure white, the frir being 
abiuidant and soft, but only half the length of that of the back. 

The tail, which is roughly half as long as the head & body, is very bushy, the very 
long hairs with which it is clothed beiiag towards the tips a rather redder brown than 
the back. There is a deep blackish-browni mark not far from the root of the tail covering 
a scent gland; and the extreme tip is also blackish-brown. 

The head, with a broad face and large eyes, comes rather abruptly to a narrow 
muzzle and is wholly dominated by the enormous, pointed ears which are broad as 
well as long. On their backs they are sandy-grey, but all around the marginal area 
on the inside there are dense long white hairs. The crown and front are sandy, but 
much of the rest of the face, surroimding the eyes, the rhinarimn, the cheeks, and the 
upper lips, is pmc white with the exception of two dark, reddish-brown lines descending 
from the inner comers of the eys to the lips. The upper parts of the limbs are, in the 
northern African specimens, reddish-sandy; but in the Air examples they are nearly 
white. 

Skull (fig. 7). The skull tapers fairly sharply from a moderately broad braincase 
and enormous bullae to a very narrow rostrum. The profde dips appreciably just 
forward of the orbits to give this narrow and low muzzle. The braincase is rounded; 
the supraorbital ridges arc well-pronoiuiced but narrow and sharp, hollowed on their 
upper surfaces, in the maiuier of Vulpes rather than Cants. There is no very marked 
intcrorbital constriction. The distance across the zygomata is wide, the maxillary 
process broad, the slender arches sharply up-curved, the circumorbital ring widely 



58 THE PAnNIVORES OF WEST AfRICA 

open. There is little or no sagittal crest and only poorlv developed supraoccipital ones. 
The bullae are extremely large. The palate is very broad betweeji the carnassials but 
narrows abruptly and becomes practically parallel-sided anteriorly. Its hind margm 
is about level with the middle of (/i-. There is nothing particularly remarkable about 
the lower jaw or the dentition apart from its very sharply cuspidate nature whicii 
probably facilitates an insectivorous diet (fig. i). 

Taxonomy. This, despite the multitude of names, actual and in permutation, by 
which this animal has been knowji to science, is pretty straightforward. It is true that 
in the early days Butfon (177O, Suppl. 3: 148) referred to it as the anonymous annual, 
and that Lesson (1S27) wrote ot it (in translation) that perhaps no animal had more 
engaged naturalists than this; they have made of it by turns a dog, a galago, or the 
type of the genus Fciiucc. 

The old coiifusion of naming has now been s\\ ept aside and ti.xed \\ ith apparent 
permanency as Fciiiicais zcrdd. But, at a very considerably narrower pitch than that 
indicated in the previous paragraph, some doubt of the animal's precise relationship 
still remains. The tennec has long and widely been regarded as a kind of fox, and is in 
fact very frequently referred to as the feiuicc fox; but the cytological researches of 
Matthey (1954) have shown tliat the chromosomes (diploid number, 2N — 64) indicate 
that the genus lies closer to the wolves {Caiiis) than to the foxes (Viilpes). 

So far, no races have been described. The West African material from Air is poor 111 
quality and meagre in amount and it is therefore not possible to draw any reasonable 
conclusions; but superficially the animals from this area seem somewhat paler and 
shorter haired than those from further north-east; but there is no significant diilerencc 
of size so far as can be deduced from the mean measurements of three h'om the one 
area and of five from the other. 

Habits. Accoimts of the.sc in the wild have been briefl^' given h\ several collectors 
and observers over a large number of years; but, in all, the information from these 
sources amoiuits to little beyond the most obvious facts of life in the desert. However, 
because of the fascination which this miniature "fox"' of charming appearance has 
long exercised over human-beings, especially as a household pet, the feniiec has been 
more closely studied in captivity than most carnivores. Good descriptions of its behav- 
iour and disposition under these circumstances have been given in recent years by a 
number of writers of whom the following arc the chief: Rcnsch (1950), l^etter (1957), 
Volf (1957), Hill (1961), Saint Girons (1962), Vogel (1962) and Gauthier-Pilters (1966). 
Though these accounts concern animals in unnatiu-al conditions thc\- nevertheless cast 
important light on instinctive behaviours, the more convincing in that many traits 
are displayed in common b)' animals of different origins at different times and places. 
It is not possible in a work of the present nature to do more than glance briefly at 
some of the recorded facts, more especiallv those which most probablv rc\eal the 
femtec's normal activities. 

Fermecs are essentially nocturnal or crepuscular in their activit)-; \ct though, like 
so many desert animals, they avoid the full hirce of the mid-day heat, they are not 
averse to sunning themselves for brief periods before the dav is far advanced. 15ut 



FENNECUS 



59 




Fig. 7. Faiiicais zada: skull. B.M. No. 1939.1746, ?, x i 



60 THE CARNIVORES OF WEST AFRICA 

very little of this insolation suffices for their daily needs and they soon seek shade — 
though they ma)' emerge for a second spell, a pattern that may be observed in many 
domestic dogs. The fennec's home is a hole in the groiuid, which it digs for itself, 
fairly deep probably to avoid overheating during the day. These "earths" are lined 
with soft material, and they have a number of different escape passages some of which 
are said occasionall)' to commmiicate with the homes of other fennecs, as these are 
sociable little animals which in suitable areas live amicably together in some numbers. 
The unit of these communities is almost certainlv a small family group; and it would 
seem from indications, though it is not certainly known, that male and female hang 
closely together for long periods. Favourable localities for femiccs are characterized 
by soft sand into which burrowing can be carried out with ease. For this reason stable 
sand-dtuies fixed by the roots of sparse vegetation are ideal, wholly barren desert having 
no attraction since food must be available in fair quantities. Given a soft soil such as 
exists in these sites, digging, carried out with the forepaws, is very rapid. All who have 
kept these animals refer to one habit that makes them difficult to tolerate as domestic 
pets: the constant and very noisy nocturnal scratching resulting from attempts to dig 
into unyielding wooden or concrete floors. That the instinct to hide is very strongly 
developed is illustrated by the fact that even in captivity fennecs always seek to shelter 
underneath furniture; or, if nothing more resembling an "earth" is available, by pulling 
over themselves such bedding as they may be provided with. 

It has already been said that fennecs dislike intense heat, particularly direct insolation; 
but they very much appreciate warmth. They are, indeed, highly sensitive to cold, 
and this is another reason for the depth of the burrow so that a wholly equable tem- 
perature can be maintained. When above ground on the hunt for food, protection 
against the night cold that develops rapidly after sunset in sandy deserts is to some 
extent provided by the dense woolly pelage; but, to judge from behaviour in captivity, 
fennecs are intermittently active for short periods and this might be explained by the 
necessity, in nature, to return from time to time to the warmth of the home. 

Another highly important reason for the avoidance of intense insolation and the 
frequent seeking of shelter in a deep burrow, in which the humidity probably remains 
at a high level, is the need to conserve water. It is imlikely that the majority of feiuiecs 
in the wild have easy access to standing water in order to drink, and very likely they 
obtain most of what they need from moisture in fruits and bulbous roots. Observations 
on drinking habits in captivity are curiously conflicting. Rensch found that his fennec 
drank little or nothing for the first two or three years except at the mating season; 
but Later in life it drank daily. Vogel found much the same thing; but he noticed, 
in addition, that the animal was averse to drinking out of a bowl but readily lapped 
up drops spilled on the floor — so much so that it was observed, itself, to scatter water 
from the drinking vessel with its forepaws or nose. Rensch mentions his fennec as 
licking drops from a tap. This would seem to indicate that even if these animals had 
access to oases, or other pools, they would prefer to get what extra supplies they 
required beyond what is provided in fruit by licking small quantities troin leaves or 
other surfaces moistened by dew or rare showers of rain. Schmidt-Nielsen (1964), 
in connexion with loss of bodily water, foimd that urine could attain a high concentra- 



FENNECUS 6l 

tion in the fennecs. When over-heated, fennecs can cool off to some extent, hke dogs, 
by panting; but as this involves loss of essential moisture from the tongue it is not an 
operation that can commonly be engaged in. 

Fennecs are seemingly almost omnivorous, and there is little difficulty in feeding 
them in captivity since they readily eat most things from roast beef to marmalade. 
Fn their natural surroimdings they live on small rodents, such as gerbils and jerboas, 
lizards, insects, eggs, small birds and a good deal of vegetable matter, of which fruits 
and tuberous or bulbous roots form the main part. Vogel observed his fennec, like a 
domestic dog, to eat grass when opportunity offered. According to Professor Monod 
of Dakar quoted in Dekeyser (1955), fennecs are very fond of the bright yellow, leafless 
parasite Cistanche phelypaea Cout. often growing in fleshy clusters at ground level on 
the roots of, amongst others, the so-called salt bush, Salvaclom persica, commonly 
occurring in the Sahel and Subdesert zones. Big pieces of food are consumed in a 
sitting posture; small ones are eaten while standing. 

In many of their ways these little "desert foxes" reflect the manners of some domestic 
dogs: as, for example, in their inquisitiveness, evinced in sniffing at or quizzically 
regarding unfamiliar objects; in hiding away surplus tit-bits of food; in turning round 
three or four times before settling down. They particularly resemble poodles in their 
ability to stand and walk upright on their hindlegs, and in the way they stretch their 
hindlimbs after sleep, or lie flat on their bellies with both fore and hind legs stretched 
out. In other ways they are rather cat-like, particularly in their manner of cleaning 
themselves by licking their forepaws and "washing" their heads. The large ears are 
cleaned on their insides by the hindfeet. Like cats, too, fennecs are very competent 
jumpers, being able to accomplish a standing more or less vertical spring of 600 mm 
or more and a horizontal spring of 1200 mm. They also, at times, can purr something 
like a cat; and when they sit on their haunches they curve their tails sideways and 
forwards like a cat, sometimes even raising the tip off the ground as cats do. Fennecs, 
too, are very competent climbers and can ascend vertical obstructions that offer some 
sort of foothold. They can also squeeze through remarkably narrow crevices. 

These animals scratch a shallow hole to defaecate and urinate in and they cover their 
droppings with sand, sometimes shovelling it over with the nose or, more often, 
flinging the earth backwards with a scraping action of the hindlegs, just as domestic 
dogs can often be seen to do. This latter ritual is performed automatically even when 
it is pointless on the hard wooden floor of a cage. Feimecs utter a variety of sounds 
from a low growl to a higher snarl ; when content and being stroked or fondled the\- 
make a purring noise similar to that of a cat. They are given to yapping at night. There 
is an annual moult. Opinions differ as to whether these animals are ever completely 
tameable. As with almost any wild creature it is probably merely a matter of inbuilt 
temperament, varying from one individual to another. One fennec, at least, has shown 
itself to have a very long memory and to be able to recognize its owner with excited 
pleasure after many months of separation — and tliis at a distance by sight, the possibilit)- 
of smell being eliminated by a glass front to the cage. These animals have a longevity- 
of about 12 years. Schmidt-Nielsen records that the young are dug out of their nests, 
fattened up and marketed for eating. 



62 1111 (AUMVOUIS (II W IS 1 AIUKA 

MatiiisT .md breeding in captivity have been observed and recorded b\ I'etter, Volt. 
and Saint Girons. After penetration the male turns round so that the two sexes face 
m opposite directions, and coition lasts i{- hours. There is apparently no attempt at 
Mibsequem copulation. The period ot gestation is so or si days. Petter found tliat after 
mating the female continued to act normally but the male grew aggressive towards 
people whom lie knew well; this became more pronoimccd after parturition and so 
he was separated from the niotjicr. By then she was so upset that she kept transportnig 
her newlv born cub from place to place and ended by thus accidentally causing its 
death. Volf records verv much the same sort of thing. From i to 3 cubs are boni. Their 
eyes are closed, their ears free though giving no hint of their future huge size. The 
pelage is fuie and short. Under natural conditions the -soung appear aKi,ays to be born 
in March or. at the latest, the begiiuiing of April. 

Gauthier-Pilters (1966) has recorded in sonic detail her observations on play in the 
tennecs made on 21 animals over a period of 11 years, all born in the wild. Juvenile 
fennccs play with each inlicr verv like fox cubs or the puppies of domestic dogs, 
biting each other in the legs and neck and rolling each other over. They also shake 
small objects as if shaking a rat to death. From the age of about 6 weeks they indulge 
ill racing plav, chasing each otlier. the pursuer with cars erect and directed forw^ards, 
the pursued with them laid back, zigzagging or making abrupt changes of direction. 
Tliis thev will do even, if necessarv, in a very confuted space. Where opportunitv 
offers the\- may plav a kind of hide-and-seek. Should play become too rough and one 
of the participants get cornered, or for that matter at an\- time or any age when offensive 
or defensive postme seems necessar\-, the back becomes arched and the ears are laid 
back in a threatening attitude. Feiuiecs have no abilitv to raise their hackles like a dog, 
but instead, the black mark at the root of the tail is erected and displayed. Solitary plav 
also takes place, using articles ot food or substitute to\-s, tossing these in the air or shoving 
them about with foot or nose. Chasing tlieir own tails is sometimes indulged in. 
Fennccs will play even w-ith strange anini.ils which thev have only just met. Gauthier- 
l^ilters (1966) records games between \'oung teiuiecs, dogs and jackals; but cats appear 
unwelcome and frightening. Should i-itlier creatures fiil thev pla\' willingK- with 
human beings. 

Inclination to pla\- appears to vary with se.isiiii. taljuig to a muiimum at breeding 
time. This observer found tlie pattern ot female beliaviour with the male to differ 
before and after parturition, exhibiting in the first place elements of greeting behaviour 
and in the second of breeding behaviour, that is tlie bringing and offering of food — 
though the male was, in the event, never allowed to take it. Most pla\' takes pLace at 
evening or .it night, hideed. witli captive tennecv. to judge from the experiences 
recorded b\- Ci.aithier-Pilters. this can be somewhat ot a tri.il tor the owners ot these 
luH turnalK- ver\- active creatures, sijice they persistently play after dark for hours on 
end. preferabK' over and luider beds, tables and chairs. This author records a T4-day 
trek with half-tamed tennecs w. hich for safety had to be tied at night to her camp-bed 
and which, despite their thus restricted freedom, spent their time in romping or, at 
interv.iK. digging deep holes in the sand. 

Sueli pla\' activities, espeeialK' 111 the voung, are recogni/abl\ related 10 beluiMour 



VULPF.S 63 

cxliibitcd 111 various situations regularly arising in later life, and for which they arc, 
in effect, rehearsals, training and strengthening the necessary muscles and inculcating 
instinctive reaction on occasions demanding offence, defence, capture or kilhng. 

Table 2: Numerical data for i'enncais zcriia 





West 


West 


Sudan 








A&ica 


Africa 


(Dongola) 


Northern 


Northern 




(Afr) 


(Air) 


single 


Africa 


Africa 




means 


extremes 


specimen 


means 


extremes 


Vcgetttion 


Subdescit 


— 


Subdesert 


Subdesert 


— 


Number in mean 


3 


— 


I 


5 


— 


Condylobasal length 


83-2 


82-8-83 -7 


82-3 


83-8 


78-7-86-7 


Basilar length 


77.9 


77-4-78 -2 


76-5 


78-2 


73-6-80-4 


Palatilar length 


39-0 


38-9-39-I 


37-3 


38-9 


36-4-41-0 


Zygomatic breadth 


45-4 


44-4-47-2 


43-8 


46-1 


43-5-48-6 


Upper cheekteeth breadth 


24-2 


23-3-2S-4 


24-0 


25-0 


23-2-25-9 


Nasals, length 


27-5 


26-2-28 ■ S 


26-8 


27-1 


26-1-28-7 


Interorbital breadth 


16-5 


16-0-17-0 


15-8 


16-2 


15-5-17-0 


Postorbital constriction 


20-3 


19-8-21-; 


17-7 


20-2 


18-7-21 -3 


Braincase breadth 


3fi-8 


36-7-37-0 


35-7 


36-5 


35-5-37-5 


Toothrow (c — III-) 


3S-I 


33-6-36-0 


35-3 


35-4 


32-4-36-4 


p* length 


7-4 


7-2-7-6 


7-0 


7-6 


7-4-7-8 


111^ + II fi length 


lO'I 


9-9-10-2 


9-4 


lO-I 


9-5-10-5 


Mil length 


8-3 


8-1-8-4 


8-1 


8-3 


8-0-8-5 


1112 ■- 1113 length 


8-0 


(7-9)8-0 


y-i 


7-6 


7-2-7-8 


Head & bodv 


355 


350-360 


360 


371 


333-395 


Tail 


170 


160-180 


190 


147 


125-187 


Hindfoot 


93 


92-94 


92 


94 


90-9S 


Ear 


90 


86-95 


90 


94 


90-97 


RATIOS (per cent) 












Tail/head & bod\- 


48 




S3 


40 




Zygom. br./condylob. 1. 


55 




53 


55 




Braincase/condylob. 1. 


44 




43 


44 




Braincase/zygom. br. 


81 




81 


79 




Palatilar l./condylob. 1, 


47 




45 


46 




Interorb./postorb. 


81 




89 


80 




P*JC— III 


21-0 




19-8 


21-4 




p^jm^ — II fi 


73-2 




74-6 


75-2 




iiilliii, - (113 


104. 




114 


109 





c;cnus VULPES Hemiiig. 1N22 
True Foxes 



I iilpcs Frisch, 1775, Das Naliirsyslcm dcr viir(ussigai Tliicre, 15. This work has been ruled to be unavailable 
I'y the International ("ommission on Zoological Nomenclature (Opinion No. ^sS of 1954). The name 
IS the Latin for a tox. 



64 THE C.ARMVOHES OF WEST AERICA 

I'lilpcs Okcii, iSif), Lihibihli da Kalnr^cschiihk\ 3, 2: 1033, 1034. Type species Vtilpcs aviiiiiunis Okcii 
(= Cam's viilpcs Linnaeus). Okeii's Lilirhiuli is similarly unavailable (Opinion No. 417 of iy56). 

I ■"//)« Fleming, 1822, The Philosophy of Zoology, 2: 1S4. Type species Catiis rulpcs Linnaeus. 

(^yimlopcx H. Smith, 1839, Jardinc's Wmiralist's Library, 25 (of the series), 9 (of Mamm.ils): 222. Type 
species Caiiis corsac Linnaeus. 

Distribution. This genus is spread tluoughout Europe, much of Asia and of North 
America, and over the more arid areas of Africa. The common red fo.x of Europe 
{I'lilpcs viilpes), in one form or another, occurs over a great part of tliis wide range 
with the exception of the more southerly parts of Asia and Africa south of the Medi- 
terrajiean region — though its identity with the American red fox, frequently alter- 
natively designated V. fnlva, is disputed. Rather less than a dozen species in all are 
recognized, of which three occur in Africa south of the Sahara, two of them being 
found withui our limits. 

Foxes are small to moderate-sized carnivores, having a long, soft, dense, often ver\- 
attractive coat and an outstandingly bushy tail — termed in hunting circles the "brush". 
This profusely haired tail in conjunction with a sharp face and prominent ears makes it 
difficult to differentiate foxes precisely from some of the typical (wild) dogs; and, 
indeed, this genus has often been synonymized with Canis. However, in general, 
ill Africa, foxes are smaller and have shorter legs — they are much more able than dogs 
to slink along with their bellies almost in contact with the ground. Nevertheless, 
despite shortness of limb they can move verv fast over long distances. A more positive 
character dividing the foxes from the t\'pical dogs is that the tail in the tormcr measures 
more than half the length of head & body, whereas in the latter the reverse is the 
case. In foxes the pupil of the eye is a vertical oval, rather in the style of a cat but 
with not so wide a range of expansion and contraction. Some, if not all, foxes have .1 
scent gland on the top side of the tail about 50 to 60 mm beyond the root; this organ 
exists, according to Anderson (1902), in at least one of our West African species. There 
are five clawed digits on the forefoot but only four on the hindfoot; the soles of the 
feet in the African species are abimdautiv hairy between the pads, riieppclli vet more so 
than piillidd. 

Skull (fig. n). No general description ot the I'lilpa skull is given here. Its form 
bears a close overall resemblance to that of Cniiis, differing, apart from a considerable 
disparity of size, in one minor point. This concerns the supraorbital ridges which in 
Cauls are dorsally smootlily convex, whereas in Viilpcs there is a slight concavity in 
the upper surface. It is this lack of any important cranial or dental disagreement that 
has made taxonomists sometimes doubtful of the validity of generic distinction between 
1 'iilpcs and Civiis. 

Habits. Little enough is known of the lives ot the majority ot species apart from the 
fact that the palaearctic common red fox has been intensively observed and to some 
degree studied over a long period of time. This species is popularly regarded as exhibit- 
ing in its appearance and behaviour the acme of slyness and cunning, and its name has 
passed proverbially into European languages as the most apt and succinct expression 
cf thc^e attributes. It has, indisputably, other more admirable qualities: courage, 
determination, endurance. It is the combination ot all these that has reiideicd this fox 



VULPES 65 

possibly the favourite and most exciting animal of the chase. And it is these charac- 
teristics, too, that have in large measure enabled these harmful predators to survive in 
considerable numbers in the face of human expansion and of intensive organized 
hunting over many centuries, when wolves, bears and other carnivores antagonistic to 
man's interests have tended towards extermination. In what degree, if at all, this 
craftiness and skill in self-preservation are shared by African species is unrecorded. 
But it is perhaps w'orth noting, from an historical point of view, that a Hunt in semi- 
English style existed some years ago at Zaria (Nigeria), its activities being directed 
against the sand fox inhabiting the neighbouring fields and hillsides in fair numbers. 
There is good reason to suppose that proximity to man so far from being distasteful is 
welcomed by foxes since any occasional danger arising from it is more than com- 
pensated by the ready availability for food of domestic animals and birds, more con- 
centrated in amoimt than wild prey and less apt at defending themselves. 

Most foxes arc in large part nocturnal, but they are also commonly active during 
daylight hours, especially in the early mormng or late evening. So far as is known, 
all species of Vidpcs shelter and breed m self-constructed "earths" in the ground or, 
much more rarely, in the protection of holes formed by rocks. Some live only in small 
family units ; others are more gregarious. Beyond these generalities it is difficult to go ; 
for nothing appears to have been recorded for African species regarding the various 
aspects of breeding, and little about feeding habits. Concerning these latter, it may be 
safely assumed that African foxes take a wide diet, not only of flesh but of fruits and 
other vegetable matter as well. Their enemies are the more powerful carnivores, 
birds of prey, the larger snakes — and, of course, diseases. These last, and their 
possible impact upon man, have not been investigated for tropical African foxes. 

Taxonomy. It has already been mentioned that opinion has differed in the past, 
from Luinaeus onwards, as to whether the foxes constitute a separate genus or should 
be regarded as wholly one with Caiiis; but the two are now generally looked upon as 
validly separable. There are no problems with regard to African species, though there 
has been at times a little confusion of thought — vide Thomas (1918), who cleared up 
points relating to nicppdli and pallida but fell into the trap of supposing that Gray's 
I'ldpcs dorsalis from Senegal was indeed a fox, whereas, as pointed out by Ellerman & 
Morrison-Scott (195 1), it was in fact a jackal. 

The two species that occur within our limits may be separated thus: 

KEY TO THE WEST AFRICAN SPECIES OF VULPES 
(previous key page 3 5) 

Tail over 300 mm, tip white; dorsal pelage with a bright reddish spinal band 
flanked with greyish; cars over 80 mm long; breadth across the outside ot 
the upper cheekteeth over 30 mm; c — • 11 fi over 45 mm rueppelli {pa^c 66) 

Tail under 300 mm, tip black; dorsal pelage darker medially but not bright red or 
flanked with grey; ears under 80 mm long; breadth across the upper cheek- 
teeth under 30 mm; c — • 01- under 45 mm . . . pallida {pa^c 70) 



66 I 1 1 I < A 11 M \ ( Ml I s 1 1 1 W I s 1 M in I A 

VULFES RUEPPELLI (Schm?) Kiipix-ll's Fox 

(.',iM/s nippclii [sn] .Scliin/. 1^:15, C'liriV; '.v 'I'liiirnitli . . . Am ilciii iV(iH^i).(i.(i//i'/i /rq' ubcncl:! . . . , 4: 
S08. Ootigola. tdiiard Riippi-ll, attcr whom this was named, explored and collected in various parrs 
ol norrli-eastcrn Africa tor the Senckenberg Nature-research Societv, Frankfort-on-Main. 

I'lViif fntitiliatf Cretzschmar, iSzy, in Riippell's Atlns ch i/cr Rcise iiii ihn<tUilicii Ajriha, Saiigethiere (dated 
1S2O): 15; pi. 5. Nubian desert and Kordotan. The name fdiiitliiiis is the Latin word tor hnngrv 01 
starred, given, apparently, in reference to the slender body and thin legs. 

Caiiii sablmr Hcinprich &: Ehrenberg, 1S33 fide a MS. note by Shcrboni, Syiuholac Physicac, sfii hoius 
ct Dcscripiiowci Mdiiiiiinlfiiiii .... dccas secunda, folio fi. Dongola. This was a vernacular name. (Not 
a synonym ot r. pallida, as given in G. M. Allen, I93y). 

I 'h//vs nippclii cnf'iti, Thomas & Hinton, 1921, Noi'il. zocl. 28: 5. South side ol Mt. Bague<!aii, 1000 
metres. Type in the British Museum, No. 21. 2. 11. 26, J; skin good, and skull m good condition 
except tor the left zygoma partly missing. The subspecitic name is the Latin tor bluish-grey and rcters 
to the dorsal pelage. Ot doiibtfiil v.ilidit\ . 

Distribution and general. RiippcH's tox (Plate 1) occurs from the Sahara in tlic 
west (Air, HoL;L;ar niouiitains, Tibcsti) to Libx'a and Eygpt in tlie north ot the con- 
tinent; and cast\vardl\' in Africa to the shores of the Red Sea, and south to Bcrbera 
on the Gult of Aden. From the Red Sea it ranches across t]ie Arabian pcinnsuhi north 
to the Dead Sea and Iraq, and cast as far as Baluchistan and Atgli.iiiistaji. Throuohout 
this wide area it confuies itseh to arid sites, ohen rocky or stoiiv rather than sandv — 
the type ot desertic terrain vernacularlv termed Haiihulii. 

In suitable locahtics these small foxes, with a head & body length ot some 450 mm 
and tail about tlorce-quarters of this, are tairly abundant. In fact. Fetter (1952) charac- 
terizes them as one of the commonest carnivores ot the Sahara. The species is plentifully 
represented in museum collections, though there are rarcK' surticient specimens from 
anv one locality or area to furnish adequate data. 

Description. The overall impression of the coat ot Riippell's tox (Plate 1) is that 
ot ,1 speckled grev back with a broader or narrower not very clearly delimited speckled 
pale red band d<iwn the spine, usually trom the neck to the root of the tail. The flanks 
are more creamy and, generally, relativeh' lightK' speckled. This creamy colour 
encroaches towards the red median band in two patches just behind the slioulders. 
The luiderparts, w itli the exception ot cases noted later, are piux white. 

This change trom a white belly, tlirough grey, to a red dorsum is brought about 
in the f<illownig \\av. As 111 other members of the family the pelage consists ot a dense 
mass ot very tuie, long and soft imderfur amongst which are stout bristle-hairs, tairly 
wideK' scattered. In the majority ot specimens the pelage below, both undertur and 
bristle-halr^. is pure white. On the Hanks, the underfur in some specimens remains 
pure white; in others, even from the same area, it starts to become darkly pigmented; 
Init 111 all cases some ot the bristle-hairs develop a few niillimetres ot black tip, and 
some become black throughout their length. This accords to the flanks a varyitig 
degree ot speckled greyness. Moving towards the mid-dorsal line the luidertiir acquires 
.1 deep mauvy-grey colour; and final] \-, m the spinal region, the terminal portion 
.issumes a bright rufous colour. The underfur here is about 2<, mm long. The bristles, 
about 3s to 40 mil) long, also become dark basalK' but iiearK- all have a pure white 



Plate 




Sand Fox (I'liZ/x-s pallida): Fcnncc (Famvais zcrda): Riippi-ll's F<ix (I'ulpes rucppclli) 



VUIPES 67 

subtcniimal liand and a narrow black tip. It is these wliitc subtcrminal zones, cacli 
about 5 mm wide, tliat come to lie together and give the coat its highly frosted appear- 
ance. The bristle ends are subject to wear and to snapping, and when this happens the 
degree ot frosting is corrcspondingh' less marked. 

Two specimens from the Hoggar Moiuitauis have the belly and lower chest a very 
striking, intense mauvy-rcd, the pelage where this occurs being, in comparison with 
more normal specimens, both scanty and peculiarly close-lying. The sanic feature, 
but in a very much paler torm, is found also in a specimen from Dongola (Sudan), 
B.M. No. 32.7.3.1. The reason for this is not apparent. It does not seem to be moult 
and the colour bears no resemblance to anything occurrii^g elsewhere in the coat. 

The neck and the crown of the head are pale reddish or sometimes buffy, the same 
colour extending down the front of the face to the rhinarium, being bordered by two 
very conspicuous dark brown bands descending trom the eyes to just forward of the 
middle of the upper lips — which are, themselves, white. Something of the same dark 
colour actuall)' encircles the eyes. The cars are pale red or bufty-red on their backs 
and have long white hairs around the margins of their inner sides. The)' are pointed 
and much larger than those of I'tdpcs pallida. The legs are pale red on their outer 
aspects, white inside. The extremely bushy tail is a mixture of buff and black-tipped 
hairs about 50 mm long, the total amount of black varying considerabK' with different 
specimens. The tip is always white. There is a gland on its dorsal surface, not far from 
the root, which according to Anderson (1902) clogs the surrounding fur with a yellow 
secretion that gives off an aromatic odour. 

Skull (fig. S). This is like a small version ofCaiiis adiistus without such well-developed 
occipital crests. The profile is much the same, having only a slight dip over the nasals 
to break the otherwise flat outline. The rostrum is relatively more slender. The brain- 
case is rounded; the postorbital processes arc blunt and narrow, their upper surface 
having the slight depression that differentiates I'ulpes from Canis. There is very little 
constriaion between the orbits. The zygomatic arches are strong, almost semicircularly 
upciirvcd in the maimer of C. aureus. The bullae are relatively large; the palate usually 
terminates at about the mid-length o(\ifi or sometimes anterior to this. There is nothing 
remarkable in the dentition; but the camassials arc of greater size in this species than 
in pallida, as Table 3 on page 74 demonstrates. 

Habits. Little is biown of this fox in the field. It has already been said that riieppelli 
often affects a more rocky, hilly terrain than pallida does. Buchanan noted on the 
label of the cacsia type from Moiuit Baguezan that the species is locally held to occur 
only ill the mountains and is never heard of at Agades in the plains a little further south. 
Petter (1952), however, writes ot a specimen captured in the region ot Agades. An 
example from Somalia is labelled as having been trapped on a hill; but the only other 
collector's note contradicts the idea of rucppclli being purely a fox of stony ground 
since the relevant specimen (from Wadi el Natrun near Cairo) was dug from a burrow 
in the sand. Indeed, it is known that the species does occur hi the same areas as the sand- 
loving fennec. RiippcU's fox is reputed to be largely nocturnal ; but like all foxes it is 
not infrequently to be seen on the move during the early and late hours of daylight. 
Petter (1952) appears to be the only author to have given some accoimt of the 



6S 



Till ( AKNivours oi wrsr Ariiu a 




Ik., s. I'lilprf iiirpprlli: skull, T\pc of iyji'.(M, B.M. No. 21.2.I i.2('i, j, -■', 



VULPES 69 

living animal, chiefly in regard to a captive specimen in Paris. He noted it as having 
exceedingly keen sight and hearing, able to detect movement or an luiusual sound at 
over 100 metres. This captive animal became very attached to the human beings with 
whom it had had close relations from an early age. It remained playful with advancing 
years; its joy at visits from well-known friends was expressed in sonorous cackles and 
long moans; it wagged its tail rapidly and would lie down for caresses with its ears 
laid back. On the other hand, with strangers it would growl and retreat, utter short 
barks and snap. Other specimens have, indeed, been known to possess quite different 
temperaments from this animal, being either aggressive or wholly indifferent. 

Fetter's fox thrived on a mixed diet, flesh and vegetable, but it never drank. This 
would accord wdth the sort of tiling to be expected in the free state, where gerbils, 
insects, roots and fruits would be the kind of food taken, and where the opportunity 
to drink would be extremely rare. The captive animal only occasionally actually ate 
out of its dish; it preferred to fill its mouth with a large quantity of food which it then 
carried away and iiid in its beddijig straw, returning to repeat the action once or twice 
before settling down to actual eating. As in other canids it went tluough the motions 
of scratching earth over what it had hidden even though this, on wood, carpet or 
cement, was in fact pointless. If given live food it would play with it for a long time, 
over an hour, without injuring it. 

Riippell's fox does not have the ever-present strong, objectionable odour of the 
common red fox of the palacarctic regions; but it nevertheless has scent glands. Should 
anything threaten to rob it of food that it is in the process of consuming it presents its 
hindquarters to the presumed aggressor, makes small menacing noises, lowers it head 
and arches its back, separates its hindlegs, which are stretched to their maximum, holds 
up its tail and ejaculates towards the unwelcome visitor a malodorous secretion from 
the anal glands. 

Almost nothing is known about breeding, the preferred period of the year, the 
duration of gestation or the size of the litter. Anderson (1902) records, for Egypt, 
a female with 3 newly bom yoiuig wliich had their eyes and ears closed. This species 
does not appear to be so gregarious as pallida. 

Taxonomy. Halt-a-dozen races have been described. If extremes are selected — as 
an exceptionally dark skin from Somalia, B.M. No. 97.X.9.9, and a particularlv red 
one from Arabia, B.M. No. 52.i487^therc is a very marked difference of impression. 
Yet, this apart and despite the animal's wide east-westerlv range, there is over all a 
considerable uniformity of appearance. Speckle-coated tropical mammals very fre- 
quently display appreciable variations of tone and pattern amongst the members of a 
single local population; and, given sufficient material, it is nearly always possible to 
pick out an individual which exliibits such variations to a degree that other specimens 
from the area demonstrate to be not properly representative. This is so in rneppelli. 
In complete contrast to the two very diverse skins just cited it would, without knowing 
their widely separated provenance, be difficult to fuid any significant distinction between 
B.M. No. 86. 10. 15. 4 from Afghanistan and B.M. No. 34.8.2.10 from the Hoggar 
Mountains in the Sahara. The latter, together with two more from this area, have 
white undersides, while two others of the same series, all collected within 10 davs, 



70 THE CARNIVORES Of WEST AFRICA 

and irrespective nt altitude, liave intense red bellies. In view ot these facts it is by no 
ine;ms sure that any real value is to be found in the attaclmient ot subspccific names in 
nicppclli. 

In regard to the reputed West Atneau race aicsiti, Buchanan collected four skins 
within a limited area, about 25 miles across, on and around Mount Baguezaii, all at 
about 1000 metres ui Air. One ot them is badly m moult (end of July) and need not be 
considered. The remainder differ amongst themselves. The type, said by Thomas & 
Hinton to be "distinguishable by the nearly complete disappearance of the ochraccous 
on the back, and its consequently more wholly grey colour", stands in this respect 
alone, the other two behig good matches for the majority of specimens from Sudan 
to the east and the Hoggar Mountains to the north. The type not only has the breadth 
of the dorsal band reduced but the tone of its red also appreciably deeper than is usual; 
and the skin, in fact, very exactly matches that dark specimen from Somalia mentioned 
in an earlier paragraph. The type of cacsia would therefore seem to be nothing more 
than an idiosyncratic exception and not, in respect ot its lesser area ot red, typical of a 
distinct race. It any case were to be made out for separating West African specimens 
racially it would seem to be more likelv to stand on their rather more silvery appear- 
ance. But even this is marginal. 

The table of measurements on page 74, admittedly derived from a most inadequate 
number of specimens, indicates that there is very little difference of any significance 
between specimens ot nicppclli from various areas, with the possible exception ot those 
from north-east Africa, which may prove to be very slightly smaller. 

VULPES PALLIDA (Cretzschmar) Sand Fox 

Canis palliJiis Cretzschmar, 1827, in lliippeH's Athis zu dtr Ri'i^c im uonlliihui AJrika, S.iugcthiere (dated 

1S26): 33, pi. II. Kordofan. The Latin pallidus means pale, referring to the light-coloured peLige. 
I'ldpcs edwardsi Rochcbrunc, 1883, Bull. Soc. pliiloiii., P{tris (7) 7: 8. Cayor, Oualo, Gandiole, in forest 

ot gum-trees (Acacia), Senegamhia. This was named alter Professor Alphonse Milne-Hdwards of the 

Paris Natural History Museum. 
Caiiis (Cyimlopex) pailidus oertzctii Matschie, 1910, Sber. Gts. nciturj. Fnniuic Bit/.: 370-371. Dikwa, 

Cameroun. Type, No. A. 165, 10,1, ,^, in the Berlin Zoologisches Museum. Named .after its collector 

Lieutenant von Oertzen. 
Vulpes pallida haitcrti Thomas & Hinton, 1921, Novil. zool., 28: 4. Takoukout, Damergou (Niger), 

1550 feet. Type in the British Museum, No. 21.2. 11.23, o. skin, good, and skull, good except that 

H13 is missing on both sides. Called after Dr. E. Hartert who collected for Lord Rothschild's museum 

at Tring, Englind. 
Canis pallidas (Mivart) Riippell, subspecies /mdv/i Zimata, iy3s, Slxi. Akiul. ll'isi. llitu, 144 (i): y. 

Gourma-B.ihrus. A lapsus calami for the above. 

Distribution and generaL The sand fox, sometimes called the pale fox or pallid 
fox (Plate i), is roughly the same size as Riippell's fox but has a relatively shorter, 
black-tipped tail and a smaller ear. The two alternative names, though apt in regard 
to Subdcsert specimens, are not properly applicable to examples from moister biotopcs 
where the pelage assumes deeper hues; and the term sand fox is also not altogether truly 
descriptive since this animal does not confuie itself entireK- to sandy places but is 



VULPES 71 

fouiid also, some distance south of the Sahara, on other soils. However, it demands a 
soft, friable soil rather than a stony one and thus differs from Riippell's fox. V. pallida 
is not of such wide distribution as nicppelli, being confmcd to the Saharan and sub- 
Saharan region of Africa between about 10° and 20° north, but having a 
considerable trans-continental range from Senegal to the Red Sea. It is most typically 
an inhabitant of the very dry zones, the Subdesert, Sahel and Sudan types of vegetation, 
but it also occurs, apparently more rarely, in the somewhat moistcr Doka woodland. 
West African specimens exist in the British Museum from Takoukout, Agadcs and 
Zindcr (all Niger); and Gonibe, Farniso, Kontagora and Zaria (all Nigeria). A. J. 
Hopson observes [in litt.) that the sand fox is common near Lake Chad and numerous 
colonies of up to 30 burrows are to be seen in Acacia tortilis var. raddiana woodland 
[i.e. Sahel) usually on the crests of sand dunes. Elsewhere it has been recorded from 
Kaycs (Senegal), near Timbuctu and Gourma Rharus (Mali); and the neighbourhood 
of Fort Archambault (Chad). In the correct localities it is a fairly common animal. 
The species has usually been divided into five races; these are discussed later. 

Description. By comparison with the strikingly coloured nieppclli this is a relatively 
dull, though elegant, little fox, with its fmely speckled sandy coat and black-tipped 
tail (Plate i). There is an appreciable range of colour from creamy-buff to pale red; 
but this will be dealt with in more detail later, the present section concerning itself 
only with a broad general description of the pallida pelage. All specimens have a 
broader or narrower, ill-defined, band of deeper tone riuming from neck to tail. The 
pelage is composed of the usual mixture of abundant fmc long underfur ajid scattered 
stouter bristle-hairs. The former is from 15 to 20 mm long, the latter from 25 to 
30 mm. In broad terms, the imderfur is white-based; it then has a paler or deeper 
sepia zone, becoming finally in the terminal third huffish or reddish. The bristles are 
white-based, then medium-brown, succeeded by a brilliantly pure wliite band, which 
passes through a progressively deepening red-brown zone until it becomes, in most 
cases, black at the tip. It is the white bands and the dark tips of the bristle-hairs that 
impart the speckled pattern to the back. The intensity of colouring in the pigmented 
zones IS considerably variable, giving rise to the different overall impressions that have 
led to the erection of races. The pelage becomes paler on the flanks and passes in the 
majority of cases witliout any very sharp lijie of demarcation into a pale underside, 
generally pure white or creamy. 

The neck and the crowii of the head are much the same as the back; the backs of 
the ears slightly darker in tone. The cheeks are pale to a ver\' varying degree in different 
forms, pure white in those from the Sahel and Subdesert. The eye is surromided by a 
dark ring, broadest at the inner corner, and contmuLng thence to the lip, sometimes 
rather indistinctly. The muzzle is sharp. The legs are lighter or darker red-brown on 
the outside, white on the inner faces. The tail, which is not so bushy as in nieppclli 
is the same basic colour as the back flecked to a greater or lesser degree with long 
black tips, which form the usual black mark not far from the root and come together 
terminally to form a very pronounced black tip. The live weight of an adult is about 
i.\ to 2 kilos. 

Skull. This does not differ to any marked extent from that of nieppclli except that 



72 THE CARNlVORhS OP WEST AIRU'A 

the bullae arc sliL;luly larger and the nasals appreciably longer. The most pronounced 
dirterence between the two species lies in the dentition, hi piilliiia both carnassials top 
and bottom, are from i to 3 mm shorter than in nicppclli: and as the length ot the 
posterior cheekteeth is much the same in the two species, the ratios which the caruas- 
sials bear to these are considerably less. 

Habits. Although these little foxes are pretty common over a broad arid zone 
stretching tor 6000 kilometres across the continent from cast coast to west coast, 
little has been recorded of their way of life beyond the bare fact that they live a fairly 
social existence in soft open sandy wastes. Like all members ot the genus they shelter 
in burrows, but no one has given any detailed account of these. Nor are there any 
kno\\ii facts relating to breeding; but a specimen is noted as having lived for nearlv 
three years in captivity. Petter (1952) noticed that these foxes do not emit the usual 
foxy odour. Rothschild & Wollaston (m de Winton, 1901) observed that they swim 
well and readily. Anderson (1902) records that Withcrbv found skulls of the species 
in the nest of a kite, near Khartoum. 

Taxonomy. No question arises as to the standing of p,illiili and its complete inde- 
pendence from nicppelli or other species. The only point at issue is its subdivision into 
races. Tlirec of these have been attributed to West Africa: hartcrti from the Sahcl 
vegetation, cihiumhi from the Sudan woodland, and ocrt:ciii tor the Doka. The nominate 
race was described from Kordofan (Sudan), its exact provenance lying possibly in 
Subdesert, possibly in Sahel vegetation. 

The question of subspeciation in paUiih is, as so often, a ditiicult one. It is not that 
appreciable differences do not exist between animals from difterent localities; there is. 
for example, a wide discrepancy ot coloration between the type ot lunterti, 13. M. No. 
21. 2. II. 23, and skins from Gombe (Nigeria), B.M. Nos. 2.S.6.3.9 &: 11. The trouble 
lies in drawing clear lines of distinction and in the danger, in the face of a good deal of 
variation, of entering into a complexity ot rather meaningless nomenclature. In 
addition there is the usual paucity ot material from which to draw conclusions. Evalua- 
tion of pelage character is rendered a little more ditticult by the different setting of the 
skins, those that are flat, owing to the exposure of the flanks and part of the belly, 
giving a much paler impression than the colour of the actual dorsum w arrants. The 
table on page 74 shows that there is not a great deal to choose in size between the tew 
specimens of each reputed race; hartcrti, as represented solely bv the tvpe, seems to be 
a little smaller than the rest in nearly every respect. 

Before considering other races the first thing is to determine the appearance of 
tvpical pallida. No precise type locality was named bevond the general description 
"Kordofan" and there is nothing in the British Museum actually so labelled, the most 
likely specimens to correspond to the original locality being those from Shend)-, 
on the Nile below Kliartoum. They do, in fact, agree very closely with Cretzschmar's 
plate. They are well matched dorsally by others from Suakiii 300 miles to the north- 
e.ist on the Red Sea, both localities reputedly 111 Subdesert vegetation (I'/i/e the map in 
Keay i-t al., 1959). Anderson (1902) said that his specimens (presumably from Suakin) 
"agree exactiv with the type of the species in Frankfort". Turning now to West 
Africa, a skin from Agades (Niger; Sahel vegetation) corresponds ver\' closely to the 



VULPES 73 

Shcjidy material, though witli somewhat more white on the face. It might thus be 
regarded as representing typical pallida in West Africa. The type of harterti, from 
Takoukout, Sahcl zone some loo km to the south, is marginally paler and browner, 
but it seems scarcely worth while drawing a nominal distinction between the two. 
It is true that, as Thomas & Hinton pointed out, harterti seems to be smaller; but this 
conclusion is drawn from the measurements of a single example compared with equally 
or almost equally, inadequate data. 

The skins from further south, at Zinder (Niger, near the edge of the Sahel/Sudan 
zones) and Farniso (Nigeria, well into the Sudan zone) are a close match for one 
another dorsally, being of a slightly warmer red-brown than those from nearer the 
desert. But there are differences below. The flanks of the two Zinder skins are paler 
than in tliat from Faniiso, but one has a pale pinkish-ochraceous wash from throat to 
aims, the other being chiefly white with an extremely pale hint of pink in one axilla 
and over the belly. The Farniso underparts are greyish, but much of this is probably 
due to staining. 

In the past these slightly warmer-coloured skins have been assigned to edwardsi. 
Rochebriuic's description of this form is far from precise. Most of it would cover a 
range ot pallida specimens; but he does state that the coat is grey belov^'. This is unusual 
and only one specnncn in the British Museum could possibly fit such a character and 
that is the Farniso skin, in which the grey colour is possibly extraneous. The only 
two measurements given by Rochebrune, for length and for height, are ill-defmcd 
and have no great significance. On the strength of the description furnished, his sum- 
ming up of the case for edwardsi seems altogether too sweeping: "Confused with 
Caiiis pallida of Ruppell, this remarkable type differs from it in size, pelage and many 
anatomical characters" '. Whether the Zinder and Farniso animals really correspond 
to this Sahcl form from Senegal is at least open to doubt. 

There remains Matschie's ocrtzcni to consider. He described this as being "isabel" 
on the back; this is a somewhat disputed colour but usually reckoned as a dirty yellow 
and is depicted as such in Ridgway (1912). No British Museum specimen could be 
likened to this. However, Matschie described the underside as reddish, the throat 
rosy-white, in contrast to the usual piure white or near- white oi most pallida. Specimens 
from Gombe, Nigeria, some 370 km to the south-west and in the same vegetation 
zone (Sudan) as the type locality, and especially B.M. No. 28.6.3.9, might be held to 
correspond to this. The mid-dorsal pelage is very much that of the Zinder and Farniso 
skins but the flanks are more brown, and the underside quite different from that found 
in any other London example oi pallida. The overall impression is therefore that of a 
much more intensely coloured animal. Specimens from Kontagora and Zaria in the 
Doka zone of Nigeria have similar dorsal coloration but their bellies are mostly 
whitish. 

It will thus be 'iccn that there arc appreciable differences between the various existing 
West African specimens, and especially between the extremes from Subdesert and 
Doka. But it would be diflicult to say with conviction which, if any, correspond to 
present named races. Out of the 1 1 West African skins, in only three cases are there 
more than one example available from a given locality (Takoukout 3, Zinder and 



74 



THE CARNIVORES OE WEST AFRICA 



Gombc each 2) ; and these show at least slight differences from each other. It would 
appear that the existing specimens arc randomly taken samples from a clinc; and no 
matter what proliferation of races were proposed it would remain impossible to sav 
where one started and another tuiished. In these circumstances, and especially ui view of 
the very limited material available, it would seem that no practical end is served by 
attempting to categorize specimens into hard and fast subspecies. The best that can be 

Table 3 : Numerical data for species of Viilpes 

riieppelli pallula 





.§ 



















< 


Z 


1 

Sub- 


Sub- 


Sub- 


Sub- 




rt - — ^ 

op;,- 

"§ £-1 


" = 

2 . 2 

rt 


Vegetation 


5 


desert 


desert 


desert 


desert 


Sahel 


Sudan 


Doka 


Number in mean 


3 


3 


4 


6 


2 


I 


2 


I 


Condylobasal length 


98-5 


102-9 


105-5 


io6-o 


92-6 


91-4 


99-6 


101-2 


Basilar length 


91-4 


96-4 


loro 


98-8 


90-8 


85-0 


93-1 


93-8 


Palatilar length 


47-8 


50-9 


53-7 


52-0 


45-6 


43-8 


47-1 


46-9 


Zygomatic breadth 


53-3 


57-4 


5S-6 


ST7 


54-^ 


50-S 


55-5 


54-0 


Upper cheekteeth breadth 


30-9 


30-6 


31-4 


3I-I 


28-2 


26-7 


28-6 


27-7 


Nasals, length 


34-1 


36-0 


39-1 


38-4 


33-6 


30-8 


34-4 


31-7 


Interorbital breadth 


i8-8 


20-3 


20-2 


20-4 


19-5 


I7-I 


19-5 


17-2 


Postorbital constriction 


2I-I 


20-6 


19-1 


20-6 


21-5 


22-1 


21-4 


19-6 


Braincase breadth 


38-7 


39-1 


40-7 


40-6 


39-6 


39-3 


40-6 


39-4 


Toothrow (f — m^) 


46-1 


47-8 


50-4 


49-6 


43-4 


41-6 


42-2 


44-3 


p* length 


10-3 


10-2 


10-6 


10-5 


8-5 


7-7 


8-3 


8-8 


ml + m- length 


12-7 


I2-I 


13-0 


12-8 


II-6 


II-2 


12-2 


12-3 


mi length 


12-0 


II-6 


II-9 


I2-I 


(9-8) 


8-8 


9-9 


10-6 


'"2 + '"3 length 


8-0 


8-9 


9-4 


9-0 


9-0 


(8-6) 


9-4 


8-9 


Head & body 


41J 


456 


462 


47^ 


420 


406 


448 





Tail 


305 


3:^3 


358 


342 


270 


273 


286 


— 


Hindfoot 


97 


106 


113 


114 


108 


89 


100 


— 


Ear 


87 


90 


97 


95 


70 


64 


66 


— 


RATIOS (per cent) 


















Tail/head & body 


74 


71 


77 


7i 


64 


67 


64 


— 


Zygom. br./condylob. 1. 


54 


56 


56 


55 


59 


51^ 


56 


53 


Braincase/condylob. 1. 


39 


38 


39 


38 


43 


43 


41 


39 


Braincasc/zygom. br. 


73 


68 


69 


70 


73 


77 


73 


73 


Palatilar l./condylob. 1. 


49 


50 


51 


49 


49 


48 


47 


46 


Interorb./postorb. 


89 


99 


106 


100 


91 


77 


91 


s,s 


p-l/f - iifi 


22-4 


21-4 


2I-I 


21-2 


19-6 


18-5 


19-7 


19-8 


p^jm^ + iifl 


8l-I 


84-3 


81-5 


82-1 


73-2 


68-7 


68-1 


71-4 


mijmz + mz 


150 


130 


127 


134 


108 


102 


105 


no 



lYCAON 75 

done is to say of intensely coloured specimens with dark bellies "near oertzeni"; and to 
call pale Sahel and Subdescrt animals "p. pallitla"; cdwardsi with its reputedly grey belly 
would seem either not to exist as such or to be applicable to a wide cliiial range between 
the two extremes. 

Subfamily SIMOCYONINAE Zittel, 1893 

hi this Subfamily the sole African genus, Lycaoii, is rather doubtfully lumped together 
with two other living genera, Speotlws, a South American dog, and Ction, from Asia, 
to form a convenient taxonomic imit. Included with these is a large number of fossil 
genera, chief!)- of uiterest as uniting the Canidae with the Ursidae, the bears. These 
matters are not pursued here and we proceed at once to consideration of the only 
genus of immediate relevance to West Africa. The name Simocyoninae is derived 
from the Greek siinos flat nosed, and kyon, kynos a dog. It has no particular significance 
apart from the fact of the muzzle in Lycaon and Speotlws being somewhat less sharp 
than in the jackals, foxes and wolves. 

Genus LYCAON Brookes, 1827 
Himting Dogs 

Lycaon Brookes, 1827, in Griffith's Cuvier's Animal Kingdom 5: 151. Type species Lycaon tricolor Brookes 
(= Hyaena picta Temminck). Lycaon was a character in Greek m)thology who was transformed into 
a wolf. 

Cynhyaciia F. Cuvier, 1829, Diclioimairc des Sciences Naturelle 59: 454. Type species Hyaena picta Tem- 
minck. This name was formed from the Greek kyon, kynos dog coupled with the name Hyaena. 

Hyenoides Boitard, 1842, Le Jariin ties Planles: 215. Type species Hyaena picta Temminck. This is the 
Greek termination -oeides, from cidos form, implying resemblance, added to the name Hyaena. 

Hyaenoides Gervais, 1855, Histoire Naturelle des Mammiferes 2: 53. This is simply an academically 
more correct spelling of the above. 

The himting dogs are the sole African representatives of the Simoc^'oninae and are 
characterized in appearance bv their varicoloured, mottled coats; and in structure by 
their feet, both fore and hind, having only four digits. In the skull, the posterior upper 
molar [iti~) is much reduced in size, being smaller than the antero-extemal cusp of the 
caniassial (/)'') ; and the palate is relatively broader than in the true canids. 

Taxonomy. The rather doubtfully justifiable association of Lycaon with Speotlws 
and Cuoii has been referred to above; but though the relationship of these to each other 
is possibly not as close as a subfamily connexion would imply the group is a convenient 
one and could only be rationalized at the expense of creating a number of smaller not 
very meaningful imits. 

As regards the organization of the genus itself, Matschic (191 5) erected a great many 
independent species, three of which were applicable to West Africa. The basis of all 
these was very slender, resting almost exclusively on variations in pelage pattern; 
and the modem view is that there is, in fact, only one valid species, pictiis, throughout 
Africa, Matschie's creations ranking, if they have any validir\' at all, as no more than 
races of this. 



76 THE CARNIVORES OT WF.bT AfUICA 

LYCAON PICTUS (Tcininmck) I kuitiiit; Dog 

llyiiiiui piciii Tciiimiiick, 1S20, .!»»/.» i;ih. Sii. pliys. Unix. 3: 54. Co.ist ot Mii,!.inibit|uc. The L.uin wcul 

I'iita mcms painted or oiii.ite ami was applied to tliis animal because ot its varicoloured coat. 
Ilyaaici rciwlica Bnrchell, 1822, Trnrch in llic lulaior of Si'uiliciii .l/r'Vn i: 456; & 1^23, 2: 229. North- 
east ot Asbestos Range, Griqualand West, South Africa. The specific name is I atin .iiul means given 

to hunting. 
Cains {Lytaoii) tiUolcr Brookes, 1S27, in (intiith's Cnvkr' s Animal KingJofn 5: iji- C^ape of Good Hope. 

The specific name is a combination ot the Latin words tri- meaning three and in/cr the colour ot tlic 

skin, given ni reference to the black, white and yellow pelage. 
Lyiaon typicus A. Smith, 1S33. 5. Afr. Q. // 2: 91, (■— "Burchell's Lycaon"). I he specific lume is Latin 

for typical. 
Lycaon piclus sluiiiais Thomas & Wroughton, 1907, Ann. Mao. nat. lint. (7) 19: ili-\7<'- Mam (not 

Maui), lower Shari River (Chad). Type in the British Museum, No. 7.7.S.74, ']'; skiii and skull both 

in good condition. The racial name is a Latinized form ot the type locality. 
Lycaon tnanf;ncnsis Matschie, 1915, Slier. Ges. natnrj. l-rcniidc Bcti: 364-366. Near Djaniiaga, north ot 

Sansaniie Mangu, near the Oti River, Togo. The name is a Latinized form ot the town Mangu from 

near which the animal came. 
L}'rrt(i» inischliilii Matschie, 191 5, Shcr. nmnrj. rrcuiulc, Bcii: }t<i>-\iv>. Northern part ot the Kete Kr.uhi 

district, Togo, probably near Bimbila. between the Oti and Oak.i Uivcrs. This was called after Us 

collector. Professor Mischlich, a German government oHicial. 
Lycaon chcrnmicri Matschie, 1915, Shcr. Cos. naturf. I'rcnmic, Boil.: 369-371. Lake Chad area, probably 

near Dikwa. This was named in honour ot the German Governor ot the Kamerun, Karl Ebermaier, 

who sent the living animal to the Berlin Zoo. 
There is a further extensive 20th century synonymy, mostly erected b) M.itschie and mostly applying 

to southern .and eastern Africa. 

Distribution and general. The luniting dog, somctmus too rcstnctcdk, ,ind 
ccrtainlv today inislLMdiiiglv, called the Cape luiiuiiig dog, is eoiitmed to Atrica hut 
has a very wide rajige throughout much ot the coiuiiieiit trom the Sahara to South 
Africa. In the former of these it is generally regarded as prohably little more than an 
occasional rare visitor; hut Heim de Balsac (1936) thinks there is good evidence in 
support of a distinct Saharan race. Two specimens \\ere captured north ot In Ouzzal 
(Adrar des Itoras) in 1927 (Lhote 1946); and it is said to occur m Tanezroutt. Tihesti 
and Enncdi. 

South ot the Sahara it is an .uninal ot the more open kinds ot grass-woodland but is 
erratic in its occurrence, the more so todav as its numlicrs are, in general, becoming 
ever less and less owing to the widening settlement and use ot the land by man and 
the consequent reduction ot available food in the torm ot antelopes. A century ago it 
might at times be seen to hunt hi packs ot up to a hundred, or possibly even more, and 
those of thirty to forty were common. Now, and especially in West Africa, the packs 
tend to be smaller, sometimes only six or eigin, and often less; but R. H. Kemp, 
of the Nigerian Federal Forest Kescarch Department, saw a pack of at least 30 in 
April 1969 in Borgu (extreme western Nigeria), halt ot them pups estim.ited to be 
five or six months old (personal communication); and \Mth general g.ime protiction 
in this Reserve it seems possible that it is there becoming again more numerous. 
There arc records of occurrence from the north ot the lvor\- Coast, the upper Volta 
area, and across this belt to Sudan. Eritrea and Somali.i. Thence the species ranges 



LYCAON 



77 



^^ 




78 "I HE CARNIVORI-S OI WEST A H) H. A 

down tlic eastern halt of the continent south to Transvaal and Portugese East Africa 
and across to Angohi and Soutli-west Africa. South of tlie Tropic of Capricorn, where 
it tornierly d\\ eh in large numbers, it is now absent or only a very occasional wanderer. 
In the British Museum there exist trom West Africa only one complete specimen, 
that trom Mani (Lower Shari River, Chad), and two partial specimens, the one an 
unmeasured skin trom Leri-u-duchi (Zaria, Nigeria), the other a skull from Lake Chad. 

The hunting dog prctcrs open coimtr;,- where vision and pursuit of quarry arc not 
much obstructed. It is therefore most commonK' to be foiuid in the Sudan and Sahcl 
woodlands where the grass is tairly short and relatively sparse; and though it certainly 
occurs in the Guinea woodland this is probably for the most part during the dry-season 
when, as a result ot the annual grass burn, this type of country becomes very open. 
During the rains, on the other hand, the ground cover in most places in this zone is 
both too dense and too high — 2 metres or more — to make hunting in the manner 
ot these dogs a practical possibility. This applies in a somewhat lesser degree to the 
Ooka zone. 

Description. Lyciuvi (tig. 9) is easily the bulkiest and most solidly built of West 
African canids, standing 61 to 66 cm at the shoulder, with a head & body length ot 
about 1 14 cm and a bushy tail of 35-5 to 3S cm. The weight of a tully grown animal 
would be some 27 to 32 kg; but the average in a piack, is probably more in the nature 
ot 20 kg. The Himting Dog is quite immistakeable with its short coat, blotched with 
black, yellow and (often) white, and its very conspicuouslv large and rounded ears. 
Though quite obviously a dog, it is, by the teaturcs just enumerated, very readily 
distinguisliable from the jackals, ^vhich are much smaller and have pointed cars. 

The pelage is harsh, of moderate length, close-lying and not at all dense; and it is, 
thus, vastly ditferent from that typical ot the jackals and foxes. It is, indeed, of entirely 
different composition since it consists solely of stiff, terete, bristle-hairs and has no 
underfur whatsoever. These bristles are, by comparison with the closely packed fur 
ot the foxes, relatively widely spaced at their roots and emerge from the hair follicles 
mostly in pairs, occasionally in tltrees, rarely singly. Instead of the colour ajinulation 
common in the pelage of other canids these bristles are to all intents and purposes 
imicolorous trom root to tip, pure white, blackish or pale yellowish-brown. These 
tints occur over sharply dcfuied areas of the skin and result in a broad mottling of the 
pelage with little ijitergraciing of one patch with the next, the fur thus exhibiting 
nothing of the intimate "pepper and salt" mixture common to so many mammalian 
coats. The proportion of one colour to the others varies very considerably, as do their 
relative positions on the body. Matschic (19T5) tried to establish fixed patterns in 
relation to various geograpihical areas, the basis of his proposed specific naming. Not 
only were the data on which this sup>position was foiuided mostly exiguous, but also, 
where in the British Museum collection more than one specimen exists from a given 
locality, constancy of pattern is seen to be non-existent. This is amply confirmed by 
photographs of packs taken bv R. H. Kemp in the Borgu Game Reserve, Northern 
Nigeria, (Howell. 196SI, and private comiiuinication) ; and also by those illustrating 
Kiihme (1965b). 

The chest and belK- arc motllcd 111 the m. inner ot the b.ick but the hair is considerably 



LYCAON 79 

shorter and yet more scanty, scarcely obscuring the skin. In the only two West African 
specimens available for study the chin and under the jaw are black; the throat and 
upper part of the chest white. The back of the neck and the top of the head to the 
forward level of the eyes are pale brown; and a broadening black stripe riuis medially 
through this region over the crown and down the centre ot the face to join the all- 
black muzzle. The posterior part of this line is sometimes faint or lacking. The muzzle 
itself is short, broad and heavy; the nostrils are well-separated and widely open. The 
large, rounded, cars, about 115 to 125 mm long, are black on their backs and to some 
extent on their anterior faces as well, though here there is also a white tuft situated 
over the lower part of the inner edge. 

The legs are variegated with the same colours as the back. In this subfamily there 
are only four toes on each foot, the short claws being deep and powerful. The feet 
themselves are strongly built. The two middle toes are widely separated but joijied by a 
narrow naked web (Pocock, 1914b). The proximal margins of the digital pads carry 
long fringes of stiff reddish hairs; but the rest of the sole between these and the plantar 
pad is scantily clad. The distance between the plantar and carpal pad is luiusually long 
and narrow. The hindfect are similar but narrower. The tail is short-haired in its basal 
third, and then has a longish-haired black and white bush, the terminal quarter being 
conspicuously piure white. The females usually have about 1 2 teats. 

Skull (figs 10 and u). The skull is very powerfully built, with strong ridges and 
arches and a massive dentition. As canid skulls in general go it is short and broad. The 
braincase, as in other members of the famih', seems unexpectedly small ; it is romided 
but carries a pronomiced, sharp sagittal crest which joins an equally prominent supra- 
occipital crest to form a backwardly-projecting pyramidal "helmet". The interorbital 
region is slightly constricted, the pointed, triangular, do\\Tiwardly curving postorbital 
processes being wcU-dcveloped, their hind margins continuing and converging post- 
eriorly across the temporal region in a diamond outline to meet the forward end of the 
sagittal crest. The inuzzle is noticeably short, wide and deep, the anterior nares very 
open. 

The zygomatic arches are broad and strong, sharply upcurved, and since the pointed 
process on the upper margin of the jugal is not very distantly separated from the post- 
orbital process the orbit goes some way towards complete encirclement, but not quite 
so near as in the hyaenas. The palate is broad tliroughout, being particularly so between 
the camassials, and shows little of that marked anterior jiarrowing between the pre- 
molars and canines that gives the more typical dogs their sharply pointed muzzles. 
Posteriorly the palate comes to an etid about the level of the hind margin o( in~. The 
bullae, though prominent, can not, in comparison with the size of the skull and their 
relative development in the jackals and foxes, be characterized as large. The mandible 
is robustly built, with deep, solid rami and strong coronoid, condvlar and angular 
processes. 

The dentition is powertul, all the teeth being relatively broad and strongly cusped. 
But the posterior upper molar, in-, is much reduced in size and is less in bulk than the 
antero-extemal cusp of the caniassial, p*. As in the other West African canids the 
cheekteeth are f^. The outer upper incisor (i^) is much enlarged and caniiriform in 



So 



till ( AIIM \ lUil S ol WIS! M UK A 




Fig. 10. Lyiium piain: skull, r\pc.' >>( iliaiim^, li.M. No, 7.7.,s,74, 



' ; l,\tcr.il view 



shape; the outer lower uicisor (/g) li also enlarged hut not to the same marked extent 
and not caiiniitorni but riattisli and bieusped hke the rest. 

Habits. The limiting dog lias, over a long period ot vears, attracted a good deal ot 
attention which has resulted m an ahiuidant recording ot the chief features of its mode 
of life. Some of its alleged habits, traditionally held and passed from writer to writer 
or by word of mouth from one naturalist to another, appear to have been directly 
contradicted by close and specialized stud\' of the species made in the field in recent 
vears. Tlie most prolonged and concentrated investigations have been carried out by 
Kiilime (i9''isa and b), Estes & Cloddard (lyfiy) and Goodall (1970), whose published 
observations contaiji a great deal of information relating to the daily and iiightK' 
habits of packs in the Serengeti plains and Ngorongoro crater (Tanzania), observed at 
close quarters over long periods. Much of what has been noted in these areas is doubtless, 
niiihitis nitilninlis. also applicable to West Africa. 

The interest and repute of hunting dogs rest, as their common English name indicates, 
on the \\a\ 111 which tlie\ secure their food; for, although sohtarv individuals arc 
from time to time observed, these di^gs generallv live a higlilv communal existence 
and hunt, skilfulK' and persistentb', m packs. These vary in number from an extended 
famiK' unit of .iboiit half-.i-dozen to a sizeable group ot 30 or so. Ainthiiig larger 
tlhiii this IS tiida\', even in East Africa, relatively rare. CertaiiiK' 111 West Africa few 
p.ieks of this magnitude are likel\- to be found, though one such has recently been 
observed, i'robably a dozen dogs might well be regarded as constituting a more 
normal pack for this region. C'oncrete information ot anv kind relating to the area 
covered bv this present work is, however, extremely scantw In East Africa packs have 
often been touiul to show .i pretlominance, 111 numbers, of males. 



LYCAON 



8l 




Fic. II. Lycai'ii picnis: skull, Tvpc of sliariiiif. B.M. No. 7.7.S.74, v, -< ' ; palatal &' dorsal views 



Hunting is normally engaged in regularly twice a day, in the early morning and 
late evening; but if the weather is deeply overcast, or for some other less apparent 
reason, it may as an exception be carried out during the intervening period. A chase 
has been known to be continued till after dark (Estcs & Goddard, 1967) ; and moonlight 
hunts have also been recorded; but such nocturnal events in ill-lit conditions are 
relatively rare since Lycaon hunts by sight rather than by smell. According to Kiihmc's 
observations it would seem that what is acceptable as a meal may vary from day to 
day, a change of diet apparently being desirable, readily available prey of a kind that 
was one day willingly taken being passed by the ne.xt in the search for other flesh. 
Estes & Goddard, however, found that something in the nature ot 69 per cent of the 
kills were Thomson's gazelle, and iS per cent juvenile wildebeest. What is pursued is 



S; Tin; ('AI!niv(ii(i:s or wkst m-iuca 

doubtk'ss dctcnnincd to sonii: cxiait b\' the size ot the pack, the more numerous tliis 
may be tlie greater tlie possibihty oi being able to overpower the larger sorts of ante- 
lopes. Li West Africa it is unlikely that anvthing as bulky as a waterbuck or roan would 
ever be tackled, though attempts would certainly be made on their calves. On two 
occasions U. H. Kemp has foiuid the kill in Nigeria to consist of a fully mature harte- 
beest in good condition. When a Land Rover stopped near one of these kills some of the 
pack made attempts to drag the hartebeest away and two dogs did, in fact, succeed in 
drawing it about 27 metres (personal commimication). Probably the main sources of 
food in West Africa are gazelles, reedbuck, cob and hartebeest; but bushbuck, oribi 
and crowned duiker, though less gregarious than the others, may doubtless also be 
taken. A bushbuck killed by hiuiting dogs was in fact observed by David Brown in 
the Borgu Game Reserve (Nigeria). Warthog are fronr time to time hunted. Hiuiting 
dogs are. mdeed, known to take much smaller fry such as cutting-grass [Thryoiioinys), 
giant Gambian rat {Cricctoinyf), hares [Lcpiis), ground squirrels [Eiixcnis) and so forth; 
but this they usually do as independenr individuals, not as a pack though others may 
be present and all on the move in search of more satisfying prey. 

A look must be taken at some of the long-held views regarding the method of 
hunting that have now been upset by close and continuous observation. Traditionally 
these dogs were believed to pursue their prey tirelessly and relentlessly over vast 
distances luitil the quarry faltered and fell from sheer odiaustion, the outcome of pro- 
longed terror and ph)sical effort. It \\ as even said that in the case of the larger and 
stronger antelopes such a chase might extend over as much as i s or 20 km and that 
during this the leading dog would be deliberately relieved from time to time by relays 
of others which were less fatigued. What in fact reaiU- happens is, to take the morning 
luuit, that soon after dawn the various dogs of the pack emerge from their sleeping 
quarters and wander slowly round defiecating and carrying out an extended greeting 
ceremony, each with ever\' other member of the pack. This consists of poking the 
nose into the corner of the other dog's mouth and of licking the lips and face. If all 
are not present a not unmusical bell-like call is uttered to assemble distant members. 
EventualK', some half an hour or so after daybreak the pack moves ofi at a trot in 
more or less open formation to look tor game. When suitable quarry has been sighted 
approach to it is made with some care, at a walk, in a slightly crouching attitude with 
head and tail both lowered. 

in the past it li.is been commoiiK' held that Lycaoii strikes far more terror tn the heart 
of the antelope population tli.iii does the presence of lions; and that not only would 
antelopes stampede at the first hint of danger but would actually entirely desert any 
district tainted by the mere existence of luuiting dogs (Bere, iys6). Nevertheless, that 
.uitelopes do not, in fact, always recognise their danger is shown m E. M. f-ang (1963) 
where a herd of gazelles is recorded as actually riuming inquisitively towards an 
approaching pack until they were no more than about 30 metres away before realizing 
that they were Ln grave peril. Kiihme, in the papers cited above, also found that antelope 
herds were cither not visibly wary of near-by hiuiting dogs or took flight only when 
an obviously actively hunting pack on the riui towards them was within some two or 
three hundred metres, thoui;h Estes & Goddard estimated the alarm distance in these 



LYCAON 83 

circumstances somewhat higher. It is, however, quite certain that a herd of antelopes 
instinctively knows when a pack means basincss, exhibiting no fear even though dogs 
may be at very close quarters but stationary, and but little alarm when a pack is merely 
walking or trotting. It is this that enables the dogs to approach fairly close before setting 
a herd in rapid retreat; or, in the calving season enables them to edge very near to a 
mother and her fawn before snatching the latter. 

Once the prey has been set in motion the dogs break into a run and follow, still in 
more or less open order. If more than a single antelope is being chased there appears 
to be no deliberate selection of any particular victim from the start, choice of that 
which becomes the ultimate prey being as much a matter of chance as an)thing, resolv- 
ing itself nito a question of the slowest member ot the herd. This is possibly a gravid 
female or the tardiest to abandon grazing tor flight, most frequently a male. At the 
start of a chase every dog follows that beast wliich is most directly in front of it, and 
hence nearest, but they tend to switch to any one that is obviously less distant from its 
aggressor, until ultimately the hunt nearly always resolves itself into the pursuit of a 
single animal. There is no deliberate taking over the task of leading dog in relays, as in 
the past supposed. If change takes place it is probably most often due to the fact that 
the quarry, failing to escape on a direct course, starts to circle; and whereas the fore- 
most dogs follow immediately on its heels others further in the rear can more easily 
follow the chord of the arc and so, pursiung a shorter hue, gain gromid over both the 
antelope and erstwhile leader. Yet this is not necessarily always the pattern; for Estes & 
Goddard (1967) found, in Ngorongoro, that the dominant dog of the pack selected 
the victim and led the pursuit. Nor does the hunt always resolve itself into a single kill. 
E. M. Lang (1963), for example, describes an incident where a pack divided into two. 

The search for prey is carried out at a trot of about 11 km an hour; actual pursuit 
is at a steady 48 km an hour, a rate which can be maintained by healthy adult dogs 
over long distances. Short spurts can if necessary be made at about 56 km an hour. 
Not all the pack can keep up with this, and laggards or young dogs may still be i or 
2 km in the rear at the time of the kill, not catching up till the leaders have already had 
5 minutes or so at demolishing the victim. So far from the dogs patiently and steadily 
rmuiing an antelope to exhaustion before killiiag it, as popularly supposed, Kiihme 
observed that the kill was preferably made as swiftly as possible. Prey given a leading 
start of 300 metres would be caught and overcome in a further 600 to 800 metres. 
This according to other observers is rather short. Estes & Goddard reckon the average 
chase to cover i.l to 3 km and to last 3 to 5 minutes; and Goodall found that the hunt, 
if not successful, was abandoned after 4 to 5 km; but he did, nevertheless, follow one 
that fruitlessly covered somewhat more than this. Since, moreover, this pack had 
travelled 8 km m the preliminary search for suitable prey before it actuallv started 
to hiuu it must in all have covered some 13 or 14 km. 

The idea that these dogs lope untiringly behind a fleeing antelope until it drops 
from exhaustion must also be abandoned. The antelope is not always captured; and if 
the chase goes on past a certain distance the dogs, as Kiihme observed, can get just as 
exhausted as their quarry. This is confirmed by Macintosh (1953) who writes of the 
fuiding of an e.xliausted gazelle, pantmg in the grass and just capable of staggering ofi". 



S4 Till 1 4, u s I \ (iHi s oi \i 1 s r Ai i(i( A 

and some iSO inctrci nr sn awav a hunting ilni; m a similar stati.-. In such cases the 
pack rests tor some s or lo minutes and then tonus up for a renewed Iruut. SimilarK'. 
it a successful hunt should tarn out to provide an insutlicient meal for all or part ot 
the pack, the late-arriving laggards tor example, then a second hunt is almost immed- 
iately entered upon, at least by those with unsatisfied appetites. 

Hiuiting is, to all intents and purposes. alwa\-s by sight, not smell, and it the quarr\', 
during the chase, is obscured by grass the hunting dogs make intermittent leaps to 
keep their prcv in view, the white tail tip at these moments becomnig very conspicuous 
to an onlooker. Opinions have diftered regarding whether these dogs liimt mute or 
not. This is dealt with later. Eventuallv, in the normal hunt the foremost dogs atter a 
while fuld themselves almost literalK- at the heels of their prev and tliev then rcpeatedlv 
jump t'orward to snap at the hindquarters, flanks or bellv. tearing away ribbons ot 
skill or lumps of flesh. At length the animal is brought to a standstill or knocked down 
and the dogs leap upon it. There is no attempt to kill it outright. It is either literalK 
torn apart bv seizing the limbs and tugging m opposite directions, or eaten alive, 
startinr; bv ripping open the bellv and devouring the entrails. It is, of course, impossible 
to sav for certain but there is some likelihood that the victim does not suffer so much 
111 this as might at first seem, being m a daze and a merciful state ot shock. All but the 
tousihest parts — heads, horns, hoofs, and sometimes skin and leg boties — generally 
disappear with astonishing rapidity. It has been mentioned above that not all hunts 
end m a kill. There are widelv diverse estimates of the success rate. Goodall (1970) 
records onlv 39 successful kills in 91 lunits. that is 43 per cent. Estes & Goddard (1967), 
on the other hand, analysing a much smaller number of hunts found a success rate ot 
over twice this. Ss per cent. 

However, success rate apart, hunts do not always result 111 a satistactor)' meal, tor 
other, stronger predators have to be taken into accoiuit. Lions are not above robbing 
hunting dogs of their legitimate kill, and thev are too powerful tor the dogs to be 
.ihle to do anything about it, even as a pack. Spotted hyaenas are another matter. 
These too. of course, are bigger and stronger than hunting dogs but they lack proper 
pack co-operation; so that the outcome ot attempts at superpredation b\- them depends 
ver\- largelv on the relative numbers ot the two contenders tor the meal. There is no 
doubt that hvaenas sometimes deliberately sit and watch hunting dogs, or w.mdcr in 
their vicinitv eating tlieir droppings, waiting for them, or at times even deliberately 
stimulating them, to set out on a himt, in which the\ then follow them. It the chase 
ends with onlv one or two dogs m at tlie kill the hyaenas can then leap m, frighten 
the dogs off and snatch a Itasty teed before the rest of the pack comes up and, b\- a 
reversal of power, establishes ownership. If the hvaenas do pluck up enough courage 
to sneak 111 again and seize part of the pre\- rlie\- are chased oft (E. M. Lang, 1963). 
.uid thev are reduced to waiting on the outskirts nf the feeding pack until the dogs have 
hiiishedand the\- can then safcK' close m and polish ofl the skin and bones or any other 
residue that ma\- be left. There is, indeed, an odd sort ot enmity between hiuitnig 
dotrs and spotted hvaenas. Sometimes the former tolerate the latter in close proximity; 
at other times thev delight in mobbing a single hyaena, biting fiercely at its buttocks 
so that ultiniatelv its onlv defence is to squat down and so protect its hindquarters 



LYCAON 85 

while snapping at dogs that come too near. But there never seems to be any question 
of the dogs, even as a pack, killing a hyaena or even doing it much visible harm; 
and it eventually slinks oft". Jackals, vultures and other birds of prey which almost 
invariably try to share in a kill arc driven off and made to wait, though golden jackals, 
in their nimble fashion, often manage to dash in from time to time and secure a mouth- 
ful here and there. 

Kiihme estimated that the pack he observed occupied a livijig area ot roughly 16 
square kilometres and hunted over an area of between 100 and 200 square kilometres. 
It will be appreciated that it is thus not normally easy or possible for a casual observer 
to follow a hunt at close quarters from its inception to its end, the more so as it is carried 
out at a fairlv high speed. Only if one chanced to be actually near at hand at the correct 
moment and suitably moimted was tliis possible in the past; and the fmer details of 
what actually takes place during pursuit were therefore witnessed by relatively few 
people. This accounts for a good deal of supposition, of inaccurate observation or 
wrong interpretation of a pack's behaviour. The problem is lessened today by the 
existence of motor vehicles that can travel safely and easily across country; and this is 
the method of observation employed by Kuhme and other modern field workers. 
Himting dogs are little disturbed by the presence of a vehicle at a reasonable distance. 
R. H. Kemp (personal communication) twice foimd these animals to pay little attention 
whatsoever to his Land Rover, coming towards it and getting out of its way very 
slowly or, when it stopped, lying down in front of it. One stood on its hindlegs to 
get a better view. Nevertheless, Macintosh (1953) records the attack by a pack on 
the tyres and wings of a slowly moving car. 

Lycaon, apart from its normally wide hunting range is, except at breedmg times, 
by nature a rover and in the course of a year may cover an extremely wide stretch of 
countrs'. It is driven to this wandering habit partly by the migrator)' nature of the 
antelopes on which it preys since these are pretty constantly on the move to fresh 
pastures. On the other hand, some limit may be set to unrestricted nomadism by the 
existence of other neighbouring packs, each claiming right over a more or less defmed 
area, though there seems to be appreciably less insistance on strict territories in hiuiting 
dogs than with many other predators. Anyhow, some degree of boundary observance 
has been noted in East Africa where the large antelope population is sufficient to 
support numerous Lycaon packs; but in much of West Africa the herds of antelopes 
arc comparatively small, the hunting dog being consequently of relatively rare occur- 
rence and the territories over which the few packs can roam correspondingly large and 
probably, for this reason, less well defmed. During three months of breeding and 
family upbringing the pattern of life is, of necessity, different since the interest of the 
entire pack is completely centred roimd the earth or earths m. which the pups are 
being raised. It therefore, apart from brief daily sorties after food, becomes something 
of a sedentary unit. Such a pattern postulates a set breeding season for all females. 
Kiihme's pack contained only two adult bitches, each with a recenth- born litter; 
it would be interesting to know what happens where there arc several adult bitches 
breeding at different times. 

It has already been mentioned that hunting dogs, though subsisting mainh' on 



S6 Tin; (,ABMVoui:s oi \\\%t amuca 

ai;tclopcs, apparently like to alternate tront one species to another; and that when 
these are not conveniently available they turn their attention to other kinds of flesh. 
Ir is virtually certain that nothing would come amiss to them so long as it was readily 
obtainable; and they certainly do not hesitate to eat the dead of their own kind, hi 
parts of Africa, and particularly in the past when they were in larger packs and of more 
common occurrence, they have shown themselves to be highly destructive of cattle, 
sheep and goats; .and they have themselves been relentlessly hunted and shot or killed 
with poisoned bait on this accoiuit. It was probably their depredations amongst 
domestic flocks that chiefly earned tor them the reputation ot ijidiscriniinately slaughter- 
ing tar more than they could possibly eat; but here again Kiihme's observations rmi 
coiuiter to this widely held belief since he foiuid that his small pack never killed more 
than was absolutely necessary to satisty tlie S adults and 15 whelps. When the Litter 
were very yoiuig and ate relatively sparingly the toragers regularly killed and brought 
home a single 30 kg antelope morning and evening; but when the pups had grown and 
become more dcmandijig they doubled their twice-daily kill. This figure is, of course, 
nothing more than a rough estimate; but at the end of the rearing period when all 
23 dogs might be assumed to eat practically adult meals this would work out at an 
average of something like 5 kg a day. On the other hand, Estes &: Goddard (1967), 
working on a pack numbering 2t, estimated that there was an average of 2-7 kg a 
day of meat available to each dog; this amoiuit would however vary with the size ot 
a pack between 2 and 4 kg, the smaller packs getting the larger anKiimt. It is interesting 
to compare these estimates with the figure given tor zoo animals of i.V kg ot meat 
daily (Dekker, 1965). Such captive specimens would, of course, get relatively little 
exercise compared with those hunting for themselves in the field and their demand 
would be correspondingly less. It may here be noted that these zoo specimens were 
given an occasional whole chicken or whole pigeon, mdicating that, in the wild, 
birds, such as francolins, may be included in the diet of Lycaoii. 

There is, however, some evidence in support ot the charge that these ajiimals kill 
more than they can eat. R. H. Kemp (personal communication) twice observed in 
Nigeria that the kill had scarcely been damaged except for one hindleg and the belly 
ripped open. At IS a.m. in the Borgu Reserve he once observed about a dozen animals 
which had just made a kill ; but the rallying hoot heard in the distance not only indicated 
that the pack was in fact larger but, strangely enough, served also to call away the 
dogs by ones and twos trom their newly killed hartebeest, \\'hich still remained un- 
touched at dusk. This must be luiusual; but the commonest expLuiation ot only 
partially eaten prey is probably that these corpses are the outcome ot second hiuits by 
imsatistied members ot a pack, too few in number to be able to consume all ot their 
kill. Hiuiting dogs, unlike many other carnivores, never feed a second time from the 
same kill. 

Kiihme never observed his pack to drink, and it is to be presumed that much ot the 
necessary water can be picked up trom blood or from licking their coats when wet 
with dew, or from like sources; tor it is known that these dogs can go tor long periods 
without drinking. But they certainly do drink on occasion. (loodall says that they will 
do so daily it possible, and quite clearly deliberateK' seek out water-holes to this end. 



LYCAON 87 

Apart from tlieir understandable fierceness in the pursuit and capture of tlieir 
necessary prey, these animals seem by nature to be of an essentially friendly and sociable 
disposition amongst themselves. They do not seem to engage in the fierce protection 
of boiuidaries practised by the spotted hyaenas; they arc not even aggressive towards 
other packs unless these approach too near to the breeding burrows (Goodall, 1970). 
There appear to be no records of any attack b)- hunting dogs on man ; and in fact, 
they seem in nature to regard human beings witli some indifference. They can, however, 
be successfully tamed but have an imwelcome drawback as a domestic pet in their 
strong odour. This is probably due to glandular exudation; but they have also attaching 
to them another smell that seems to permeate their very bones and persists for many 
years after their death. A cupboard containing thoroughly cleaned skulls of Lycaon 
has a particularly nauseating odour, shared, it is true, with lu'aenas and some other 
flesh-eaters. 

Amongst their own kijid these wild dogs are playful and rarely quarrel, and then 
only briefly. They share the same shelters, even at breeding time, and peacefully look 
after each other's young. They have, according to Kiilime (1965 a & b) developed a 
classless social system that involves the ready sharing of both food and labour. Such a 
view is diametrically opposed by Goodall (1970) who, while agreeing about the sharing 
of food and labour, fmds that there is a recognized social order not only throughout 
the pack as a whole but m each of the sexes separately as well. This conclusion was 
reached as the result of long and close observation of a single pack, and was derived 
trom the interpretation of submissive body postures. The pack was normally led in 
the hmit by an old male; but even when on one occasion this dog fell into fourth place 
he was still foimd to influence the direction taken by the leaders. Estes & Goddard 
(1967) found that there was a pack leader, cither male or female, but no other rank 
order. 

It has been noted earlier in this present work (on page 48) that Wyman (1967) 
observed a food-begging ritual amongst the jackals. An exactly similar ritual had 
previously been discovered in the hunting dogs by Kiihme. That is to say that on the 
arrival home of the toraging party those dogs that have remained on guard, and those 
young which are ot an age to require a full meat diet, excitedly thrust their noses at 
the lips of the returned hunters luitil these latter, in response, regurgitate some of the 
meat with which they have recently, and partly with this end in view, been filling 
their stomachs. This performance is repeated until those that have remained at home 
have received enough food to satisfy their hunger. The yomig, in their imrestrained 
impatience, often go beyond the mere ritual of nose-thrusting and become aggressive, 
actually biting the adults in their lips or legs. It is this voluntary sharing of food that 
enables labour also to be parcelled out between those that hunt and those that undertake 
the vitally important duty ot guarding the helpless young against marauders. It is 
extremely interestijig to note that the males take a tremendous part in the raising of the 
young. On their return trom himting they make a point ot regurgitating food for 
them in preterence to the females ; and, it was observed in one pack, that when, through 
death, no adult females were left the feeding and bringing up of the orphan pups was 
willingly and successfully undertaken by the males alone (Estes & Goddard, 1967). 



SS illl. (:A1(M\(MMS (11 W 1 s I AlKliA 

N(i niotlicr, thcrctorc, is essential attcr the first slmrt pennd oi milk feeding. Care 
of tlie voting, botli in tlie steps taken to guard them and in the matter of seeing that 
tliey are properly ted. seems to plav a dominant role m pack behaviour; tor later in 
lite wlien the yoiuig dogs reach the point of accompanying their elders on the hunt 
thev arc given absolute priority over the kill, all the adults retiring until the pups have 
first tully satisfied themselves. This, according to Kiihme, is the onl\- kuul of pre- 
cedence recognized within the pack. 

This begging ritual at the homecoming has been extended to a similar, but un- 
rewarded, greeting ceremonv carried out between adult members of the pack m the 
earlv morning and late afternoon before setting out on a hunt, and especially between 
those about to remain behind and those who sallv forth after food. Any ot the voung 
who attempt to take part in this ritual m which the adults tonnallv register with each 
other their fellowship ot the community are snarled or snapped awav; and it is not 
until they are accepited without protest at this twice-daily ceremom of mutual recog- 
nition that they can be reckoned as tully adult members of the societ\'. 

In the course ot their hunting, feeding and the ripping ot their still living prey to pieces 
these dogs become well splashed and smeared with blood; but thev take good care 
later to lick themselves quite clean. This is. of course, easier with their short, sparse 
bristle-fur than it would be had thev the dense tangle ot long tuie undertur possessed 
bv the foxes, hi this type ot pelage, too, there is less shelter tor fleas and other ecto- 
parasites; but, ncvertlicless, a good deal ot time is spent scratching at their tiu'. 

Like other dogs, these utter a variet\' of sounds according to mood or circumstances. 
Their most famous note is the ott-repeated bell-Iike "liooo-liooo" used as an assemblv 
call, either to gather the resting pack together tor hunting or, when c]uartering tm' 
game, a find has been made and there is immediate need to concentrate all forces on 
the line of assault. It is also used should a dog tuid itself separated from the others. The 
cogenc\' of this call is indicated bv the observation made b\' R. H. Kemp, noted above, 
that it was sutticientlv compelling to induce a dozen dogs to abandon a newly-made 
kill without feeding. Although it is soft and musical it carries a ver\' long distance, 
2 or 3 km. 

It seems that most ot a chase is earned out in Mlence, but the dogs break into guttural 
or raucous velps when closing in tor a kill and emotion rises to a pitch. This becomes 
an almost bird-like chirruping at the kill. Tins last sound is uttered alsii at other mom- 
ents ot high iuiticipatorv excitement such as when about to set forth on a hunt, or 
when mobbing hyaenas. The normal cr\' ot alarm at possible danger is a growl breaking 
into a short, deep, harsh bark which, however, is not really much like that of a domestic 
dog. A softer bark is used as a greeting; and there is a good deal of whining as an 
aeccnnpaniment to begging for food. 

ComiiKiiiK', hunting dogs, leading a nomadic lite, he about during the da\ or 
sleep at night simply curled upi on the surface in whatever slight cover the vegetation 
affords; but when it is a question of breeding, the whole life ot the pack becomes 
centred round a burrow or burrows made in the abandoned holes ot aardvarks, giant 
pangolins or other excavators. These luiderground shelters have not been described 
111 aii\- detail but appear to be lined with grass. Though their main purpose is for 



LYCAON 89 

parturition and tlic protection of the young they may also be used by any aduk as an 
occasional refuge against the n^iddax- heat or inclement weather or tor sleeping at 
Jiight. Not infrcquejitly two or more bitches litter in a single earth (Brand & CuUen, 
1967) or in closely proximate liolcs, cither without quarrelling or with at most minor 
disagreement. Indeed, as part of the commiuial life and communal sharing out of food 
and duties they look atter, and even suckle, one another's yoiuig. 

The abscjice of precedence within the pack appears to extend to the matter of 
mating, the males not contending for the females as they commonly do in otjier groups. 
Coition takes place several times over a matter of a couple of days. The period of 
gestation is from 69 to 73 days; and there may be any number between 2 and 19 pups 
in a litter (Goodall), probably 7 or 8 being an average size. The pups, which weigh 
350 to 380 grammes at birth, arc suckled for from 10 to 12 weeks, their eyes opening 
between the loth and 14th days. This is a tricky period and should anything happen 
to disturb the mother she may, as has been found many times in captivity, eat her 
yomig. If she is upset or thinks them in danger she carries them in her mouth from 
place to place. The presence of the male seems to be necessary at the period immediately 
following birth (Cade, 1967; Crandall, 1964). Crandall foimd, indeed, that the mother 
did not clean the new-born puppies, this being done by the male, whose active duties 
in comiexion with the litter were then over. But though the close presence of the father 
becomes for a period tuiwelcome he must remain within easy distance. 

From the time of the openijig of the eyes onwards the pups, though very much 
centred on the nest, quit it more and more frequently and for increasingly longer 
periods. Much of the suckling, if not all after the first couple of weeks, is carried out 
above groiuid, actually in or not far from the entrance to the earth. Diuring this act, 
which may take place either at the direct vocal invitation of the mother or in response 
to whining requests from the yoiuig, she may either lie down or stand. In the latter 
case the pups stretch up on their hindlegs steadying themselves with their forefeet 
against the mother's belly. Suckling sessions last only from 2i to 3 minutes (Goodall), 
in sharp contrast to the lengthy bouts indulged in by young spotted hyaenas. 

Though breast feeding is continued for nearh' 3 months the pups are in fact given 
meat after about a fortnight. This is specially prepared for them by the female chewing 
to small size some of the meat she has begged from the returned himtcrs and regurgitat- 
ing it for the second tune for her offspring. The male parciu, too, may take some part 
in this, as he does, now, in the general care of the litter. Gradually in this wa)- 
the young are weaned luitil they take large pieces of flesh for themselves, regurgitated 
by the hunters ; for from this stage the eiu'nc pack evinces a lively interest in the welfare 
of their next generation ; and, as has already been said, should, through accident, no 
adult female remain the males competently see to the upbringing of the litter. The 
whole process of raising a family to the point where the young dogs can take care of 
themselves and join in the himting occupies about three months. Throughout this 
period the breeding-groiuid is never, or only exceptionally, left unprotected, a varying 
Jiiunbcr of dogs and bitches always remaining on guard duty while those in search of 
food are absent. Dining the period of upbringing the site of the home burrow is 
changed, over a short distance, from time to time, especialK- it any danger has tlircatencd. 



90 THE CARNIVORES OF WEST MUKA 

To do this, in tlic early stages, tlic pups arc carried in tin- mouth to tlic new site, other 
teiiiales besides the actual motlier sometimes assisting. 

It is interesting to note that in order to escape very pressing danger a jiunting dog 
lias been kno\\ii seeiningl v to teign death with complete success though it is not clear 
whether this was, in tact, merely very convincing acting or an actual coma brought 
about by an attack of very powerful domestic dogs that had been so closely pressed 
home as to cause real death, by worrying, of a companion. The hunting dog in 
question maintained its apparent demise though dragged over the ground tor some 
distance, and for some further time atter its aggressors had gone away ard left it lying. 
Nothing is definitely known ot the expectancy ot life of these animals in their natural 
enviromnent but one has been known to live in captivity to an age of about lo years. 

Taxonomy. It has alreadv been mentioned, in connexion with the general des- 
cription of the Lyme;; pelage, that Matschie (1915) used the differing proportions of 
colours and their varving situations on the body to create a large number ot indepen- 
dent species. The material c-in which this study was based was extremely limited for 
each of the separate species proposed; and the characters on which these species were 
based, moreover, were such as might be suspected as intrinsically iticonstant. Further, 
apparent diflerences of size cannot be ascribed to entire populations on the strength of 
single skulls possibly ot varying ages and of unknown background. Holz (1965), 
as the result ot statistical study, which embraced skull measurements, came to the con- 
clusion that there was absolutely no foundation tor Matschie's proposals at specific 
level, and that valid separation even at racial level was doubtful as ajiy distinctions 
appeared to be clinal. It is, of course, possible that, at some future date, given a statisti- 
cally significant number of specimens from each of a variety ot localities, some valid 
racial separation may become evident. But any present attempt to maintain such dis- 
tinction on the basis ot the tew, mostly single, specimens that now exist seems to 
pretend to an luidcrstanding ot the situation that dots not. in fact, exist, and is without 
real meaning or value. Subspccitic names are therefore ignored in this present work. 

The accompanying table shows the measurements of the onlv two British Museum 
West African specimens, together with those two of Matschie's proposed species for 
which measurement data arc given in his paper. 



91 



Table 4 : Numerical data for Lyiacii piclus 





Mani 




Djannaga 


Kete Krachi 




(Shari River) 


Lake Chad 


(Togo) 


(Togo) 




Type of 


B.M. No. 


Type of 


Type of 




sharicus 


1937.7.10.1 


mangucnsh 


mischichli 


Vegetation 


Sahel 


Sahel 


Sudan 


Guinea 


Number in mean 


I 


I 


I 


I 


Condylobasal length 


189 


210 


— 


— 


Basilar length 


173 


192 


C.I 90 


— 


Palatilar length 


8S 


98 


— 


— 


Zygomatic breadtli 


122 


135 


125 


128 


Upper cheekteeth breadth 


68 


77 


7^ 


74 


Nasals, length 


67 


71 


65 


76 


Interorbital breadth 


37 


46 


49 


49 


Postorbital constriction 


37 


42 


— 


4a 


Braincase breadth 


69 


72 


69 


69 


Toothrow (c — in-) 


n 


89 


82 


87 


p* length 


19-4 


22-2 


— 




ml + Ml- length 


21-5 


25-6 


— 


— 


nil length 


22-5 


27-8 


— 


■ — 


(112 + H13 length 


14-9 


i6-2 


— 


— 


Head & bod\- 


900=1 





1 150'' 


1130" 


Tail 


325a 


— 


375 


410 


Hindfoot 


220* 


— 


200 


215 


Ear 


C.90 


— 


120 


125 


RATIOS (per cent) 










Tail/head & body 


36 


— 


34 


36 


Zygom. br./condylob. 1. 


64 


64 


— 


— 


Braincasc/condylob. 1. 


36 


34 


— 


— 


Braincase/zygom. br. 


57 


53 


55 


54 


Palatilar l./condylob. 1. 


46 


47 


— 


— 


Interorb./postorb. 


100 


109 


— 


116 


p^jc-iifi 


25-2 


24-8 


— 


— 


p*jm^ + nfl 


90-0 


87-0 


— 


— 


"'l/'"2 + '"3 


151 


172 


— 


— 



*. As given on the labels; Thomas's type diagnosis figures differ. 
''. These appear to have been taken from the prepared skin. 



9i THE CAnNIVORES OF WEST AEKICA 

Family MUSTELIDAE Swainson, i(S35 
Polecats, Weasels, Otters, etc. 

General. The Mustelidac constitute aai important family of very wide distribution 
throughout the world except in Australasia, Madagascar and the polar regions, being 
most common in the northern temperate belt. The family contains animals with such 
\\ell-known English names as polecat, stoat, ferret, mink, marten, badger, ratcl, 
skimk, weasel, otter and others. These are carnivores ot from very small to mediinn 
size, often with strikingly patterned coats and mostly with a greater or lesser degree of 
objectionable odour, the secretion of special glands, in some cases invohmtary and 
constant, in others deliberate as a defensive act. 

The mustelids may be terrestrial, arboreal, rivcrme or lacustrine; and there is an 
cxtralimital genus tjiat lives in tlie sea. In general diey have long relatively slender 
bodies and markedly short legs for their size; there are five digits on each foot, webbed 
more or less deeply and armed with claws that are non- retractile. Progression is, in 
African mustelids, mostly digitigrade. hi the Mustelid.ie as a whole development ot 
an aural bursa is variable; but as far as all the West African representatives ot the 
family are concerned it is always absent. 13y nature these animals are fierce hiuiters, 
being responsible for the destruction of large quantities of small mammals and birds, 
and in some cases a certain amoiuit of insects. To the great majority of Mustelidac 
any kind of flesh is welcome, preferably freshly killed by themselves but they will 
consume that killed by others if it is readily available, and even carrion it they are 
hard pressed. The main prey is rodents, hares, slnrews and birds — in Africa up to the 
size of a francolin or the young of larger breeds. Eggs, too, are welcome, and, less 
commonly, reptiles and amphibians. More exceptional diets are tound in the river- 
haimting otters, which live on tish, and in the ratcl. which often robs wild bees' nests 
for honey and grubs. 

Skull. The rostrum is short and blunt; there is no alisphenoid canal. There may 
be 32, 34 or 36 teeth, each of these totals occurring in West Africa. This is due to the 
different arrangement of the cheekteeth in the subfamilies: 3-j or 'j-^ or jy,. There 
arc always 3 incisors and i canine, top and bottom on each side. 

Taxonomy. Five subfimilies are now generally recognised (Simpson, 1945). The 
position, liowever, is by no means as straightforward and clear-cut as this might 
imply. Palaeontological evidence of atlinities and present structural morphology are 
both complex and in the past the tamily has been grouped and regrouped in a number 
of diff^erent ways. Pocock (1921c) gave a useful summary of the variously held opinions 
between Gray in 1869 and the time of his own paper, hi this last he assessed the differing 
views, gave an accoimt of what he regarded as relevant external characters and evolved 
yet another, virtually independent, classification in which he recognised no less than 
fifteen subfamilies in respect of living forms alone. Simpson (1945) once more reviewed 
the position, in the light of both living and extinct forms; and, regarding Pocock s 
treatment as "excessively inflated", reduced the subfamilies of ext.int mustelids to the 
five uo\\' widck accepted. 



MUSTELINAF. 93 

Only three of these subfamihes occur in West Africa, the two that do not exist 
there being the Melinae (true badgers) and the Mcphitiuae (skunks). The tlircc sub- 
famihes relevant to this present work may be distinguished by the following key. 

KEY TO THE SUBFAMILIES OF MUSTELIDAE 
(Previous key page 28) 

1. Dorsal pelage very long and with a well-defmed longitudinal pattern of black 

and white bands; total length of skull luidcr 70 mm; cheekteeth 3-5. 

Mustelinae {pai^c 93) 
Not like this 2 

2. Fur coarse, the upper side typically whitish or pale grey from the crown ot 

the head to the root of the tail (but this area often reduced or, not infre- 
quently, entirely black like the rest or the pelage) ; claws of the forefeet 
very long and powerful; ears with no free external shell; skull with 
very little iiuerorbital or postorbital constriction; bullae large and 
subglobular; cheekteeth 1^ .... Mellivorinae {p(,oc no) 
Fur soft and sheeny; postorbital constriction fairly marked; bullae rather 
flat and subtriangular; cheekteeth ^ (but ;)i often missing). 

Lutrinae {page 128) 

Subfamily MUSTELINAE Gill, 1S72 
Polecats, Weasels, etc. 

General. This subfamily contains most of the smaller mustelids, fulfilling the 
important function of holding the rodent population in check. They, ot course, kill 
other prey too, mostly birds, insects and some reptiles. Their bodies are often long, 
slender and lithe, the whole effect being added to by an unusually lengthy neck; but in 
the African species these features are not so pronounced as amongst the European 
and American forms. There is no bursa on the ear piruia. The tail is short to moderate, 
that is to say at most never quite as long as the body; and it is sometimes showy. 
The pelage is often striking ; and in the case of the mink {Miistcia, subgenus Liitrcola) 
is so much sought after as the most luxurious of commercial furs that it has become the 
subject of a flourishing farming industry and of selective breeding experiments directed 
to the production of distinctive colourings. Some members of the subfamily inhabiting 
the colder northern regions undergo a pelage change in the winter, the brown coat 
becoming white; in the case of the stoat {Mustcla enninea) the tip of the tail remains 
black and the fur is much sought after on this accoinit, most notably as a decorative 
trimming to official robes. Another well-known member of the subfamily is the 
ferret, a domesticated, and often albino, form of the Asiatic polecat, bred, and to 
some extent trained, as a hunting animal to flash rabbits or other subtcrrestrial species 
from their burrows. 

The t;eiuis MiistcLi penetrates into the Mediterrane.in region of northern Africa; 



94 II" (AUNUDiti.s oi \\i:sr aiuila 

but ill so tar as the true Etliiopian region is concerned tliere are three Atrican nnistehne 
genera, laoiiyx. Piwilictis and AuTi/iu'j/e, ot which only the first two are found witluii 
the area deah with in this present account. PoccUcfJiilc is essentially a soutliern African 
genus but extends as tar north as Tanzania and parts of the Congo and Uganda; and 
there seems no obvious reason why this animal, which closely resembles htoiiyx in 
external appearance except tor a tar more slender bodv, should not have spread turther 
westwards across tlic continent. Tlie two genera witji \\hich, however, we are here 
concerned iiiav be distinguished as tollows: 

KEY TO THE GENERA OF MUSTELINAE 
(Previous ke\' page 93) 

White marking on the tace continuous across the trout; tips ot tlie ears only 

very slightly white; total length ot skull under so mm Poccilictis (/)ii^'c 104) 

White facial marking broken above the eyes; tips ot the ears conspicuously white; 

skull ss mm or more ....... Ictotiyx (jhit^c g^) 

C.ciuis ICTONYX Kaup, iN.^s 
Atrican Polecats 

Zi'iillii Okcii, 1S16, Li'liiiiuch dcr Natin\^fscliklilf, 3, Zocil. pt. 2; m. Okeii's work has been ruled to lie 
imavailablc by the International Commission on Zoological Nomenclature, Opinion No. 417 ot 1956. 
This name is the Spanish zoiilln, a diminutive ot zona a fox. 

ZoriUa I. GeofFroy, 1826, Dii'lioiiiiaiif iliusitjHLit'Histoin- Kalnirllc, 10: 215. This name has been suppressed 
by Opinion SiS of the International C'omimssion on Zoological Nomenclature and placed on the 
Olticial Index ot Rejected and Invalid Names in Zoology with the Number 1912. {Bull. :ool. Noiiiciicl. 
M- 1.S3-154)- 

hlpiiyx Kaup, iS^s. Ddi 'fliiincicli . . .. i: 152. liy Opimou MS ot the International Comnnssioii this 
name has been placed on the OHicial List ot Generic Names with the Name Number 1759. Type 
species. b\' inon<itvp\, as recorded in the same C~)piuion, Ictotiyx aT/u'/i.s/.s" Kaup f— liiiulyjnis snitUiis 
I'errv). The name was compounded from the Greek iili<. iiiiJos, a weasel, and ii;;)'.v claw, tlie latter 
referring to the stronglv cl.iwed toreleet. 

Rhnlidoi^ale Wicgm.imi, iN^S, Arcli. Naturt^csili., 4th year, pt. 1: 27K, tootnote. Type .species Briulyinn 
stn'iitiii Perry, selected in Lllcrmau, Morrison-.Scott S. Hayman, 195.1: in. The name is from the 
(rreek rh,ihlo.<, rod or w.iiid, i/i.ifc/iVi'.v striped, and ijfl/e a weasel, referring to the pelage. 

htidofiyx Agassi/, 1X40, Si'iin-iit-liiiitiii Zoi'lottici Index I'liii'irsulis: ssS. An emendation ot Lioiiyx Kaup 
in accord.nice with tlie (Jrcek trom which it was derived. 

Li:'my< Roberts, I9;6, Ann. Traiisv. Miif. 18: 22!>. A Inpsm calami tor liioiiyx. 

This is a moiKispecitic genus the description ot which is, thus, adequately tunii.shed 
by the accoiuu ot the species which immediatelv tollows. There has been very con- 
siderable, and heated, argument m recent years over the tjuestion ot whether the 
name laony.x shoulci replace the commonly used ZorilLi. The matter, now resolved by 
Opinion XiS. w .IS .1 higliK complex one the various aspects ot which can be toujid bv 
those interested ^et out 111 1 11111.1 (ii)(ti. mjCis and n/iCi) ; Van Cn Idcr {ti)<>(t) ; and I Icrsli- 
kovit/ f K/i^ and igOd). 



ICTONYX 95 

ICTONYX STRIATUS (I'crry) African Striped. Polecat or Zorilla 

Bradypiis sirialiis I'cny, 1810, Aruviit or ihe Miisvuiii oj !\'alnral History, part 1 1, pi. (41) and text. Cape ot 
Good Hope. A note on the validity of this name is given by Hollister (1915); and the name siriatus 
Perry has now been placed on the Official List o( Specific Names in Zoology with the Number 2202. 
The name siriatus is the Latin adjective meaning marked with channels or bands, given in reference to 
the dorsal pelage. The generic name is fronr the Greek hraJys slow, and pons foot, meaning slow- 
footed. 

Mephitis capviisis A. Smith, 1S26, A Dcscripiiir Cataloi^iic of tin- Soiiili Ajricaii MiistKiii, pt. i. Mammalia: 
20. No locality named. This generic name is the Greek word tor a disgusting smell, and refers to the 
body odour. 

Miistcia zorilla p. Si-ncgalciisis ]. B. Fischer, iS2y, Synopsis Aiaiiinialiuiii: 219. Senegal. The generic name 
is the Latin for a weasel. 

Mustcia zorilla Smuts, 1832, Disscrtalio Zoolo^ica, eiiuiiicrationviii Maiiinialiuiii Capcusinm (omincns, 12. 
Not Viverra zorilla Schrebcr ( = Spilogalc). 

Mephitis africana Lichtenstein, 1S3S, Ahli. preiiss. Akad. IViss. for 1.S36: 284. Cape of Good Hope, Scne- 
gambia, Abyssinia, Barbary. 

Rhabdo^ale iiinsteliiia Wagner, 1841, in Schreber's Saufttihierc, Supplement, 2: 219. Cape of Good Hope. 
The specific name is Latin for weasel-like. 

Zorilla striata var. seiiegalaisis Gray, 1865. Senegal. Type in the British Museum, No. 50.7.8.38, }; skin 
good except for the terminal part of the tail probablv missing, skull good. 

Distribution and general. The African striped polecat is very ohen, both in 
literature and speech, termed the zorilla or zorille, sometimes zoril, a Spanish name 
having no real application to Ichviyx. The steps by which this came about are involved; 
but reduced to their simplest terms the position stems from confrision arising in the 
early 19th century between the African polecats and the central American spotted 
skunk {Spilogale), of rather similar appearance and of which zorilla was the local 
Spanish-American name. 

This is one of the most striking ot small Atrican mammals, with its conspicuously 
banded black and white coat; but it must not be confused with the closely similar, 
though very much smaller, striped weasel which may occur in the same locality. 
When the animal is at rest in grass or other cover this bold pattern doubtless serves to 
disrupt its form and conceal its true nature; but such coloration is just as likely, in the 
open into which the animal must often enter, to act as a warning to rapacious birds 
and other beasts ot prey, cautioning them to avoid attacking a victim capable of 
emitting a nauseating fluid. Couspicuousness in these circumstances is added to by 
the absence from this entirely black-bellied animal of the usual protective coimter- 
shading. 

Ictoiiyx striiitiis ranges over the drier zones of Africa from the Cape northwards to 
Sudan and Ethopia on the cast, and thence across to Senegal in the west. It also occurs 
in South-west Africa and parts of Angola. This is an animal chietlv ot the drier open 
woodlands, mostly (as tar as West Africa is concerned) the Sudan and Sahel, but 
penetrating also into the Doka zone. It would seem to be pretty common in the southern 
and eastern part of its range; Shortridge (1934) characterises Ictciiyx as one of the most 
ubiquitous mammals in Southern Africa, in all kinds of vegetation, down to the sea 
coast and even in town gardens, [t also occurs up to 2000 metres on mmuitains. It seems 



yd Till- CAUMVOHliS (.U WLST AlUICA 

to W nukli less ccmniuiii in tlic area dealt witli in diis present account — tliougli this 
apparent relative rarit\' in West Atnea ina\' be due merelv to lack ot collecting. Onl\' 
S skins and s skulls exist in the British Miiseuin Iroiii tins area: from Senegal (2), 
larniso. Nigeria {2) and Zaria (Nigeria). 

Description. Tlie African striped polecat (tig. 12) varies a good deal in size, and 
such dirterences have been used, at least in part, diagnostically in tlic erection of local 
races. The males are generally held to be appreciably larger tlian the temales; but this 
distinction is not evident in tlie scant data available tor West Atrica — one male skull 
compared w ith tour females. Tlie pattern of the pelage is broadly the same thr(Highout 
the animal's wide range but the proptortions of its coniponcnt elements, that is to say 
black and white, var\ , seeminglv trom place to place and so commonly constitute the 
m.ijor basis ot proposed subspcciation. The lengtli and quaht\' ot the fur also var\. 
In West Atnca this polecat attains a ]icad cV body length ot between 320 and 350 mm, 
the narrowly bushv tail being some 250 to 2N0 mm. The body is tairlv, but not 
markedlv, slender, and the legs short, tlie animal standing abi-nit lis to 130 mm at the 
shoulder. Tjie weiglit ot an aduh may be about j kg en- sometimes a little more, 

Tlie dorsal pelage is tairly long and dense but loose in texture. It generally has a 
sheen. The underparts are considerably shorter and often sparser, sometimes almost 
naked over the bells and chest. On the back it may be regarded, purely tor convenience 
ot description ot West Africa specimens, as basically white marked with three highly 
conspicuous longitudinal black bands. (Tlie converse of this prob.iblv more accurateU- 
expresses the true position). The centre one ot these black b.ands, running along the 
spine, starts trom the crown ot the head; the two lateral ones originating a little short 
ot this, at the tore part ot the neck. These three lines are here narrow (roughly s mm 
or less broad) and they riui approximately parallel to not quite the middle of the 
back. There the two outer ones broaden somewhat (to roughly 12 to is mm) and 
diverge, curvinc; outwards to the top ot the flanks and then in again to the root of the 
tail, where they meet or nearly meet, thus enclosing a regular elongated ellipse. The 
medial black line continues its path to tlie tail through this \yliitc ellipse but at about 
tJie midpoint broadens into a black blob, roughly ot diamond shape and 30 mm wide 
at its broadest point, then narrows again posteriorly, meets the two outer bands and 
continues on to the basal part of the tail. The sides of the neck, the shoulders and the 
entire tmderparts, together with the legs and teet are wholK black. The long-haired 
but nevertheless rather narrow tail is an intimate mixture ot black and \s'hite, the 
latter colour dominating in West Atrica (except in the basal 70 mm) though not alw.iys 
elsewhere. 

The pelage is composed ot long tine undertur .uul (.oiisulerabK longer and stouter 
bristles. These elements are selt-eoloiired throughout, either all-black or all-white; 
and. in general, the two colours occup\' clearly defined areas ot skin without inter- 
mixing. 13ut there are exceptions. I lere and there one or more isolated pure white 
bristles occur amongst black patches; and sometimes there are small islands of white 
m the black .neas; and, more rareb', in some (extr.ilimital) specimens or forms, bristles 
(dlouri'd .It the b.ise ami arising 111 the bl.ick arcis. beconic. .l^.llnsl the prewiiling 
character ot pi^meiit.ition, white in their termin.il li.ill. Sneli aberrations, ,\\](.\ especialK' 



ICTONYX 



97 




,xlw 



Fk;. 12. African Striped Weasel (Pocciliclis liby<a) 
Afiic.m Striped Polecat {hUviyx slriaHn) 



<jS 



■rill l.MlN'UiiHIS (M W I.ST AliniA 




Fli:. I}. liloiiyx siritidis snu[,;tih-iisir. skull. 'l'\pi.\ li.M. Ni^. 50.7.X.VS, f, ;■- 2; l.itcr.il vi( 



this Inst, lead, in cases whore thev arc more marked, tlian citliers, to a certain bhirnnp; ot 
colour and ot niargijial definition in tlie pattern. The hve available West African skins 
are ot nnitorm external appearance and can be distinguished trom forms which occur 
elsewhere; and there would seem to be possibh' a soinewliat soinider toiindation tor 
local races 111 tins species thaji is often tlie case, hi West Africa tlie underfur is iS tii 
11 mm long, the bristles 32 to 3s mm; but some of the solitar\' \\hite bristles run up 
10 3K mm. 

The head m this species is small for the gener.il size of the biul\'. shallow-skulled and 
narrow, the muzzle bluiuK' pointed. Basically, the whole face, chin and throat are 
black; but this is interrupted b\ ver\' prominent white markings. These consist ot a 
broad white patch which runs, on each side of the face trom below the ears inwards 
to |ust above the outer corners of the eyes; separated from the inner limits ot these is a 
central white frontal patch, lying between and backwards of the eves. The separation, 
broader or narrower in different forms, is generallw (.piite clear-cut; but in two West 
African specimens there is considerable tendenc\ to tuiite — anil thus, incidentally, 
obscure a distinguishing character bet\\een Ictonyx and Poccilictis. Fn some specimens 
the cheek patches spread below the ears a little onto the throat. The size and extent 
of tJie white facial markings have been used in racial diagnosis. The ears are small aaid 
semicircularlv rounded, the top portion being narrowly but vcr\' conspicuously white. 
The forefeet are larsjer than the hind and are furnished with yer\ long, deep, slightly 



ICTONYX 



90 




Fig. 14. kloiiyx slrlatus seueijalctisis: skull, Tvpc, B.M. No. 50.7.S.38, ?, X 2; palatal & dorsal views 



100 I 111 1 A1IM\ (11(1 S Ol \\ 1 ST A Mil ( A 

iui\cd (.l.iws .lb. nit tlnw llnu■^ as lug as tliosc ot tlic luiultccc. The soles (if Ivitli 
tore aiul liiiiLit(.'ct arc naked. Albinos arc not imconinioii. 

Skull (tigs. 13 and 14). Tliis is \'cry solidiv built and it is rarcK- possible to detect, 
even in moderately yoimg ones, anv trace ot the sutures since tliese become completely 
ossil-ied trom aji early age. In form, the skull tapers from a fairly broad braincase to a 
remarkablv short, bkuit rostrum. There is very little or no sagittal crest but a well- 
developed supra(Kcipital one, often prominently framing the whole posterior part 
ot t]ie skull. There are narrow, sharply pointed postorbital processes, and there is a 
distinct post.-rbital constriction; but trom this point forward there is relatively very 
little narrowing ot the skull up to the widelv open iiares. The zvgomatic arch is strong 
at both its anterior and posterior roots but decreases centrally to an unexpectedK- 
slender, weak Jugal. The palate is very broad between the caniassials and extends, 
alter a sudden contraction, narrowlv far beyond the toothrows. The bullae are, like 
the rest ot the skull, \erv stroiigK- built, long but relatively shallow, considerablv 
strengthened posteriori}- hv overgrowths ot the mastoids and paroccipital processes, 
and connected in front to the pterygoids bv slender processes. The palatal foramina 
arc small. 

The dentition is reduced but compact and powerful. In the upperjavv the six incisors, 
deep from front to back, are closeb set in a more or less straight transverse row, the 
outer one on each side larger than the middle four. The canines are not very tall but 
.ire strong. There ,U'e only three premolars on each side with no space between them 
or even slightly imbricate; p- is about iialf the size ot p''; p^, the carnassial, is well- 
cusped, having also a small anterior hook on the cingulum. There is only one molar 
ni' ; this IS ver\' narrow laterally but extends almost parallel-sided, well into the palatal 
region, hi the manner of this subfamiK' it is deeply excavated between the outer and 
inner edges. 

The mandible matcjies the upper jaw in its shortness and extreme strength ot build, 
the rami being deep, solid and curved. There is very little angular process; but the 
condylar process appears luiusnally wide and strong. The straightly transverse uicisors 
reflect those ot the upper jaw in the compact solidity of their setting. There are three 
premolars and two molars. Of the latter, the carnassial, ;»i, is very sharply cusped but 
III-2 is greatly reduced 111 size and is scarceK* as big as the anterior cusp of iiii. It is, in tact, 
more or less tuiictionless since on occlusion it lies posterior to the upper dentition, 

its front edge barel\ touching the rear margin of m'. Tlie dental formula is thus 

- 1.1. .•'.I 

-1.1..'!. 2 "" J-1" 

Habits. Considering how conmion Iilonyx is, at least in parts ot its range, and how 
often it has been kept as a zoo exlnbit or a domestic pet it is strange that so little is 
known ot us way ot lite beyond the broad outlines. This is in some measure due to 
the tact that the striped polecat is mainly nocturnal, though it may sometimes be 
observed at dusk or dawn. During the day it lies up in holes in the ground, wjiich it 
excavates itself with the aid ot the long and strong claws of the forefeet. It also digs 
t(-) get at sheltering rodents, such as gerbils, or at large insects. Besides subterranean 
burrows, Icioiiyx takes advantage ot naturally occurring shelters amongst boulders; 
and. ill South Africa, it is well-known to live in, or in holes beneath, farm outhouses. 



ICTONYX 



Striped, polecats seem to live more or less solitary lives except at mating time when 
pairs arc seen together. They live on rodents, hares, ground-squirrels and other small 
mammals, insects of the larger sorts, birds, eggs and probably lizards and frogs. They 
are said to be deadly enemies of even large snakes. Roberts (195 1) records that these 
carnivores sometimes prove a nuisance to the field naturalist by following the line of 
his traps and consuming all the rodent specimens that have been caught. They not 
infrequently take up residence near villages or farms and then sometimes become 
persistent fowl thieves. However, they appear to be very easily trapped (Thomas & 
Hinton, 1923). The normal gait is a trot with the back slightly hunched; but observers 
differ as to whether the tail is held up or out. When chased by a dog the striped polecat 
can run fast. Though mostly terrestrial it can also clnnb into trees; and it is said to be 
able to swim well if forced to. 

But Ictonyx does not always take to flight, or only for a short distance to where it 
may reach a rock, log or tree affording a slight eminence whence it may with greater 
advantage against a larger enemy bring into play its best kno\«i response to fright or 
attack. This is to present the rear quarters to the source of danger, lift the tail vertically 
and ejaculate fluid from the anal glands into the face of the enemy. This is not only 
nauseating in its smell but also very persistent and difficult to eradicate from fur or 
clothes even with washing. Roberts (1951) records having received a discharge full 
in the face. Apart from the sickening stink, this caused a painful burning in the eyes. 
He found that the application of benzine quickly neutralized the offensive fluid except, 
of coiu-se, in the eyes to which he was unable to apply the remedy. It gradually oozed 
from these throughout the night causing a continual renewal of the stench. This 
defensive action has, understandably, been most commonly observed as a reaction to 
chasing by dogs. These do not seem to be much put off at the time though it is said 
that once one has experienced the polecat's ejection it becomes chary of going into 
attack on future occasions. Such an effect is the essence of warning coloration. 

All observers are agreed that as well as being, like most of its close relations, a very 
fierce hunter and killer, Ictonyx is correspondingly stout-hearted and tough in its own 
defence. It utters angn,' shrill screams and erects its hairs thereby doubling its apparent 
size. When pressed to its limit, and threats or anal glands have failed to deter the 
attacker, it feigns death, for half an hour or more, luatil it judges that the danger is 
past, when it gets up and trots oft. There is an interesting manuscript note made by 
Sir Andrew Smith in his owai interleaved copy of his Catalogue (1826) in the British 
Museum: "This is also called by the Farmers Duckatongat in consequence of the 
size of the anus — It is extremely tenaceous [sic) of life and will oftai after being bitten 
and shook by dogs for a long time and left apparently nearly dead recover and run 
about in a few hours as well as ever". 

It may be added that after such an encounter the dogs would be completelv unfit 
for admission to a house since they would inevitably be unbearably smelly. The coat 
of the polecat, too, would from force of circumstances be in a similar condition; 
but normally these animals are entirely without smell. Further, they are easily tamed, 
even if nearly full-growai when captured; and it is a remarkable fact that they then, 



THE CAIiNIVOUP.S OF WIST A 11! I ( A 



luidcr domestic circumstances, rarely it ever use tlieir defensive odour. Tliey become 
not only tame but actively friendly. 

Despite their one-time common appearance in zoos few accurate observations seem 
to have been recorded of mating, gestation, number of young, and general behaviour 
in relation to bringing up a litter. It is said that there are generally 2 to 3 yoimg at a 
birth, bom naked and blind. A. J. Hopson (in a letter) thinks that in the Lake Chad 
area, where it is occasionally seen, the species may breed during the rains since yoiuig 
ones are to be observed from September to November. It is recorded that one lived 
for almost .s.\ years in the London Zoo. A malarial parasite {Plasiiwdhiin) was recorded 
from one of five specimens caught in traps in Senegal (Leger & Bedier, 1923). 

Taxonomy. It was stated above, in the opening ot the accomit ot the genus, that 
Ictoiiyx is monospecihc; yet Roberts (1951) was convinced that there were four recog- 
nisable species in South Africa alone. He arrived at this conclusion largely from a 
consideration of the coloration ot the tail ; two, of differing but overlapping sizes, 
with the tail wholly white; one with the tail wholly black; and one with a white 
tipped black tail. These were further divided into races, twelve m all for southern 
Atrica,on such characters as the relative widths of the black and white stripes, or whether 
the black stripes reached as tar forward as the neck or the back ot the forehead. It is 
doubtful whether the character of tail coloration is of a specific luture; and, indeed, 
Ellerman, Morrison-Scott & Hayman (1953) rejected it, suggesting that it was of no 
greater significance than the well-known phasal occurrence of black or white tails in 
the mongoose Ichuctiiiiia alhicanda. They likewise attached no taxonomic importance 
to the alleged size diiierence, concluding that the genus consisted of a single species, 
stridtus alone. These are matters extraliniital to this present work, cited to show that 
the sailing in laonyx is not as plain as it might seem and that complete tuianimity of 
opinion does not necessarily exist. In as far as they concern West Africa it inav be 
said that the five specimens from this region in the British Museum do, in fact, corres- 
pond to Robert's external diagnosis o{ striatiis and, in large measure, to the skull sizes 
he quotes for this. 

The question of the validity of subspecies is quite another matter. It has already 
been said earlier that variety of pattern certainly exists, though it lies within fairly 
narrow limits; and though unquestionably a good deal of minor individual variation 
occurs some overall constancy seems to attach, at least broadly, to series from different 
regions. Significant differences of size also seem to occur. There, therefore, appear 
to be rather more satisfactory grounds for the recognition ot local races in /. siii<itii< 
than is the case with many other genera. A practical stumbling-block exists m the 
difiiculty, the inexactness and the general inadequacy of verbal description. Such 
expressions as "much whiter above" or "black lines less clearly defmed" convey little 
of positive value to either museum taxonomist or field worker devoid of a range ot 
study material. A plate of comparative photographs is about the only really satisfactory 
answer to this difficulty. 

One described race certainly and one possibly concern West Africa. Fischer's dia- 
gnosis of the first of these, scucgalciisis — half-a-dozen Latin words amounting to "with 
white marks, especially above, almost confluent with the tlanks" — is an extreme 



ICTONYX 



103 



illustration of the inadequacy of verbal description in relation to variations of such a 
pattern as characterises striata. Gray's (1865b) description of his sciicqaleiisis (in com- 
parison with the South Africaji form) is not much better: "White streaks broader 
leaving only very narrow dark dorsal ones; tail whiter." Nevertheless, it can be taken 
that a form conveniently identifiable as senegaleiisis does exist in "West Africa from 
Senegal to at least Lake Chad; but whether some of the more easterly animals validly 
differ from it in pattern, though they possibly do in size, is open to question. 

The second possible West African race, sudanicus, was described by Thomas & 
Hinton (1923) a male from Jcbcl Marra at 1200 metres, the pelage being noticeably 
longer than in senegalensis, the pattern, though broadly a good match, rather less 
intensely black, with the three parallel dark lines over the neck yet slimmer and far 
less distinct. It has been suggested (Dekeyscr, 1955) that possibly this form may reach 

Table 5 : Numerical data for Ictonyx striatus 



Vegetation 
Number in mean 
Condylobasal length 
Basilar length 
Palatilar length 
Zygomatic breadth 
Upper cheekteeth breadtli 
Nasals, length 
Interorbital breadth 
Postorbital constriction 
Hraincase breadth 
Toothrow (r — m') 
p^ length 
!»' length 
"U length 
"12 length 

Head & bodv 
Tail 

Hindtoot 
Ear 

RATIOS (per cent) 
Tail/head & body 
Zygom. br./condylob. 1. 
Braincase/condylob. 1. 
Braincase/zygom. br. 
Palatilar l./condylob. 1. 
Intcrorb./postorb. 

Hll/»l2 





senegalensis 






(West Africa) 




<J 


$ means 


Total range 




Sudan and Dole; 


a 


I 


4 


5 


64-8 


62-6 


Sg-j-66-2 


58-2 


54-1 


53 •1-58-2 


30-3 


29-2 


26-o-30'3 


37-7 


37-2 


35-3-40-7 


23-2 


24-0 


21 •7-25-2 


17-0 


17-3 


I6-0-I9-0 


14-0 


15-2 


I4-0-I6-7 


29-9 


29-5 


28-2-3I-0 


20-3 


19-8 


I9-2-20-3 


7-2 


7-0 


6-9-7-2 


3-5 


3-1 


2-3-3-7 


7-8 


7-3 


7-0-7-8 


2-3 


2-6 


2-3-3-2 


— 


323 


320-327 


— 


262 


248-277 


— 


51 


50-53 




25 


24-25 


58 


81 
59 


— 


46 


47 


— 


79 


79 


— 


47 


47 


— 


121 


114 


— 


35-5 


35-3 


— 


206 


226 


— 


339 


281 


— 



sudaiiiais 

(Jebel Marra) 

<? 

?Sahel 

I 
66-5 
6o-o 
30-7 
38-8 
21-9 
13-6 
17-8 
i6-i 
30-1 

20-0 

6-5 
2-9 

7-2 



355 

255 

58 

24 



72 

58 

45 

78 

46 
no 

33-0 
228 
257 



104 Till (.AHM\()llls Ol WISI MIIKA 

tromjcbcl Marra as tar west as Lake t'liad; but m all probability it dwcs its shaggiiicss 
of coat to its high altitude habitat and is thorctore absent troni the low-lying country 
between its type locality and Nigeria, and thus cannot be looked upon as being a 
contiguous torm to sciici^iilciisis. It is omitted from the checklist ot know]i West A&ican 
forms given on page i; but it might none the less tm^n up, discontinuousK , 
on some ot the West African highlands. The size ot the only known specimen, the 
tvpe in tlie British Museum, is not markedly different from sciic(;alcnsis. as shown iji 
the table on page 103. A slight ditticulty exists over the comparison ot sizes in this 
genus; the males arc generally reckoned as appreciably larger than the females. There 
IS only one male ot satcgalcnsis to compare with other reputed torms or with the four 
females in the British Museum. This last reveals no significant sexual difference, the 
t\pe, a female, being almost as large as, or in some respects larger than, the male. 
Tliis is shown in Table s. 

Ictonyx striatus senegalensis (FiscIut) Senegal Striped Polecat 

The description ot this torm may be taken as that given on pages 96 to 100. Specimens 
exist in London from Senegal (Thics .and unspecified), and from Nigeria (Faniiso ami 
Zaria). 

Genus POECILICTIS Thomas & Hmton, 1920 
Striped Weasels 

AvYi7ii7ii- Thcnuas & Hmtoii, Hjio, Ann. Mtii<. tun. Hisi. (y) 5: 3'>7-j6S. Tvpc species IWciliciii liliyni 
(Heinprich & Elireiiberg) [-- Mmlchi Ubyca HcmpricK & Ehreiibcrg) from Libya. This name is trom 
the (Ireek jh^ihi!o< v.iri-cttloured or pied and Jt7/V weasel, in reterence to the black and white pelage. 

Distribution and generaL The striped weascK are, at least in West Africa, ver\- 
closely similar in external appearance to the striped polecats though ot much smaller 
size. For a long time the two were, indeed, reg.arded as congeneric; but there are 
certain small external and cranial differences which led Thomas & Hmton to erect 
for the weasels a separate genus. Seven species have at one time or another been named 
but these are all now regarded as either synonyms or merely forms of a single species 
lihyui. The range of this, and hence of the genus, is restricted purely to the northern 
portion ot Africa, from about 12"'' Nortli, that is within the Sudan woodland zone, to a 
little short of the north coast from Morocco to Libya and lower Egypt. Poccilictis is 
thus essentiallv a dry or very dr\' country animal, about half-a-dozen local races being 
recognised. The general generic diagnosis and the points which distinguish Poccilictis 
froin Ictonyx can be sufficiently gathered from the description of Uhyca which follows. 

Taxonomy. Whether Poccilictis merits separate recognition from Ictonyx as a genus 
IS open t<i doubt. There arc a few small external dirterences of patteni and of such 
minor matters as hairy or naked soles. Thomas & Hinton, wlio erected the genus, 
characterised the Poccilictis skull as truncated, greatly expanded across the mastoids, 
.md having livpertrophied bullae. Verv little support tor the fu'st two of these is affordetl 



POECUICTIS 105 

by comparative measurement of the two genera. The bullae are indisputably larger m 
Poecilictis; but this alone caimot justify a genus or even a subgenus. Since its creation in 
1920 Poecilictis has been accepted, by most authors, including Simpson (1945) and 
Ellerman & Morrison-Scott (1951). However, Dekeyser (1955) rejects it — probably 
justifiably, though this course has not, in fact been adopted in this present work. 

POECILICTIS LIBYCA (Hemprich & Ehrenberg) Striped Weasel 

Mustek lihyca Hemprich & Elircubcrg, 1833, Synibolae Pliysicae sen Icoiics ct Descriplioiies Mammalinm . . ., 
decas secunda, folio k verso. Libya. Mustela was the Latin name for a weasel. 

Rlwbdogale mukiviuMa Wagner, 1841, in Schreber's Die Saiigethicre, Supplementband 2: 221 f.n.; Supple- 
mentband 3, pi. 133B. Upper Nile. The specific name is from the Latin inuhum much, many and 
vitla a band, with reference to the striped pelage. Valid as a race. 

Iclonyx freiiata Sundevall, 1S43, K. svenska VeleiiskAliaJ. HaiiJI. for 1842: 212, pi. 4 f.i. Senaar. The 
specific name is Latin for bridled, given in fanciful resemblance of the pelage pattern to a horse's 
harness. 

Poecilictis rothschildi Thomas & Hinton, 1920, Novii. zool. 27: 316. Farniso, Kano, Northern Nigeria. 
Type in the British Museum, No. 21.2. 11.29, t; skin only, in good condition. This was named after 
Lord Rothschild under whose patronage and for whose museum at Tring this and many other specimens 
were collected by Captain Angus Buchanan. The skull, which was originally believed to exist, has 
never come to light. 

Distribution and general. The little striped weasel (fig. 12) is not as wide-spread 
and is apparently much less common than its bigger relative the striped polecat. So far 
as is known its distribution is restricted to the edges of the Sahara atid the contiguous 
arid zones from Senegal and Morocco in the west to the Red Sea coast in the east, 
but nowhere further south than about the 12th parallel. Both Ictonyx and Poecilictis 
occur in the same area — witness the British Museum specimens of each, caught at 
Farniso on the same day. Althotigh, to judge from museum collections, the weasel is 
much less abundant than the polecat it is apparently locally common. Six of the eight 
West Afi-ican specimens came firom Farniso, five of them caught on the same day or 
almost the same day. 

Description. Although the pelage pattern in Poecilictis is ftuidamcntally the same 
throughout the species there is a wide difference of detail between forms from difterent 
localities. The description which immediately follows is therefore couched in broad 
terms, the fuier points of diagnosis of West African animals being reserved for the 
later account of the race iiiultirittatci. 

The striped weasel has a head & body length of from 200 to 280 mm and a long- 
haired, more or less bushv, tail of from 100 to 170 mm. It stands some 60 or 70 mm at 
the shoulder ; but the tliift uig out of the fur in excitement makes a good deal of difference 
of appearance iji size. The pelage, which is composed of dense, fme, fairly straight 
luiderfur and abiuidant, much longer, stouter bristle-hairs, is of very diverse lengths in 
the various races, from moderately long to very long. The dorsal pattern is of dark 
blackish-brown and white. In West Africa it consists clearly of alternating longitudmal 
dark and white bands, closcK' recalling that o( Ictonyx; but. though this is doubtless 
the basic pattern throughout the species, some of the extralimital forms have it so 



I06 TIIF. CARNIVORES OF WEST AFRICA 

broken up into incgular spots and patclies, and tlic dark so niucli overlain by tlic 
lengthy white Iiairs, that the resemblance to the striped polecat becomes relatively 
remote. In most cases the dark element is not so black as in Icionyx. The imderparts and 
legs, as ill tliat genus, are very largely blackish, but there are generally a few rather 
isolated white spots, especially on the belly. 

The facial markings offer a fairly constant distinction between Poccilictis and Ictoiiyx. 
In tlie former tlie white patch that nuis across the front is continuous as a broad band 
from ear to ear uistead of being interrupted above eacli eve as in the latter. (In one 
Pi\'cilictis skin tliere is a very narrow interruption; and in a tew Icionyx there is a naiTow 
connexion between the medial frontal spot and the lateral patches). In Poccilictis this 
white band is continued bclo\\' the ears onto the chiai where it forms a, usually, verv 
conspicuous forwardly-pointing V. The upper lips in tliis genus are always more or 
less white; but the upper rim of tlie ears, luilike the polecat, has only very little wliire 
on it. 

The tail is whiti.sh, being covered with very long bristles which are broadly white 
at both base and terminal third, the medial band being dark brown. The claws ot the 
forefeet arc only slightly longer th.m those of the hindfcet, and thus offer no close 
resemblance to the long, stning claws of Icionyx. The soles of: the teet, too, differ 
from that genus in being hair\- between the pads. There are well-developed anal glands 
capable of ejecting a malodorous fluid. 

Skull (fig. 15). This, though in most cases ot markedly smaller size, bears a very 
close overall resemblance to that of Ictonyx. The bullae are relatively larger, being 
somewhat more domed and, m absolute terms, almost as long as those ot the larger 
animal. 

Habits. Even less is known ot this animal than ot Ictonyx. Petter (1959) seems to 
be the oiilv author to have observed and recorded these small carnivores in captivitv. 
There are no field notes iif anv kind on collectors' labels. The general habits would 
appear to be quite similar to those ot the polecat; the weasel is nocturnal, lives on small 
rodents, insects, eggs, yoimg ground birds, and lizards; and it shelters throughout 
daylight hours iji subterranean burrows. These it digs for itselt in sott sand, either down 
into level surfaces, or straight into the sides ot dimes. They are ot very simple torin, 
and for the purposes of breeding consist of a single gallery ending in a chamber which 
has been dug deep enough to reach compact soil. This chamber is lett uulined, the 
yoiuig being deposited directly on the soil itselt. 

Once these refuges have been detected it is easy enough to dig out the inliabitants; 
but there is, in northern Atrica, a strong prejudice against doing this because of the 
virtual certainty of being subjected to the ejaculation ot nauseating fluid trom the 
anal glands. Unlike the polecat the weasel stinks at all times without remission, a charac- 
teristic which places it in quite a different category as a domestic pet. Moreover, 
according to Fetter its temper is quite unlike that of the friendly polecat; for, although 
living peaceably with others of its kind, his captive specimens proved aggressive to 
anyone who disturbed them. Handling was therefore difficult except that it was found 
possible to do so when the animals went to sleep since they took some time to wake up 
again. Wlien angr\', striped weasels erect their fur and growl or hiss. 



POECILICTIS 



107 




Flu. 15. Poecilictis libyca: skull, Type ot rolhsdiiUi, B.M. No. 25.5.12.26, <J, X 2 



loS THE CARNIVORKS OF WEST AFRICA 

So tar as is known, there are generally two or three yooiig in a litter. It would seem 
that the period of gestation might be as little as 37 days, a remarkably short period, 
imique in camivores; but this is by no means certain, and as Pctter, to whom this 
observation is due, points out it might be as long as about 1 1 weeks. The yoimg are 
born blind but their ears are not scaled down. They are not absolutely naked but are 
covered, except at the shoulders, with very short (1-5 mm) hair. Some very yoiuig 
w easels brought to Petter in Morocco were entirely covered with white hair, the skin 
being invisible. The dark pattern developed at about 3 weeks. It seems that the eyes 
open at about }\ weeks. As so often happens when breeding conditions are not exactly 
right in captivity, the three young boni to Fetter's specimens were unceasingly carried 
firom place to place by the mother mitil they died from this treatment. The adults, 
however, were still alive at the end of five years or more. 

Taxonomy. Although all the forms now included in Poccilictis were originally 
described as independent species the genus has long been regarded by taxonomists as 
inonospecitic. There is a vast difference in size between North African skulls, from 
Morocco and Tunis, and the normal run of lihyca. The same applies in a rather lesser 
degree to those from the Red Sea coast at Suakin; and there would seem to be a strong 
probability that two, if not three species arc really involved, that is lihyca, vailiauii 
Loche and oralis Thomas & Hinton. However, it is accepted here that West African 
specimens without doubt belong to the first of these. 

Of the 7 named subspecies one, rothschildi, was created from West African material; 
and the 7 British Museum skins, all from practically the same restricted area, all clearly 
accord with Thomas & Hinton's diagnosis. But whether rothschildi differs validly, 
or indeed in any significant respect, from Wagner's iiuihivittata is another matter. 
These authors do not make clear with what reputed nniltivittata specimens they made 
comparison for the purpose of their diagnosis ; but it would seem fairly clear, firom 
their references to the length ajid colour of the pelage, that they must have taken 
examples from the lower Nile as representing this race rather than from the upper 
Nile, whence Wagner had described it. For present purposes comparison has been 
made with tliree Sudanese specimens from Omdurman, Tuli Island (Khartoum) and 
Jebel Hadoza — those three specimens, in fact, which Setzcr (1956) took as typical of 
nniltivittata. From these the West African specimens differ most obviously in one minor 
point: that in si.x out of the seven specimens there is a small, sometimes very small, 
black mark on the tail tip; whereas in the Sudan animals this, though detectable, is 
reduced to the blackening of the extreme tips of a few hairs. The frontal white band in 
Sudanese examples may be a trifle narrower; and possibly there is somewhat less white 
on the upper lips. But these differences are so slight and, with more material, may well 
prove inconstant that it scarcely seems worth while retaining two races. The table on 
page no shows that there is very little divergence of size between the West African 
and Sudanese specimens. This present account therefore rejects rothschildi. 

According to Hoogstraal (1964) it seems likely that some races are less widespread 
than they used to be, and are even, through lack of adaptation to more severe com- 
petition, either extinct or on the way to becoming so. 



POECILICTIS 109 

Poecilictis libyca tnultivittata (Wagner) Soutlierly Striped Weasel 



Tliis form is found over a zone stretching from Senegal to the Nile, covering for 
the most part the Sudan and Sahel vegetation zones but extending a little south into 
the Doka woodland in Nigeria and a little north into the Subdesert in Sudan. How 
continuous its distribution is throughout this belt is imknown; but it must be locally 
comnion since Buchanan obtained four adult specimens on a single day at Faniiso 
and another three days later. At the eastern end of the range specimens exist in the 
British Museum from Omdurman, Tuli Island (Khartoum) and Jebcl Hadoza; while 
in the west there arc 6 specimens from Farniso (near Kano), one from about 150 kilo- 
metres further north-west at Dan Kabba, and one (a skull only) from an unspecified 
locality, all these places lying in Nigeria. Dekeyser (1955) records the race from Air 
and the region of Timbuctu. It has been reported verbally as plentiful at Sokoto 
(Nigeria) ; but whether correct distinction was made between Poecilictis and Ictonyx 
is not certain. 

This is a small form, the head & body some 200 to izo mm, the tail about 100 to 
130 mm. The pelage is oi moderate length, but not nearly so long and loose as in 
Egyptian specimens. The undcrfur is about 10 to 14 mm long, the bristle-hairs some 1 5 
to 25 mm, some of the vcr)' longest white ones rumiing to 30 mm or more. The pattern 
of blackish-brown bands on white is very distinct. Starting from the back of the neck 
there are, as in Ictonyx, three parallel dark lines, separated, except sometimes m the 
case of the middle stripe, from the dark band extending across the crown. These dark 
lines are not so clearly defmed as in Ictonyx and are, indeed, often confluent and always 
somewhat broader than those of J. striatus. At just before the mid-back they begin to 
diverge, as in that species, the outer ones in a very similar fashion, but sometimes 
broken up, curving in again to the root of the tail. But the middle one is quite different 
from Ictonyx since it divides into three, the median portion continuing in a direct line 
down the spine to the root of the tail, the two outer ones describing an ellipse roughly 
parallel to the two extreme outside stripes, joining together again about the level of the 
thighs. There are thus, at the small of the back, five dark bands alternating with white, 
not three as in Ictonyx. The dark colour is continued from the median stripe on to the 
base ot the tail the rest of the tail being whitish, though obviously overlying a dark 
colour. This is due to the banding of the constituent bristle-hairs as described earlier. The 
tip has a narrow black mark, some 30 to 40 mm long, often much smaller, along 
one side. 

The top of the head is blackish; the white band wliich runs across the tace trom side 
to side above the eyes is about 10 mm broad. It continues under the corners ot the 
jaw forward across the throat to meet at a sharp angle at the chiji. The upper lips are 
fairly broadly v,'hite, especialK- near the rhinarium; the lower lips narrowly so. The 
cars have only small white tufts at their extreme apices. The limbs are totally black; 
and so is most of the underside except for small irregular spots of white on the belly. 

In some specimens the dark pattern is considerably overlaid by long white bristle- 
hairs so that it becomes greyed and relatively obscured. 



110 



Tin: ( ARMVOHI S (H WEST AlHUA 



r.iblc fi: Numciic.il d.it.i tur I'ocdiiilis lihym 







"roihschiltli" 




intthii' 


illaia 






(West Africa) 




(Sud 


an) 




J means 




Total range 


o" means 


Range 


VepcMtion 




Sudan 




?Siibdesert 




Ninnbcr iii mciii 


6 


I 


— 


3 


— 


Ccndylobas.il Icii'jtli 


4S-0 


47- S 


4fro-4y8 


46-7 


46-0-47-2 


H.isiLir length 


43-2 


43-'' 


4I-2-44-S 


42-0 


41 ■2-42-6 


Pabtilar length 


20-2 


21-0 


19-3-21 -6 


19-5 


I9-O-20-I 


Zxgomatic lircuith 


2';-2 


27-9 


27-7-29-S 


26-2 


26-1-26-4 


Upper cheekteeth breadth 


17-2 


iS-4 


lrt-5-18-4 


16-5 


I6-3-16-9 


Nasah. length 


— 


— 


— 


9-7 


S-2-IO-9 


Interorbital breadth 


12-S 


12-7 


I2-I-I4-0 


11-4 


n-3-11-5 


Postorbita! constriction 


II-I 


n-0 


10-3-11-7 


10-2 


9-6-10-8 


liraincase breadth 


22-4 


22-Ci 


21 -9-22-8 


22-3 


21-8-23-0 


Toothrow (r — nt^) 


15-5 


16-3 


15-0-16-3 


15-3 


15-1-15-6 


;!■' length 


5-4 


ys 


5-0-5-S 


y3 


5 -2-5 -4 


»i' length 


2-6 


2'7 


2-5-2-9 


2-4 


2-2-2-6 


mi length 


s'5 


5-5 


5-2-5-8 


5-rt 


5-6 


111-2 length 


2-0 


2-2 


1-8-2-2 


2-0 


I-7-2-1 


Head "^' bod\' 


209 


2I.S 


204-222 


210 


— 


Tail 


I OS 


121 


96-127 


135 


— 


Hindfoot 


29 


2S 


27-32 


30 


— 


Ear 


IS 


17 


17-20 


15 


— 


RATIOS (per cent) 












Tail/head & bodv 


50 


5^' 


— 


64 


— 


Zygom. br./condylob. 1. 


rtl 


59 


— 


56 


— 


Hraincase/coiidylob. 1. 


47 


4S 


— 


48 


— 


Uraincase/zygom. br. 


77 


Si 


— 


S5 


— 


I'alatilar l./condvlob. 1. 


42 


44 


— 


42 


— 


Intcrorb./postorb. 


lis 


ii.-i 


— 


112 


— 


p'^/i: — lii"^ 


34-X 


33-X 


— 


34-7 


— 


;.'/'» 1 


20X 


202 


— 


221 


— 


ntilfii2 


-7i 


250 


— 


2S0 


— 



Subraniily MELLIVORINAE Gill, 1S72 
Ratels 

This subf<iinil\' consists ot a single living genus containing a single species: there is 
therefore Httle need to give any separate accoiuit ot it. Fiu^thcr, only one fossil genus 
lias so far been included here. However, at one time the rate! was, together with the 
striped polecat and striped weasel, included in the same subfamily as the skiuiks and 
rlic common European badger, the Melinae; but it is now generally recognised that 
there are obvious differences which are ot more tlian generic standing. As far as the 
three African subfamilies are concerned it is easv to differentiate between the fairly 
bulkv, stockv, harsli-furred animal now to be deah with and the small, loose- furred, 
striped animals ot the Mustelmae on the one hand and the sleek, aquatic Lutrinae on 
the other. 



MELLIVORA III 

Genus MELLIVORA Srorr, 1780 
Ratels 

MelUvora Storr, 1780, Prodroimis mcthodi Mammalium . . . : 34 aiid Tab. A, Mamm. Type species Viverra 

ratcl Sparrman (= Viverra capciisis Schrcber). This name is from the Latin iiic! honey, and voro to 

devour. 
Ratelhis Gray, 1827, in Griffith's Ciii'ier's Animal Kingdoin, 5: 118. Type species Viverra capiiisis Schreber. 

A Latinization of the Africaans name. 
Ratelus Bennett, iSjo, Tlic Gardens and Menagerie of the Zoological Society, Quadrupeds: 13. Type species 

Ratehis nicllivoriis Bemiett. A variant speUing of Ralellns Gray. 
Ursilaxns Hodgson, lS}6, Asiatic Researclws, Ip, i : 61. Type species I'rsitaxns inanrilns Hodgson, an Indian 

race. The name is from the Latin for a bear, nrsns, and for a badger, taxiis. 
Melitoryx Gloger, 1841, Ceineinniilziges Hand- nnd Hilfsbuch dcr Nalnrgeschichte, 1; 57. Type species 

(as selected by Pocock, 1941, The Fanna of British India . . . : 454) Viverra f(i;)f«sij Schreber. The name 

is compounded of the Greek nieli, nielitos honey, and oryx a tool for digging, referring to the animal's 

habits. 
Lipotus Sundevall, 1842, K. svenska VetensAkad. Handl.: 211-212. Type species Ursus niellivorns Cuvier 

{= Viverra capensis Schreber). This name is formed from the Greek leipo to be wanting, and oiis, otos 

ear, with reference to the almost complete absence of any ear conch. 

The range of the rate! is very wide: from Sierra Leone, upper Senegal and the 
southern fringe of the Sahara in the west, throughout most ot the Ethiopian region 
to the extreme south of the continent; and across the Arabian peninsula and Iran to as 
far east as central India. Though most commonly an open-woodland animal it is known 
to occur from the high forest to the subdesert. Despite the vast area and variety of 
vegetation, climate and terraine involved in this, Mcllivora is generally reckoned as 
monospecific; the characters of the genus, therefore, are implicit in the following 
account of this single species. 

MELLIVORA CAPENSIS (Schreber) Ratel or Honey Badger 

Viverra capensis Schreber, 1776, Siingelhiere, pi. 125; and 1777, 3: 450 and 58S. Cape ot Good Hope. 
Viverra ratel Spamnan, 1777, K. svenska VetensAkad. Handl. 38: 147. Cape ot Good Hope. Ratel was 

the Africaans name. 
UrsHS mellivorns G. Cuvier, 1798, Tableau elenienlaire de I'histoire natnrelle des Aninianx: 112. Cape of 

Good Hope. Ursus is the Latin for a bear; the specific name was derived from the Latin niel honey, 

and voro to devour. 
Ratelliis lypicns A. Smith, 1833, S. Afr. Q.J/, 2: 83. South Africa. 
Mellivora leuconota P. L.Sc later, 1S67, Proc. zool. Soc. Loud.: y8, pi. S. West Africa. The specific name 

is from the Greek leucos white, and nolon the back, in reference to the striking pelage pattern. 
Mcllivora cotloni Lydekker, 1906, Proc. zool. Soc. Loud.: 112, pi. 7. Ituri forest, Congo. This was called 

after Major P. H. G. Powell-Cotton who secured the specimen. 
Mellivora concisa Thomas & Wroughton, 1907, Ann. Mag. not. Hist. (7) 19: 376. Type ui the British 

Museum, No. 7.7.S.62, S'. skin and skull, both good. The specific name is the Latin adjective for cut 

off or short, given because the grey colour of the back does not extend as far as usual. 
Mellivora signala Pocock, 1909, Proc. zool. Soc. Lond.: 394, pi. 6i. Type in the British Museum, No. 

9.7.:9.I, sex ?; skin, good, and skull, good except for one nasal nrissing. The name is from the Latin 

signo to mark out, probably with reference to the white head and neck, contrasting with the rest of 

the pelage. 



112 Tin. CAIIM VOIIIA Ol WIST AIIUCA 

.\l<Hii\'rii lnhlhVhiiii 'I'liom.is, iy::5, Ann. M,i-;. uiil. Hitl. (y) l6: igo-iyl. 'i'\ |'c in the liiitisli MuscLim, 
No. ;.S-5-i---?. J, juvenile", skin and skull, hotli good. Tliis w.is named in honour ot the collector, 
(."aptain .^ngus lUuhanan. 

Distribution and general. The ratol (tig. i S) is unique amongst Atriciui carnivores 
hoth 111 appearance and liabits. As regards the tormer, it exactly reverses the usual 
niainmalian colouring ot dark on the back and pale beneath : and as concerns the second. 
It is tiie onlv t^csli-eater that regularly sets out to procure honey as a major part ot its 
diet. The extreme looseness and tougluiess of its skin are also out of the ordinary. This 
animal has therefore excited a good deal of interest in the past 200 years since accoiuits 
were first given ot its habits. The distribution from extreme westeni Atrica to India 
as a single species is also exceptional, almost rivalling that ot the leopard. The often- 
used alternative name honey badger is not, like so many common English names, 
altogether inappropriate; tor in its thick-set, short-legged body, its lumbering, flat- 
tooted trot and burrowing habit it quite closely recalls the true badger of Europe. 

Mcllii'or.i is a not uncommon animal especially in the more easterly and southerly 
part of its range. T. S. Jones (personal communication) characterises it as rare or rarely 
seen in Sierra Leone. In 14 vears he never came across a skin; but he did, however, 
see a living animal crossing a road in car headlights about so km south of Njola 111 
invasive woodland; and was told ot other sight records at Bo and Mano. The ratel. 
indeed, is not ot'ten seen in the wild. This is because it is to a large extent nocturnal: 
and though it is sometimes active during daylight hours it seems to be pretty war\' 
111 this respect and generallv to avoid exposing itself near human habitation. Neverthe- 
less, in its commonest tonus it is a striking animal even bv moonlight, its silver-grc\' 
hack picking up and reflecting the pale moonshine, the rest of the bodv being an 
indistinguishable shadow. This, of course, does not apply to the all-black .animals 
w hich exist in the high forest, especially in West Africa. Whether these are represen- 
tatives of a valid race or simpiv melanos is not clear and will be discussed later. 

Description. Indeed, the animal could be regarded as timdamentalK black, this 
colour being relieved b\- the intrusion ot white over a greater or smaller area ot the 
dorsal surface. In what is generalK' looked upon as the most typical form the whole 
upper surface trom the forehead, just above the e\ es, to the root ot the tail is pale 
grev or even, at times, pure white. This pale colour extends as a broad parallel-sided 
band, the width of the head, across the crown and the top ot the neck as tar as the 
shoulders; it there broadens and curves down deeply over the flanks, narrowuig again 
above the thighs, terminating at the rump or actual!}- extending onto the tail. There 
is a sharp line of demarcation between this dorsal patch and the black, the latter covering 
the front and sides of the face, the torelcgs, shoulders, low er flank, hindlegs and thighs. 
The entire luiderside is also clad with black hairs; but, except sometimes tor the throat, 
so sparselv that the skin is clearly visible. However, this general picture is liable to a 
good deal of variation. Even at its most "typical" the colour is generally more inteiisek 
white over the head and neck, becoming more .nul more adulterated with black 
|iosteriorl\-. It is the diminisliing of the white iiiriueiice, always progressiveK' from 
I town to I. 111. iliai >'ives rise to ditfereiit "tonus". Soinetniu's 111 these dorsal variations 



MELLIVORA 113 

tlic divisoii bct\\ccn black and wkite is more or less clear-cut; in other cases tlie black 
is '"frosted" with white, verv lightly over the hindquarters, more densel\- over the 
shoulders and neck. 

The pelage, though contorninig throughout the species to an overall basic pattern, 
nevertheless exhibits some variety in the fuicr details of its composition. Two elements 
are present: abundant long, strong, straight, flat bristle-hairs, which dominate the 
pelage; and much finer, curly, terete underfur. Both are most commonly cither all- 
black or all-white. The white bristle-hairs arc approximately 0-15 mm in width, their 
length varying between 2Z and 30 mm; the black arc a trifle less broad, about 0-13 mm, 
but their length is generally somewhat greater and may, indeed, reach as much as 38 
to 40 mm, especially on the lower back. There is, in fact, some tendency for the fur 
to lengthen from neck to rump. The underfur is conspicuously fmer though, at 004 mm 
diameter, still pretty coarse by comparison with softer-furred carnivores. In the majority 
of cases bristle- hairs and underfur arc of one colour throughout their lengths — all-black 
or all-white; but si^iiata is exceptional in that some of the white bristle-hairs have black 
tips and some ot the black bristlc-hairs a white subterminal band, not sharply detmcd 
but about 5 mm broad. 

The white areas of pelage ma^• consist purely of white bristle-hairs and ■\\'hite imder- 
fur; but t]uite commonly the latter is black though, sijice it is short and well overlain 
by the white bristle-hairs it has little or no eifect on the overall colour. However, 
when such an area is invaded bv black bristle-hairs as well, the total effect is grey. The 
black areas are generally pure black; but sometimes there are scattered white bristle- 
hairs amongst it. The underfur is normally about 18 to 20 mm long and is generally 
relatively sparse and so plavs little part in the overall tcxtiurc or appearance of the iur; 
but in some cases it is distinctly more abundant, more undulate and much longer, 
reaching 24 to 27 mm. In the form known as si<^iiata the basal portion of the bristle- 
hairs, as well as the underfur, tends to curlincss. 

The body of the ratel is fairly long — some 60 to 70 cm — but distincth- thick-set 
and broad across the back. The skin appears to be remarkably loose; and many field 
observers have maintained that it almost seems as though it were independent of the 
body and that the latter could turn and twist freely inside it. The head, for this bulky 
body, is unqcstionably small, rather flat, and with a short muzzle, the whole tront of 
the face being black. The eyes are not very large; and the ears, though shaped and 
folded rather in the manner of human ears, have no independence of the head but 
appear as little more than ridges of skin. They have no bursa. The legs are short and 
sturdy, the animal standing some 23 to 28 cm at the shoulder. The feet have 5 toes each, 
armed with very strong claws, those of the hindfeet short, but those ot the front 
remarkably long, the central three of much the same length (25 to 30 mm) and forming 
a unit connected by short webs, the ist and 5th digits lying posterior to these. The soles 
are thickly padded and naked to the wrist, the posture being semi-plantigrade. The tail 
is short, naked below at the base but clad with long hairs for the rest, though narrow 
in form since these are fairly close-lying. The circumanal region is hairless; there are 
two very large scent glands situated within the anus and these according to Pocock 



11-1 



TlIF, CARNIVOUI-.S or W'fST M I! 1 ( A 




Fin. i6. .l/i7/ii'iira uipciisii: skull, li.M. No. 26.S.7. 



I ; lateral \'icw 



([920) "discliargc cnpiousK a surtot.itmg Huid cxactK- as in tlie Skunks . . .'" There 
are two pairs ot abdominal mammae m tlie temales. 

Skull (tigs. 16 and 17). The specimens in the Bntisli Museum sliow. irrespective ot 
age, a wide range of size and some variety ot detail. All, however, are very solidK' 
built; the sutures fuse at an early age, and in tully adult skulls quite literally leave 
absolutely jio trace of independent bone structure. The braincasc is relatively rather 
broader than m the dogs. There are slight, pointed postorbital processes but little iji 
the way ot interorbiral constriction and usually not much postorbital narrowing, the 
trontal aspect riuming almost parallel-sided from the temporal region to the rostrum. 
The last is ver\- short and broad, the nares ^videly open. The same applies to the posterior 
narcs, the intervening nasal structure being Imely complex ijidicating a highly developed 
sense ot smell. The zygomatic arch is strong, the glenoid fossa deep. In fully mature 
skulls there is a slight sagittal crest which joins a very marked, though shallow, supra- 
occipital crest. The whole posterior aspect ot the skull is one ot firm solidity of build, 
including the foramen magnum and its adjacent condyles. The anterior palate is short 
and broad: posteriorly it continues, parallel-sided, well back of the toothrows. The 
bullae are large though not highly domed, and like the rest ot the skull extremely 
strongly built, being fused posteriorly to large paroccipital processes. The mastoids 
arc well-developed and prominent. 



MnLMVt)liA 



"5 




Fig. 17. Mcllivora capcnsis: skull, B.M. No. 26. X. 7.1, j, x \ ; palatal & d 



orsal views 



The dental formula is jyyy = 32- The teeth seem often of rather irregular develop- 
ment, one or more being exceptionally small, set at an unusual angle, or lacking. 
The type o( si(^iinta has a second lower molar on the lett side but not the right. The 
carnassials appear to be slow to erupt; and in several skulls, even with wcll-ossified 
sutures, two sets of teeth exist, what is presumably a very adult-looking milk dentition 
being replaced by the fuial set. The cheekteeth are in the young skull very sharply 
cuspid; but, for an animal reputed to live to a great extent on relatively soft food, all 
the teeth seem to wear very badh'. The incisors are often pretty irregular; but when 
properly developed form, top and bottom, a very compact, strong cutting unit, the 



M(' Till ( AIINUIIIU s Oi WrST A I R 1 r A 

upper OIK'S being ei>n>Klcralil\' Luger tli.m the lower, the outer pau" iji eacli case beijig 
somewliat bigger than tlic others. Though it is not always detectable, in the lower jaw ii 
is bitid, i-2 tritid, and 13 generallv unnotched, though it occasionally shows faint traces 
ot trifidv. The canines arc stout but. tor carnivores, relatively short. 

Habits. One ot the foremost characteristics of the ratel is its extreme bravery and 
general toughness. It appears to be quite without fear, and when flight seems of no 
avail will turn savagely to attack man or anv other creature. It can take any amoiuit of 
punishment and is so tireless in combat that it has been known to e>draust and overcome 
tar larger .animals. Indeed, one is said to have killed a buttalo in the Kruger National 
i'ark. One ot the things that makes it so ditiicult tor dogs or anything else to deal with 
is the e.xtraordinan.' tougluicss and looseness ot its skin. The tirst makes it well nigh 
impossible tor teeth to penetrate; the second enables the ratel. in spite of being held, 
to twist and tuni so readily that it can inflict a fierce bite on any aggressor that has 
seized it in hand or mouth. Sikes (1963) tound that the only sate grip was by the skin 
on the back ot the head; anvwhere else, including the scrufl of the neck, was highly 
dangerous. How impenetrable the skin is is illustrated bv an accomit given by T. 
Rawson-Shaw (in Fleetwood, 195N) in which he relates t];at blows from a matchet 
■"which would have cut any other animal of that size in half merely bounced ofl", 
leaving a shallow gash on his hide, and it took about ten of these and tour .22 bullets 
to kill it". 

Others have tound much the same; and it is commonly held that a direct shot m 
the head trom a tairly powertul rifle is the onlv certain way ot killing a honey badger. 
In the Bauchi area of Northern Nigeria it was well knowii that arrows and spears were 
almost useless and that the best, indeed reputedly onK', way ot certainly killing one ot 
these animals was to club it over the back ot the head. As opposed to these views the 
plain tact remains that though a tew are damaged a very high proportion of the skulls 
m the British Museum have perfect craniums. In tussles with dogs it is usually the ratel 
that succeeds in sinking its teeth into its opponent, hanging on tirelessly, with jaws 
clenched like a vice, despite being banged on the ground or against trees or rocks, and 
tmally, ^\hen the dog is completely exhausted, makhig ott apparently none the worse 
tor the experience. Sikes (1964a) tound that, m pla\', a captive ratel being swiuig 
round hanging on to a sack actually appeared to enjov being bumped roughK up and 
down on the ground. 

Caught young, m tact, these animals take very kindh' to captivitN and become not 
only docile but activcK- triendK' (Sikes, 1963 and 1964a; Hoesch, 1964). They are 
attectionate and very playtul, even inventing games (Sikes, 1963). They recognise 
tnends and voices, distinguishing various tones in sumntons, command, warning or 
reprimand. Hoesch's pet used to nibble in a restrained tashion at hands or clothes m 
triendship. But in captivit)', as 111 the wild, ratels can with some abruptness work 
themselves up into an ecstatic tury. During such a bout the hair stands on end, the 
.mimal toanis at the mouth and becomes almost literally blind to any external calming 
influence. Such moods, however, quickly vanish. Sikes (1964a) supposed that these 
rages, which ot coiu'se are an invaluable reaction to aggression in the wild, w^erc 
brought on or auirmented b\- unusualK' larije releases ot adrenalin into the blood 



MELLIVORA II7 

Stream. Hoesch (1964) found liis captive animal to be clean in its habits. Sikcs (1964a) 
observed hers to urinate or defaccatc into a suitable hole, which if not of approved 
size or shape would be industriously altered. The droppings were rarely covered. The 
scent emitted by ratels affects different people in different ways, some finding it 
extremely repulsive, others (Sikes) after a time not unpleasant. All agree that it differs 
from the ordinary musky smell of other mustelids. 

Ratels are mostly solitary animals but may at mating time hunt in couples, or a 
mother may be seen with her two young ones. When not active in the search for food 
they shelter in holes, either in the ground, in tree trunks or logs, or amongst boulders or, 
in suitable country, in a cave. Earthen burrows may be dug by themselves, this being 
one of the functions of their powerfully clawed forefeet; but, like so many other 
subterranean dwellers, they make full use of aardvark holes, warthog dens or termite 
mounds. No accoiuit has beeai given of the extent or nature of these underground 
homes, whether they have side passages or any kind of lining. No one appears ever 
to have attempted to dig out a ratel; and, mdced, it can be gathered from the previous 
paragraphs that this might well prove a hazardous undertaking. 

The ratel is generally recorded as being purely nocturnal, a reputation doubtless 
deduced from many a midnight raid on fowl houses and the fact that it seems to be 
relatively rarely encountered by day. However, not unusually it is active at dusk or at 
da\vn ; and in certain conditions it is undoubtedly on the move in full daylight. Probably 
these animals suit their habits to prevailing circumstances. In areas where they are 
liable to be much harassed by man they make constant use of the cover afforded by 
darkness; but in remote districts of low human population they probably become 
bolder. Certainly, captive in zoos, they fmd no difficulty in adopting a tully diurnal 
habit; and their known behaviour in coimexion with honey seeking, detailed later, 
would seem to postulate pretty regular daylight activity. 

When on the hunt ratels move at a steady jog-trot with the fore toes characteristically 
turned in. Tliis at best is not so fast as a young man can rmi, and it has been often said 
that they can keep tliis pace up tirelessly; but this is notliing more than a guess hazarded 
on the leisurely nature of the gait and of the known general stamina of the animal. 
No one has, in fact, ever followed a ratel very far; as soon as it becomes aware of 
being chased, as it soon must on open groimd where it can itself be seen, it makes for 
cover in undergrowth in which it is quite impossible to keep track of its direction. 
Nor have any experiments been carried out with marked animals to determine, 
without continuous observation, how far they travel. During the course of its hunting 
or other excursions a ratel defmes or redefines its territory by means of its scent glands, 
squatting at intervals to press its anal region on the groimd or against strategic trees 
or rocks. As no specific field investigation of the ratel has as yet been carried out the 
extent of the hunting range is unknowii. 

Despite the name honey badger these animals are by nature essentially carnivores, 
and the pursuit of food therefore concerns itself a good deal with flesh of various kinds. 
Almost certainly a hungry ratel would take any sort that came conveniently to hand, 
including carrion; but its normal range of food embraces small rodents, birds, eggs, 
insects, lizards, tortoises and frogs. Some, at least, of the rodents it may dig out from 



IlS THE CAUNIV01U;S Of WhST AflUCA 

their burrows — gcrbils, groinid-squirruls aiul otlicrs; the birds arc sucli as (.iwell and 
nest upon the ground or pay it trequent visits: trancolins, bustards, plovers, quail, 
doves and so forth. Frogs, too, arc sometimes dug up: Angus Buchanan (1926: 251), 
writing ot the race ot Mclliuora from Air named after him refers to "a regular warren 
ot holes scooped out in the night by a ratel 111 search ot dormant trogs buried in the 
sand a toot or two beneath the surface"". 

Ratels have sharp, backwardly-pointing papillae on their tongues which enable them 
to deal etlectively with tough foods. Tortoises present no difficulty to this animal 
which can readily crush the shells with its very powerful jaws. Snakes also arc taken, 
including highly poisonous ones; and that they are not always of small size is shown 
by an extraordinary record in Aliican IViU Life, 1964, 18 : 37 which tells ot a ratcl 
tighting, killing and feeding on a python some lo or 11 feet long. The begiiuiing of 
this exceedingly noisy and energetic combat was not seen, but the battle must have 
continued tor some half an hour, at the end of which the snake "was so mutilated that 
it looked as it it had been run over by a train". The ratel has, at the Asiatic end of 
its range, earned a reputation tor excavating yet another kind of food: inadcquatelv 
buried human corpses. Though doubt has often been cast upon this, Pocock (1941 : 465) 
says that the truth ot the belief has been proved beyond doubt. Like other small carni- 
vores ratels welcome vegetable food such as berries and other fruits, roots and bulbs 
as a regular part of its diet. They inquisitively lift stones or tear flaking bark from 
trees. 

The ratel's partiality for birds very often leads it into becoming a pest in farmyards 
or private compounds where domestic fowls are kept. Such is its strength and per- 
sistence that it is difficult to keep out. It has been recorded (Fleetwood, 1958) as ripping 
thick planks from a strongly built hen-house, or burrowing beneath stone foiuidations. 
A ratel is known to have climbed the mud wall of a hen-house to a small window 
covered with wire netting fastened down with staples, ripping this obstruction aside 
to gain entr)'. Another, unable to get in in any other way, tiumelled under the wall 
and up through the floor, which was paved with large stones set in mud mortar. The 
sturdy, muscular legs and long, strong claws of the forefeet enable tiuuielling to be 
effected speedily even in hard groimd. Once entrance has been effected to a fowl- 
house it often happens that everything inside is slaughtered and eaten, nothing beyond 
a few scattered feathers remaining. In one episode recorded by Fleetwood (1958) 
17 Muscovy ducks were lost, and in another 36 halt-grown pullets; and the same sort 
ot thing has happened to many others. Rawson-Shaw, to whom these figures are due, 
reached the conclusion that a ratel may have an ancillary following of other carnivores, 
such as civets and jackals. For just as the lion's kill may soon be made use ot by a waiting 
band of scavengers, so other less adventurous and able carnivores may well be at hand 
to take advantage of the courage, strength and determination ot this pioneer burglar. 
It should perhaps be made clear that in many such cases it has been quite defniitely 
proved, by shooting, trapping or other means, that it was indeed a ratel responsible 
for the entry and at least initial damage. It must be added that visits may be repeated on 
successive nights either until there is no further attraction or the raider has been killed. 
Sikes (1963) found that in captivity a young ratel would eat a whole dove, but that 



MELLIVOKA 



119 




i::o rm c ai!Ni\ ours (ir wr.si .\i kka 

an adult requires a hill-growii towl. Skiu, liau% tcatlicrs and bones ot a vietini arc all 
eaten as well as the flesh. The food is held down by the forepaws. 

But the tood and the robberv for which Mcllimtra is most widciv hinicd concern 
lu>ney and wild bees' nests. Since about the middle ot the i6tli century, when a mission- 
ary priest wrote an account of Portuguese East Africa, ojie of the travellers' talcs that 
must have circulated in Europe was ot a bird that for its own ultimate benefit deliber- 
ately guided men, by song and a clapping of the wings, to stores of wild honey. It was 
not luitil some 200 years later, however, that this story was extended to include the 
ratel as possibly the bird's chief medium tor breaking open the nests and laying bare 
mutually beneticial supplies of honey, wa.x and grubs. The story of this strange co- 
operation, it not symbiosis, between animals ot othcrv,'ise quite different ways ot 
lite, and wliich set the natural history world wondering tor tuo centuries, was due to 
the writings ot a Swedish naturalist, Sparrman (17S6, 2 : iHo-i8i) who accompanied 
Captain Cook on his second voyage to the south seas and explored the interior ot 
South Africa on the way back. His historic, though certainly not reliable, account 
make both amusing reading and a good starting point tor the examination ot this 
reputed partnership: 

"The itilcl, a sort ot weasel or badger, by nature destined to be the adversary of the 
bees, and the unwelcome visitor ot their habitations, is likewise endued with a particular 
faculty tor discovering and attacking tliem within their entrenchments. His long 
claws, besides assisting him in digging the dark subterraneous passages which serve 
him for an asylum, are likewise ot use to him in the occupation he is frequently 
employed in of undermining whole colonies ot bees. Now, as a man placed at the mast- 
head can easiest descry a sail or land at a great distance about sun-set, so probably 
this time of the dav is the most convenient for the ratel to look out for his supper; 
for he is likewise said to be particularly attentive to his business about sim-set, when he 
will sit and hold one of his paws before his eyes, in order to modify the ra\-s ot the 
sun, so as to render them inoffensive to his organs of sight, and at the same time to 
have a distinct view of the object of his pursuit: and when, ui consequence of peering 
in this manner cin each side ot his paw opposite to the sun, he sees any bees fly, he kjiows 
that at this time they are going strait forward to their own habitation, and consequently 
takes care to keep in the same direction as that in which the\' fly, in order to tnid 
them. He has besuk^, as well as the Hottentots, the Caffres, and the peasants ot the 
Cape, the sagacitv to follow a little bird, which flics on by degrees with the alluring 
note o( clu'ir, clicir, clwn, and guides its followers to the bees" nest . . . 

"Those bees' nests which are built up in trees, are m no danger whatever from the 
ratel. In the fust transports of his rage at having sought after these bees in vain, he uses 
to bite or gnaw the trunk of these trees; and these bites arc sure marks for the Hotten- 
tots, that a bees nest is to be found up that tree. I should myself have entertained many 
doubts eonceniing all these properties attributed to the ratel, had I not obtained various 
accoiuits of this curious animal, entirely corresponding with each other, from many 
experienced farmers and Hottentots living in different parts ot the coiuitry". 

Tlie bird to which Sparrman made reference is known as the greater honey guide 
(Jihlicdtor iiulicdtor Sparrman), a representative ot a family, the liidicatoridae, com- 



MHLLIVOKA 



prising four genera, largely African but extending their range to Malaya and the East 
Indies. Nearly all of the eleven species are known to eat wax as part of their diet; 
and all, so far as is known, arc parasitic, laying their eggs cuckoo-fashion in the nests 
of other birds. Some two or three species without question seem to guide, or to 
attempt to guide, men to bees' nests; but whether similar behaviour could in truth be 
related to ratcls as well, as Sparrman had asserted, was not so clear. It so often happens 
in natiu-al history that early held beliefs or superstitions, especially when they concern 
the unusual, are copied from work to work, year after year, without closer inquiry. 
They make good reading. Like others, the story of the ratel and the honey guide was 
incessantly repeated, both verbally and in written accounts; yet critical enquiry showed 
it to be nothing more than hearsay, no wholly reliable and competent observer ever 
having openly laid claim to actually witnessing the alleged behaviour. The story, 
in fact, might have been on a par with the pelican pecking its breast or the porcupine 
shooting its quills. 

At length, however, Friedmaiui (1955), Curator ot Birds in the United States 
National Museum, spent many years making an extremely interesting and detailed study 
of the Indicatoridae and their habits. He came to the conclusion that while Sparrman's 
account was in many respects inaccurate, the reputed association of the ratel with the 
honey guide was true. He arrived at this opinion only as the result of a great deal of 
sifting of old records and circumstantial evidence. No one has, in fact, ever witnessed a 
complete sequence of events from the initial attraction by the bird of a ratel's attention, 
through the stages of guiding, to the discovery and breaking open ol a nest. It is merely 
that McUii'ora and Iiulicatorhavc been seen together in a number of isolated circumstances 
that are capable of being joined together into a coherent series of events. Whether the 
behaviour, either in its observed parts or presumed whole, has been correctK' inter- 
preted is altogether another matter. 

Briefly, what takes place, with a ratel as with humans, is that the bird, perched not 
very high up on a tree, utters a great deal of unceasing chatter, climiii, chiirra, cinirra, 
to attract attention; and having achieved this end flies, still chattering, a few yards 
ahead and settles on another perch. The ratel responds to this chatter with occasional 
grunts or growls. This is continued imtil within soimd of a bees' nest the bird becomes 
quite silent. Such a course may cover a few yards or half a mile or more; and, corres- 
pondingl)-, a tew seconds to half an hour. The route taken, except when extremely 
short, is never direct but circuitous, meandering or even criss-cross. Having arrived 
in tlie vicinity of a bees' nest the honey guide will remain silently perched quite 
patiently for a very long time luitil the store of honey has been laid bare and pieces of 
comb are left lying about to feed on. 

This is no place to discuss in any detail the alleged guiding habits of Iiidicalor except 
in so far as may be necessary to draw attention to the fact that the apparent co-operation 
between bird and manunal is not to be lightly accepted at its face value. The uncritical 
animal lover may all too easily follow others into a trap. When this apparent partner- 
ship was first reported it seemed to be nothing other than a case of an intelligent and 
hungry bird discovering a store of honey, seeking out a better equipped ally, and 
deliberately c.xcituig this essential participant to curiosity and leading him or it to the 



THE CARNIVORES OF WEST AFRICA 



mutually desirable spoil. All of this is very nnich open to question. It has been amply 
shown that hiuiger is not tlic operative excitant snice birds exercise the guiding ritual 
even with tull stomachs. It is very possible that there is, indeed, no deliberate seeking 
out ot a helper but merely that the chance sight of certain animals leads to an agitated 
reaction in the birds. The complex path taken to reach an apparent goal suggests that 
no specific goal was actually in sight at the commencement but that by criss-crossing 
the bush over a sufficient distance a nest is eventually hit upon. It is a common fact 
that in much ot the African woodlands there arc plenty of wild bees of one knid or 
another and tliat by quartering the bush it is not excessively difficult even for a human 
being to discover some sign ot a nest in a tree or a bank. There may, therefore, be no 
true guiding, the expedition being, in fact, little more than a voyage of mutual dis- 
covery. 

If the bird stimulates the ratel to action with its chatter, tlie ratel in its turn keeps 
the bird in a state of excitement by responding from time to time with its deep growling 
grunt. African hunters who know this utter the same sound with a view to keeping 
the bird interested. Guiding does not necessarily take place to a nest well stocked with 
honey; it may be to an almost new one; and it is knov^ai that a long-deserted nest 
containing good stores of honey holds no attraction for the honey guide. It is the 
sight or soiuid ot numerous bees that brings the bird to a halt and causes its chatter to 
cease. This is in the vicinity of, rather than actually at, the nest; but the assertion that 
the bird will, if necessary to an unintelligent helper slow in detecting the desired honey, 
go further and point to it with its bill is, as might be expected, nothing more than a 
picturesque embellishment. 

These arc some ot the many aspects ot so-called guiding brought out by Friedmann's 
classic study, which make it clear that it is not the simple purposetul act that early 
naturalists assumed it to be. A tew other points may be glanced at in conclusion. 
Firstly, it is interesting to speculate on how this parmership between bird and beast 
evolved; what can have been the original stimulus, and the route by which it came to 
its present state. The co-operation serves no vitally essential purpose as it does in 
true symbiosis; honey badger and honey guide can each flourish without the assistance 
of the other. There are, indeed, signs that the habit is, troni force ot changing circum- 
stances in modern Africa, on the decline. We are, in this present work, strictly concerned 
only with the habits o{ Indicator iudiccitor in relation to the ratel; but it may be briefly 
added that it is possible that the bird occasionally leads baboons to honey, and may 
make, apparently wholly unsuccesstul, attempts to interest other animals. A caution 
should perhaps be given that the course taken by the bird may not always lead the 
follower profitably to a bees' nest; for more than one man has foimd himself suddenly 
face to face with a dangerous animal. Such an occurrence, once regarded as deliberate 
on the part of the bird, is most probably a purely accidental outcome ot quartering 
the bush in its search for a concentration of bees. Finally, for those who may tuid 
themselves in proximity to a noisily excited bird in the grass-woodlands, the greater 
honey guide is of moderate size, some 120 to 150 mm long with an additional tail of 
about 60 to 80 mm. Its coloration is variable with age, sex and other factors but in 
general is dark drabby or olive brown above, the wing coverts and some of the tail 



MELLIVORA 123 

feathers white-edged, and with a small chrome-yellow bar on the shoulder. In the 
young the breast is bright yellow. 

Sparrman asserted that the ratcl was quite unable to climb and that bees' nests in 
trees were therefore unattainable. This is not strictly true. A ratel may not be able to 
climb up a siurfacc that others no foothold, or very far up a vertical smooth-barked 
trunk; but given a reasonably rough surface, and particularly one at a slight slope, 
such as so many open- woodland trees have, it is a pretty competent climber. A mud wall 
or a coarse-barked tree can be scaled with some ease ; and any African bee farmer who 
fails to set liis hive sufficiently high and along a sufficiently narrow branch is liable to 
fuid it robbed. Since a fair proportion of wild bees' nests are made in hollow trees this 
ability to climb is useful if not important. Ratels can climb wire-netting or expanded 
metal without difficulty, bite through wire, scratch ajid gnaw to pieces strong posts, 
and burrow through hard earthen or wooden floors so that a cage to retain a captive 
animal must be of very solid construction. 

Little is known of mating or breeding despite the many specimens which have been 
kept in zoos. The gestation period, not very accurately determined, is said to be about 
six months. There are commonly two yoimg at a birth, bom blind. They are moved 
from place to place if necessary in the mother's mouth. They utter a plaintive whme. 
Details of development are unknown ; but there are two or three records of ratels 
having lived in captivity for approximately 24 years. Hoesch (1964) found his tame 
animal to drink only rarely; but it used its drinking bowl for keeping cool. 

Taxonomy. Since its fu-st alleged variant was described in 1792 M. capciisis has 
been something of a favorite for the naming of subspecies. G. M. Allen (1939) lists 
II currently recognised for Africa, and Ellerman & Morrison-Scott (195 1) 5 more 
for Asia. No less than 5 races are, indeed, said to occur in West Africa. The points 
v^'hich have been taken into consideration in the diagnosis of races are the extent of the 
pale dorsal colour; its degree of whiteness of greyness; the annulation of or absence 
of it from, the hairs; general size. Differences based on these are often clear enough in 
single specimens; but whether a variation has any constancy throughout a population 
and hence constitutes a valid local subspecific distinction can be established only by the 
existence of adequate material. Of the 12 West Africaji specimens available in London 
from which to confu-m the validity of 5 reputed races, 2 have no skins and are therefore 
racially indeterminable; and of the 10 remaining, only in three cases (Air, Tarkwa and 
Abenasi) is there more than a single specimen from a given locality. Of the 8 specimens 
which have skulls, only 4 can be regarded as mature; and of these, 2 are badly mutilated 
and incomplete. From such material it is quite impossible to reach any imequivocal 
conclusion. 

In the matter of pelage pattern, it is dirticult to believe that Thomas & Wroughton's 
concisa, with merely a small area of the white lacking from the lower back, is anything 
but a minor individual variation, the more so as signs of a comparable reduction is 
to be seen amongst skins from other regions. The possibility, though unlikelihood, 
of this occurred to the authors; but they held, if such did prove to be the case, that its 
smaller size was alone sufficient justification for their proposed race. As the type of 
omisa is a young male w ith quite unworn teeth, the sutures not fully ossified, and no 



124 THE CARNIVORES OE WEST AFRICA 

development ot crests, size comparison is scarcely a reasonable basis for argument. 
Size plays the major role, too, in Thomas's huchanani; but the type of this is quite 
juvenile, with unfused sutiurcs, teeth not tully erupted with some of the milk dentition 
still in place. It is true that Thomas cited a paratype which, though described as a 
fully adult female with worn teeth, has skull measurements (as given by Thomas 
since this skull is no longer available for assessment) that differ only minutely from 
those of the juvenile type, as shown in the table on page 127. It is, ot course, possible 
that the ratels of the inhospitable Subdesert. feeding largely on frogs and lizards, 
are truly of a smaller size; but argument based on present study material is not altogether 
convincing. 

It is tempting, and not entirely unreasonable, to suppose that differences in the grey 
dorsal patch have little to do with locality and are nothing more than the sort of 
individual variation that one might expect in a pelage of this kind. But, looking 
outside West Africa to districts from which a much wider range of study material 
exists, there docs often seem to be a strong overall uniformitv in pelage coloration 
from a single area — as, for example, in the dozen skins from Somaliland. The argument, 
however, is weakened by the frequent ability to match patterns from widely separated 
regions, a striking example of this being a skin from Suakin on the Red Sea (No. 
4.8.2.27) having a verv distinctive pure white flank stripe precisely similar to one from 
Namaqualand in South-west Africa (No. 4.2.3.54). The hiicliaiiaiii skins vary somewhat 
amongst themselves; and when the reputedly mature female is put with the Somaliland 
specimens it is seen to differ very little from them except for size. 

The situation might he explicable by the postulation of ecological rather than geo- 
graphical races. This brings us to the consideration of cottoiii, the all-black form. It has 
been suggested, and is very possible, that this is nothing odier than a melano which, 
in the nature of these things, might turn up in any population. Yet the fact that the 
onK- 5 knowii Mcllivom skins from Ghana are all of this type miglit suggest that the 
black coloration has some kind of racial connotation rather than that ot an occasional 
mutant. The form was originally described from the Ituri forest, over 3000 kilometres 
away on the other side of the continent; and it is also known from Cameroun. The 
factor which these three regions have in common is high forest; and it would seem that 
cottoiii might, indeed, be an ecological form typical of wet, closed-forest conditions. 
Yet this, too, would appear to be negatived by the fact that of Bates's 4 Cameroim 
specimens, 3 are all-black and 2 have very white backs presumably typical oi Icnaviota. 
It is true that onlv one of these four specimens, a white one, is associated with a definite 
localit\', Bitve; but practically all of Bates's mammalogical collecting was carried out 
in a restricted area of high forest. It is recorded also (J. A. Allen, 1924) that partly white 
forms (which, incidentally, appear from photographs to be excellent matches for the 
West Afncait signata) occur in the Congo in the same region as the black form. 

Age, too, mav play its part in the question of blackness. The three black Ghana 
skins accompanied by skulls are all very old animals with extremely worn teeth; 
and Pitman, quoted in Shortridge (1934), expressed the opinion that "As the Ratel 
gets older the white patch on the back becomes gradually darker, until in some very 
old males it is onlv just distinguishable ". On the other hand, if this ^\cre so one \\ oiild 



M K L L I V O H A 



125 



have supposed that attention would have been drawn to the fact in coiuicxion with 
the several specimens known to have lived in zoos to a ripe age. 

The situation, therefore, is confused. The present writer is of the opniion that many 
of the variations of pelage are purely individual and may turn up in widely separated 
places though there is possibly a tendency for some to be commoner in certain ecological 
conditions than others. They are, thus, rather "varieties" than regional or truly eco- 
logical races. They are retained herein as such and described below for what they are 
worth. 





Fig. 19. Mcllivom capaisis: dorsal coat patterns, showing the extent of white in tlie v.irioiis 
tonns: a. laiconota, h. biicliaimiti, c. coticisa, d. sif;iialii, c. coiioiii 

A table showing such measurements as are available for West Atrican specimens 
will be foimd on page 127; but since for the reasons given above these figures necessarily 
embrace a mixture of ages from juvenile to very old they are not reliably comparative. 
The diagrams in fig. 19 will aid in the identification of the ditfereiu forms; but 
there seems no reason why all manner ot intermediate patterns should not tinn up. 



126 THE CARNIVORES OI WEST AFRICA 

Mellivora capensis leuconota P. L. Sclater White-backed Ratcl 

Tliis was originally said to conic troni "West Africa" without closer definition; 
but it has since been held (e.g. G. M. Allen, 1939) to range from southern Morocco 
to the tormer French Congo. No specimen definitely identified as Iciicoiiotn occurs in 
the British Museum; but it is to be assumed that two of Bates's Cameroun specimens, 
one from Bityc, arc this form. Though onh- the apparently smaller of the two has 
external measurements (given on page 127) it appears to rank among the largest of all 
ratcls. 

In this torm the entire upper side troni the hice to halt way along the tail is pure 
creamy white with a minimal and almost undetectable admixture of black hairs. The 
white hairs may have narrow black tips, but these, agaiii, are almost invisible. 

Mellivora capensis cottoni Lydckker Black Ratel 

This form occurs in the liigh forest, or at its edge, in various parts from at least 
Ghana to the north-eastern Congo. The known West African specimens all come from 
Ghana: Tarkwa, Abenasi and Oda. 

As its name implies this form is topically entirely black though some skins liave an 
extremely small number of scattered white bristle-hairs only detectable on close in- 
spection. But there is one exceptional partial skin from Abenasi in which there is an 
even scattering of white or white-tipped bristle-hairs throughout the whole of the 
deep brown, rather than black, dorsum. The coat is thin and harsh, but the whitish 
hairs become rather denser over the top of the head. 

Mellivora capensis concisa Thomas (^' Wroughton Lake Chad llatcl 

Only one specimen of this is known, from Lake Cliad, but it is almost certainly only 
a minor variant of the ratel that occurs eastwards throughout the Sahel and Sudan 
zones, as tar as Somaliland and there called hrockmani Wroughton & Chccsman. 
The dorsal coat consists of a preponderance of very long pure white bristle-hairs 
amongst long fine black undcrfur and a minority of black bristle-hairs. The feature 
which is held to distinguish concisa is the fact that, for all practical purposes, the white 
bristle-hairs are entirelv lacking from an area starting on the spine at about the small 
of the back and broadening a little to the tail, which is also all black. 

Mellivora capensis signata I'ocock Speckled lUtel 

The t\pe, from an unspecified locality in Sierra Leone, via the London Zoo, is the 
onlv specimen known in the British Museum; but according to Monard (1940) tlie 
race occurs at Catio, Portuguese Guinea. The characteristics of the type are that although 
the pelage is the normal dense white over the crown, the pale colour starts to thin out 
over the neck and shoulders, and continues thence to the rump only as scattered white 
"ticking " amongst essentially black fur; and also that the white in this part of the coat 
is not due to wholly white bristle-hairs, as in other forms of ratel, but to subterminal 
white anniilation of otherwise black bristle-hairs. The "ticking", already scanty. 



MELLIVORA 



127 



tadcs out entirely just torw arc! of the rump, the whole hindquarters and tail being 
completely black. The skull has an extra lower molar (nio) on the left side. 

Mellivora capensis buchanani Thomas Air Ratel 

The question of the small size of this form, known ojily by three specimens from 
Elmcki River and Tarrarc River in Air (Subdescrt), has been discussed above. Apart 
from this matter of size it is doubtful whether this form differs essentially from the 
Sahcl and Sudan zone ratcls referred to above under coiicisd. It is a little whiter over the 
head but this may be due to the young age of the specimens now to hand. For the 
rest, it is essentially a grey-backed form, the pelage being an intimate mixture of long 
pure white bristle-hairs amongst fine long black undcrfur and a fair proportion of 
black bristle-hairs. The white mixtmc spreads to the basal part of the tail, one of the 
specimens having a small white tip as well. 



Tabic 7: Numerical data for Mcllirora capciisis 













hucliaimiii 








atiicisa 


siiiimia 


(Type & 




kiicoiiota 


colloni 


(Type: 


(Type: 


paratvpe : 




(Bitye) 


(Ghana) 


Lake Chad) 


Sierra Leone) 


Air) 


Vegetation 


Forest 


Forest 


Sahel 


?Forest 


Subdesert 


Number in mean 


I 


3 


I 


I 


2 


Condylobasal length 


— 


142-3 


134-8 


125-8 


toS-5 


Basilar length 


— 


129-8 


120-7 


120-1 


97-2 


Palatiiar length 


(67) 


59-1 


57-5 


52-2 


46-3 


Zygomatic breadth 


S7-0 


8i-i 


71-0 


66-7 


62-3 


Upper cheekteeth breadth 


49-1 


49-2 


42-y 


44-0 


38-6 


Nasals, length 


25-2 


— 


::3-4 


23-7 


21-6 


Interorbita breadth 


36-7 


35-6 


30-5 


29-6 


29-2 


Postorbital constriction 


30-3 


30-5 


32-1 


28-5 


30-2 


Braincase breadth 




66-3 


58-: 


58-5 


52-8 


Toothrow [c — nO-) 


41-8 


40-7 


38-8 


33-5 


32-5 


p^ length 


14-4 


13-5 


12-6 


13-0 


II-O 


"ji length 


4-3 


4-1 


4-3 


4-- 


4'i 


"H length 


15-4 


(i3-y) 


14-4 


13-0 


I2-I 


Head & body 


S40 


736 


590 





4S9 


Tail 


240 


240 


156 


— 


140 


Hindtoot 


135 


120 


95 


— 


85 


Ear 


40 


3.S 


ty 


— 


29 


RATIOS (per cent) 


- 










Tail/head & body 


29 


33 


26 


— 


29 


Zygom. br./condyiob. 1. 


— 


57 


53 


53 


57 


Braincase/condylob. 1. 


— 


47 


43 


47 


49 


Braincase/zygoni. br. 


— 


82 


82 


88 


85 


Palatiiar l./condylob. 1. 


— 


42 


42 


4^ 


43 


Intcrorb./postorb. 


121 


116 


95 


103 


97 


p^jc—m^ 


34-4 


33-1 


3-'5 


38-8 


33-8 


p^iiA 


335 


3^y 


-94 


310 


262 



128 TUF. (.AKN'IVUKI.S Ol WEST AI-RICA 

This is one ot the tew specimens with any field notes: accorduig to Buchanan it 
liiirro\v>; tor deeply buried frogs, the stomach being found to contain remanis of 
these and ot li/ard'-. 

Siibtainilv LUTRINAE Baud. iSsy 
Otters 

Distribution and general. The third and last West African subfannly of the 
Mustelidae has a wide distribution not only throughout much ot Atrica but also in 
North and South America, and in Asia to southern China, japan and some of the 
larger East hidia islands. Five or six genera, covering a score ol species, are recognised, 
all ot them at least semi-aquatic, mostly riverine or lacustrine, but one, liiiliyilia 
Fleming, marine in the northern Pacific Ocean. Two of these genera occur in Afric.i. 
including the region dealt with in this work. 

Vegetation is of little moment to the otters. They occur m all kinds, their mam 
criterion for existence being the availability of ample stretches ot water harbouring 
adequate supplies offish and other food materials, and preferably fringed with enough 
plant growth to atford secure cover. This last, however, is not absolutely essential, 
tor they are sometimes tound along relativelv open stream-banks. Otters, thus, mainly 
trequeiit rivers, lakes, swamps and. since they do not object to brackish water, estuaries 
and tidal creeks. This does not mean that they are to be foiuid everywhere in such 
situations; for they are for the most part shy, generally confining themselves to quiet 
localities little frequented bv humans. For this reason there are large stretches ot poten- 
tially suitable water quite devoid of these animals, hi West Africa otters are known to 
range trom Senegal to Cameroiui and from Lake Chad to the mangrove swamps. 
They appear to be commonest in the rain-torest, but there arc not a great many study 
.specimens trom which to draw more than very general conclusions regarding either 
abundance or distribution. More than one species may exist in a single locality. 

Yet though they are largely associated with and mainly dependent upon water it 
must not be thought that they are in any sense wholly aquatic. Otters, indeed, spend 
a good deal ot their lives on dry land, coming ashore to eat, to sleep, to detaecate or 
urinate, to breed or merely t(i indulge in play. More surprisingly, they occasionally, 
and tor no very obvious reason, travel long distances overland and may be unexpectedly 
come across many miles from water, sometimes at night picked up by the headlamps 
ot a car. 

Description. Otters vary a good deal in size, hi West Atrica they may range trom 
.1 weight of about 30 kg to one of five or six times as much. Similarly the head & body 
length may range trom some 60 cm to nearly 100 cm with, in addition, a tail of roughly 
halt these lengths. But even the largest ot these is small in comparison with some extra- 
liinital species, one trom the Amazon basin being said to attain a weight ot nearh' 
3> kg. The very short-legged body is long and beautitulK' streamlined in shape, 
increasing gradually in diameter trom head to hips. It is intensely muscular and solidk 
built but extremely lithe, hideed, the extent to which an otter can, and regularly does, 
lU'x its bod\' 111 all directions must be seen to he belie\ed. 



LUTRINAn 129 

The head is broad, sliort-inuzzlcd and markedly flat; the neck ot approximately 
equivalent diameter. The facial vibrissac are an extremely pronoiuiced tcature, long, 
strong and abundant. There is a tuft below the chin and particularly large tufts on the 
lower checks. The rhinarium differs in size and shape in the various species, being quite 
diagnostic (Pocock, 192 rb) ; the nostrils arc usually widely flared on land but arc capable 
ot closure on submersion. The roundish eyes are of moderate or even small size; 
yet, by their wide-open brilliance and large roiuid pupils, tliey contrive to be a very 
conspicuous feature, even at night when they pick up and reflect ligiit from the lapetiiin 
liicidiiiii — in varying colours, ruby-red, cmcrald-green or amber according to species 
(Harris, 1968). The oval ears are small, set well down and back on the side of the 
head, and are devoid of any bursa. They, too, can be sealed by the closure of two lips 
diu^ing diving. The tail, in hunting circles usually termed the "pole" or "rudder", 
is a very notable structure. It is roughly half the length ot head & body, often rather 
more or sometimes slightly less, highK' muscular, thick and compressed dorso-ventrally 
at its base but tapering thence continuously to a narrow tip, thus completing the 
overall streamlining. The short legs arc powerful, the paws terminating in five digits, 
both front and rear. These are webbed to a greater or less extent, in some species to 
the extreme tips ot the digits, in others to only about half-way or less. In the most 
typical otters the toes are armed with fairly powerful, sharp, non-retractile claws; 
but in a few, including two of the West African species, the so-called clawlcss otters, 
there is nothing beyond small vestigial nails, and even these arc absent from the majority 
of the digits. The palmar and plantar pads arc various in shape, sometimes rather poorly 
develofied (Pocock, 1921b). The hind teet arc always larger than the toreteet. 

On the uiaderside of the tail, near the root, there arc paired scent glands which 
abimdantly exude a quantity of milky fluid with a powerful musky odour but not so 
repulsive as that of some other mustelids such as the polecat. Their purpose seems to 
be rather for recognition and the demarcation of territory than tor dctence. Females 
may have, at most, three pairs ot mammae, abdominal and latero-pectoral, but some 
ot these are often lacking or obscure. 

Otters arc noted for the character of their pelage. The tact that this is dense and 
handsome has given it some commercial value, both locally and in the world fur 
trade, with consequent adverse effects upon the survival of various species. The colour 
differs, ot course, in the difterent species but in general is, dorsally, lighter or darker 
scpia-browii, though it is in some cases distinctly pallid, and pure white areas occur in 
others. Albinos are known. The dorsal fur is sometimes slightly "trosted"; and it 
always, even when quite dry, has a sparkling sheen. This is due to the composition ot 
the pelage, which is remarkably constant throughout the genus. It is short, of miusually 
uniform length over the entire body, as much below as above; very soft to the touch 
and exceedingly dense. The hairs are of two kinds. The short, crinkled underfur, 
about 8 mm in length, is extremely closely packed and is of outstanding slenderness, 
being, at o-oi mm diameter, amongst the finest ot all hair. Despite this, its surface, 
seen luider high magnification, is rough with long and relatively coarse scales. In the 
majority of cases it is white for most ot its length but becomes lighter or darker brown 
near the tips, so that if an otter's coat is rubbed up the wrong way it is seen to be 



130 Till ( AliNlV01)l:S Ol «EST AfKlCA 

wliitc-bascd- latlKT surprisingly smcc tlic extreme densit\ ot the tur iiornialK' com- 
pletely obscures what lies below the closelv-packcd tips. 

Mixed amongst this crowded underfur stand much longer hairs, by comparison 
widely spaced yet in sum forming a continuous coat that entirely overlays and conceals 
what lies beneath. These bristle-hairs, which arc i6 to iS mm long, are very slender and 
whitish in the basal halt so that they are there indistinguishable by eye from the under- 
hir; but beyond the reach ot this latter they expand into a thick, usually densely pig- 
mented blade. This is not concavelv guttered as in some mammals but has a slightly 
convcxiy oblong section: and though by comparison with the rest ot the tur relatively 
broad, it is. in fact, only about o- 1 mm wide with a length ot some 6 or 7 mm, tapering 
terminally to a sharp tip. It is these blades that glisten and give the coat its characteristic 
sheen. Because of their long slender stalks the hairs have nothing harshly bristly about 
them, hi the newly-born otter, at least in so tar as the West Atrican Liitia maculicolli^ 
is concerned, the pelage is quite ditterent. There is no sign at all ot dense undertur, the 
coat being almost entirely composed ot long, straight, rather wiry hairs. A somewhat 
older specimen ot Aoiiyx aipciisis shows the development of the dense coat well under 
way but the ultimate outer cover of glossv bristles onh' beginning to be visible amongst 
the. as yet longer, wiry baby hair. It is possible that the reluctance ot young otters to 
take to the water may be due to the inadequacy ot their protection against cold. 

The composition of the belly tur is precisely similar to that ot the back; and, what 
is more unusual, so is that ot the tail too. It has been said that this pelage is waterproof; 
but though this is not absolutely true, it is a fact that the rightly packed underfur renders 
Its penetration by water extremely difficult, and the hard, glossy bristle-hairs, though 
they bimch together when wet, arc very rapidly dried out. An otter gets rid ot the 
water in its coat not so much bv vigorous shaking, as a dog does, but more by rolling 
or nibbijig against a dry surface. This is possibly because its legs are too short to give 
itselt a really effective shake. It may be added that the skin ot an otter is tough, rather 
like that ot a ratel; and similarly it is also prettv loose so that the animal can twist and 
turn dangerously if held by it. 

Skull (tigs. 20, 21, 23 and 24). There is some ditierence ot style as well as ot size 
between the three otters ^ylth which this book is primarily concerned. There are also 
variations due to age and sex, chietly to do \\ ith development of the crests. Detailed 
description is reserved tor the accounts ot the different animals given later, the tollowing 
being some ot the more general features of the lutrine skull. 

The typical otter skull has a broad, flat bramcase which contrasts strongly with the 
narrow, often parallel-sided mterorbital-intcrtcmporal region. The muzzle region, 
^ltuated anterior to the orbits, is extremely short, being even more abruptly truncated 
than 111 Xkllivofii, the anterior nares \yidely open. In the most typical otter skulls there 
are prominent sharp postorbital processes, the posterior ridges ot whichjoin the sagittal 
crest, helping to outline in this interorbital area a tairly well defined pentagonal plate. 
In tully developed males of most species there is a well-developed, sharp sagittal crest 
extending from the postorbital region to the equally pronounced, upcurved supra- 
occipital crest, which continues as sharp ridges round the sides of the braincase forward 
to the very large mastoids. In females, though the supraoccipital crests may be present. 



LUTRINAn 131 

the sagittal line generally carries nothing more than a very low ridge which in really 
old skulls may rise to a brief crest posteriorly. The zygomatic arches arc broad and 
strong; the maxillary zygomatic process divides clearly into two branches which enclose 
an exceptionally open intraorbital canal, obviously transmitting muscle as well as 
nerve, nothing at all similar occurring in other West African Canoidea. The palate is 
continued, parallel-sided, well beyond the back of the toothrows; the bullae are 
unusually flat and elongated, relatively insignificant. The bone structure of old animals 
is dense and solid, fusion of the sutures leaving little or no trace of original independence. 

The dentition is often very irregular through the loss, or whole or partial failure to 
develop, of a nimiber of cheekteeth. These, when all arc present, are jr^, giving a 
total for the whole mouth of 36 teeth; but this not uncommon deviation from normal 
should be kept 111 mind since it may give rise to some doubt or difficulty in the recog- 
nition of lutrinc skulls. The most frequent tooth to be lacking is the very small anterior 
upper premolar (pi), which, when not missing is generally sited in close proximity 
to the inside face of the canine, p- may also be in contact with the canine, as though, 
through the great shortening of the muzzle, there is becoming less and less room to 
accommodate these teeth and they are in the course of being eliminated. The exterior 
upper incisor [P) is much larger than the two interior ones (i^ and /'-) ; in the lower jaw 
the distinction is not so marked and the incisors as a whole are often irregular and not 
in line. 

Habits. Otters in Europe, since they are the subjects of regular organized hunting 
for sport, have been under close observation for many centuries ; and, the world over, 
they have commonly been kept as domestic, and often very intimate, pets. Many 
books and articles have been written on this latter aspect, especially in recent years, 
and a good deal, therefore, is known of the behaviour in captivity of several different 
kinds of otter. Yet there are often unwelcome gaps in field knowledge of the more 
hidden aspects of daily or yearly life; and this is particularly true of the three African 
species dealt with in this work, one of which, indeed, is scarcely knowni except for 
skulls. But even in the less secret matters of everyday biology the views of specialists 
arc sometimes strangely at odds about such things as senses of sight, hearing and smell, 
body odour, the use of glands, or readiness of the yoiuig to swim. Nevertheless, there 
remains wide agreement regarding broad general behaviour; and doubtless individual 
otters differ from one another in their physical and mental attributes, as other animals 
do, and care must be taken m assuming the general from the particular. So much, 
indeed, has been written of specific individuals in relation to sometimes rather unreal 
circumstances that one difficulty is to know what, in a brief account, can be taken as 
representative of the class as a whole in its purely natural environment. The most 
comprehensive account of otters in general, with descriptions of all the species, together 
with their habits in so far as they are known, is to be found in Harris (1968). It must 
be emphasized here that though in all likelihood most of the account of habits that 
follows is applicable to African species these last have, in fact, been relatively little 
studied in detail in the field and therefore it may be discovered that behaviour in the 
three species with which this book is primarily concerned differs in certain aspects 



\}1 TUF (MINIVoniS (ir WIST AIUK A 

trom tin- i^ciicr.il pitturc givcji. This may be especially sd as regards breeding. Doubtless 
the three species dirtcr amongst tliemsclves in this as in (ither respects; and it is, indeed, 
vaguely recognised that tlie only two of the three that have been observed in lite do 
live somewhat difterently, though this is more a matter ot general impression rather 
than ot tirniK' established tact. 

Otters are active by day or by night depending to some extent upon local circum- 
stances as well as prctcrcncc. It there is much danger ot human interterence they become 
rather carctui of exposing themselves to view and are more inclined to torage nocturn- 
ally, particularly on moonlit nights. In any case their hunting tends to be more especialK- 
.\n early morning or late atternoon occupation. However, both species and individuals 
ot them ditfer somewhat in their degree ot sh\-ucss, and otters have been known to 
come into towns at night or to raise families in drains near or under houses; and once, 
even, beneath the floor-boards ot a living room. 

For the most part otters dwell near water, most commoiiK' tresh, that is to say 
rivers, lakes or inland marshes; but they are by no means averse to brackish or even 
salt water and so not infrequently are to be found in creeks or estuaries or along the 
coast. They sometimes travel overland to a considerable distance trom water, but there 
IS no evidence that they spend long m such situations. Though they like to live near it 
thev do not pass all, or even the major part, ot their lives in water; they mostly take 
to it primarily tor the business ot seeking tood, but this vital business, as will be seen 
later, is tempered with a good deal ot sportive activity. Sleeping, breeding, detaecation, 
urination, and nearly always tceding are essentially dry-land occupations, as well as 
exteirsive periods ot out-ot-water play, and it will thus be seen that, though some species 
are more acjuatic than others, otters m general spend at least as much time ashore as 
they do in the water. It is customary to reter to the males as "dogs" and the females as 
"bitches"; but in spite ot this analogy with dogs the young are generally spoken ot 
as "cubs", not "pups ', though the term "whelps ' is sometimes used. 

On shore, an otter centres itselt on a set shelter tor the purpose ot sleep or breeding. 
This is nearly always subterranean but may occasionally be in dense vegetation. These 
dens, or "holts" as they are technically termed, may be self-constructed or taken over 
from other animals or adapted trom naturally-occurring holes suclr as those amongst 
riverine tree roots, amongst boulders, or even in a cave. Generally they have an under- 
water entrance — but extremely little is known of the exact nature ot these homes in 
Africa. It at least seems highly likely that the African clawless otters would find some 
difficulty in digging except in the softest of swampy mud. The shelters are situated 
well above flood or seepage level and are frequently lined to some extent with vege- 
table matter, grass, leaves, reeds or moss; tor, although they spend a good deal ot time 
in water, otters like very much to be warm and dry. 13rying ofl ot the outer coat, 
by rolling or rubbing, is in tact one ot an otters tirst urges on leaving water. Custom 
in nest-lining, however, varies; pcssibly when the soil itselt is of a warm and dry natm-e 
no nest lining is called for. Such conditions would seem to exist over much of tropical 
Africa, except possibly in the forest belt at the height of the rains. 

Holts are occupied sometimes by a lone otter, sometimes by a tamily party. Such 
parties usually consist ot a bitch and her offspring up to the subadult stage; the dog otter 



LUTRINAE 133 

is, as tar as the holt is concerned, generally kept at a distance, and occupies a separate, 
though neighbouring, shelter. 

Some observers have maintained that otters combine and co-operate in fishing 
parties; but it is doubtful whether such collaboration, if it exists at all, ever extends 
beyond the immediate family. Most otters are by no means gregarious creatures and 
holts are generally fairly widely separated; and it would seem probable that each unit 
has its own recognised stretch of water, especially when one remembers the powerful 
scent-glands which otters possess and their probable use for demarcating territory. 
Yet a fixed territory implies some degree of permanence, and otters have a considerable 
reputation as wanderers, never dwelling long, perhaps only a night or two, in one 
place except at actual breeding time. Such nomadism or semi-nomadism may be for 
safety; or it may be that after a while a given stretch of water begins to become ex- 
hausted of food, or the fish, by constant hunting, accentuatedly wary. Whatever its 
motive, this continual moving renders the field study of otters a matter of some 
difficulty except at breeding times since the observer often does not know where to 
fuid his animals from one day to the next. 

It is commonly thought that otters feed solely upon fish but this is far from true. 
Fish may form a large or even the major part of the diet but a good many other things 
living in or around the water are eaten, such as crabs, crayfish, prawns, mussels, snails, 
large water-beetles, and, on land, groimd-birds, water-fowl, eggs, rodents and possibly 
other small mammals, lizards, including geckos, snakes, frogs, toads, grasshoppers, 
crickets, mole-crickets and other large insects. These are generalities, for each species 
has its preferences and each locality a differing availability of food materials. For 
example, it would seem that the two commonest West African otters differ in their 
feeding demands, Liitra maailicollis being predominantly a fish eater, Aonyx capensis 
consuming more crabs. But no set work has been done on this or on the kinds offish 
preferred by each species in West Africa, though it is known that the latter species, 
at least, consumes, in Lake Victoria, a fair number of Lungfish {Pivtoptcms) and probably 
clariid catfish as well. Eels, which form a considerable part of the food of otters in 
Europe, do not occur in the rivers of West Africa except possibly in cstuarinc waters. 
These considerations arc of some importance since it has been held in Europe and 
elsewhere that otters cause a great deal of economic harm both in fisli-liatcheries and 
amongst the more valuable "game" fish. Farmers join with tishermen in their con- 
demnation, saddling otters with poultry thefts of which, often, they are not guilty. 
Such charges have, on the other hand been strongly denied; and there is a growing 
appreciation that otters may, in sum, effect far more good than harm in many ways, 
not least by getting rid of much fish that is diseased or otherwise tmdesirable. 

Small articles of food taken in the water, such as snails, mussels and other mollusca, 
are then and there swallowed with a bite; small fish may be eaten in an upright position 
while the otter is treading water with head and neck above the surface, or while the 
otter is floating on its back; but anything large is brought ashore to be consumed. 
It may, if really large, be held down by the paws and eaten thus on the ground; but 
moderate-sized fish are held up to the mouth between the hands and eaten gradually 
from end to end. It has often been asserted that a start is always made at the tail end. 



134 II" <'AHN1\()UES or WKST AlKICA 

tlic licad tuially bcuig discarded; but tlirs is by no means always so. Food is invariably 
eaten fresh at once or discarded, never stored, and, in tact, otters are usually believed 
not to eat ;uiytliing but the tood they have just caught; but Stephens (1957) records an 
opinion that, at any rate in Europe, the\' will, and regularly do, consume a large amount 
of carrion. 

Ability to catch tish implies a superior ability to swim. The adult otter is certainly a 
past master in this art. On the surface, with the head partly above water, an otter 
progresses in a rather leisurely fishion, using its teet alone in a sort of dog-paddle; 
but once it has dived in earnest the whole demeanour immediately chajiges and the 
animal becomes a lively and remarkably dexterous hunter. Propulsion is mainly by 
powerful thrusts of the hindlegs, used as a pair, aided sometimes by a sinuous and 
rhythmic up-and-down movement ot the body. The forelegs arc occasionally but not 
always held back against the belly. Tlie tail acts as a rudder, hi this way a moderate 
speed is achieved over a short distance; but it is rather the otter's facility, with its supple 
bodv, to equal or outdo its prey in swiftness oi turn that makes escape for the fish a 
matter of some hazard. None the less, it seems that otters as a rule prefer to go after 
the slower moving kinds ot fish; and certainly the two categories mentioned earlier 
as being commonly eaten in Africa are bcuh sluggish. 

The eyes are kept open dm^ing submergence and probably tunction more etticiently 
luider water than on land. The ears and nostrils are scaled; but no one seems to have 
determined how, when an otter opens its mouth to seize a tish, it avoids its liuigs 
being swamped with water. Otters swim on their backs as well as on their bellies, and 
they sometimes float thus idly as though half asleep. How faultlessly streamlined tlie 
body is is well demonstrated not only by the easy progress ot the submerged animal 
in rapid pursuit but, as well, by the manner in which an otter habitually slides into the 
water with scarcely any disturbance ot the surface and consequent alarm tor the lish. 
To all intents and purposes the only sign ot an otter's diving lies in the trail ot rising 
bubbles, probably from air trapped in the fur. 

Otters, being mammals, must, of coiurse, come up from time to tune to breathe 
otherwise they would drown. Normally they remain submerged tor a matter ot a 
few seconds; Mortimer (i9''i3), observing the speckle-throated otter in Zambia, found 
the commonest time to be from 10 to 1 s seconds, and he never wimciscd a dive lasting 
for more than 45 seconds, and that only on one occasion, but when they so wish or 
it is necessary they can remain imder water tor some four or five minutes; and Ma.xwell 
(1961) timed ari Asian otter for just on six. This, in human estimation, is a very long 
time though it is brief in comparison with some other aquatic animals — a quarter ot 
an hour for a manatee or a seal, and an hour or even two for some of the whales. 
Though no investigation appears to have been carried out on submersion actually m 
the otters it would seem likely that the mechanism by which prolonged abstention 
from breathing can be achieved is much that found amongst other groups. This, in 
brief, is a marked reduction of oxygen consumption brought about firstly by a very 
considerable slowing down of the heart-beat, and secondly b\ a constriction ot the 
blood vessels in less important areas ot the body while maintaining a full flow to the 
brain. There are contributory mechanisms such as, during preliminar\' breathing. 



LUTRINAE 13s 

a nearly complete change of air in the luiigs, a greater ability in the blood to hold oxygen 
and less sensitivity of the respiratory nerve centre to increased concentration of carbon 
dioxide. Ability to stay under water has nothing whatsoever to do with extra lung 
capacit)' which, indeed, by increasing buoyancy would be more of a hindrance to 
diving than a help. Harris (1968) suggests that because of this adaptation to prolonged 
submersion otters are diiiicult to anaesthetize; but Stephens (1957) refers to two cases 
within her experience without hint of trouble. In one, morphia and chloroform were 
used; in the other, nembutal. 

Whatever the precise mechanism an otter can, in five minutes, swim a long distance 
luidcr water; and though fishing and other underwater foraging is most commonly 
carried out at shallow depths it has been shown, by the capture of otters in traps, that 
at times dives of at least 18 metres are luidertaken. Ability to remain below water is 
to some extent dependent upon age, young otters being incapable of as long periods 
of submersion as adults. Despite outstanding mastery of water it is, indeed, generally 
held that yomig otters arc most luiwilhng to enter it and, if they do, prove themselves 
to be nervous and highly inept swimmers. There seems little doubt, from numerous 
cye-\s itness accounts, that in most cases the mother has to teach her yoiuig both how to 
swim and how to catch fish. To judge from Maxwell (1961) it would seem as if this 
might well apply also to the clawless otter which occurs in West Africa; but doubtless 
idiosyncrasy plays its part in this as in other matters, for Mockford (1967) fomid two 
young captive otters of this species to take to water quite naturally and easily at an 
age estimated to be no more than six or seven weeks. This is the more remarkable in 
that Aony.x is incieasingly regarded as markedly less aquatic than other species. 

It seems likely that the exceptional development of the facial vibrissae may assist 
in the avoidance of rocks and other obstacles in cloudy water; and, by their sensitivity 
to the currents produced by even slight disturbances, aid in the detection of moving 
tish. Though both ears and nostrils are closed under water, hearing and smell neverthe- 
less probably play their part too; certainly on dry land both these senses are tairly 
acute, smell, to judge from the complex internal nasal structure in comparison with the 
poorly developed auditory bullae, much more so than hearing. Cdours, indeed, must 
play a constant role in the otter's life, as m those of most mustelids. Signals are con- 
stantly being left on rocks or on deliberately gathered bunches of vegetation; and the 
odour lett by otters in normal passage over the ground is strong and very persistent, 
it being well known to those who hiuit these animals in Europe that hounds may 
often be deceived into picking up a trail already a day or two old. The faeces, doubtless, 
have their own particular odour and convey their message to others. Otter droppings, 
which are most commonl)- black, rather liquid and slim)-, are technically termed 
"spraints"; and it is well known in Europe that set spraintmg places arc used, some- 
times over a period of very many^ years. Whether this applies to the same extent in 
Africa is not recorded; but Mockford (1967) fomid young Aoiiyx in captivity always 
to use dcfmite spots and, therefore, to be scrupulously clean. E)Te (1963), on the other 
hand, foiuid an otter of this species awkward to keep about the house since it persisted 
in registering o\\ nership in this way. E.xpericnce of those who have kept otters as pets, 
as well as the abundance of spraints in the wild, indicates that dcfaecation, always 



136 THE CARNIVORES OF WEST AFIUCA 

sinuiltancously accompanied by urination, is frequent; and tlierc is good evidence, 
troni observation and troni examination oi the nitestines, that digestion and passage 
ot food through the gut are rapid. 1 lowever, tlie spraints always contain some indica- 
tion ot the food taken, in the form ot undigested iish bones and scales, or chitiuous 
fragments from Crustacea. Excretion is performed witii straddled hindlegs and tail 
erect, and is accompanied by a constant dancing movement. 

Though otters so obviously, in nature, thus constantly leave traces of their presence 
or passage it is a remarkable fact that no one ot the very many who have kept them 
as pets has ever expressed objection to their smell, despite the animals' sometimes 
habitual use ot carpets, chairs and otten beds. In fact, otters have been recorded as 
having no smell at all imlcss kept out of water tor several days ; and some have described 
the faint odour ot the dense fur as very pleasant. 

Besides the evidence of their spraints otters trequently leave indications ot then- 
existence in an area in the torm ot rejected or discarded portions ot food. Feeding, 
except of small, readily consumed objects, always takes place on shore and very often 
at detmitc sites such as a secluded strip of river bank or a conveniently flat rock. Such 
favorite feeding spots arc often littered with the heads or tails of fish, or even whole 
bodies with a single bite out of them, for otters are sometimes wanton killers of more 
than they need. One ot the most frequent signs in Africa is the remains ot crab shells 
or claws, since ^4();))'.v appears to have a preference for these arthropods above fish. 
Another sign of the recent presence of an otter is the footprints in soft sand or mud. 
These, amongst those who specialize in himting otters, are knowni as "seals", a technical 
term that has been in use for at least three centuries. These tracks differ amongst the 
West African otters, since some feet are clawed, others are without them ; some are 
webbed between the digits, others not. 

Although they are fiuidamcntally land animals otters are not there so completely 
adept as they are in water. Their walk, though brisk enough, is a kind of slight waddle 
from one side to the other with head low and back humped. Yet some are known to 
cover considerable distances, up to 20 or 25 kilometres, at this pace, and apparently 
without tiring. When put to it they can gallop moderately fast, "humping" along fore- 
teet and hmdteet alternately. However, apart from their not very expert gait they still 
remain highly flexible, swift and versatile in movement, rolling, twisting, turning, 
scrambling over rocks, logs and other obstacles. They can climb to a certain extent, 
and, in fact, have been recorded (Stephens, 1957) as sometimes sleeping up in low 
trees. They often stand upright on their hindfeet, using their tails as a third support. 

Otters have several different sorts of call or other vocal soimds according to circum- 
stances. Animal noises are always difficult to express in writing; moreover, they differ 
quite a lot in the various kinds of otter and are not very well recorded for the species 
occurrhig in West Africa, if at all. Not much, therefore, can be said on this subject. 
The shrill nocturnal whistle of the European otter is, apparently, never uttered by 
African species. Harris (1968) describes the soiuid made by .4('/))'.v when suddenly 
alarmed as "a strongly aspirated and explosive 'HaliV ". The same author says that 
this species uses "a querulous moaning wail" to express anguish, apprehension or, 
sometimes, greeting. Shortridgc (1934) says he has frequently heard African otters 



LUTRINAE 137 

"barking"; and, quoting Stevenson Hamilton, gives the sound made by Aoiiy.x at 
bay as a tliroaty "Kwa-a-a, kwa-a-a". Yoiuig otters, as those of so many other kinds 
of mammals, continually utter bird-like twitterings and plaintive squeaks. 

Though otters may occasionally idle in water they do not often do so, wakingly, 
on land. Unless they arc actually asleep in their shelters they are generally busily active 
and rarely, like so many other animals, lie merely basking in the sim. When they are 
not engaged in eating or excreting they inquisitively examine all the details of their 
surroundings ; or when this eager inspection palls they indulge in play, either alone or 
with others of their family. Otters have earned a remarkably wide reputation for 
play; and they arc, without doubt, amongst the most dcdicatedly playful of all animals, 
their activities in this field having every mark of real play and utter enjoyment without 
hint of anything more piu-poseful behind it. The stories of those who have had the 
good fortune to watch young otters m nature or who have kept them as pets are 
endless and are pervaded by a bewitching charm that is, at best, no more than palely 
reflected in even the most persuasive accounts of other animals of any kind. This is no 
place to record these, and only brief generalities of behaviour can be given. 

Play takes place equally in water and on dry land, and as much by a lone otter as 
between two or three. It not infrequently involves the deliberate selection and use of a 
toy, and the otter's capacity for deriving continuing amusement from the simplest of 
sources is a matter for some amazement. In water, play may consist of delighted and 
quite aimless gambolling in the form of dives, twists and turns, or of endless rapid 
revolution about the longitudinal axis. It may consist of dropping a stone and diving 
to catch it in its descent; or it ma)- be a walk on the bottom, shuffling a shellfish along 
with its nose. One otter is recorded (Maxwell, 1961) as taking a ping-pong ball down 
to the water to enjoy, and expertly appreciate, for lengthy periods its lively antics 
when released at a depth. A ball has its fascination also on land, being struck along with 
the nose or dribbled football fashion with the feet or flung from the mouth over the 
shoulder. In nature stones are used for this exercise in place of a ball; and an otter may 
occupy itself with a smooth pebble for a long time. The young may sometimes play 
a species of "tig" with one another. One of the pastimes for which otters arc most 
widely famous, either alone or in company, is sliding. This is easily done in winter in 
cold climates on snow or ice, set slides being deliberately made and repeatedly used; 
but when or where these elements do not exist mud slides down steep river banks are 
made. These are by no means uncommon in Europe, but whether such slides exist any- 
where in Africa seems never to have been definitely recorded. Water-chutes are similarly 
used for play where rivers fall over flat rocks in rapids. 

The amusing and interesting things that otters can do in the artificial surroiuidings 
of a human habitation caruiot be gone into very deeply here, the more so as circmn- 
stances must differ in every case. But it may be recorded that the otter's forcpaws, 
which are so fashioned that they are expert at holding things, arc rarely idle, exploring 
the possibilities and uses of anything within reach. This is particularly so with the West 
African Aoiiyx in which the fingers are imhampcrcd by claws and the structure is not 
unlike, in appearance and manageability, a monkey's paw. So dexterous are these 
hands that Aoiiy.x can open tins or bottles (Eyre, 1963) or peel hard boiled eggs, and 



138 Tiir, cARNivonns of west atrica 

amongst other tilings has shown itself to be an expert pickpocket. When indulging in 
such an operation, or on the rare occasions when they may attempt slyly to rob a 
companion of some tit-bit, these otters have the engaging habit of distracting attention 
from their act by turning their heads away and gazing abstractedly into the opposite 
direction. 

It may be added here that otters have proved themselves to have most excellent 
memories both for persons and situations, recognising tnends after prolonged absence 
(Pitt, 193S). They are long-lived, one being recorded as existing tor 23 years in captivity. 

There is no doubt that otters, caught yoimg. make highly intelligent, enchanting, 
and indeed lovable, pets, being not only extremely entertaining but affectionate and 
loyal as well. Nevertheless, a warning must be given. Thcv are ccrtainlv not for every- 
one or for those who like to preserve a well-equipped and tidy house. Their insatiable 
inquisitiveness, their sense of play resulting in mischiet or destruction, their deter- 
mination to get their own way, and their constant liability to soak clothes, furniture 
and beds calls for an even temperament and a forbearance not possessed by all. Add to 
this, occasional lapses into bad temper resulting in painful and sometimes serious bites, 
and it will be seen that the care of a babv otter is not a task to be entered upon lightly, 
cspeciallv if it is accompanied bv admission to the house. 

Something must now be said of breeding, though remarkably little that is detuiite 
is know-ii of this even as regards the long-observed European otters. Copulation takes 
place in the water, lasts an hour or more and is repeated several times over the course 
of a day or two. The period of gestation is very much in dispute. Probably it is normally 
in the nature of 9 weeks or a day or two less, but this period may be very much length- 
ened by delayed implantation of the fertilized ova in the womb. There may be any- 
thing between i and 5 cubs at a birth; but both these extreme figures arc unusual and 
the normal expectancy is 2 or 3. The newly-born cubs are covered with fme, short 
fur but the eyes arc closed and appear to remain so for a matter of 5 weeks, though 
opinions differ largely as to this. Probably all the figures here given for these matters 
will be toiuid to be only approximately correct for species in West Africa. The young 
are, if necessary for safety, carried from place to place by the bitch using the loose 
skin at the back of the neck. They are even said, still as blind juveniles, to be transported 
thus actually under water from one river bank to another; which would appear to 
indicate that not only do the nostrils and ears close reilexly but that, also, the oxygen- 
conserving submersion mechanism described on page 134 conies automaticallv into 
operation at a very early age. 

It has been said (Stephens, T9S7) that the yoiuig are born toothless; but two juvenile 
skulls of West African Lulra and .■lc)//)'.v in the British Museum are both well supplied 
with teeth. Otter bitches make very good mothers, caring for their litters assiduously 
and protecting them fiercely. They alone bring up the family, at least in the initial 
stages, the dog otter not being allowed to come near the nest though lie remains in the 
vicinity, occupying a holt of his own. Tlie cubs seem to develop fairly slowly and do 
not leave the nest for several weeks. TJiereafter they have still to be taught swimming 
and diving and to acquire proper proficiency in hiuiting. At this stage the father may 
return and take part in the training. The cubs probably remain with the mother untij 



LUTRINAE 139 

they are sexually mature, which is after about two years. There seems to be generally 
at most one litter a year, possibly less. Even in the diversely seasonal climate of Europe 
litters may occur in any month of the year, winter as well as summer ; in the tropics it 
would seem that season, for water-haunting animals, must matter even less. 

Otters have been found to harboiu: a number of different kinds of internal parasites, 
both in the gut and the bloodstream, of all the usual groups of wonns, flukes and 
protozoa; but nothing is known of this in connexion with African species. They have 
also been recorded as carr)'ing ticks, and it has long been believed that otters visit the 
sea from time to time in order that the salt water should rid their coats of these and 
other ectoparasites. In the normal course of events otters spend some time cleaning 
their fiu: by scratching, rolling, rubbing, biting and possibly licking. The last does not 
seem to be a common habit; but as otter fur is at times a constituent of the droppings 
(Stephens, 1957) they must at least occasionally clean themselves, or their cubs, in this way. 
Nevertheless, external parasites seem to be less common than on some animals, possibly, 
sea-cleansing apart, because the extreme densit)- of the underfur makes life amongst 
it and passage through it difficult. The hazard of drowning, too, must always exist. 

Apart from disease otters have two main enemies, man being the chief. There is, 
indeed, reason to believe that the otter population of the world, and especially of 
newly expanding areas such as Africa, is markedly decreasing, partly by direct persecution 
and partly because their once quiet haunts arc more and more being uivadcd and diver- 
ted into economic use. The other enemy, in tropical waters, is crocodiles. Otters have 
often been observed in places infested with these reptiles and though they show no 
visible awareness or signs of fear it is obvious that they must continually be on the 
alert, ready to make a swift avoiding tiurn or to flee to the bank. They may also be 
aware that a crocodile's periods of entry to the water have something of a set daily 
rhythm dictated by the demands of their cold-bloodedness and virtual lack of bodily 
heat control except by alternate smming and immersion. It is possible, also, in the 
tropics that pythons fmd their \\ ay into otters' holts; and though an adult dog or bitch 
would give a good accoimt of itself in a fight with such a snake — though it might well 
be hampered in a restricted space — juveniles would easily be taken. And there is always 
the danger, on land, of being surprised and sprung upon by a leopard or other of the 
larger felines. 

Taxonomy. The position of the Lutrinae as a distinct and valid subfamily of the 
Mustelidac has never been brought into question since Gray (1865) fu'st suggested this 
classification. Within the subfamily itself the genera are fairly clear except that Paraonyx 
is often regarded as no more than a subgenus o( Aoiiyx since it was so treated by Eller- 
man, Morrison-Scott &: Hayman (1953). This course is adopted here. Some of the 
characters picked on by Hinton in the type as diagnostic of Pi!M()/!]'.v are seen not to 
be valid in other and better skins; and while there is, indeed, a remarkable difference 
in the teeth it is one of size rather than of anything phylogenetically more fundamental; 
and the rest of the characters, both cranial and external, denote an undeniably close 
affinity with Acnyx. 

The second West African genus is the almost world-wide Lutra, which ranges over 
much of Europe, Asia, Africa and America. In consequence there have almost inevitably 



140 THE CARNIVORES OF WEST AFRICA 

been attempts to subdivide it eitlier into independent genera or at least subgenera. Tlie 
sole African species of this genus, macuUcollis, was first gcncrically separated from the 
others by Gray as HytJrogalc. This name was later found to be preoccupied and the 
proposed genus was eventually restyled Hyihiais by Pocock who, by reason of a 
number ot small dirterences, both cranial and external, supported Gray's interpretation 
ot the position. Once again, the distinctions drawn appear too minor to warrant full 
generic separation and Hydrictis has commonly been reduced to the status ot a sub- 
genus. It is so dealt with herein; but see page 141 for further comments on this. 

The fullest study of the taxonomy of the Lutrinae is a long monograph by Pohlc 
(1920) dealing with all the genera, species and races. Pocock (1921b) described some 
aspects ot external characters. 

Tlie two West Atrican genera may be told apart thus: 

KEY TO THE GENERA OF LUTRINAE 
(previous key page 93) 

Cheeks, lips, throat, sides of the neck and the entire chest wholly white or cream; 
forefeet without long claws and with only slight webs; skull length more 
than 115 mm; mastoid projecting very prominently behind the aural oritice 

Aonyx {pai^c 148) 

Wliite or cream area contuicd to the mid-throat and upper chest, and then to a 
greater or less degree irregularly blotched with the normal dark pelage colour; 
a similar spotted patch often on the after part of the belly; toreteet with well- 
developed claws and webs; skull less than 115 nnn; mastoid not remarkably 
prominent ......... Lutra [\h^gc 140) 

Genus LUTRA Brisson, 1762 
Typical Otters 

Lutra Brisson, 1762, Ri'^tuiin Aniiiialc in classes IX distrihnlnm . . .. 2nd edit.: ij|. Type species Miistcla 
lutra Limiaeiis, Sweden. Lulra was the Latin name for an otter. (It is doubtful whether Brisson's Ri\;iuiiii 
Atiiiiiale is properly available under the hitemational Code; but certain ot his names, including Lutra, 
were proposed to the Commission tor validation, see Ellcrman & Morrison-Scott, 1951 : 3, though no 
action has been taken on this and the matter has been dropped. Even in the case of rejection the name 
still stands as of Briinnich.) 

Lutra Briinnich, 1771, Zoologiae Jundamcnta . . . Grumie i Dyrctocrai. Type species Mustila lutra Luni,aeus. 

Hydrof^alc Gray, 1865, Proc. zool. Soc. Loud.: 131. Type species Lutra utaculUollis Lichtenstein, South 
Africa. This name was preoccupied having already been used twice, by Kaup, 1829, and by Pomel. 
1X48, in each case tor a shrew. It is derived from the Greek liydor, liydr-, water, and ijfi/c, a wcisel. 

Hydrictis Pocock, 1921, Proc. zool. Soc. Loiid.: 543. Type species Lutra macuUcollis Lichtenstein. Tlie 
name was compounded from the Greek words hydor, hydr-, water and ictis, a weasel. Valid as a sub- 
genus. 

There is further synonymy ot little concern in African litcr.uure. 

General. As already stated above, the typical otters are widely spread over a good 
deal of four continents. Many of the chief characters and habits of the genus have been 



LUTRA 141 

given in the general introduction to the subfamily and there is no need to elaborate 
further what has been said there; the more important distinctions between it and Aonyx 
are set forth in the key just given and will become more apparent ni the detailed 
descriptions which follow. It remains only to consider briefly in what characters the 
African section Hydrictis differs from the most typical, Palaearctic, members of the 
genus, which, beyond this, are of no other taxonomic concern to this work. 

Hydrictis was, following Gray, proposed by Pocock as generically sepiarablc from 
Lutra on the groimds of a reduction in the size of the rhinarium and a simplification 
of the external ear; larger, more fully webbed and hairy-solcd feet, the palmar and 
plantar pads being less well developed. The skull was distinguished by "many cranial 
characters, especially the shortness of the muzzle, length of the orbital floor, and the 
generally immatiu'e aspect of the skull owing to the feeble development of constrictions, 
crests and prominences". This last is a very apt description, and even the oldest Hydrictis 
skulls are in no way so robustly built as those oi Lutra Intra. The most noticeable feature 
is the narrowness of the interorbital region and complete absence of postorbital pro- 
cesses, as shown in figure 21, and the consequent lack of the usual lutrine interorbital 
pentagonal plateau, observable in figure 24 o£ Aonyx. The sagittal and occipital crests 
are relatively insignificant; the mandible and its condylar structiure weaker. There is, 
indeed, quite a strong case for regarding Hydrictis as valid at full generic level. 

Subgenus HYDRICTIS Pocock, 1921 
African Long-clawed Otters 

Since Hydrictis consists of a single species, maculicollis, no description ot the subgenus 
is called for beyond the differential characters just given above and the account of the 
species which follows. 

LUTRA MACULICOLLIS Lichtenstein Speckle-throated Otter 

Lutra maculicollis Lichtenstein, 1835, Arch. Naturgescli. I: 89-92, pi. 2, f. i. South Africa (Bamboos Moun- 
tains). The specific name was derived from the Latin words macula, a spot, and collum, the neck, in 
reference to the markings on the throat. 

Lulra grayii Gerrard, 1863, Catalogue of the Bones of Mammalia in the . . . British Museum: loi. Port Natal 
(= Durban), South Africa. A nonten imdum. This was named after Dr. J. E. Gray of the British Museum. 

Lutra malschiei Cabrera, 1903, Boln R. Soc. csp. Hist, nat., 3: 182. Rio Muni. Named after Professor 
Matschie, German zoologist. Possibly subspecifically applicable in West Africa. 

Lutra maculicollis nilolica Thomas, igii, Ann. Mag. nat. Hist. (8) 8: 726. Malek, upper Nile, Sudan. Possibly 
applicable in West Africa. 

Distribution and general. The name speckle-throated otter is that most generally 
accepted for this animal; but it is a little misleading in that the picture conjured up by 
"speckling" is of a much fmer order than that occurring in this animal For that reason 
it has sometimes been termed the spotted-necked otter. This, though in one respect 
more accurately descriptive is in another less so since the back and sides of the neck 
have no markings. The most correct name would be spotted-throated; but this has 



14- Tin; (AKN'IVOIMA OI WEST MIUC.A 

never Ihx'h used — possibly Ixwuiso it is less euplionious than the term most eonimonly 
.ipplicd. 

The speckle-throated otter is widely spread throughout Africa trom near the Cape 
northwards to Ethopia in the east and Liberia in the west. Specimens have been recorded 
from most West African territories but in the British Museum exist only from Sierra 
Leone (Waterloo, and near Kenema); Nigeria (Oban, Maiduguri); and Camcrouji 
(?Ndop): six skins only, two of them juvenile, and two skulls, one alone of which is 
adult. Nevertheless, maculicoUis is thought to be not luicommon in suitable localities 
tliough, being secretive and wary, it is not often come across by collectors. A.J. Hopsoii 
observed it to be frequent on tlie shores of Lake Cliad. It is also plentiful aroinid Lake 
Victoria (Proctor, 1963); and it is possible that the species is more at home in lakes 
and similar wide expanses of water than in rivers. Skins are not commonly seen on sale 
in local markets. Indeed, T. S. Jones (personal communication) thinks tliat, at any 
rate as far as Sierra Leone is concerned, tliis is a much scarcer species than Aouyx. 
Amongst about 30 otter skins lie examined during his years in that coimtry only one 
was Lntid — that cited above from Kenema. Kuhn (1965) gives this species as occurring 
in the following places in Liberia: Farmington River, Gbanga, Kpeaple, Biadatou, 
Deaple. Harbcl, Kahuple, Siron, Tappita, Towaitown and Zwcdru. 

About 10 different races have at one time or another been named, 6 of them recog- 
nised in G. M. Allen's Checklist (1939). None of these is actually from the area dealt 
with in this book, but it is possible that matschici Cabrera described from lower Camer- 
oun applies to otters of this species from the forest belt of Nigeria; vN'hile iiiloiica 
Thomas may cover specimens trom the Sahel zone. The position, however, is by no 
means clear. Harris, indeed, cites Liberia, Nigeria and Cameroun as comitries of 
occurrence of the nominate race though these areas are, of coiu-se, geographicallv, and 
probably ecologically, most distantly separated from the type locality. 

Description. Liiiii'i inaciilicollis (Plate 2) is the smallest of our three otters, \\ itli a 
head &; body length of about 65 cm and a tail which is rather more than halt this. The 
full-grown weight is about 4-5 kg for a dog otter or 3-5 kg for a bitch (Mortimer, 
1963). These figures arc the outcome of actual weighings of living animals; the often 
quoted "not more than 20 lb (9- 1 kg)" sems to be nothing more than a visual estimate. 

The pelage of the speckle-throated ottt r is almost entirely a deep rich red brown — 
intense sepia — both above and below. The base of the fur is, as in all otters. \\'hite. 
The only exception to tins deep brown lies on the throat and sometimes the chin, the 
fore part ot the chest and the after part of the belly which are white, spotted or blotched 
with the normal dark ground colour. The size of these mottled areas varies a great 
deal from individual to individual, being sometimes verv much reduced or, particularly 
as regards the belK' patch, lacking Tlie upper lips arc also narrowly white, but this is 
not very noticeable It is the extent of these white markings which best serves as a means 
of field recognition; for although the other two West African otters are bulkier and 
have a rather paler pelage, age mav affect size, and wetness the apparent colour of the 
tur. In the species now under discussion the white is verv strictly confined to the lower 
surface ot the bodv, whereas in both Aoiiyx and P<ir<uviy.\ it uninterruptedly covers a 
iiuuh larger area, rising to the le\i'l of the eves and the sides of the neck and is thus 






Spccklothrontcd Otter (Liiira mnailicollis): Cape Cla\vlc« Otter [Amyx dipeiisis) 



LUTRA 143 

readily observable in various postures and at a distance. Albinos or partial albinos are 
known to occur. 

Other external features by which uhicitlicollis difiers from .4(i;/)'.v are only detectable 
in the dead animal, or at least at very close quarters. One of these is the rhinarium, 
which is much narrower than in the clawlcss otter, consisting virtually of two narrow 
wings enclosing the nostrils without any broad central, dog-like, dorsal area. The 
second lies in the feet, which, unlike those of Aonyx. have short but very distinct, 
sharp claws on each of the five digits of both fore and hindfeet, and the toes fully 
webbed practically to the tip. The palmar pad is a small four-Iobed structure confined 
to the proximal part of the foot; the plantar pad is poorly defmed and consists of three 
narrow lobes (Pocock, 1921b). The undersides of the digits have long hairs; the webs, 
sparse hairs. The eye is orange-red in adults (Proctor, 1963). 

For the description of an imusual skin see the taxonomy section below. 

Skull. As only one adult skull, and that broken and of a not fully mature animal, 
exists from witliin the boiuidaries set tor this work, the following description is com- 
piled largely from extralimital material. The Lutm (Hydrictis) skull (figs. 20 and 21) 
is not only much smaller than those of the other West African otters but also ot an 
altogether less robust build. The braincase is ovoid, either without any trace of a sagittal 
crest, or in fully adult males a low one, often more m the nature of a ridge than a 
crest. Only the oldest females have anything of this kind at all ; and it is, in fact, so unusual 
amongst a wide range of specimens as to raise doubts about the accuracy of the sexing 
in these few cases. Much the same applies to the supraoccipital crest, which even in the 
oldest males is not very prominent and only rarely markedly upcurved. 

The braincase narrows sharply anteriorly, and forward of this the intertcmpoial- 
intcrorbital reigon becomes very conspicuously narrow and roughly parallel-sided to 
the deeply depressed, sharply trtuicate rostrum. Ver\' noticeably absent are any signs 
of postorbital processes; and it is this peculiarity which most clearly separates Hydrictis 
from Lutrn sciisu stricto. The anterior nares are very wide. The zygomatic arch is fairly 
slender but none the less strong, its anterior root widely divided to form a large oval 
intraorbital foramen. There is a sharp clearly-defuied spur on the upper margin of the 
jugal marking the lower posterior limit of the orbit. The palate is continued well 
beyond the toothrows. The bullae are of moderate size and rather less flattened than 
in other otters, including the typical European Liitra; the mastoids are not very pro- 
nounced — as they are in Aonyx and Pcinioiiyx. The mandible is moderately strongly 
built. In mature skulls the sutures fuse leaving no trace. 

The dentition is initially sharply cuspidate and, in fact, retains some of this character 
even in old skulls. The compact toothrow is crowded towards the canine, both p^ 
and p^ being closely approximated to this tooth; the former of these very small and 
sometimes deciduous, though much more rarely so than in other species. The incisors, 
both top and bottom, form two straightly transverse rows, tightly packed and strong. 
The size and form of the posterior cheekteeth in relation to those of the other African 
species are shown in figure 22. 

Habits. There is not a great deal of particular application to this species to add to 
the general review ot otter habits already given. Detailed studies have been made of 



144 



( AUMVOKl.S Ol WIS I AlKKA 



/,. iihuiiluollis by I'miti-r (1963) ni tlic ticld 111 1 anzjiiia, and by MortiiiKr (I'X)}) in 
/ambia, mostly im a ilonicsticatcd female. Moth accounts tend U> show tliat tlic specklc- 
tliroated otter corresponds in liabits and beliaviour tairly closely to other Liiira species 
in other parts of the world. 

rishing may take |ilace at any hour of the day, or sometimes night, although on the 
relativelv undisturbcxl sliores of Lake Victoria the species a]ipears to be chiefly diurnal; 
it is commonest soon after sunrise; but the otters may actually alrc-ady have been swim- 
ming about for some time before this in the scmi-darkness. From Procter's account 
it would seem that maculkollis is markedly more social than most. He records that 
parties of from 1 to 6 are fairlv common, larger numbers than this getting progressively 




Fig. 20. Lutra maaiUiollis: skull. B.M. No. 23.1.22.44, se.\ 



I ; lateral view 



rarer; but he once observed a schoiil totalling about a score. Small parties are mostly 
bitches; the larger ones dog-otters, often voung. The two se.xes arc mostlv difficult 
to distinguish apart in the water, though full-grown males are appreciably more 
heavily built than females. On shore, identification is easier, particularly at times of 
excretion since in the dogs the urine is projected forwards, in the bitches towards the 
rear. Defaecation is invariably accompanied bv urination, but not I'l'a' versa; and, 
as in other otters, there are recognised sprainting places. The spraints are mostly of the 
common, blackish slimy, otter type; but drier faeces of artliropod shells and fish scales 
are also dropped. In excretion the body crouches low, almost touching the ground; 
a bitch holds the tail out stiff horizontally, the dog vertically. Passage of food through 
the gut may be extremely rapid. Hoase-cleanluicss is evident in captive specimens. 

Like other otters, macuhcollis indulges m play both in and out of water. Amongst 
other things it is not infirequently given to teasing its prey, flinging fish back and forth 



I.UTIIA 145 

liom l.iiul 10 w.iirr .uiil |nirMiini:, it, sonu'lniu's giviiii; U a sliglu nip In ilisaMc it 
p.irli.illy ,ukI slow down us iiiovciiK-nts. (!r.ibs .nul smh like ,ur Inittcil alioiit with the 
nose, i'roeier, tluniii; loni; observation oftiie speikle-tliroated otter arouiul tiie shores 
oil ake VieliMi.i never saw nukl-slKles like tiiose reeorileil as made elsewliere hy other 




Fig. 21. I.iilrii iiiiUiiliiollir. skull, 11. M. No. ij. 1.22. 44, sex ?, X I ; p.ilatal t\ 



ili»rs.u N'lcws 



species; Init Mortimer notieeii what might well have been one ol these, the more so as 
there were otter tracks around it. Incidentallv, the tracks (".seals") lett hy this otter 
can be readily distingiiislied troin those ot the two otlier West Atricaii species by tjie 
clear prescjice otclaw marks and, on very soft mud or sand, the taint imprint of webs 
reaching to tlie ends ot the digits. 



146 rill-. CAUNIV ORES or WKST AlUICA 

Wlu'ii diviiii; to tisli. the liod\' is arLlu'd .ilinost clear ot rlic \\ate\ and tlic subscqiioiit 
dive is luarlv vertical. This otter, as other species, exhibits astonisjiing grace and 
agilit\-, especially 111 the rapidity ot its turns: but there is not much actual speed m 
swimming. Mortnicr gives measured figures tor short distances which indicate a rate 
ot" progress through the water ot only about 4 km an hour. Both tore and hindlcgs 
are used in swimming, the main thrust coming from a vigorous use ot the latter. 
Apart trom swimming, diving is also graceful and skilled, being earned out trom 
heights of up to 6 feet, leaving scarcely a ripple on tJie water. One ot these otters, 
however, is not above an ungainlv scramble into the water with a large splash wlien 
suddenly alarmed. 

On land the gait is rather awkward but progress somewhat taster, f-or ordinary 
purposes a rather clumsy walk is used in which the feet are moved alternately or the 
hindlegs as a pair — this latter a gait sometimes used by domestic dogs as an alternative 
to the normal trot. A taster speed can be achieved by a run, the legs moving alternately. 
Mortimer's measured speed tor this is about 5.2 km an hour; and tmally there is a sort 
of gallop, the animal ■"humping" along, moving fore and hindteet in pairs alternately. 
This was found by Mortimer to be at the rate of about -j-z km an hour, though it is, 
said that over a short distance one ot these otters can keep up with a running man, 
which would be at least twice as tast as this. Mortimer's experience with a young 
tame otter made him doubttul whether this species could travel long distances overland 
as this one, at least, appeared to become very soon overheated. Clambering up and 
over rocks is effected without diiiicultv; and Procter observed tliat a jump ot at least 
one metre could be made across a gap. 

Although various toods, including crabs and moilu cs, .nv taken tish seem to con- 
stitute the chief diet in this species. Procter tound Hiiplochivniis oi from 10 to 20 cm to 
be a favourite on Lake Victoria, a genus that occurs, but much more rarely, in West 
Africa. TiLipia is readil v acceptable but often proves too tast to catch. Mortimer observed 
that a numiier offish would be landed betore the meal was actually commenced; but 
this IS certainly not always so. His tully-grown bitch ate trom 4-1 to 4-7 kg a week; 
and, as otten popularly held tor other species, invariably started at the tail end. Smaller 
tish or molluscs are eaten in the water, either upright while treading water or while 
tlie otter is floating on its back. Vegetable tood seems not only acceptable but desirable; 
tor Procter observed one of these otters to eat a small quantity ot sedge, and Mortimer's 
tame one deliberateK robbed the garden ot carrots, beans, potatoes and peas. 

When not in the water these otters spend a good deal ot time grooming their tur. 
On emergence from the water the head is shaken but not the body, the drying ot wliich 
IS aided bv the usual otter practice ot rolling or rubbing. Vision on land is obvious!)' 
only good over very short distances; hearing and smell are better developed. Mortimer 
observed the latter to make immediate response to seemingly quite insignitic.uit quan- 
tities of water in flower vases or other small vessels. A pungent musky smell is emitted 
as a reaction to fright, in yoimg animals at least. The soiuids uttered are various. Procter 
describes four. The commonest call is "a shrill chikkering", used when playing or 
scolding. The second is "a prolonged mewing ya-a-ii-ii', probably a challenge. Another 



LUTRA 147 

is "a squealing whistle", made when excited, as in play-figliting; and he once heard 
an "ic-yaiig" call but never anytliing resembling a bark. Mortimer described the 
commonest sound uttered by his iiuiadicollis as a high-pitched squeak which, in certain 
circumstances, changed to a high-pitched trill. 

Very little has been found out about breeding in the speckle-tlnroated otter. It 
would seem that a litter of 3 is a common number. Procter thouglit the period of 
gestation might be a little over 2 months. Mortimer, who observed and recorded growth 
over a period of nearly two years, found the weight of his female to be 1-36 kg at 
li months; 2-5 kg at 5 months; 3-3 kg at 6i months; and 3-5 kg at 7A months, at which 
time it was apparently full grown, for the weight thereafter remained stationary at 
this figure, at least up to 20 months when the weigliing ceased. The body length, too, 
remained luialtercd. Teeth were still being cut at 6 months. 

Taxonomy. Nine forms have been named from different parts ot Africa, six ot 
them currently recognised in G. M. Allen's Checklist (1939). These have been diag- 
nosed, often as independent species rather than races, on the characteis ot size or colour 
or markings, and mostly from single specimens or otherwise quite inadequate material. 
Colour is notoriously uiu-eliable and in iiiMulicclIu is known to change with age (Procter, 
1963). The same author indicates that the spotted white underside markings are even 
more luidependable since even among the limited population on Lake Victoria "there 
is a very great individual variation in the amount of white on the throat. Every gradation 
is seen firom no white spotting at all to the whole throat, chest, belly, front and sides 
of the fore-limbs and part of the hind-limbs being almost completely white". It has 
been suggested that both mat<chki Cabrera, from Spanish Guinea with very dark 
pelage and relatively large teeth, and iiilotica Thomas trom Sudan, reputedly of yet 
larger size with larger teeth, may occur, in their respective habitats, in West Africa. 
But it will be obvious that, even it these races should pirove to be valid, it is impossible 
to tell from existing West African material — one broken subadult skull and no skin 
furnished with measurements — whether this is so. 

The anterior halt of a very interesting skin exists in the British Museum, No. 34.9. 16. i 
trom Maiduguri in the Sudan woodland of Nigeria. This is dorsally of a golden-brown 
colour rather lighter in tone than the usual run of rich chocolate inaailicollis pelage ; 
but the chief interest lies in the throat and chest area, normally pure white spotted with 
groimd-colour. In this case it is merely a lighter golden colour without the least in- 
dication of any spotting ; and, moreover, this paler area reaches higher than is usual in 
maculicollis, in tact to about the level of the eye and ear, the side ot the neck and shoulder 
much as in Aonyx. Indeed, the specimen, which is a flat one, at first glance gives the 
impression of being an Aonyx skin that has been smoked and so discoloiurcd; and it is 
in this and other respects not far removed from the type ot Waterhouse's Aonyx 
pocnsis, No. 55.12. 24. 414. But it is not smoked; and the only foot now remaining on 
this partial skin, the left foreleg, bears three distinct claws, proving it incontrovcrtiblv 
to be Lutra. 

The measurements in the table on page 159 are derived from extralimital 
specimens of Bates's together with the yoimg adult skull from Oban, south-east 
Nieeria. 



148 



THE CARNIVOUFS OF WEST AFIUCA 



Genus AONYX Lesson, 1827 
African Clawless Otters 

.-loiiy.v Losoii, iS:!7, Mmnul lic maminaloiiic, oil hisloirc nalurcUc dcs iiiatiiiiiijcrcs: !$"■ Type species Aoiiyx 
dilalaiiM Lesson ( -= Ultra capinsis Schinz), from the Cape ot Good Hope. The name is the two Greek 
words a without, and onyx claw, with rctercnce to tlie tect. 

Aiiahysli-r Murray, 1R60, Proc. R. phys. Soc. Edinb. 2: 157. Type species Aiialiyslcr culaharkm Murray 
(= Liimi capcnsis Schinz) from Calabar, Nigeria. This name is given by Murray himselt as meaning 
belonging to an estuary, since Calabar lies in the estuary of the Cross and Calabar liivcrs, but whence 
such a meaning was derived is not clear. 

Piiriiiiii/.v Hinton, 1921, .4iiii. Mag. nat. Hist, (g) 7: 194-196. Type species Paraoiiyx philippsi Hinton, 
Uganda. The name is a compound o(Aoiiyx with the Greek prefix para meaning near. Valid as a sub- 
genus. 



There arc two groups of clawless otters, one wholly African, the other wholly 
Asian; the former arc assigned to the genus Aouyx, the latter to Ainhlouy.w cither as a 






Fig. 22. Lutrinae: posterior cheekteeth; top row, right upper ;''' and 111^; bottom row, leit 

lower Hii and iih, ■ 2: a. Litlra iiiaciilicolhs. B.M. No. 23.1.22.44; b. Aoiiyx capci:si<, 

B.M. No. 10. II. 25. 2; c. Aoiiyx {Paraoiiy.\) coii(;ica. B.M. No. 193S.9.29.8 



AONYX 149 

full genus (Pocock, 1941 ; Simpson, 1945) or as a subgenus ofAonyx (Ellerman & Morri- 
son-Scott, 195 1). Tlic Asian animals are much smaller than the African and have their 
feet webbed to the ends of the digits, whereas the African ones fall short of this. The 
African clawlcss otters arc themselves divided into two groups, once considered to be 
distinct genera but now usually dealt with as one, though split into two pretty clear 
subgenera. These two sections have many points in common; but in spite of this it 
seems best to proceed at once to their separate consideration. The important diagnostic 
character lies in the dentition but the two subgenera may be told apart thus: 

KEY TO THE SUBGENERA OF AONYX 
(previous key page 140) 

Dorsal pelage only slightly "frosted"; the crown width o( iii^ about 13 to 15 mm, 

andof/di 10 mm or over ...... Aonyx (^page 149) 

Head and shoulders often with a well "frosted" mantle; the crown width of »|i 

about 9-5 to 10-5 mm, and of );!i 6-5 to 7-5 mm . . Paraonyx (page 154) 

Subgenus AONYX Lesson, 1827 
Typical African Clawless Otters 

Aonyx comprises a single species, capensis Schinz, and is sufficiently differentiated 
from its companion subgenus in the above key, and no further subgencric details are 
here necessary. 

AONYX CAPENSIS Schinz Cape Clawless Otter 

Lulra capensis Schinz, 1821, in Cuvier's Das Thiarckh . . . aus dcm Framosiihcn frey iihcrsctzt ... I: 214. 

Cape of Good Hope. 
Lutra immgiiis F. Cuvier, 1823, Diclionnaire des Sciences iiaturelle . . .,27: 247-249. Cape of Good Hope. 

The specihc najnc is the Latin for without a claw. 
Aonyx delalandi Lesson, 1S27, Manuel de mammalo^ie oil histoire iianirelle des mammifercs: 157. Cape of 

Good Hope. This was called after Mons. Delalande, who provided the information on which the 

species was founded. 
Lulra poeiisis Waterhouse, 1838, Proc. zool. Soc. Land.: 60. Fernando Poo. Type m the British Museum, 

No. 55.12.24.414, sex?; skin only, in poorish condition. 
Aiiahyster calabariciis Murray, i860, Proc. R. phys. Soc. Ediiib. 2: 157-158. Calabar, Nigeria. Type in the 

British Museum, No. 63. 12. 17.5, 5; skull only, in fair condition but 3 upper incisors, the right m^, 

and 1112 from both sides missing. The original B.M. number was 836b. 
Lutra ganibiaiiiis Gray. This was listed as a s)'nonyin by Gray himself under Aony.x lalandii in Proc. zool. 

Soc. Loud, for 1865: 130, as having been published m "Gray, Cat. Mamm. B.M. in (skull, B.M.)". 

In Gray's own handwriting in a MS. catalogue the reference appears as "Lutra Gambianus Gray 

Ost. Cat. p. 141, 1847". Both references appear to be quite untraceable, and no diagnosis seems, in fact, 

ever to have been published. The undescribed intended type, however, is in the British Museum, 

No. 46.11. 2. 12, o, (original No. 836a); skull only, in moderate condition, partly broken and with a 

number of teeth missing. 
Lutra lenoiri Rochcbrune, 1888, Vert. Nouv. Afr. Occid., ser. 3, p. 9 (according to Trouessart, 1897, Cat. 

Mamm. I'l'i'. Foss. cd. i, pt. 2, p. 285); but this reference has not been traced — nor, seemmgly, was it 

by G. M. Allen (Chcckhst, 1939) cither. 



I>0 THE CARNIVORES OF WEST AFRICA 

Distribution and general. This is possibly the most widespread, it not the com- 
monest, otter in Africa. Most otter accounts, both ot wild and domesticated animals. 
seem to conccrr. th;5 species rather than Ltara. Described originallv irom the Cape of 
Good Hope. :? somewhat misleadingly todav referred to as the Cape clawless 

otter, it has .^^.^c j^cn collected or reported from a large number of places in most 
countries ixom South Africa northwards to Ethiopia in the east and Senegal m the 
west. It is as much a forest animal as one of the more open and arid woodlands ; and 
it is knov^Ti to occur on mountains at over 2000 metres. G S. Child (private communica- 
tion) records it from Kainji Lake and the Borgu Game Reserve in the Doka bush of 
e."ctreme ■western Nigeria. It is said to inhabit, sometimes, the same localities as Liitra; 
and its unmbtakeable tracks have certainly been observed by A. J. Hopson in the 
sand some 350 metres from the shores of Lake Chad, on which ituiculici^llu is kno\\ii 
to be tairly common. Moreover, skins of .-iLi/iy.v are frequentlv exposed for sale in 
Malamiatori market, not far ir -ke. In Sierra Leone Aonyx is. according to 

T. S. Jones (personal commun:,:^:.-.. , plentiful in all areas, being frequent in the 
mangrove swamps. Skins are, there also, commonly offered for public sale. It seems 
possible that Aonyx sensu stricto is entirelv replaced in certain areas bv the subgenus 
Paruonyx. 

Nine 'West African specimens exist in the British Museum from: an unspecified 
locahtv- in Fernando Poo; Calabar (forest belt), Zaria (Doka woodland) and Maiduguri 
(Sudan w-oodland), all Nigeria; Ashanti and Asikum near Oda (both forest) in Ghana; 
Sierra Leone, Bonthe and an unspecified localirv- but almost certainly forest; Bolama 
Island. Porruguese Guinea (mangrove?) ; and Gambia, locality unkno\\ni. Two are 
complete juveniles; of the remainder onlv two have skulk, and there are no external 
measurements whatsoever. 

Description. The Cape clawless otter (Plate 2) is a bulky animal. At 16 to 20 kg 
it is tar larger than L. inacuUcoUis; and even bigger specimens than this have been 
recorded in South Africa, up to nearly 30 kg. The head & body length is from about 
750 to 925 mm and the tail about 110 to 5S0 mm, but precise measurements for West 
Africa do not exist. Indeed, of all the Aonyx skins in the British Museum only three 
have measurements, one from Kenya and two from Ethopia. The coat is variable in 
colour from a deep red-brown, almost as dark as L. indLiilicollis, to a distinctly paler 
mid-brown, the spinal zone being a little darker than the sides. In some cases the 
hairs have pale tips, panicularly over the shoulders and neck, but these are not so white 
as in Puraoiiyx and the overall effect not nearly so "frosLcd". The feature which, apart 
from size, visually distinguishes this otter from maadkollis is the wholly white or 
cream, quite unspotted, chin, throat, upper chest, side of the neck and of the face to 
the level of eye and ear. The margins ot the ver\- small ears are white. 

It is the feet, however, that constitute the most important ditFerence berw,-een Aonyx 
and LutTd. They are quite devoid of the usual carnivore long, sharp claws though in 
some cases furnished v\.-ith rudiments. These last are to be found only on the hindfoot, 
usually as tiny flattish nails on the 3rd and 4tli digits. The digits themselves are finger- 
like in appearance; and, indeed, the whole paw is very similar to a small hand, the 
more so as the interdigital webs are much reduced on the forefoot. The structure is. 



AONYX 151 

in fact, used for grasping much in the maimer of a hand and is far more deft than the 
usual mammalian structure apart from the primates. The palmar and plantar pads 
are better developed into a miited central pad than in Liitra. The rhinariimi is broad 
and rather dog-like. 

Skull. The mature Aoiiy.x skull is an incomparably larger, stronger and more rugged 
structure than that o( Ltitra maculicollis. The rather flat braincase is not only broader 
but has also, in the males, a sharp, deep sagittal crest and pronoiuiced, upturned supra- 
occipital crest which continues round the sides of the braincase to exaggeratedly large 
and prominent niastoid processes. There are also large, slender paroccipital crests. 
In the females this cresting is not so higlily developed ; while in the young of both 
sexes the cranium is roiuided and quite smooth. 

The long intertemporal and short interorbital legions are abruptly narrow and more 
or less parallel-sided ; there are sharp postorbital processes and slight ridges joining 
them to the sagittal crest, thus marking out the pentagonal plate usual in this subfamily. 
The arches are strong with well-developed jugal hooks reaching up towards the post- 
orbital processes. The rostrum is extremely short and falls away almost vertically; 
the anterior nares very large. The palate is rather narrow and continues, parallel- 
sided, well to the rear of the toothrows. The bullae are small and flat. The condylar 
hinge is deep and strong; the mandible ver\' powerfully built, with a tall coronoid 
process, much excavated at its base in the ramus to accommodate a large muscle. 

The dentition is powerful. The incisors of the upper jaw form a nearly straight, 
serried transverse row, those of the mandible being often a little more irregular. The 
canines are exceptionally long and sharp, suited to penetrating deeply and holding 
securely a peculiarly slippery and active prey; the fust tliree premolars, above and 
below, are relatively small, p'^, when not lacking, and often p- as well, being practically 
in contact with the inner face of the canine. The result of the smalhiess of these teeth 
is that, though there is not strictly what is technically known as a postcanine gap (see 
page 18), there is enough space between the two toothrows anteriorly to accom- 
modate a fair amoimt of flesh and thus increase the firmness of the canines' hold on a 
fish. The cheekteeth as a whole form a short compact series; the upper caniassial has 
one long pointed cusp, the lower three, set in equilateral fashion, but soon wearing 
do\\ii and together with the very broad posterior molars forming more of a crushing 
than sectorial imit, well suited to dealing with the soft flesh of fish and molluscs. In 
spite of the general impression of strength the teeth as a whole seem ver\- liable to 
both wear and damage. It is the great breadth of the posterior cheekteeth wliich 
chiefly serves to diflerentiate Aoiiyx seiisti stricto from the subgenus Paraonyx (fig. 22). 
The maximum crowii width o( m^ in the former is of the order 13 to 15 mm as con- 
trasted with 95 to 10-5 mm in the latter. 

Habits. Aoiiyx has often been kept in capti\'ity, and if procured at a sufficiently 
early age makes an agreeable, interesting and amusing pet, friendK" to both human 
beings and other animals. Its behaviour in domestication has therefore been well 
observed and corresponds closely to that of other otteis in similar radier anificial 
circumstances. But its way of life in nature is a different matter. Although the species 
has often been wxitten about most of these accounts consist, in fact, of a good deal of 



152 THE c:aknivoiu-s or \vi;st afiuca 

surmise aiid generalities and remarkably little that is actually positive. Statements, 
indeed, seem sometimes to be diametrically opposed; as when {fide Shoitridge, 1934) 
Lancaster says that Aoityx capcnsis is solitary as a rule, while Moseley asserts that it 
hunts in small companies. Both, of course, may be true, the latter applying, as in any 
other otter species, to a bitcli and her yoiuig. 

The African clawless otter may, in remote streams, lakes or swamps, sometimes be 
come across fishing or sporting in the water at high noon; but where it has reason to 
tear man it confines its activities to the early and late parts of the day, and possibly 
becomes largely nocturnal. A great drawback ot having to adopt a nocturnal existence 
must lie in the increased difficulty of detecting and capturing tish in the dark. Possibly 
crabs are easier. Aony.x is always held to be much less ot a fish eater and more of a 
crab eater than Liitra iiuuulicollis. It is a fact that the spraints contain a good deal ot 
crab shell, and the feeding sites, constantly revisited, are littered with claws and other 
remains of these crustaceans. The species is said to take, also, the usual list ot other 
foods — molluscs, lizards, small rodents, birds and so forth. Eyre's tame ^4ci;/j'.v greatly 
appreciated freshly-killed moles. It has in southern Atrica a great reputation as a 
fowl and egg thief; but no clear evidence of this seems to be forthcoming and any 
such robbery might as well prove to be the work of a ratel. Certainly any interest in 
eggs does not appear to lie in their attraction as food; tor Eyre (1963) records an 
occasion when a domesticated otter, having ot itselt discovered two dozen eggs in a 
cupboard, instead of treating them as a desirable meal, hurled them, presumably one 
by one for amusement in the usual otter fashion of dealing with a pebble, all over the 
floor. Zammarano (1930), as the result of first-hand experience in the field, decided 
that the clawless otters were endowed with very acute senses, and he characterized 
them as cautious and cunning in the highest degree and the hunting ot them, con- 
sequently, as anything but easy. The species has otten been observed on the sea coast; 
it is, or was some years ago, to be seen in the large lagoons, not far from the sea, between 
Tiko and Douala (Cameroun). T. S. Jones obtained a specimen in a comparable situation 
of mangrove and sea at Bonthe (Sierra Leone). 

Clawless otters make good pets if taken young and brought up on the bottle. In 
captivity these animals, though demanding in time and patience, e.xert considerable 
charm. They display quite unexpected intelligence, solving problems which they 
would not come up against 111 nature such as opening tins, bottles and cupboards with 
amazingly dexterous fuigers. The paws are used almost as hands, not only holding 
things but also to throw small articles (Eyre, 1963). This author describes a variety of 
sounds made; a high-pitchcd squeak, demanding attention; a purring growl with a 
warble in it when pleased; a whine of frustration; a hiss and a growl when trightened; 
and a "most luiearthly scream, most trightening, when he is in a temper". The explosive 
"Hah I" uttered bv .4i'/i)'.Y in the wild has already been referred to in the general obser- 
vations. 

Nothing defuiite appears to have been recorded of breeding. The normal litter 
seems to be 2 or 3, but as many as 5 has been estimated trom a tield observation of a 
bitch and cubs. T. S. Jones (personal communication) had numbers of young brought 
to him in Sierra Leone and from these the litter size appeared to be invariably 2. The 



AONYX 153 

period of gestation is unkiiowji but has been guessed at as about 2 months. There is 
probably no set breeding season in West Africa but T. S. Jones reckons that it is nor- 
mally between September and October, the whelps being born in a hole in the bank. 
On the 5th October he obtained a very young animal which had its eyes open but which 
he thought was still being suckled. The skull of this specimen is, in fact, very thin and 
there is no sign of the permanent dentition. The skin has the correct adult coloration 
though the fur, as noted earlier, has not its fmal composition. On the other hand, two 
newly-born specimens exist, both wholly white, above and below. Eyre (1963) says 
that in a young animal the round eyes are a deep navy blue. As with other otters, the 
young have to be taught to swim by their parents — a fact that must be borne in mind 
with captive whelps. They are often somewhat reluctant. 

Taxonomy, hi spite of the very v/ide range of this otter there is little doubt of there 
being but a single species of ^4o»)'.v in its restricted, subgeneric, sense. The question of 
races is more obscure. Pohle (1919) recognised four, With, poensis Waterhouse in addition 
as a separate species; G. M. Allen (1939) six, none of them from West Africa. There 
have, however, been four attempts to relate special forms to this region, three of the 
proposed types resting in the British Museum. In 1838 Waterhouse named pocnsis 
from a skin which, from its size and the nature of the pelage, appears to be that of a 
young animal. The diagnostic characters given, largely colour, are without much value ; 
the skin has no feet and it is therefore not absolutely certain that it is Aoiiyx at all. 
In fact both Thomas (1889) and Lonnberg (1910) considered it to belong to Lutra 
maculiaillis. However, the throat, neck and sides of the face follow the pale, unspotted, 
pattern of Aonyx, but instead of being white or creamy they arc, as Waterhouse 
described them, "of a rich deep golden yellow with a faint brownish hue" — though in 
the past 130 years this has become obscured by London grime. Pohle (1919), disagreeing 
in respect of synonymy with Gray, Trouessart, Thomas and Lonnberg, accepted this 
unusual colouring as having full specific validity. He further entirely rejected Thomas's 
and Lonnbcrg's opinions regarding the genus of this specimen on the grounds that the 
throat was luispotted, a character that never occurred in the niaculicollis group. That 
Pohle himself seems to have been wrong in this is demonstrated by the iiiaciilicollis 
skin, B.M. No. 34.9. 16. i, with indisputable claws on a forefoot but without spots, 
and not very dissimilar from the specimen now under discussion. The possibility 
therefore yet remains that the poeiisis type is, in fact, a Lutra; and nothing further can 
be decided regarding its generic, specific or subspecific standing until more and better 
material emerges from Fernando Poo. 

The next proposed form, calaharicm Murray from Calabar in south-eastern Nigeria, 
was foimdcd on a skull very poorly diagnosed by Murray, not only as a species but as 
the type of a distinct new genus as well because it had one fewer premolar in the 
upper jaw than Lutra vulgaris of Europe. This absence oi p^ is now known to be a not 
unusual dental variant, and J. A. Allen (1924) regarded the species as "practically indeter- 
minable" from Murray's diagnosis. The skull, now in the British Museum, appears to 
dirter in no material respect, except possibly its slighter size, from other typical capetisis. 
Next, never referred to in literature except by Gray himself, comes his gamhianns, of no 
date but prior to 1865. The skull of this reputed species also is in the British Museum 



1_S4 THE CARNIVORES OF WEST AFRICA 

and, apart from its somewhat larger size, docs not differ from typical capcnsis. The 
name is a iwincn uihlimi; no description of the species seems ever to have been attempted 
and it appears, with untraceable reterences, merely in the list of synonyms ot Aoiiyx 
laliwdii Lesson given in Gray 1865: 130 and Gray 1869: 109. 

Finally there is Rochebrune's supposed kiioiri from Senegal. The work in which 
this is reputed to have been described is said (Thomas, 1889: 196) to have been privately 
printed and hence never validly published. However, he apparently actually examined 
Rochebrune's description and remarked, in a footnote, that the so-called "diagnosis" 
surtered the usual lack of all diagnostic characters. This, therefore, is also a ncmcn 
nudum ; and in this case there is, moreover, no biown existing specimen of the reputed 
species whatsoever (Lonnberg, 1910: 3). 

Whether or not there arc forms that might constitute valid West African races is 
quite indeterniinable from the study material at present available. In the erection of 
such races considerable caution would be called for in view of the knowii idiosyncratic 
variation that occurs in otters — illustrated earlier in this account by Procter's obser- 
vations oiLntra macnlicolUs. The situation, in fact, has changed very little in the half- 
century since J. A. Allen (1924), with commendable understanding and restraint, 
wrote of the then reputed forms of ^4(';i)'.v capcnsis: "As these five forms appear to have 
been described in each case from a single specimen, without flesh measurements and 
in some instances from poorly prepared material, none of them can be said to rest on 
a very satisfactory basis. The differences in coloration indicated by the descriptions of 
these forms are more than covered by the range of variation in the present Lang- 
Chapin series of some twenty specimens from a single locality (Faradje), while the 
individual difference in size is more than covered by the twelve adults. The status of 
these various forms should be held more or less in abeyance until a good scries from 
each type locality has been studied and compared. Under such circumstances, it seems 
better not to add another name to the list . . . ." 

Such skull measurements as are derivable from the present very poor West African 
material in the British Museum are given in the table on page 159. There are no external 
data. 

Subgenus PARAONYX Hinton, 1921 
Small-toothed Clawless Otters 

The chief characters which distinguish this subgenus from Aoiiyx scnsti stricto are 
sufficiently shown in the key on page 149. The main point of difference certainly lies 
in the dentition, the contrast in size of the posterior cheekteeth being so marked, 
as shown in figure 22, that in spite of the otherwise strong overall general resemblance 
of the skulls it is easy to tell the two subgenera apart at a glance. 

Three species have been described: couq^ica Lonnberg from lower Congo, philippsi 
Hinton from Uganda, and micradon Pohle originally from lower Cameroim but 
subsequently reported also within the limits set for this account. These have been 
pretty generally accepted as independent species; but it appears to the present writer 



AONYX 155 

that the evidence in support of this, based as it is on insinifigcant differences of size 
argued for the most part from single skulls, is very slender indeed. The three "species" 
can at most be reckoned as races; and the value of even such a reduced status is, on the 
few data now available, of doubtful worth or validity. Lonnberg's description ot 
congica, the earliest form now assigned to Paraoiiyx, was based largely on a broken skull 
from which very limited measurements could be taken. Because of tliis, PolJe, in 
later erecting micmdon, was able to make only three cranial comparisons, though in 
respect of teeth he made ten. None of the differences he cites seems to have much 
importance. Hinton in diagnosing philippsi as the type species of his new genus Para- 
onyx appears to have been quite unaware of PohJe's earlier creation. The three forms 
are here treated as a single species, congica. 

AONYX CONGICA Lonnberg Small-toothed Clawless Otter 

Aonyx capensis congica Loiinbcrg, lyio, Ark. Zool. 7, No. 9: 1-8. Lower Congo. 

Aoiiyx iniaoioii Pohle, 1920, Arch. Nalurgcscli. for 1919, sect. A, pt. 9, 85: 145-147. Bomsc, Nana River, 

Cameroun. The specific name is from the Greek micros small, and odon tooth. 
Paraoiiyx philippsi Hmton, 1921, Ann. Mag. nat. Hist. (9) 7: 196-200. Lake Bunyonyi, British Ruanda 

(= Uganda). Type in the Biitish Museum, No. 21. i. 22.1; skin in good condition, skull with the 

arches broken and partly missing. Called after Capuin J. E. Philhpps, M.C., District Commissioner 

at Kigezi. 

Distribution and general. No specimen actually assigned to congica or emanating 
from the Congo exists in the British Museum; but there arc four skulls and five skins 
of philippsi, all from Uganda. The position as regards the form most closely associated 
with western Africa, microdou, is in some ways the best. The existence o£ Paraoiiyx in 
West Africa itself was not suspected until in 1938 Dr. M. D. W. Jeffreys obtained 
several skulls from the Nun marshes. Two others were procured later by the present 
writer from the same area, so that there arc now 9 skiills and a skin in the British 
Museum collection, a much better representation than there is of the other forms and 
of the more widely distributed Aonyx capensis. 

It seems impossible today to fix with precision the whereabouts ot the village of 
Bomse whence micivdon was originally described; but it was probably at about 5°55N., 
I5°I5E. The Nana River is a tributary of the Sangha and flows eventually into the 
Congo. It is thus just extralimital to the area covered by this present work. The Nun, 
or Noun, River flows into the Sanaga and thus lies well outside the Congo basin to the 
north. Its upper reaches pass through the Ndop plain and are there extensively marshy, 
the centre of this swampy area being at about 5°55N., io°25E., that is to say some 
40 km directly east of Bamenda. The surrounding area is today open country with 
scattered trees of a rather Guinea woodland type, though in the past it was very probably 
closed forest, rehcts of this vet remaining. The altitude of the Ndop plain, though lower 
than that of the siurroundmg mountains, is probably still of the order 1300 metres. 

No specimens assigned to microdon from elsewhere in the area herein treated as 
West Africa seem ever to have been recorded; but Pohle cites a number of localities 
of occurrence in lower Cameroun. These, however, are for skins imsupported by 



156 



THE CARNIVORES OF WEST AFRICA 



skulls, and their identification as what we now call Paraonyx rather thaii as typical 
Aonyx is, as Pohle himsclt pointed out, somewhat uncertain. 

Through the courtesy of Dr. P.J. H. van Brce of the Zoologisch Museum Amster- 
dam it has been possible to examine a skull procured by him from Loa-Loa near 
Makokou, North-cast Gaboon. This lies in a geographical situation distinct from those 
of the other specimens imder consideration, since it is neither in the Shari and Niger 




Fig. 23. Aonyx (Paraotiyx) amgica: skull, B.M. No, 1938.9.29.4, 5e.\ ?, ■■ i ; Literal view 



drainage systems, taken in this work as constituting the major part ot West Africa, 
nor in the Congo basin but in the westwardly sloping area emptying into the Atlantic 
via the Ogowe River. The locality is some 700 km south of the Nun marshes. Dr. van 
Brce's specimen ditfers in certain respects from the iiiicwdoii material from this last 
area; but it is less toothworn, and the basisphenoid suture is still detectable, as it is not 
in the British Museum material. Moreover, it is a female, whereas none of the London 
examples is sexed. These facts may account in some measure for the differences which 
arc apparent: there is no sagittal crest, no sharp supraorbital processes, the rostrum at 
the level of the canines is appreciably narrower, the mastoid processes are far less bulky, 
and the coronoid process more slender. But even it these distinctions held for further 
specimens from this area they are racial rather than spiecitic. 



AONYX 



157 




F.G. a4. Ao.y. iParaony.) co„sica: skull, B.M. No. Z938.9.39.4, sex ?. x .; paUal & dorsal 



I.SS THE CARNIVOKIiS OF WEST AFRICA 

Description. In general overall appearance Pciraoiiyx is extremely similar to Aciiyx. 
The body bnild is much the same, possibly a little larger; the pelage colour and pattern 
basically the same, that is to say deep sepia-brown on back and belly but the chin, 
throat, cheeks and chest wholly creamy-white without spots. The most obvious 
external dirtcrcncc lies ni the "trosting" of the head, neck and shoulders due to the 
bristle-hairs being white-tipped, A little of this may be present in A. capcnsis but it is 
usually either lacking or not very marked; in A. coti^ica, on the other hand, it forms a 
conspicuous feature extendnig from just back of the eyes to a little past the shoulders. 
In some cases it is pretty concentrated, in others well diffuse but none the less obvious. 
The margins ot the small ears, too, are more conspicuously white. The forefeet arc 
un webbed; the hindfeet only to the second joint, Hinton said that the facial vibrissac 
were weakly developed in comparison with capensis; but there is not much evidence 
that this is so except possibly in so far as those situated above the eyes are concerned. 

Skull (hgs, 23 and 24). Apart trom the dentition, which is comparatively shown in 
figure 22, there is not much to distinguish the mature Paiaonyx skull from mature 
Aotsyx except tor a rather broader interorbital and intertemporal breadth. The rami, 
instead of being more or less straight-sided, curve in towards one another. 

Attention has already been abundantly drawn to the relatively small size of the 
posterior cheekteeth, the most important diagnostic feature of the subgenus; the average 
crown width ot iii^ is only about <)h mm as contrasted with at least 50 per cent 
more in capensis. The canines arc rather shorter and slighter; and in 8 out of 13 skulls/)^ 
was never developed. 

Habits. Absolutely nothing is known ot the habits ot l\iraouyx. Lonnberg (1910) 
and Hinton (1921), however, both made some interesting surmises from the nature 
ot the tect and the reduced size of the posterior cheekteeth, supposing from the one 
that this species would probably prove to be yet more terrestrial than capensis; and 
from the other that it was not well adapted to dealing with hard-shelled Crustacea but 
rather with small terrestrial vertebrates and eggs. In this connexion it is of interest to 
note that the Ndop plain, which microdon inJiabits, has numerous small streams and 
other water-courses but no largish rivers such as otters normallv like to frequent and 
swim in and which would be much better stocked with fish. The swampy nature of the 
terrain makes it ideal for amphibian lite, and it would thus ccrtainlv seem possible that 
this otter consumes a high proportion ot frogs and relatively little tish. 

Taxonomy. Something has already been said ot this. There seems to be no good 
reason to suppose that three independent species exist. The question boils down to 
whether the named forms do in truth constitute valid races. The data are slender and a 
little confused. There is no specimen ot coiigica in the British Museum from which to 
make direct measurements and comparisons; in respect ot tins, then, there exist only 
Lonnberg's figures from a broken skull (incidentally, misquoted in Hinton, 1921). 
On this very fhmsy evidence it does, however, seem to be slightly the largest. The 
table on page 159 gives the impression that microdon is a little larger than pliilippsi; 
but the 3 skulls of the latter are all females; the microdon skulls are unscxed but from the 
appearance of the sagittal crest nearly all seem to be males. The evidence, therefore, 
IS quite inconclusive and a third name is, at an\' rate tor the time being, best omitted. 



LUTRINAE — MEASUREMENTS 



159 



Table 8: Numerical data for subfamily Lulrinac 





Lutra 




iiiacuUcollis 


Vegetation 


Various 


Number in mean 


4 


Condylobasal length 


105-7 


Basilar length 


94-4 


Palatilar length 


41-4 


Zygomatic breadth 


62-0 


Upper cheekteeth breadth 


32-6 


Interorbital breadth 


i6-o 


Postorbital constriction 


i6-6 


Braincase breadth 


5I-I 


Toothrow [c — ml) 


33-0 


p* length 


II-7 


ml length 


8-1 


ml breadth 


9-5 


mi length 


13-3 


ma length 


4-5 


Head & body 


648 


Tail 


378 


Hindfoot 


109 


Ear 


18 


RATIOS (per cent) 




Tail/head & body 


58 


Zygom. br./condylob. 1. 


59 


Biaincase/condylob. 1. 


48 


Braincase/zygom. br. 


83 


Palatilar l./condyloh. 1. 


39 


Interorb./postorb. 


96 


p'^jc—m^ 


35-5 


p^jm^ 


144 


mi/;n2 


296 



Aoiiyx capciisis 

calabariais gambiaims 

Type Type 

Forest PGuinea 



I 


I 


(130) 


(140) 


114-4 


I25-I 


59-0 


65-4 


86-7 


93-8 


(47) 


(49) 


26-3 


32-3 


24-5 


26-1 


66-5 


69-0 


42-4 


46-6 


14-4 


(u-7) 


13-2 


13-3 


14-5 


iS-6 


19-8 


I9-I 


(6-6) 


6-2 



(67) 
(51) 

77 
(45) 
107 

34-0 
109 

(300) 



(f'7) 
(49) 

74 
(47) 

123 

31-5 

no 
308 



A. (Parac 

inicrodon 
?Guine3 

9 
125-0 
113-2 

54-1 

93-4 

39-5 

32-1 

31-9 

70-2 

37-9 

9-8 

8-6 
10-3 
12-4 

4--I 



75 
5<S 
7S 
43 
loi 
25-8 
114 
302 



nyx cotigwa 

philippsi 

?Sudan type 

3 

118-0 

104-9 

50-2 

84-2 
36-6 

26-6 
29-4 
66-7 
34-9 

9-2 

7-4 

8-3 
11-8 

3-8 



71 
62 

79 
43 
91 
26-4 
124 
310 



l60 IHE CARNIVOHES OF WEST AFRICA 

Superfainily FELOIDEA Simpson, 193 1 

The morphological distinctions between this group and the corresponding super- 
family Canoidea have already been set out and discussed on page 29. Superficially 
the Feloidea arc a somewhat less heterogeneous assemblage. There arc only three 
families, all well represented in West Africa. The Felidac, or cats, may be taken as 
most typical of the group and, large or small, arc of unitorm and unmistakable 
appearance. A major section of the second family, the Viverridae, though with certain 
very obvious differences, bears in many respects a sufficient overall external resemblance 
to the Felidac to have earned the widely used general designation "bush cats"; though 
another section, the Hcrpcstinae or mongooses, stands fairly clearly apart. The third 
family, the H)-aenidae, provides for the layman something of a mystery in that these 
animals which bear considerable external resemblance to dogs should, in fact, be 
grouped systematically with the cats. However, the fossil record shows them to be 
closely connected with the Viverridae, of which they are a late offshoot (Simpson, 

1945)- 

As already made clear m connexion with the Canoidea there is no single character 
that indubitably defmes this superfamily other than the septate bulla; and that, as will 
later be seen, is deceptive in the case of the hyaenas, as Pocock (1916c) made clear. 

The three families, as they occur in West Africa, may be separated by the following 
key: 



KEY TO THE FAMILIES OF THE FELOIDEA 

(previous key page 27) 

Cat-like in form; head roundish with a short muzzle; ears (except m F. lihyca) 
wholly or partly black on their backs; skull with the rostrum abruptly 
deflected and short; braincase mostly rather roiuidcd but sometimes with 
a low, abrupt sagittal crest and well developed supraoccipital crests; 
cheekteeth only §-j or |-| Felidae [pa^ic m) 

Somewhat dog-like in form with long legs but the back sloping downwards 
from shoulder to rump; muzzle long but blunt, the jaws very powerful; 
braincase long and remarkably narrow, merging gradually into a sharp 
sagittal crest but with a poorly developed supraoccipital crest; rostrum 
long; cheekteeth j-| or ^ //yaenu/ac (/i./i^v 341) 

Usually of small or smallish size, legs short, muzzle pointed; coat often spotted 
or speckled; braincase long and oval, with a pronounced supraoccipital 
crest but at most a low sagittal ridge; rostrinn long; premolars 5-57-4, 
molars ^-^ Viverridae {page lOi) 



VIVERRIDAE l6l 

Family VIVERRIDAE Gray, 1821 
Civets, Genets, Mongooses etc. 

Distribution and general. The Viverridae are a large family divided into 6 sub- 
families and about 35 genera covering roughly 75 species. They are to be found in 
Asia, Africa and in a limited fashion, to the extent of one mongoose and one genet, 
in Europe. They are in a way specially associated with the island of Madagascar in 
that 4 of the 6 subfamilies occur there, 2 of them, embracing one-fifth of the viverrid 
genera, being endemic to the island and, indeed, the only wild carnivores there. Three 
subfamilies appear in West Africa; and in these there are 13 genera containing 23 
species. 

Most of the Viverridae are of smallish size; but in West Africa the civet is a relatively 
bulky creature running to as much, sometimes, as 14 or 15 kg. They live in all kinds of 
terrain from rain-forest to near-desert and up to an altitude of about 1 800 metres. 
Some are wholly terrestrial, some largely arboreal; none is wholly aquatic but some 
live near streams or swamps and swim well. Most of them are, except at mating time, 
of a solitary nature; but some of the smaller mongooses hunt in parties. Some are in a 
small and mosth' local way valued for their skins; one, the civet, has over the centuries 
played a more important and widespread commercial role in providing a basic in- 
gredient ot many perfumes; but all in an indirect way exert an influence upon civiliza- 
tion and the balance of nature in general by reason of their impact upon rodents, 
reptiles, amphibians and other creatures that man regards as pests and which would, 
in fact, get out of hand without the constant controlling influence of these small 
carnivores. 

Description. In general, the Viverridae have long, slender, somewhat cylindrical 
and often cat-like sinewy bodies, but there are a few of rather heavier build. The tails 
are mostly long or very long, sometimes highly mobile, always well-haired and often 
bushy. In some it is conspicuously ringed; but never, except in one extralimital genus, 
prehensile. The pelage is of varying length, generally rather on the short side but always 
dense, spotted, striped, speckled or, in rare cases, black. The legs are short, the feet, 
in African species, mostly digitigrade, the sole pads being highly characteristic. In 
West Africa there are with one exception always five digits on each foot (though this 
is not so elsewhere) and the claws in the genets and palm civets are partly retractile. 
The head is small, the muzzle long and sharp, the eyes large or small, and the ears 
sometimes, if not large at least very prominent, sometimes inconspicuous. A bursa 
may or may not be present on the outer rim of the pinna according to subfamily. 
The females have 2 to 3 pairs of abdominally situated mammae; in all, scent glands are 
well developed especially the so-called anal ones; but there may be others in the 
perineal region, between the base of the penis and the scrotum in the male, or between 
the vulva and anus in the female. 

Skull. The skull is elongate, with a braincase that is much longer than broad, ovoid 
in shape, smooth except sometimes for a low sagittal crest, and always with a prominent 
upcurvcd, sharp supraoccipital crest. There are marked, pointed postorbital processes 



i''i2 THE CARNivonts or wnsT atrica 

which ni some genera are short but vvhicli in otliers nearly approach or in some cases 
join thejugal process to form a complete circumorbital ring. The postorbital constriction 
IS generally marked. The zygomatic arch is moderately stoutly built, the maxillary 
process being yery short and penetrated by a tair, but not large, intraorbital canal. 
The rc'>strum is usually long, but less so in the Hcrpcstinae; the anterior nares wide. 
The palate is mostly rather narrow but broadens somewhat posteriorly and is con- 
tinued in a narrow- parallel-sided extension considerably to the rear of the toothiows. 
The bullae arc large and bulbous, showing anteriorly a more or less clear constriction 
indicating the internal septum dividing the structure into tw-o chambers; but Nandiiiia 
is extremely unusual in having the posterior one of these merely cartilaginous and thus 
disappearing from all but very carefully prepared skulls. The paroccipital processes are 
sometimes virtually absent or indistinct, but arc mostly well developed and closely 
adherent to the posterior aspect of the bullae. In old skulls the fusion of all sutures is 
very complete. 

Flower & Lydekker (1891) wrote that the second lower incisor was raised above the 
level of the first and third; but, at least as far as West Africa is concerned, this is wholly 
untrue. The six incisors of both jaws form a compact more or less straight transverse 
row, the outer ones being somewhat or pronouncedly larger than the others; in some 
cases the lower incisors arc bilobed. The premolars arc ot the usual form; the upper 
caniassial has an anterior lobe despite Flower & Lydekker 's assertion to the contrary, 
though mostly small and sometimes, with wear, indistuict. The upper molars are 
mostly narrowly transverse and are usually markedly smaller than this tooth, but m 
some of the mongooses j/A does approach />' in size; j;i-, when present, is generally 
much smaller, and in XiVidiiiiii is reduced to a peg. The dental formula is very variable 
even within a single subfamily, the total number of teeth being 36, 38 or 40. Since the 
incisors and canines are always '^ this is due to diversity in the premolars and molars 
of both jaws, these cheekteeth numbering 7-^, r^, j-,, or ^. 

Habits. Since these are vastly different in the different subfamilies, or even genera, 
little that is ot general application to the family can be said here. The Viverridae are 
by nature fierce, both in the pursuit ot their pre\' and in self-defence; but some ot them 
it captured at an early age prove themselves to be interesting if not charming pets. 
Many people have kept, and have derived pleasure and even benefit from, the civet, 
genets and several different kinds of mongoose; but some of these, if not all, exhibit 
far more independence of their homes and keepers than members of the Canoidea or 
even of the Felidae or true cats. 

The Viverridae are nearly all essentially nocturnal or crepuscular but some of the 
mongooses can be come across by day. They arc all basically carnivorous, consuming 
various kinds of smaller creatures, rodents, birds, reptiles, amphibians, myriapods, 
crabs and insects; but many, if not all, also consume fruits. They mostly live in small 
holes or crevices; but, in fact, extremely little beyond broad generalities is known of the 
way of life of the majority of them. 

Locomotion in certain African viverrids has been photographically investigated, 
analysed and sometimes figured bv Taylor (1970) in relation to these species occurring 



VIVERHIDAE 163 

ill West Africa : Civcttictis civcttii, Gciictta spp., Nainhnia hinotata, Atilax paludinosus, 
Herpestes ichneumon, Galerclla sanguinca, Ichnewnia albicaiuhi, and Miingos iitungo. The 
motions studied arc walking, trotting, running, galloping, jumping, climbing, swim- 
ming and burrowing. He found that the only form of locomotion common to all 
was walking; trotting is the gait most usual to the bigger open-country viverrids, 
probably in association with the larger territories there customary. 

Taxonomy. Of the three major groups into which the Feloidea arc traditionallv 
divided there is little disagreement regarding the clearly separate identity of the Felidae 
and the Hyacnidac and, at the same time, the close affinity existing between the animals 
of which each of these families is composed. The Viverridae, however, are more open 
to doubt. The six apparently natural groups of animals, comprising nearly 40 genera, 
therein lumped together may seem a somewhat heterogeneous collection to form a 
single family. Pocock, basing his ideas largely on external characters, gradually veered 
away from the standard conception of the Viverridae as a close-knit unit and came 
ultimately to elevate several of its six subfamilies to full family rank. All three West 
African groups were concerned in this more disassociated conception of Pocock's, 
the civets and genets becoming the Viverridae in a restricted sense, the mongooses 
the Herpcstidae (originally Mungotidae), and the West African palm-civet the Nan- 
diniidae, independent of the Asiatic palm-civets with which it had previously been 
associated. 

Simpson's (1945) classification, however, rejects this and retains the wider conception 
of the Viverridae; and this grouping is adopted herein. It has a long history of accep- 
tance by taxonomists, dating back, to all intents and purposes, a century and a half to 
Gray. During this period the majority of systematic mammalogists, no matter what their 
renaming or rearrangement of major cadres may have been, have seen little to disagree 
with in the close association with one another of all the small carnivores under dis- 
cussion. 

As regards the taxonomy of the Viverridae in general it must be added that Diicker 
(1957) from observations on instinct and behaviour reached the quite divergent con- 
clusion that the Herpestinac are, in fact, related rather to the Canoidea than to the 
Feloidea, and most particularly to the primitive Mustelidae. 

The three subfamilies may be separated by the following kcv, external distinction 
being easy but cranial differentiation more difficult. 

KEY TO THE SUBFAMILIES OF VIVERRIDAE 
(previous key page 160) 

I. Coat speckled, or unicolorous blackish, sometimes with a transverse pattern of 
bands, but never spotted; tail bushy but never ringed; posterior part of 
the bulla with one exception (Galerclla) much more inflated than the 
anterior part and subglobular or rather taller than it is long; auditors 
meatus with posterior and anterior lips which meet below in a horizontal 
V, thus forming a narrow gap which often extends as a slit into the floor 



l64 THE CARNIVORES OF WEST AFRICA 

ot the bulla; postorbital aiidjugal processes long and nearly or actually 
joining to form a complete circumorbital ring . Herpestinae {pa'^c 239) 

Coat spotted or blotched, tail at least partially ringed; the whole bulla inflated, 
oval and longer than it is tall; or [Paradoxurinac) partly cartilaginous 
and usually mostly missing in prepared skulls; the auditory meatus a 
shallow rmg without projecting bony lips; postorbital processes short 
(except Paradoxurinae) and far removed from a very short jugal process, 
leaving a widely incomplete circumorbital ring ..... 3 
3. Two pale spots on the shoulders (sometimes indistinct); coat colour deep red- 
dish brown; bulla cartilaginous or missing; postorbital process well 
developed ....... Paradoxurinae [pngc 229) 

No pale shoulder spots; coat reddish but not deep; bulla normal; postorbital 
process short ....... Viverrinae (/Jiijjt' 164) 

Subflimilv VIVERRINAE Gill, 1872 
Civet, Genets, Lmsangs 

Distribution. The first of the three subfannlies to be dealt with in this work con- 
tains S genera. These are divided by G. G. Simpson (194.S) into 2 tribes, the Viverrini 
and the Prionodontini, of which only the former is of West African interest, the 
latter, with 2 genera, being wholly Asiatic. The Viverrini also comprise 2 genera that 
are wholly Asiatic, besides 3 wholly African and one that occurs ni both continents 
and extends its range into extreme south-western Europe as well. These figures, 
however, are subject to dispute according to the breadth ot conception of some of the 
genera. Very considerable doubt and controversy e.xist also in regard to speciation in 
Genctta: but in this present account the 3 relevant genera of the subfanuly are held to 
cover a total of 12 species. 

There are a few apparently very rare and extremely local viverrines but the majority 
of the genera are common; and no part of West Africa, from coast to Subdcsert, 
is without some representative ot the subfamily. Extralimitally the group is widespread 
throughout the whole continent. 

General characters. The West Atrican animals included in the Viverrinae are 
very diverse in size, ranging from the bulky civet, scaling perhaps 17 kg, to the slender 
linsang, which can scarcely weigh more than 2 kg at most. All members of the sub- 
family dealt with herein have a prominent spotted or blotched pattern to their coats, 
never being merely speckled like the Herpestinae; and their mostly long or very long 
tails arc at least partly ringed, and with few exceptions more or less cylindrical in 
form. The face is sharp, the ears mostly fairly large, rounded and conspicuously up- 
standing above the level of the crown — in this, as in many other ways, the subfamily 
differing clearly from the Herpestinae. The ears differ further in always possessing a 
marginal bursa on the pinna. The limbs are short or shortish, ending in 5 digits armed 
with curved claws that are mostly to some extent retractile but in the civet scarcely 
at all. With the exception ot the pads, and sometimes the area immediately around 



VIVERRINAE 165 

them, the soles of the feet are hairy to the heel. The pelage is composed of fme underfiir, 
often very dense, and of bristle-hairs of roiuid or very slightly flat section, which in 
some genera arc very long and dominate the coat, in others of subordinate hnportancc. 

The scent glands in the Vivcrrinae are perineal, situated between the scrotum and 
the prepuce in the male, and between the anus and the vulva in the female. They are, 
besides being of different form, thus quite differently sited from those of each of the 
two other West African subfamilies. The glands open externally into a medial pocket 
or pockets; and in one case, Civittictis, the waxy secretion collecting therein has been 
commercially exploited for many centuries. 

Skulls. The skulls are narrow; even in the case of the very large and in some ways 
exceptional Civctlktis the braincase is little broader than that of markedly smaller 
herpestines. The rostrum is long and narrow, ver)' distinctly longer and narrower 
than that of any mongoose except Liberiictis. The orbital ring is far from complete, 
the postorbital processes being short, the jugal processes often little more than rudi- 
mentar)'. The postdental palate is short, nowhere near so long as it is broad. The large 
bullae are long and ovoid; the anterior chamber well-inflated, not depressed and 
reduced in size as in almost all Herpestinae. 

There may be 38 or 40 teeth, the cheekteeth numbering ^ or ^. Again with the 
exception of Cii'cttictis the dentition is by comparison with the mongooses relatively 
light, the anterior premolars being narrow, almost linear. It is clearly of a sectorial 
nature, not of an insectivorous type found in many members of the other subfamily. 

Habits. Although nearly all the Viverrinae are common animals their lives in the 
wild have not been much observed. This is because they are nocturnal, solitarv and 
secretive, spending much of their time in trees or dense undergrowth. All but the 
civet are arboreal, though they often come down to the ground to hunt. They are 
basically mainly carnivorous, the genets fairly strictly so, probably preferring small 
mammals and birds to other things; but the civet at least, and possibly genets too, 
consume reptiles, eggs, termites, beetle grubs and so forth if opportiuiitv offers. They 
also from time to time eat fallen fruits. 

So far as is known all viverrines are hole-dwellers, either on the ground or in trees. 
Breeding habits and periods var)' and the little that is known of them will be found 
under the different genera, below. Both the civet and the genets have been kept as 
pets; but while they tolerate domestication during early life they remain more aloof 
than the mongooses and soon seek their independence in the wild. 

Taxonomy. There is little doubt of the distinctness of the Viverrinae from the 
Paradoxurinac and Herpestinae, though the relationship with the former is much 
closer than with the latter. In fact, as explained later in the introduction to the Herpes- 
tinae, it is open to considerable doubt whether the two groups are really as close to 
each other as G. G. Simpson's classification puts them. General form, pelage, feet, 
claws, scent glands, skull characters and teeth are all quite distinctly different. 

Within the subfamily the genera are clear, and there has not been that confusion 
which arose in the Herpestinae. But there is argument regarding the closeness of 
affinity between African and Asiatic representatives of the subfamily which affects 



I66 Till (AHNlVdlUS 1)1 WIS! AIUHA 

two ot tlic throe genera covered by tliis present aeeouiit. (jrciiiciis and I'oiaua. Tiie 
tormcr is \\idelv held to be synonymous with or at most a subgenus ot the Indian 
I "ivcna; and the latter to be closely allied to tjie Asiatic Imsangs, Prioiiodon and Panhcils. 
The present audior follows Pocock in considering Cii'ctliais to merit generic standing 
of its own tor reasons given later in the account ot this genus; and he further believes 
superficial resemblances between Atrican and Asiatic linsaiigs to be misleading and 
probably more the outcome ot convergence than of particularly close relationship. 

KEY TO THE GENERA OF VIVEKRINAE 
(previous kcv page 163) 

1. Size large, head ^ body about 800 mm; tail only about halt as long; a white 

or pale patch, sometimes obscure, bordered with black on the side ot the 
neck, and a black mask across die face trom cheek to cheek. Adult 
skull about 140 mm; postorbital constriction not marked; dentition 
powerful ........ Civettictis {pdi^c id?) 

Size much smaller, head & body not more than 550 mm, otten much less; 
tail at least three-quarters as long as, or longer than, head & body; no 
conspicuous black and white pattern on the side ot the neck or mask 
across the face. Adult skull rarely as much as 100 mm, mostly mucli less; 
postorbital constriction well marked; dentition light .... 3 

2. Very slender; head & body about 380 mm; tur very short and velvety; mostly 

without any clear dark spinal stripe; spots always small; tail otten widi 
taint intermediate rings. Adult skull about 70 mm; cheekteeth 5-^;, the 
posterior lower molar being minute .... Poiana [pa^c 21'j) 
Of heavier build and mostly larger size, head & body 400 to 550 mm; pelage 
longer and coarser, otten with a distinct dark spinal stripe; the markings 
may be small spots but are more trequcntly larger blotches; tail without 
intermediate rings. Skull 75 to 90 mm; cheekteeth ^ Gcrietta (pd{;c 177) 

Cenus CIVETTICTIS Pocock, 191 5 
Atrican Civets 

I'lrom Liim.icus, 175X, Syni-iiui i\\iliii\h\ lotli cJ., I: 43; m pait, ot v.irious .uitliors, cspcci.illy bctorc 
1915. Type species ("both by elimination and selection", Tlioma'., lyii) I'/Vcnii ci7)c(//ii l.inn.ieus, 
Bengal. I'ivcria was the Latin name for a ferret. 

Civettictis Pocock, 1915, PiOi. :oo\. Sec. LoniL: H4- Ivpe specie-. \'iviiiii <irtltii Schreber, (luinea. The 
name is a compoiiiul of the specific name civelin, c/.r., w ith tlie (Ircek iV;/.i, a weasel. 

Distribution and general. Civets arc sometimes referred to more fully, but nus- 
leadingly, as civet cats because of the superficial resemblance ot their colour and 
markings to certain domestic cats. Actually, their much larger size and pointed muzzles 
might be thought to justifv rather more the description often bestowed on them by 



CIVETTICTIS 167 

comitry-brcd Africans as bush dogs. Neither, of course, has any shadow of accuracy, 
though the relationship, as members of the supcrfamily Feloidea, is somewhat closer 
to the cats than to the dogs. They occur only in the Old World between western Africa 
and eastern Asia. 

Two other, relatively minor, genera apart, the civets proper all belong to the two 
cited above ni the synonymy, the former of these being wholly Asiatic, the latter 
wholly African. Until 191 5 the two geographical groups were regarded as congeneric; 
and this is not infrequently still the case (e.g. Ellerman, Morrison-Scott & Hayman, 
1953), the African civets being looked upon as little more than specifically different 
from the Oriental — a question ot which something more is said below in the taxonomic 
section. Understood in this broadest sense, civets occur, in Asia, from the east coast 
of southern China to the Indian peninsula ; and, in Africa, south of the Sahara from 
Senegal to central Angola and the northern parts of South Africa. The range is, thus, 
discontinuous, with a gap of sometliing in the nature of 4000 kilometres separating 
the civets of Africa from those of Asia. 

In the area dealt with in this present accoimt, African civets arc known to occur 
from Senegal to Cameroun. They arc equally at home in forest or open country, 
at least as far inland as the Sudan woodland; but possibly do not exist in the Sahcl, 
and certainly not in the Subdcsert. T. S. Jones has noted them at an altitude of about 
1000 metres on Bintamane Mountain, Sierra Leone. African civets arc pretty abundant 
on the gromid, except at the extreme inland vcgetational limit of their range. As 
there is only one species, the matters of description and habits will be dealt with under 
the specific head; but a brief glance must be taken at the taxonomy of the genus. 

Taxonomy. Civcttictis is grouped by Simpson (1945) in the viverrine tribe Viverrini 
together with the three other African genera Gciictta, Poiana and Oskmiiais, and the 
two Asiatic genera Vivcrra and I 'ivcrricula. Distinctions between it and the first two 
are indisputable and are made clear in this present work; OsboriiictisJ. A. Allen, 1919, 
the so-called water civet or aquatic civet is recorded only as a very rare animal from 
eastern Congo and differs from all other known viverrines in its plain, unspotted, 
pelage. It is the separation of Civcttictis from the Asian Viverra that is most commonly 
open to question and, consequently, of chief interest to West Africa. 

Pocock, who as Superintendent for many years of the London Zoo had exceptional 
opportunities for the careful and comparative examination of deceased as well as 
living animals from different parts of the world, carried out detailed studies of many 
external features that had hitherto received scant attention or none at all. On these he 
based a number of taxonomic conclusions, including the generic disseverance of the 
African civets from those of the Orient. This last, though not wholly rejected by Hollis- 
ter (1918: 116) was considered by him as better regarded as subgeueric until such time 
as further confurmatory evidence was available; and, following this, Ellerman, Morrison- 
Scott & Hayman (1953) expressed the positive view that Pocock's characters could not, 
at most, be rated higher than of subgcncric value. 

The differential characters cited by Pocock were: firstly a marked distinction between 
the forms of the perineal scent glands; secondly, clear differences in the feet, the soles, 
apart from the actual pads, being entirely hairy in the Asiatic forms but naked anterior 



I6S rili: CMIN'IVOIIl s (II WIST AIIIKA 

to the central pads m tlie Atricau. In tlicsc latter, also, there is a metatarsal pad, absent 
from the former: and the claws are long and, more significantly, only slightly retrac- 
table as compared with short and more hilly retractile m the Oriental civets. He cited, 
as well, other more mnior points concerning the form ot the plantar and palmar pads 
and also of the rhinarium. Pocock was well aware ot the considerable tendency that 
exists to depreciate the taxonomic significance of external characters, and "since most 
contemporary mammalogists will probably consider cranial and dental characters ot 
more value in the discrimination ot genera than the external features here made use 
of. . ." he added supporting ditlerences to be observed in the skull and teeth. In regard 
to the former he claimed that the bullae and the partially enfolding paroccipital pro- 
cesses were more prominent in Civctticth than in Vivcna. This, however, is not always 
clearly so. On the other hand, his dental distinctions arc manifest in that //(i, itfi and 
in-2, are always of a very much greater size in the African species than in I 'it'cini. To 
this he might have added that the lower carnassial in Civcitictis is ot a type appreciably 
further evolved from the basic sectorial form, being far broader and somewhat more 
complex posteriorly to bear on the enlarged inner section of id' . The present author 
goes along with Pocock in regarding all these points as justifying generic distinction 
between the African and Oriental civets. G. Fetter (i96ij) reached, from dental con- 
siderations alone, the same conclusion. 

CIVETTICTIS CIVETTA (Schreber) African Civet 

Vii'crra civctta Schreber, 1776, Stiiigahicrf, pi. Iii; 1777, tc.\t 3: 419-420. (hiiiic.i; but Congo, Cape ol 
Good Hope and Aethiopia are also mentioned. The specific n.iine is a Latinization of the French 

linY/r, Itself derived troni tile Arabic Ciifcrti/ tor one ot tliese animals. 

Distribution and general. The African civet (tig. 35) is distributed over most of 
the continent south of the Sahara, in the west from Senegal to central Angola, and 
111 the east from Sudan to central Mozambique, that is to say roughly between about 
15" North and about i.S" South. This area covers a variety of vegetation from rain- 
forest to open grass-woodlands, certainly as tar inland as the Yobe Valley, near Yo, 
Lake Chad in the Sahel zone (A. J. Mopson, private communication). There seems 
little point in recording the names ot the large iimnber of places tliroughout West 
Africa from which civets are known; they are pretty common .ininials, more especially 
in the moister parts ot their range, though thcv have also been reported as plentiful at 
Macina on the upper Niger, Sudan zone. 

Description. These are easily the bulkiest representatives of the Viverridae in 
West Africa, weighing 10 to 17 kilogrammes and with a shoulder height of around 
380 mm, that is to say over twice the size of a large genet. Civets are, nevertheless, 
for their size short-legged animals, their bellies not far off the ground. When on the 
hunt they trot purposefully through the bush, their long-faced, somewhat canine 
heads pointed straight before them and their relatively rather short, shaggy tails 
stretched out, but slightly drooping, behind. The impression is that of a bulky, greyish, 
heavilv spotted, somewhat dog-like animal with a black tail. 



CIVETTICTIS 



169 




> 

■3 



lyO THE CARNIVORKS OF WEST AFRICA 

T]ic dorsal pelage is coarse and rather wiry; it varies in length, being in some speci- 
mens fairly long and in others unexpectedly short. It is composed of rather imdulatc 
and tangled undertur in which are mixed fairlv abundant, verv slightly flat bristle- 
hairs, strong distally but tapering basally. These bristles are markedly longer than the 
imdertur, which they dominate not only by reason of this but because, also, of their 
much greater diameter throughout most of their length. They average about 40 mm 
in length but sometimes attain 50 mm, the underfur measuring only some 20 to 30 
nmi; and the\' may be wholly black, or white with longer or shorter black tips. Along 
the mid-dorsal line ot the back from the shoulders or the hinder part of the neck to 
the root of the tail runs a long, black, erectile crest, the stout bristles of which arc 
everywhere appreciably longer than the main pelage but increase in length posteriorly 
where, in the region of the hindquarters, they may attain 100 mm or more. These 
spinal bristles are dark througliout most of their length but always have longer or 
shorter glossy jet-black terminal portions. When, in anger or alarm, this crest is erected 
the animal assumes an appearance of even greater size than it really has. 

The coloration of the African civet is widely variable. The dorsal ground-colour 
ranges between near-white, through creamy, to a slightly reddish-buff. On this is 
superimposed a pattern of dark spots or blotches, sometimes deep-brown but generally 
black. The size, shape and independence of these markings vary. Most typically the 
spots are large, of rather irregular roundish or quadrate shape, clear-cut, and nearly 
all separate from each other. They are not set in more or less regular longitudinal 
lines, as in some of the genets, but something in the nature of half-a-dozen rows can 
be made out on either side of the spinal crest. Occasionally the spots, especially those 
near the medial line, tend to coalesce into longitudinal bands; but in some specimens 
they display a tendency to rmi together laterally; and in some thc\- are relatively ill- 
defined. Over the torequarters the bold spotted pattern fades out and the pelage becomes 
an intimate mixture ot light and dark hairs, any maculation in this area being very 
obscure apart from a number of spots low on the shoulder. 

On the side of the neck is a pronounced longitudinal black band which is separated 
below by a conspicuous white band from a second black one which broadens and 
passes below across the throat. The underparts are somewhat variable, the chest being 
always black or blackish; but the belly may be white, or blackish, or a mixture of the 
two. The face is strikingly marked with a black mask which crosses from check to 
cheek and just encloses the eyes. It is bounded anteriorly by a white area around the 
rhinarium and on the lips, and posteriorly by a whitish frontal area between the eyes 
and the ears. The continuity ot the mark across the face is interrupted in some speci- 
mens, though rarely, by a medial whitish band joining the anterior white to the frontal 
grey area. The ears are rounded, low but none the less conspicuous with their w hite 
rims. A bursa is always present; its convex posterior flap is continuous above and below 
w^ith the rim of the pinna; its anterior flap is widely but not deeply emarginate. 

The upper portions of both fore and hind limbs are obscurely spotted; but the 
lower parts and feet are wholly black. There are five digits on each foot, the claws 
fairly long and only moderately ciurved, and very slightly retractile. In both feet tlie 
central plantar pad has the small pollical lobe separate from the main body ot the pad 



CIVETTICTIS 171 

(Pocock, 1915a). In the forefoot the metacarpal pad is bilobed and is joined, in most 
cases, by two narrow strips of naked skin to the anterior naked sole area that lies 
between the central and the subdigital pads. The hindfoot, also, has this anterior sole 
area naked. The tail is little more than half the length of head & body. Dorsally it is 
wholly black; but the basal half has about 5 partial white rings, passing from side to 
side below. The tail is coarse-haired with very long bristles. 

The perineal glands have been dissected, figured and described by Chatin (1874) 
and Pocock (1915a). Since these two accoiuits differ somewhat in their details the 
following is founded mainly on that of Pocock. He characterized the organs of the 
male and female as "tolerably similar". The glands are situated between the scrotum 
and the prepuce in the former, and between the anus and the vulva in the latter. Exter- 
nally they appear as paired swellings separated medially by a slit-like orifice about 
25 mm in length. When these two glandular lobes are drawn apart a moderately large 
oval orifice can be discerned on the inner face of each, and this leads into a large hair- 
lined sac or pouch which extends forwards, backwards and upwards within the gland. 
The inner walls of these two sacs secrete into them a very thick whitish or )-ellowish 
substance with a powerful musky odour, the "civet" of the perfiimery trade; and this 
then makes its way tlirough the two oval orifices into the intervening space between 
the two glandular lobes, wliich thus becomes a storage reservoir conununicating with 
the outside world by the slit-like orifice first mentioned. Besides these perineal glands 
tliere is a pair of anal glands situated on tlie wall of the rectum. These secrete a foetid 
yellow liquid which imparts to the faeces a characteristic odour which may be of 
service in the demarcation of territory but, in view of the regular defaecation habit 
mentioned later, is more likely to be of defensive significance, as in the case of other 
small carnivores. 

Complctch^ black specimens are recorded from time to time; and in these it is 
impossible to detect any trace of the basic pattern, at any incidence of light, as one is 
often able to do in other melanos. One British Museum skin, referred to by Sanderson 
(1940), is quite different in its coloration from all the rest. Li this, the ground-colour 
of the entire pelage, including the normally white areas of the neck and face, is a 
bright ochre-red. This may, of course, be natural, due to some form of erythrism; and 
if it is it is, indeed, remarkable. But the skin is obviously a locally prepared one, and 
Sanderson thought its luiique appearance probably due to its having been cured in 
smoke, a not uncommon practice. The colour, actually, is not so much that yellowish 
tone produced by smoke as, rather, one which could be caused by steeping in palm-oil. 
This would have been a not improbable proceeding in the area from which the speci- 
men came (Ogoja, extreme eastern Nigeria); and the soft pliability of the skin, as 
contrasted with the usual brittle stiffiiess of African sun-dried pelts, lends some support 
to this idea. The fur would, of course, have been subsequently washed to rid it of oil, 
but the residual red-dyeing effect of the palm-oil in the unpigmented hairs would 
resist this. Against all this is another fact: that though the long glossy black crest and 
tail are quite typical, the spots are smaller and more obscure than usual, as if there were 
something fundamentally exceptional about the whole skin. 

Skull (figs. 26 and 27). The skull is strongly built, long and rather narrow, with a 



172 



THE CARNIVORF.S OF WF.ST AIIUCA 




Fig. 26. Cii'cttictis civctta: skull, B.M. No. 59.^41, V, 



lateral ■ 



long ovoid braincase and not very marked intcrorbital and intertemporal constrictions. 
The supraorbital processes, though clear, are blunt, deflected, and not very extensive. 
All adult skulls, male and female, have a well-developed sagittal crest, joining a sharp, 
flange-like supraoccipital crest posteriorly in a T, The zygomatic arch is strong, the 
juga! process pronomiced but blunt and short. The bullae are moderatelv large and 
oval, the posterior section iar exxecding in size the anterior. The paroccipital processes 
are closelv applied to the whole posterior aspect ot the bullae and extend ventrally 
beyond their limit in long, strong points. The mandible has deep, strongly built rami; 
and, indeed, the whole skull and dentition constitute an instrument fashioned for 
powerful biting. 



The dental torniula is 



3 14: 



40. The incisors are strong and set in a transverse 



curve: 111 both jaws the outer ones arc larger than the inner; but this is most pro- 
noimccd in the upper jaw, where fl is separated trom the others by a small gap, is con- 
siderably larger and, being pointed and curved, begins to resemble a small canine rather 
than the normal chisel-edged incisor. The canines themselves, though strong, arc 
not particularly large, and their outer faces lack the fme, shallow furrows characteristic 
of the other West African viverrids. All the cheekteeth, from ;)i to \n~ and pi to »i-.>, 
are strongly built; m^ and ;ii- areparticular]\' large tor this subfamily; m\ is exceptionally 
broad posteriorly, and the whole tooth has more of a crushing than sectorial function. 
The posterior lower molar, also, is uncomnionlv well-developed, with a flat, far less 
acutely cuspidate, grinding surface than is usu.il. 



CIVETTICTIS 



173 



One British Museum skull. No. 10.6.19.1 from Okigwi, eastern Nigeria, has an 
extra, large, premolar on each side of the lower jaw, set between p4 and 0/1. 

Habits. Everyone is agreed that the African civet is an extremely secretive animal. 
This fact combined with its more or less strict nocturnal habits is responsible for the 
almost complete lack of field observations. It may seem strange that this should be 




Fig. 27. Civctlkus civctia: skull, B.M. No. 59.941, j:. 



pah 



&d 



orsal views 



SO in a species that is pretty common in museum collections and which tairly regularly 
appears in zoos; but it is a not very intelligent animal and one that is readily trapped; 
and so, though not often observed in nature, becomes easily killed or captured. Some- 
times a civet may be seen trotting up the road in the headlights ot a car: and occasionally 
in the very early morning one may dart swittK' across a path, possibly disturbed by a 



174 T"I- CARNIVORES OY- WEST AIIUIA 

Jog troni its tcinporar\ resting place in the thick undergrowth. But civets can also 
walk stealthily; and even it one should be about in the halt-light it is not easy to detect 
Its grev spotted coat aganist checkered vegetation. With so little known about their 
habits it is impossible to be certain, but civets are generally believed to lead entirely 
solitarv lives except at breeding time. They are certainlv almost completely terrestrial, 
their shape and teet being little adapted to climbing in the manner of cats, though 
where there are positive tootliolds thev can scramble up short distances, particularly 
betore thev have attained their adult bulk. Thev are said to be able swimmers, not 
atraid ot taking to the water. Normally, during daylight hours they lie curled up in a 
temporarv "torm" in any dense vegetation; but when breeding they use a tlxcd nest, 
usually a hole in the ground made bv some other animal, or amongst tangled roots. 
It seems probable that such nests are not lined in any wav. 

When it comes to the question ot tood, information is rather more certain because 
ot the data yielded by droppings, stomach contents, and behaviour in captivit}'. 
African civets seem to be more or less omnivorous. Antelope meat, cane rat, guiiica- 
towl, eggshell, carrion, termites, beetles, grass and leaves have been found in stomachs 
(Bothnia, 1965), carrion being predominant. But thev are also said to eat other kinds 
ot insects and their grubs; rats, snakes, lizards, frogs, crabs and millipedes. T. S. Jones, 
in a private comnuuiication, savs that civets not uncommonly hunt in mangrove 
swamps, presumably to obtain crabs and mud-skippers. Me has trapped them widi 
stock-fish or dead rats; and Verhcyeii (1951), indeed, gives them almost the character 
of complete scavengers. At the right seasons they consume an exceptional quantity 
of berries and other truits. T. S. Jones observed that thc\- caused a good deal of trouble 
amongst young low palms just starting to truit at Njala (Sierra Leone) by gnawing 
otf the oilv mesocarp. Astley Mabcrly ([9S5) records that they commonly cat the 
truits ot Briddia and Zizyi'liiis; and since civets mostly deposit their droppings in one 
fixed spot until they torm a heap, such places otten abomid in dense seedling growths 
of these trees. Verheyen (1951) found one such excremcntal pile to measure 60 cm in 
diameter and N to is cm thick. Civets, like other viverrids, arc known to raid poultry 
houses at night; and Shortridge (1934) recorded that they sometimes killed )-ouiig 
lambs. The widely catholic tastes in tood of African civets, and their piartialit}' to truit, 
has been well demonstrated in captivity; specimens, within the present writer's know- 
ledge, consuming almost anything put down for them, including bowls of pawpaw 
and banana. That some food should jiave excitant properties is of interest; Mallinson 
(1968) records that it has been found inadvisable to feed freshK' killed chicken or rabbit 
to viverrids with newly born yoiuig; female civets, and others, so fed have been 
found to become highly excited and subsequently to turn on and kill their litters. 
Some ot the older writers on natural history stated that the amoiuit of "civet" secreted 
bv animals maintained in captivit\' tor this purpose depended on the richness and 
abiuidance of tood given. 

The African civet can become verv tame and docile in captivitN' it caught \ouiig 
enough; but Menzies (1966) foimd those of 6 weeks old already savage and to remain 
quite intractable. However, without question, reaition to .ittempts at domestication 
will varv with different indniduals. One lu-t civet kept, or at least encouraged, about 



CIVETTICTIS 175 

the house used to follow its owners like a dog for evening walks; and at dinner time 
would circulate roiuid and imder the table giving unsuspecting guests playful nips in 
their fingers should they by chance hold their hands down. Eventually, as sexual 
maturity is reached, civets, like so majiy imconfuied pets, disappear into the bush 
never to return. An African civet has been known to live for almost 14 years in cap- 
tivity. 

Civets make a variety of sounds. They hiss when alarmed and utter a deep warning 
growl when they feel they are likely to be attacked. If they do get into a fight with one 
another there is a great deal of excited growling and yapping very like a dog-fight. 
Some years ago residents in the environs of Lagos often heard such quarrels at night, 
being sometimes awakened by the tremendous caterwauling and yapping. 

There is little positive information with regard to breeding; but Mallinson (1968) 
has recently published a few facts gathered from animals in captivity. From two 
observed copulations he deduced that the period of gestation might be either 45 or 
60 days. Nine births, covering 20 young, showed the litter size to lie between i and 4, 
the commonest being 2. The mothers frequently killed, and sometimes ate, their 
new-born young; and it \sas foimd less worrying for the female and less likely to lead 
to traged)' if the father were removed. Walker (1964) stated that there are generally 
two litters a year; but Mallinson, on the contrary, found with 5 females that the 
average interval between litters was approximately 12 months, and never less than 
288 days. Immature specimens in the British Museum suggest that there is no favoured 
season for breeding in West Africa since ver)' juvenile animals have been taken, in the 
forest belt, in mid-October, December and January, that is to say the beginning and 
middle of the dry-season ; and also in mid-June and early July, well into the rains. 
According to Vcrheyen (195 1) the newly-born young remain only about a week in 
the nest. 

For many centuries the civets have been chiefly famous as the source of a widelv- 
used ingredient of perfumes. This is an off-white or yellowish, very thick, greasy 
substance with a powerful musky odour secreted into the perineal pouch as already 
explained earlier in this account. Like many other basic materials, vegetable as well as 
animal, employed in the production of scents "civet" in concentrated form is apt to 
be nauseating or even highly repulsive; but when used in minute traces becomes not 
only tolerable but attractive. The trade, which was once extensive in Europe, North 
Africa, the near and middle East, has now considerabl)' diminished, at least in Europe ; 
but formerly African civets were kept imder cruel conditions in very small cages in 
which they could not turn round, so that the animal could be easily controlled and was 
constantly available for the product to be scraped, with little effort, from the glandular 
pocket. This took place about twice a week, the animal being forced back against the 
back of the cage and its tail and hindlegs seized and held through the bars while the 
"civet" was collected, using a small spoon. The yield was about 2 or 3 grammes a 
time. Abyssinia was recently the chief source of supply, the substance being exported 
packed in hollow antelope horns; but earlier, Amsterdam was the centre of a big 
trade, both for the import of the substance to Europe and for its production on the 
spot by captive animals. The Dutch trader William Bosnian, who was resident m 



I7f> Till: CAKNIVORIS or WF.ST A Ml 1 C A 

r.ible y: Numencil dat.i tor Cii'cliiiti< uiif/ii 

West 
A tVic.i 
general 

Vegetation Various 

Number in mean 12 

Condylobasal Icngtli 14s 

Basilar length 134 

Palatilar length 73-6 

Z\gomatic bieadth 77-1 

Upper cheekteeth breadth 47-4 

Interorbital breadth 29-2 

Postorbital constriction H'O 

Braincase breadth 43-0 

Toothrow (f— »!-) 57'7 

p^ length 1 3 '0 

ml breadth 13 '7 

mi length I3'5 

mi length S-2 

Head i\ body S26 

Tail ' 432 

Hindfoot 123 

Ear 57 

KATIOS (per cent) 

Tail/head & bod)- 5- 

Zygom. br./condylob. 1. 53 

Braincase/condylob. 1. 30 

Braincase/zygom. br. 56 

Palatilar l./condylob. 1. 51 

hiterorb./postorb. 122 



p^c- 



--'> 



"l/"'2 ~ '"3 165 

West Africa towards tlic end ot tlic 17th century and wrote a nuniber ot letters des- 
cribing the various countries, tlicir customs and animals, li.nd, in translation, this to 
sav of the civet "at present so well known in Holland that 1 need oiiK' acquaint you 
tliat tliev are brought to be sold to us verv yoiuig. and then we give about eight or 
nine Shillings sterling for one. A large share of Trouble and caretul Attendance is 
requisite to breed tliein up: Their Food is Pap boiled or made ot Millet, with a little 
Flesh or Fish. They produce Civet when even very young; ot which that ot the Males 
IS better than that of the Females, because the latter cannot avoid urining into the 
Civet Bag, which spoils it". 

Besides its obvious wortli to the scent trade, "civet" was also tornierix- imicli valued 
for a wide varictv of reputed, but now long discredited, medicinal properties. It was 
emploved as a diaphoretic; or as an emollient in such eruptive skin conditions as small- 
pox, measles and scabies; as a balm against apoplew; as either a soporitic or an aphro- 
ciisiac; or even as an ingredient ot tooth powder. 



GENETTA 177 

Taxonomy. Other races besides the noniiiiate one have been nained, G. M. Allen 
(1939) recognising two, cotij^ka Cabrera, from north-eastern Congo, and schivarzi 
Cabrera, troni eastern Africa. If these arc accepted, then the correct citation of the 
West African civet must be Civcttictis civetta civctta (Schrcber). But the species shows 
such a degree of nidividual variation that there is a tendency today amongst authors 
to be sceptical ot the existence of identifiable races. 

Genus GENETTA G. Cuvier, 1816 
Genets 

Gciktla Okeii, 18 16, LcUrhnch Ar Nalnrgcsclikliu; 3, 2: loio. Type species Vivcrra genelta Linnaeus. Oken's 
Lehrhnch is ruled to be unavailable by Opinion No. 417 of 1956 ot the International Commission on 
Zoological Nomenclature. This name is a Latinization of the Old French (jo/rt/c, itself derived from the 
Arabic name for these animals, janiail. 

Gaiclia G. Cuvier, iSi6'/jVc Sherboni), Le Rcgne Animale. I: I>i6. Type species \lverra (fcuciiii Linn- 
aeus. 

Distribution and general. Genet is a simple enough name; nevertheless it is one 
that is not often used by English-speaking people in West Africa, who usually refer 
to these animals as bush cats. Not that the average person often, or ever, sees one of 
these nocturnal creatures in the flesh; but he may sometimes become aware of them 
through a raid on the fowlhouse, or occasionally through being brought a motherless 
baby ior sale as a pet. Genets are, none the less, widespread and fairly common, occur- 
ring through practically the whole of Africa except the Sahara, and, in the region dealt 
with in this book, ranging from the Subdescrt to the coast. They are known to occur 
at high altitudes in eastern Africa but were not taken at am- great height on the Camer- 
oun Mountain by Eisentraut (1963). 

Genets, though all conforming to a general casily-recognisable pattern, occur in a 
wide number of different forms. How many of these, and indeed how many basic 
species, are valid is a much-disputed question. In this present work nine different species 
arc described but, as will be gathered from the taxonomic section which follows later, 
some of these may not be valid or, as the outcome of a greatly-needed pan-African 
review of the genus, may eventually be shown to be merely western representatives 
of other, extralimital, species. 

General description. The genets (Plates 3 and 4) are all smallish mammals with 
pointed muzzles and large ovoid ears in which a large bursa seems to be always present 
ill West African species. The posterior flap of this arises, both above and below, behind 
the pinna, the shape of the anterior flap being variable in the different species. The 
slender, spotted body is some 450 to 550 mm long, the ringed tail somewhat shorter. 
Genets stand about 120 to 150 mm at the shoulder. An average weight of an adult is 
around 2 kg. The pelage, both of the body and the tail, consists of a mixture of terete 
or slightly flat-sectioned bristle-hairs and fine woolly luiderfur, the absolute and 
relative lengths of these varying from form to form. There is often, but not always, 
a media! dorsal stripe from about the shoulders to the root of the tail, dark, mostly 



lyS rill ' AUNivoius in west aiuk a 

lil.ick, toiisistiiig ut soincwli.it. or iinicli. lojigcr briitlcs tliaji tlic rest ot tlic dorsal 
pelage, and apparciul\' erectile. On cither side of this spinal stripe run more or less 
parallel rows ot spots, the number i^t such rows varying from 4 to about 7, though the 
hues arc apt to become contused on the Hanks and difficult to coiuit precisely. The 
spots themselves vary in number from comparatively few to many; their spacing from 
relativTly wide to close. They range considerably in size from tairly large to pretty 
small; in shape from quadrate to oval or roiuidcd; and from full independence of one 
another to longitudinal coalescence into longer or shorter bands, more or less parallel 
to the mid-dorsal crest. Such concurrence usually concerns only the line or two lines 
nearest the spine, and commonK' only the region ot the lower back. The spiots may be 
wholly ot one colour, black or some shade ot reddish-brown ; or they may be annulate, 
having black margins with larger or smaller brown centres. 

The back of the neck is basically marked with 5 more or less parallel longitudinal 
lines; a fme central one which posteriorly becomes the mid-dorsal stripe; on either 
side ot this a slightly broader, but still narrow, line which joins the innermost series 
ot spots; and, on the outside ot these, two broader, and otten very pronounced lines 
which posteriorK' curl round and down over the shoulders. But some or all of these 
nuchal lines may, ui certain specimens, become wholly or partly obscure. The belly 
is usuallv paler than the dorsal ground-colour and may be almost white, well-spotted 
or with very tew spots. The thighs are spotted, the uppier piart ot the toreleg often so; 
the lower parts ot both legs mav carrv small spots or be practically plain. Thev ma\' be 
light or almost black, this, at least as tar as the torelegs are concerned, being otten 
idiosxncratic and hence taxonomically unreliable, although in the past it has often 
been assumed to be diagnostic. There are tive toes on each toot, digit J being separated 
from the others; each is armed with arcuate, tine, verv sharp claws, which are retractile 
but not so completelv as in the cats. Beneath each digit is a naked hemispherical or 
ovoid pad; a larger, composite one surrounding a central depression is situated 111 the 
middle of the toot; and another posterior to this, very long and biturcate on the 
hindfoot. l1ic taec carries certain general markings, sometimes very clear, sometimes 
obscure though almost aKva\s detectable. The most prominent is a pale, otten white, 
spot below the eves; this is separated by a dark patch trom a similar pale or whitish 
region on the upper lips aiouiid the rhinarium. There is a pale area between the eyes, 
parted mediallv b\' a dark line; and another pale spot above each eve. 

The tails, though subject to some individual variation, are more constant 111 their 
basic char.Kters than the doi^al markings are, and sufticientU' distinctive to torm the 
easiest and most dependable diagnostic character. The nature ot the tail derives firstly 
trom the actual length ot the hair with which it is clothed and secondK from the 
relative length of the bristle hairs and underfur. In {^ciicttii, tor example, the tormer are 
not oiiK- ver\' long in themselves but also tar exceed in length the undertur, which 
plavs a verv seccindar\- role, the whole structure being ot relatively rough appearance. 
In scrvaliiid. on the other hand, the bn-.tlcs .ire short, and the abundant undertur only a 
little shorter, plaving a major role in producing a sott, smooth, subcylindrical structure, 
hi addition to this, the number and nature ot the annulations are usetullv diagnostic, 
though bv no means so unswervingU' constant as was tormerK- thought. 



GF.NKTTA 179 

C'lCiicts arc tuniislK'(.i w itli mciu glands secreting a nmsk\' ddoiir. Tlrcsc arc situated 
in tiic male between the scrotum and prepuce, or in tlic female between the anus and 
vulva. Pocock (1915a) furnished a full description of these, both external and internal. 
Briefly, in his own words, they "consist of two elongated eminences covered with hair 
both extcnially and intcrnall)-. When luidisturbed the two lobes are closely apposed, 
their line of contact being marked by a longitudinal sulcus whicli is Y-shaped anteriorly, 
that is to sa\-. just bclnnd the vulva or prepuce". Internally, the glands arc mostly 
tripartite, being imperfectly divided into three compartments by transverse ridges of 
integument. This description applies to the males oi tigriiia, partliiia, nibiginosa, gciietta, 
(kiiigolana (^ senegalciisis) and fcliiia, and to the females of the first three; but the 
females of the second three have a simpler, undivided pouch. Pocock regarded this as 
indicating two groupings of relationship within the genus. 

Skull. Unfortunately Gciictta skulls and teeth exhibit very little constant diflercnces 
of shape and so are not a great deal of help in distinguishing forms. Some of the charac- 
ters picked upon by Schwarz (1930) are misleading, being not certainly applicable to 
all members of a species-group. 

The Gciurra skull in general (tig. 29) is long in comparison with its zygomatic 
breadth, with a long, ovoid braincase, and pronoiuiccd postorbital processes separating 
marked intertemporal and interorbital constrictions. Yet though the whole skull is 
long and tapering the actual rostrum itself is remarkably short. Posteriorly, in the 
fashion common to the Feloidea, the braincase contracts and curves broadly out again 
to meet a very well-developed, flange-like supraoccipital crest. There may or may not 
be a sagittal crest over the main body of the cranium; but there is always a wcll- 
devclopcd, sharp posterior portion joining the supraoccipital crest in a T. The forward 
continuation or absence of this over the braincase has been regarded as a valid specific 
character (Schwarz, 1930); the evidence from 29 mature West African skulls in the 
British Museum indicates that this mav be broadly so but that its development is a 
function of age, and its absence save from really old specimens thus possibly mis- 
leading. There is, for West Africa, no evidence relating to sex, comparison being im- 
possible since there are no old female skulls o( pardiim, in which species the crest is 
normally present in old males. A superficial look at some dozens of Gciietta skulls 
from other regions revealed only three reputed females with developed crest. The 
bullae are fairly large, long ovoidal in shape. There is some evidence pointing to diff- 
erent relative developments of the anterior and posterior portions in different species, 
the anterior chamber aroiuid the meatus being largest in gciicna and iliicrryi. less in 
pardiiui and least in cristata (fig. 28). 

The ratio of the postorbital constriction to the interorbital breadth is fairly reliably 
diagnostic but not absolutely constant. In {ifiwtliudcs 19 skulls had the former appreciably 
less than the latter; but in 2 skulls the breadths were subequal. The atim/i'im group also 
has the postorbital width less, one externally quite typical cristata specimen, No. 10. 6. 1.2 
(Oban, Nigeria) alone having it greater. In a few exceptional cases the postorbital 
constriction is extremely marked. In one gciicltoides specimen. No. 46.397 (Oda, 
GhaiKi), it is as little as .S-4 mm in contrast toa mean of 11-5 mm; the difference between 
it and the interorbital breadth being about 6 nnn as compared with a mean of the 



iSo 



Till. ( AHNIVOKHS OI" WFST AIUK.A 




Fig. :;S. GcMifdi; left bullnL-, 3: .1. ('• 'yiuiM (? ii/rii), B.M. No. i;.ii.2.34, 
b. (,'. i,'nii7;i'ir/i.f, li.M. No. 4'i.39'', ;-, c. G. iTisiiUii. B.M. No. 39.323, ; 



order ot z niui. For such atypical incaMircmciit^ there is no apparent explanation. 
The postorbital constriction is wider than the interorbital breadth in the gciu'tta group 
and tliienyi. Citing the postorbital processes as diagnostic Schwarz stated that they were 
absent or sliglit in scrvaliihi; this assertion would appear to be whollv untrue, except of 
juveniles, in both scii'ijUiuj and cristam. 

The dental formula is j-pj-i = 40. The mcisors form tightly packed transverse 
rows, the outer teeth, both above and below, appreciably larger than the others, the 
cutting edges ot the lower set benig slightly bitid or trihd. The canines are strong and 
arc chiefly notable tor bearing verv narrow, shallow, longitudinal striations on their 
outer faces. The lower canines are abruptly curved in the middle ot their posterior 
taccs and thus have the appearance of being bent upwards. The premolars are ot the 
usual form except that jfi in some cases has an additional cusp at the base of its inner 
tace. When present it is, most commonly, well developed and quite obvious; but it is 
sometimes reduced m size. Tliis tcature is in some measure diagnostic but is not com- 
pletely reliable: in 16 :iduh gciicticiidcs skulls this supplementary cusp was present and 
large in 9. small in 3, and absent from 4. Other species show something of the same 
irregularity ot occurrence except tliienyi where the cusp occurred in all 6 available 
skulls. There is nothing especially remarkable about the remainder of the teeth except 
that 111- IS always the smallest m the two jaws and may sometimes be very much 
reduced. Thomas & Wroughton (1910), writing ot stuhhnatmi, detected a marked 
sexual ditference m the size of the upper carnassials, the temales being i mm or more 
less in length. The comparative material trom West Atrica available in the British 



GENETTA l8l 

Museum is too sparse to furnish reliable data relating to this; but such as it is it tends 
to confirm these authors' fmdings. 

Habits. With the confusion that has existed over the different species of Geuetta 
it is not practicable to relate the few field observations that exist precisely to named 
forms, and the general account which follows, a compilation from various sources, 
must take the place of the individual notes normally appropriate to specific sections. 
It can fairly safely be assumed that all genets are broadly similar in their habits except 
for such differences of detail as may arise from dissimilarity of habitat, high-forest 
species being possibly rather more arboreal in their way of life than those from more 
arid inland vegetation. It is, in some way, curious that so widespread and not uncommon 
creatures should have had so little recorded of them in nature. This is partly due to 
their secretive nocturnal lives; but they have long been kept in zoos and, within the 
present writer's experience, as domestic pets; yet though there are fairly plcntifid 
records of their littering nothing beyond the most meagre observations on other 
aspects of their behaviour and way of life has, apparently, ever appeared in print. 

Although exceptions, of course, always occur genets are pretty strictly nocrumal 
and are therefore rarely seen except in the headlights of a car by or those who range 
the bush at night with lamps. That they are from time to time seen and shot by local 
hunters is obvious; and they may occasionally be accidentally flushed from some 
da)time resting place in a tree or dense undergrowth; but the general run of skins and 
skulls in the British Museum gives the impression that the animals must for the most 
part have been taken in traps rather than killed by gun-fure. Experience with animals 
maintained about the house shows that they do, under these circumstances, put in an 
appearance about dusk; but even should they happen to be on the move in the forest 
while there is still daylight their excellently disruptive coloration combined with 
their stealthy caution would nearly always render theni exceedingly difficult to detect. 

Since authentic first-hand records of these animals in the \\dld are almost entirely 
lacking it is virtually impossible to make defmite assertions. However, they appear to 
be rather solitary creatiu'es except possibly at mating periods, when they doubtless 
forage in pairs. Even, at a later stage, the mother and her young do not openK' associate 
for long, as they do in many other animals; for young genets, once past the earlv 
helpless state that confmes them to the nest, quickly become capable of leading an 
independent existence. 

Genets are to a fairly considerable extent arboreal, their sharp, curved claws being 
almost as well adapted as those of cats to scrambling up tree trunks. Some of their 
foraging is regularly carried out in trees, for birds, possibly their eggs and certainly 
their fledglings, for arboreal rodents, bats or the helpless young of larger tree-living 
species. They are commonly found in oil palms. However, a good deal, possibly even 
the major part, of their hunting is on the ground for groimd rodents and reptiles. 
Like any other animals endowed with the power of swift climbing, genets take to 
trees as a ready and dependable means of escape from threat on the groimd. Yet it 
would seem that, the first instinctive urge apart, they appear on reflection to feel safer, 
and with wider chances of escape, on the ground rather than limited to the trimk 
and branches of a single tree; for after their first hurried rush up the bole they frequently 



TS2 the carnivores of west AFRICA 

make tor the ground again. Sanderson (1940) records that he onl\- once saw a genet 
{crisutta) in a low tree; alarmed at his presence it returned to the ground in an attempt 
to make its escape. He formed the opinion that the species was quite definitely terres- 
trial. Stevenson-Hamilton (1947), too, said of a South African species that when pursued 
it almost invariablv took to a tree but on being approached generally leapt to the ground 
again to make oH with bounds ot remarkable length. He cited an instance of a genet 
leaping from a height of 5 metres and landing 7 metres away. 

In consonance with their at least partial arboreal existence they have been recorded 
as sheltering or breeding in hollow trees or holes in trees. Loveridgc (in Lawrence & 
Lovcridgc, 1953) tells ot shooting a genet in Mozambique "as it lay curled up beside a 
hole high up on the trunk of a great baobab". Nevertheless, the majority of breeding 
observations — and they arc not many — refer, rather, to holes in the groiuid. Possibly 
this may be due to the greater likelihood of accidentally coming across a nesting 
place in or on the ground than at some distance up a tree. Yet it is strange that those 
collectors who have spent a considerable time in West Africa "smoking out" or felling 
hollow tree stems in the search for bats or flying-squirrels seem never to have recorded 
the incidental discovery of genets. Possibly the sort of restricted holes that might appeal 
to genets as favourable breeding sites, such as small natural cavities wheie branches 
have fallen away, or woodpecker borings, or honibill nests, are overlooked. The 
tccKnique of smoking nevertheless commonly succeeds in South Africa (Stevenson- 
Hamilton, 1947). 

However, holes are not always essential and, in the dry-season at least, nests may 
sometimes be exterior. One such was come across in February by an agricultural 
labourer near Freetown and described to T. S. Jones as "a bundle of sticks". From this a 
female genet had made her escape carrying one of her young in her mouth; the second 
baby was then captured in the nest and subsequently successfully reared. That such a 
breeding shelter was constructed by the genet itself seems improbable, and it is most 
likely to have been a fortuitous conglomeration ot rubbish or, very possibly, a deserted 
squirrel drey. Jones also came across another unusual oft-thc-ground refuge, namely 
the roof of a house in Samaru (north Nigeria) from which he observed a family of 
4 or s genets make its exit down a verandah post v^'hen it was nearly dark. This he 
was told happened ever)- evening and that an adult pair dwelt permanently in the roof 
(private comminncation). This home would appear to have been naturally discovered 
and occupied by genuinely wild genets; but the present writer knew a more domes- 
ticated single animal that, having been captured yomig and hand fed, also discovered 
the roof as a warm shelter and spent its days concealed in a corner of the grass thatch, 
descending regularly at dusk to the sitting-room to be fed. 

flenets, then, at least sometimes, occupy holes in trees; and according to Stevenson- 
Hamilton thev may rest stretched out along a branch. But the majority of records 
refer to sleeping or breeding quarters on the ground; and in observations made on 
captive genets in which they were afforded a choice of sleeping quarters Carpenter 
(1970) certainly found them to prefer a hollowed out ant-heap on the ground to a 
covered box fixed high up. In suitable terrain terrestrial shelters may be amongst 
rocks; but thev are more usuallv in the smaller kinds of "earths", either naturallv 



GENETTA I83 

occurring holes or those made or improved by other animals — ground squirrels, 
giant rats, porcupines, termites and so forth. They are almost certainly never self- 
constructed, the genet legs and paws being ill-adapted to excavation. Besides earthen 
holes, those amongst tree roots, or actually in rotted bases of trunks, or fallen, hollow 
logs arc used; and temporary sleeping sites, though probably not breeding quarters, 
are made ni thick grass, reeds or other dense low vegetation. The only actual first-hand 
description of a nest appears to be that of Loveridge (1922) who came across one in a 
hole in the ground consisting of a circular chamber with two bolt holes, unlined but 
perfectly clean. 

If events in captivity can be used as a guide, breeding takes place pretty regularly 
twice a year. This, according to Verheyen (195 1) is probably at the beginning and 
end of the rains — a supposition that is to a certain extent confirmed by juvenile West 
African specimens oi genettoidcs in the British Museum collected in March, April, 
June, July, September and early December. The usual litter size is 2 or 3 ; but i is not 
imcommon, and 4 have been noted on several occasions, having occurred in no less 
than three successive births in the Duisburg Zoo in Germany (Remy & Condc, 1962). 
These latter authors have gathered together a considerable number of records, cliiefly 
relating to the European genet; but the longest-continued observations of living 
animals arc those made by Volf (1959 and 1964) in the Prague Zoo. He records that a 
single female genet (geiictta) produced 32 young in 7J years, of which 20 were born 
in the last 4 years of her life, the litters being almost without exception two a year. 
This animal attained sexual maturity at the age of about 4 years and died, of old age, 
at 13 years Such an age for genets in captivity is not uncommon, the record being 
about two months short of 15 years (Crandall, 1964). 

The period of gestation has been put at 10 to 11 weeks (Volf, 1959); but Remy & 
Conde (1962) mention an American figure (Smithsonian) of only a little over 8 weeks. 
A well-authenticated period observed in Paris was 68 days, that is, just on 10 weeks. 
The young are born blind and with the auditory meatus covered; Volf (1959) observed 
the ears to open on the 5th day and the eyes on the 8th. According to the same author 
the newly-boni kittens are well-haired though this covering is of a deeper, wholly 
greyish, hue than that of adults, to whose rather paler and warmer colourmg they do 
not approximate until after the second month. In the British Museum there is a very 
dark, almost melanistic juvenile oi genettoides. Writing of the European genet Volf 
said that at birth there are only 3 rings near the base of the tail, the whole dorsal aspect 
of the tail being dark, the underside paler. With advancing growth the pale colour 
encroaches on the dark and at the same time forms more rings, which reach 6 in 
number at three weeks. Weight at birth in this species is from 61 to 82 grams. This 
increases vciy rapidly, especially in the first few days, doubling itself in 12 days and 
quadrupling in a month. The growth curve is more or less a straight line; and at two 
years the fully adult weight is of the order of 2 kilograms. This may increase slightly 
with greater maturit)'. It has been found in captivity that immediately after birth the 
male must be removed for two or three weeks as he may otlierwise kill the young. 
But according to Verheyen (195 1) in the wild the male actually provides food for the 



lS4 THE CARNMVORES Of- WEST AFRICA 

resting and nursing motlacr. Killings of their ncvvlv-borji kittens arc, for unknown 
reasons, not infrequently carried out bv the females. 

Volf (1959) provides data tor the eiuption ot teeth, given below. It is, however, 
not clear whether reference is to milk or permanent dentition; it may be presumed to 
be the former, yet the premolars which emerge subsequently to the molars on the 
44th and .SI St days would seem to be replacements of an earlier milk set. Volf observed 
tlie upper incisors to appear on the 23rd day, at which date, also, crawling commenced 
and the ability to raise the head. All the incisors and both upper and lower canines 
enipted by the 30th day; the molars appeared on the 37th day; the upper premolars 
on the 44th, and the lower premolars on the STst. The young animals attempted at 
5 weeks to detend themselves bv spitting; and after N weeks bv biting. At this age 
they were fully active and weaning took place, and thev were capable ot taking care 
of themselves. A certain amoimt ot solid tood is in tact accepted trom about 6 weeks. 
Loveridge (1922) foimd that a babv he attempted to bring up would not look at meat 
but only drank milk sweetened with sugar to the consistency of treacle; and it became 
very fond of jam and would eat egg, after shelling. Mrs. Soniajetfrey (private com- 
munication) toiuid that a baby genet could be successtully reared on tinned imsweetened 
milk with an occasional mixed vitamin tablet. This animal showed an interest in red 
meat when it weighed 255 grammes and ate its first mouse three weeks later when it 
weighed 310 grammes. 

Although the shift to solid food normally takes place at about 2 months. }-oiuig 
genets, if given the opportunity may continue to suck up to the age ot 6 months. In 
captivity, Volf records, the yoiuig of one litter were usuallv removed trom the mother 
before her next pregnancy. Where this, on one occasion, was not done the mother, 
through over-long continued suckling, reached a state of exiiaustion in which the 
foetuses failed to develop properly and the yoiuig were either still-born or died soon 
after birth. It is interesting to note that although Volf estimated that the iiciicna tcmalc 
which mothered 32 young must have been about 4 years old before she became sexually 
mature, he found the oifspring to attain this state after about only 2 years. 

It has sometimes been held that genets, unlike many other flesh-eaters, are very 
strictly carnivorous in their diet, firmly rejecting all kinds of vegetable toods. This is 
not so. Dr. Gordon Corbet shot a genet in Kenya that had its whole intestine crammed 
with a mass of orange-coloured berries, besides the remains of a .spider and insects 
(private commimication). From the description it would seem that the truits were almost 
eertanily those of the so-called Jujube Tree, Ziziplms iiuwriiiaihi Lam., widely spread 
ill the more arid woodlands of tropical Africa. Another very interesting observation 
concernmg teeding on vegetable material has recently been made by Sonia Jeffrey in 
Ghana (private communication). She found that something like 70 per cent of the 
contents of one stomach consisted of rotten wood, some of the pieces being up to 18 mm 
long. She supposed that this indicated that the animal had really been alter some more 
tasty delicacy buried in the wood. Such an explanation is very possible since beetle 
grubs, sometimes ot large size, frequently abound in dead trees, as woodpeckers well 
know. The fact that genets are often connected with palm trees leads to the speculation 



GENETTA 185 

that they may, at least sometimes, be attracted not only by the fruits but also by the 
very large grubs of the giant palm-weevil foimd in decaying stems. 

As concerns the more usual, carnivorous, foods Shortridge (1934) wrote that genets 
prefer fresh meat; but Lovcridge found tinned corned beef to be quite acceptable; 
and Stevenson-Hamilton (1947) says they arc partial to carrion and readily enter 
baited traps. Besides rats, gcrbils, squirrels and hares they take insects, especially locusts 
and beetles, and will tackle birds of the size of guinea-fowl and domestic breeds. They 
also eat lizards; while Loveridgc (in Lawrence & Loveridge, 1953) found the remains 
of a burrowing snake and a large yellow scorpion in the stomach of one. Whether 
they eat amphibians and molluscs seems to be imrecorded. They reputedly eat shrews; 
but Sonia Jeffrey kept a young genet that would eat most kinds of meat but would 
not look at a shrew. Although one of her captive specimens would not touch a freshly 
killed snake another one readily ate them. Stomach contents of genets examined by 
her included the legs of skinks and insects, and orthoptera wings ; and one contained 
two mice \\hich may, from the red colour of the fur, have been Lophiiwmys sikapiisi. 
One of her live specimens pounced on and ate a bird that flew into the house. 

In their attacks on fowl-runs they show, according to Stevenson-Hamilton, con- 
siderable cunning, generally taking a single bird at a time, returning for another when 
this has been consumed. A bird of this size is attacked at the throat, the blood sucked, 
and then the upper part of the breast taken. Carpenter (1970) found some evidence 
that males were more often poultry thieves than females; a:id that once a genet had 
acquired a taste for domestic fowls it tended to disregard more natural prey. In the 
absence of such an acquired taste rodents were nearly always preferred as food to birds, 
being pounced upon and killed forthwith, while birds were only killed and taken later if 
the rodent supply were insufticient. On two widely separate occasions Carpenter foimd 
genets to capture bats in roofs, the genera concerned being Eptesiais, Scotophilus and 
Rhiuolophus. Over a period of five nights one genet took an average of six bats a night. 

Genets are not normally offensive; but when alarmed spit and bare their teeth, 
and if hard-pressed can give a severe bite. Like other animals they vary a good deal in 
temperament and it is difficult to generalize. They can be tamed, especially when 
captured very yoiuig; and some become fairly trusting and even milldy friendly, 
though never to the extent of mongooses and other small mammals. They seem to be 
ver)' much one-man pets and to resent strangers. The majorit)', at least imdcr con- 
ditions of semi-freedom in Africa, as a rule soon manifest independence; and though 
they may appear in the evening to take what food is offered they mostly remain rather 
coldly self-sutiicient, exhibiting the aloofness of a cat with none of its friendly domes- 
ticity. Genets are also doubtful assets; for while it is true that they may keep the house 
free of rats and possibly snakes they have no loyalty when it comes to a question of 
robbing their owners of meat from the pantry or effecting a midnight entr)' to the 
fowl-house. 

Genets emit a musky odour, though not so strong as that of some other viverrines. 
Gray (1830), in his description oi iiiaaihua as a living animal in the Tower of London 
menagerie, said that it had an exceedingly strong musky smell and was continualK- 
placing its two hindlcgs against the wall of its cage, and pressing the subanal glands 



l86 THE CARNIVORES OF WEST AFRICA 

against it, leaving two chocolate brown musky streaks. These were far larger than 
those of a similarly captive common civet, which performed the same action; and it 
seemed, theretore, that more secretion was emitted by the genet than by the bigger 
animal. Both were obviously attempting to demarcate their territories; and it may be 
assumed that something of this sort is regularly carried out under natural conditions, 
though, in the wild, urination and defaecation arc probably the most usual method 
iUid adequate to the purpose. Genets have not usually been reckoned amongst the 
animals that use fixed latrine sites but Carpenter (1970), working in Natal, found these 
to be not uncommon, generally in dry river beds or thick bush, some of them obviously 
in use for a considerable time. As regards territorial behaviour and the instinct and 
ability to return to the home range the same author has made some very interesting 
observations and deductions in relation to the rusty-spotted genet in South Africa. 
From these it would appear that females are more attached to a given territory than 
males, which are more inclined to wander. By repeated trapping and marking, these 
genets have been shown to return to their original home ground trom distances up to 
about 30 kilometres. 

Taxonomy. Forty years ago Schwarz (1930) in his introductory remarks to what 
has proved to be the last published general review ot this genus expressed the opinion 
that its systematics were obscure. This is, to say the least, somewhat ot an under- 
statement. As it stands today Gciwtta is a confusion of forms, with the definitions of the 
classic species uncertain and the status of later creations — at times recognised as valid 
species, at others bandied about as mere races betwixt one reputed species and another — 
a tangled skein with no clear guiding threads giving hope of any gcnerallv acceptable 
unravelment Over the years no less than 44 discrete species have been described in 
Africa, apart from several forms rated from the start as racial. Most of these have now 
bceji either reduced in status or synonymised, and G. M. Allen in his Checklist of 
African Mammals (1939) recognised only 6 species and 27 subspecies. This is far from 
being the last word. 

There is something inherently wrong in a situation wherein authors so frequently 
disagree over the status or validity of forms and their opinions regarding their essential 
diagnostic characters; or have changed their own minds in these matters; and in 
which nmseum labels so manifestly bear witness to uncertainty in determination. 
Two plain facts emerge: the original descriptions have, in the light of greater know- 
ledge and more abundant material, often been pitifully inadequate; and the taxonomic 
significance of the characters chiefly relied upon for diagnosis has been greatly over- 
valued since most suffer the disability of considerable idiosyncratic variation. As long 
ago as 1910 Thomas & Wroughton, writing o( stithlmanni and siiahelica, two of Mat- 
schie's creations, said: ". . . we have a series of over a dozen, the extreme individuals of 
which are easily separable; but these extremes are linked up by the intervening in- 
dividuals in such a way that after most careful examination of both skins and skulls 
we have been obliged to acknowledge that we cannot fmd any constant character by 
which these forms may be separated . . .". 

It is possible to compile a long list of characters that have been used taxonomically, 
but thev can be briefly categorised as the nature of the pelage, its colour in various 



GBNBTTA I87 

situations on the body, and the colour, size, number and shape of the markings. These, 
and others, can build up into a very large number of combinations; and if, as has been 
the tendency, each of these is regarded as constituting a valid form there is considerable 
scope for taxonomic chaos. The difficulty is to decide which characters can, in fact, 
be relied upon, and to what degree. This can only be determined from long series of 
restricted provenance. One such, and the best for which published information exists, 
is the series of 46 specimens, 24 adults and 22 immature, collected by the American 
Museum expedition to the north-east Congo, and assigned to G. pardina ficldiana. 

How careful the taxonoinist must be, and how far more limited than was originally 
supposed are his morphological resources in Gaiclta, are indicated by J. A. Allen's 
(1924) analysis of this collection: "This species, like its congeners, presents, when 
adult, a wide range of purely individual variation, not only in size and coloration but in 
cranial characters and in the teeth, especially in the size and form of m^ ... In coloration 
the variation from the norm is toward, on the one hand, an extreme gray phase with 
blackish markings, on the other, a rufous phase with deep brownish buff instead of a 
gray ground color and dark brown markings (black strongly mixed with rufous). 
The dark tail-rings are black or blackish in both; the light tail-rings are much lighter 
in the gray extreme than in the rufous extreme, being white or whitish in the former 
and strongly suffused above with pale rufous in the latter. The light tail-rings are 
usually seven, but vary in number from six to eight, besides the terminal half ring, 
broken by the black of the upper side of the apical portion of the tail. The light rings 
are usually much broader on the sides and under surface of the tail than on the mid- 
dorsal line, where in some specimens they are nearly obsolete, especially beyond the 
fourth firom the base. The light rings are occasionally as wide as the adjoining dark 
ones, but usually somewhat narrower, and frequently only about half the breadch of 
the dark ones. The black tail tip varies in extent (measured from the last full ring on 
the dorsal side) from 70 to 150 mm (in one specimen 220 mm) ... It is unnecessary 
to describe in detail the irregularities in the size, number of rows and the arrangement 
of the spots on the sides of the body, since they are more or less different in each speci- 
men, and often different on the two sides of the same specimen. Neither is it necessary 
to more than note that the relative width of the light and dark tail-rings is exceedingly 
unstable and hence has no taxonomic value. Yet such inconstant features were once 
made the basis of an elaborate synopsis of the species of the genus Getietta, noteworthy 
mainly for its puerility and pernicious results ..." The synopsis referred to, in which 
over thirty different species are distinguished, was that of Matschie (1902). Apart 
from such idiosyncratic variation as demonstrated by this Congo, and other, series, 
the intrinsic mutability and untrustworthincss of external characters may be further 
judged from the experience of Schwarz (1930) who, in 20 years' study of the genet 
question, had observed animals in captivity to manifest, as the result of enviroimiental 
change, such alteration of appearance as might rank of specific worth. 

Matschie aside, since the majority of forms have been named purely firom external 
characters and firequently from single specimens, not always abult, it will be appreciated 
that there has been considerable scope for confusion and misunberstanding. Moreover, 
there has been some tendency to assume that because a specimen berives firom the 



l88 THt CARNIVORES OF WEST AFRICA 

locality broadly luuncd as that troni which the type was reputed to come it was in 
tact that species, and on this basis invest the species with characters not truly pertijient 
to it. And authors have not infrequently fallen into the trap of assuming that characters 
present in one or two specimens, such as black forclesrs or an inner cusp on p'^, arc 
diagnostically valid. 

The tindings of Schwarz in his 1930 review ot the genus, have, to all intents and 
purposes, formed the basis of all subsequent accounts. Unfortunately his diagnostic 
characters do not always stand up to examination, being, in the light of more abundant 
material, sometimes misleading, sometimes frankly inaccurate. These matters are 
dealt with in the course of the species descriptions which follow. Meanwhile, it mav 
be said that it seems almost impossible to build any lasting structure on the existing 
shaky foiuidations — shaky because of the ill-definition of manv of the reputed forms 
including some at least of the early classic species. An attempt has been made in the 
present account to go back to first beginnings and to draw attention to some of the 
difficulties and misconceptions which have since arisen. In the consequent arrangement 
ot West African forms use is made ot species-groups, not trom any great conviction 
but as a convenient way of indicating probable relationships at the upper levels whilst 
leaving the field clear at lower levels for the subsequent indication of colour and other 
variations which imdoubtedly often exist in the groups here designated species and 
which may later be thought worthy of nominal distinction. Two subgenera, besides 
the nominate one, have recently been proposed, Psciido^ciictta Dekeyser and Paragcncita 
Kulm. The former has here been rejected; the latter retained, though no cranial material, 
on which its validity depends, has been examined. This leaves the provisional organiza- 
tion of the genus in "West Africa as 8 discrete species divided between 4 species-groups 
in the typical subgenus, with a 9th, single, aberrant species in the second subgenus. 

The following facts and opinions are here given for what they are worth in casting 
possible light on relationships in this genus. Sympatry, and hence presumed specific 
isolation, is, so far as can be assumed from non-existent or limited habitat data, demon- 
strated by British Museum specimens between senegalensis, pardina and thiciryi in the 
Wukruni Hills (Nigeria); between pardina and cristata in the Mamfc district of Camer- 
oun; ajid between poaisis and johiistciii in eastern Liberia. However, crosses are known 
to be possible in the genus, though no investigation has been made into the fertility 
of the offspring. Zuckerman (1953), recording the known cases of deliberate hybridiza- 
tion in zoos, lists a cross, in 1858, between gcnetta and sencgaleiisis; in 1859 between 
(>cuctta and tiq^riua, 3 young being born; and in 1885 between gcnctta and pardina, 
resulting in the birth of a single kitten. Finally, Pocock (1915a), from study of the 
anatomy of the scent glands, came to the conclusion that two groups of related forms 
could be recognised: furst tigrina, pardina and rnhifjiiiosa; and second (,'t7)(7fi;, dongolana 
(= senciialensis) and fclina. 

It may be as well to offer a word ot warning. In view of the diversity of opinion 
held over the years amongst authors, zoo superintendents or museum taxonomists 
regarding the identity of different forms, the naming of some, if not all, of the above 
species should be regarded with a certain amount ot caution. This applies to the 
literature m general. 



GENETTA 



KEY TO THE SPECIES OF GENETTA 
(Previous key page i66) 

1. A medial dorsal stripe present, frequently longer-haired than the neigh- 

bouring pelage, mostly intense black and nearly always (except pociisis) 
contrasting by reason of its continuity and/or colour with the, usually, 
browner and broken rows of spots. Condylobasal length of adult skulls 
over So mm. (Exceptional skins giving rise to doubt will not conform to 
the following alternative key characters) ...... 2 

Spinal stripe either obscure or, more generally, present but never black, very 
narrow but nevertheless nearly always more or less clearly split longi- 
tudinally into two by a streak of pale hairs ; the whole, usually clear-cut, 
dorsal pattern of spots or lines individually very narrow (mostly under 
10 mm wide), and generally ginger-brown on a huffish ground. Cond)'lo- 
basal length of adult skull under 80 mm . . thierryi (page 214) 

2. The back of the neck with a small crest ot erect hairs, or the hairs on its anterior 

part directed forwards and parted posteriorly from the backwardly 
directed remainder by a whorl. Tail soft and woolly, the bristle-hairs not 
much exceeding the underfur in length, clearly ringed, both above and 
below, from root to tip. Skull with no sagittal crest on the anterior part 
of the braincasc; the postorbital constriction nearly always less than the 
uiterorbital breadth; an internal cusp mostly present on p^ . . . 3 

No nuchal crest or other abnormality of tlie fur of the neck ... 4 

3. Back of the neck with a short erect crest; background colour of the dorsal 

pelage mostlv pale huffish. ..... cristata {page 198) 

Back of the neck without a crest but with the hairs directed forwards, and a 
whorl; background colour ot the dorsal pelage a warm golden-brown 

bini (page 200) 

4. Tail densely woolly and soft, the abundant underfur dominating and being 

almost as long as the rather narrow, relatively inconspicuous bristle-hairs; 
ringed pretty clearly to the tip. Mandible rather flat, with the incisors and 
canines directed forwards; p* about 6-6 mm long, in^ about 5-6 mm 
broad ........ johnstoni (page 217) 

If the tail is ringed clearly to the tip it is not sott and woolly. Mandible not 
flattish, teeth larger .......... 5 

5. Tail ringed to the tip and clad with ven,' long bristle-hairs which completely 

dominate and far exceed in length the relatively insignificant underfur, 
and is hence rather harsh. The postorbital constriction greater than the 
interorbital breadth; a complete sagittal crest mostly absent ... 6 
The soft tail, in which the abiuidant underfur is not completely dominated 
by the bristle-hairs, has at least 75 mm, and generally far more, of the 
distal end entirely black or with only slight indications below of pale rings. 
Skull with the postorbital constriction almost invariably less than the in- 
terorbital breadth; a complete sagittal crest present at least in old males 7 



190 THE CARNIVORES OF WEST AFRICA 

6. Larger and grcvcr; head & body about 490 nun; condylobasal length about 

93 mm; dorsal ground-colour rather cold whitish-grey; pale rings of the 
tail with relativelv little sandy-brown adulterating the white; top of the 
hindtoot very pale grey, sharply demarcated from the intense black of 
the underside, which spreads round the back of the heel genetta {page 191 ) 
Smaller and butl'ier; head & body about 440 mm; condylobasal length about 
S5 mm; dorsal ground-colour a warmer buft; pale rings of the tail 
almost entirely sandy-brown above, pure white below; hindtoot almost 
pure w^hite above, black below but without the black spreading con- 
spicuously romid the back of the leg . . . senegalensis {pdgc ig}) 

7. Size small, head & body estimated at about 450 mm; spots fairly large and 

separate, brownish on a pale whitish-buff ground-colour; tail clearly 
ringed white and black with only a relatively short all-black distal 
portion ......... pardiiia (page 209) 

Size larger, head & body about 500 mm; tail much more obscurely ringed, 

the pale annulations being often much adulterated with darker colour, 

even those near the base; a long, sometimes very long, distal portion 

wholly black or indistinctly marked below with partial pale rings. . 8 

N. Spots large and few, wholly black or annulate, mostly separate, in about 

4 rows, on a dull butfy-grey ground-colour getiettoides (piigc 210) 

Spots small and numerous, in about 5 or 6 rows, wholly black, those adjacent 
to the spinal stripe coalescing into longitudinal bands; ground-colour 
much as in ijt';R'»()i'(/«; tail predominantly black . . poensis (pa(;c 212) 

A. gennetta Group 

General. There are two members of this group occurring in West Africa: what 
appears to be qciictta itself and saicgalcnsis. The latter is undoubtedly closely related to 
the former, sharing with it a number of fundamental characters; but it is almost 
certainly wrong to regard it as nothing more than a race. Many years ago, Riippell 
(11S36) expressed the firm opinion that a single species ranged from southern Europe 
to the Cape, differences in the size, shape and colour of the spots being nothing else 
but climatic variations of one and the same animal, the skulls all being the same. 
While it is not possible to make the sweeping inclusions that Riippell visualised there 
IS nevertheless a basic truth in his assertion ; and though they are of no immediate 
concern to West Africa it is perhaps worth recording the opinion that both the extra- 
limital ligritia and fcliiia, in so far as they can be identified from their poor diagnoses, 
are more closely allied with the group now under consideration than with any other. 
At any rate, specimens from southern Africa in the British Museum which have in 
the past been ascribed to fi'liua seem, with very little question, to he (leiictta. 

Dcscriprion. The characters which define this group, and which very readily 
differentiate its members at sight from the others are these. The dorsal pelage is very 
long, though in the tropical forms it is less so but nevertheless still appreciably longer 
than in the other groups; most typically it is 30 mm or more long; but if the forefmger 




Senegal Genet {Ccncu, ^cncf.la,^i^: Hansa tienet (Ccucu., ,hn-rry,): Crested Genet (Cccm mst.,.,) 



GENETTA GENETTA IPI 

is run up backwards through the pelage the latter is seen to be at least as long as, and 
usually considerably longer than the depth of this digit. From about the mid-back to 
the root of the tail is a contijiuous, yet-longer-haired, spinal stripe, almost invariably 
black but sometimes dark browai. This is erectile. 

The tail is vcr)' long-haired, generally noticeably wider at the root than at the tip; 
it is almost always clearly amiulatcd throughout its length, the margins ot the rings 
not sharply dcfmed owing to the partial overlap of the long bristle-hairs of the next 
proximal ring: there arc 8 to lO dark rings, and the tip in West African forms is white 
or with only a slight sprinkling of black. The dorsal aspect of the tail has the pale rings, 
either wholly or at least along a medial line, washed with red-brown; but below they 
arc more or less pure white. The general appearance is of a rather rough, wir)' structure 
due to the fact that the very long bristle-hairs far exceed in length the relatively in- 
significant underfur. In this the group differs abruptly from pardina and servalina. 

Neither the colour nor the spots are diagnostic of the group. The groimd-colour 
varies from a pale creamy white to a colder light grey; the spots may be light brown 
or almost black of uniform colour or, in the case of the darker ones, slightly dusted 
over with red-brown. They vary in size from very small (lo x 5 mm) to moderately 
large (40 x 30 mm), oblong or roundish, set in five rows on each side of the spine, 
the fifth, or lowest, row being frequently much less well defmed or incomplete. The 
spots are for the most part clearly separate, but sometimes, especially over the rump 
and in the line or two adjacent to the spine, tend to run together longitudinally. 

The forelegs are, in West Africa, pale above but sometimes darkened on their inner 
and lower surfaces. At least the foot and sometimes the whole hindleg is pale on the 
upper side and dark below. 

Skull. In the skull, the intertemporal constriction is either about equal to or, more 
generally, greater than the interorbital breadth. There may or may not be a sagttal 
crest, p* generally has a large inner cusp, but this is sometimes lacking. 

Taxonomy. Two forms are thought to exist in West Africa. Though most authors 
concur in thinking that the various forms o( genetta are closely related there is some 
disagreement regarding the taxonomic level of naming. Schwarz (1930) dealt with 
them as conspecific but separable into (imdefmed) subspecific groups; thus ajra and 
senegaletisis, though assigned to different groups, become nomenclaturally G. genetta 
afra and G. genetta seiiegalensis. On the other hand Cabral (1966), while admitting that 
this may be the case, has suggested that the G. geiwtta group is in fact a superspecies 
comprising three independent species, the above forms becoming ascribed to G. 
genetta and G. sencgalensis. Something of this latter view is adopted here, using the term 
species-groups rather than superspecies. 

The two West African species may be separated by the key given on page 189. 
It should be noted that the characters there given are not necessarily of extralimital 
application. 

GENETTA GENETTA (Linnaeus) European Genet 

Viverra genetta Linnaeus, 1758, Systema Naturae, loth ed. i: 45. Spain. 

Genetta afra F. Cuvier, 1825, in GeofFroy, E. & Cuvier, F. Hisloire NatureUe des Mammijcres . . . , pt. 52, 



192 nil; CARNIVORES OF WEST AFRICA 

pi. 195 and pt. 51, text. Barbary. The specific name is Latin tor African. Here taken as a valid race. 
Gi'iictta i'»(i;(iri.': Lesson, 1S27, Mamie! df Manimalc^iv, 011 hisloire natmcUc des Mammijires: 173. A renaming 

of V. ^vih'llu Linnaeus. I 'i/Zijiir/s is the Latin for common. 
Giitctta barbnra Hamilton Smith, 1842, in Jurdittc's Naturalist' s Library: 35(13 of Mammalia): 171. Barbary, 

which the Latin specific name indicates. This is most probably svnonymoiis with afra, a form of which 

Hamilton Smith was, apparently, ignorant. 
There are other names ot purely extralimital mterest. 

This genet, most typically from the Mediterranean region and originally described 
trom south-western Europe is, in all likelihood, an introduction into the latter area 
trom northern Africa. The species is of disputed range within the latter continent. 
Ellermaii, Morrison-Scott & Hayman (1953), for example, following Schwarz (1930) 
regarded Iclina and other forms from the south as races of (jtvu'Wii, whereas Roberts 
(1951) excluded this species altogether from his account of South African mammals, 
according it no mention whatsoever. 

However, the existence oi genetta as a species in the southern halt ot the continent 
is ot no immediate concern to this present study. Its range in northern Africa is. It is 
generally agreed that gciictta, in the form afra, occurs throughout the Mediterranean 
fringe of Africa trom Libya to Morocco; and it is known also that its range extends 
down the Atlantic coast to the south-west at least as far as Mogador (Cabrera, 1932). 
How much further it ranges down this coast does not seem to be recorded. 

The assumption in this present work that it extends at least as far as Thies in Senegal 
IS based on a single British Museum specimen, No. 9.1 1.2.34. f" view of the pro- 
venance of this specimen it was labelled saicoalcfisis. In fact, it bears extremely little 
resemblance to Cuvier & Geoftroy's plate and description trom which Fischer derived 
his diagnosis of this species; on the other hand it does correspond fairly well to their 
plate of afra and to other British Museum specimens of this deriving from northern 
Africa. 

Genetta genetta afra F. Cuvier North Atrican Genet 

The range ot pattern and size ot this animal m western Africa cannot, ot course, 
be told trom the single above mentioned specimen available. The pelage is long and 
rather harsh, its superficial ground-colour pale grey adulterated with very light brown 
and a little black, both derived, rcspectivelv, trom the subterminal zones and the 
tips of the bristle-hairs. The markings superimposed on this background arc deep 
blackish-brown with a slight over-wash ot yellowish-brown hairs. There is a very 
deep brown, almost black, spinal crest of long hairs from just short of the shoulders 
to the root of the tail. On either side of this the spots are set in four clear longitudinal 
rows with a faint indication of an incomplete tifth. The spots of the row next to the 
spinal crest are rather irregular in shape, on the whole rather narrow, and tend to join; 
all the rest of the spots are to all intents and purposes independent. Those ot the second 
row are the largest, the anterior four being 15 to 20 mm wide and the biggest some 
40 mm long. The base of the fur is darkish grey; and there is a great abmidance of very 
long, fine underfur (about 13 to 15 mm) amongst which the long (38 to 42 mm) 



GENETTA SENEGALENSIS 193 

bristles tliat form the outer contour of the coat stand fairly widely spaced at the base 
though they come together closely distally to form a continuous cover to the whole 
body. 

The tail is rather shorter than the body, long and coarse haired, with eight dark rings, 
the basal two narrow and rather indistinct ajid a very small black tip. The bristles 
measure about 50 to 60 mm near the root; the undcrfur only 14 to 17 mm. The long 
hairs of one ring overlie the bases of the hairs of the next more distal ring for some 
considerable length so that the lines of demarcation between the rings arc not at all 
sharp. 

The forelegs are pale, only slightly spotted; the hindfeet are whitish above in very 
sharp contrast to the black of the outer aspect of the toot and the back of the ankle 
joint. The face has the usual genet markings: a white patch under the eyes separated 
by a blackish patch from the white area beside and below the rhinarium. The chin is 
white centrally but is bordered by blackish lower lips. 

Skull. This in the single specimen available, a moderate-aged male, has a low, sharp 
sagittal crest extending the whole length of the cranium. The intertemporal con- 
striction is a trifle wider than the interorbital breadth, the postorbital processes short, 
with sharp points, p^ has a large, sharp, internal cusp; the antcro-extemal cusp o£ p* is 
bifid; 1112 is fairly large, quadrate, with 4 clear cusps. The bulla is shown in fig. 28a. 

The measurements are given in the table on page 219. 

GENETTA SENEGALENSIS (J. B. Fischer) Senegal Genet 

Viveira sciu'galcmis]. B. Fischer, 1829, Synopsis Mammatiuitr. 170. Senegal. 

Vivcrra don^olana Heniprich & Ehrenberg, 1833 [fide Sherbom), Symholac Physkae scti Icoiies et Des- 

criptioncs Mainmaliiim . . . , dec. 2, folio k: 2. Dongola, Sudan. 
Viverra kptura Reichenbach, 1836, Rcgnuin Aiiimalc, I: 23, f. 270. 

This has been variously regarded as a discrete species or as merely a subspecies of 
genetta, a matter that has been discussed above. The exact, or even approximate, 
locality in Senegal from which this animal was first described is imknown. 

Taxonomy. Fischer seems fairly obviously never to have seen this genet but to 
have framed his diagnosis from Cuvier's description and plate (in Geotiroy & Cuvier's 
Histoire Naturelle dcs Alammiferes . . .) published in December 1821. It was called by 
them the Genette du Senegal but not given a binomial Latin designation. The scientific 
name and description must therefore officially date from Fischer, 1829, though, as in 
the parallel case of G. tigrina, the diagnosis is not first-hand but extracted from an 
earlier work. 

Since Fischer's diagnosis is nothing more than a Latin description ot Gcoifro)' & 
Cuvier's plate the characters originally ascribed to senegalcnsis can best be gathered from 
this last and its accompanying notes. The species is shown as a considerably paler 
animal than ajra, almost cream in colour, the black spots being in only 4 rows either 
side of the spinal stripe. Each of the innermost two rows consists of 4 spots only, 
long and very narrow and quite separate, the spots of the two outer rows being shorter 
and more rounded. The tail is shown as having 10 black rings and a white tip; the 



194 THE CARNIVORES Of WEST AFRICA 

hiiidlcg as having a sharply cut oft' black mark above ajid on the angle. The chin was 
said to be black; the coat not remarkably long; but the hairs of the tail were, making 
it look bigger than it really was. In Cuvier's original description of the Senegal genet 
(Livraison XXV) the tail was given as the same length as the body (i.e. about 500 mm) ; 
but later, in the second description of ajra (Livraison LIl) it was mentioned incidentally 
as being about 175 mm longer, the proportions in the two species being thus very 
different. 

Although at one time and another a number of specimens in the British Museum 
have been ascribed to scncgalensis there is none that at all convincingly corresponds to 
the very pale animal with widely separate, sharply defined, dead black, linear markings 
illustrated by Gcoftroy & Cuvier. The nearest seems to be a doiigolana specimen 
from Shendy (Sudan), No. 1939. 1768, in which the spots are brown and roundish 
but slightly confluent, making them appear long and narrow The tail-rings are 10, 
not very clear-cut; the tip white, though mixed with a little blackish. The black mark 
on the hindleg is fairly abruptly defined. The pelage is not markedly long but there is a 
long shiny black crest from the mid-back. 

Hemprich & Ehrenberg's description ot dong^clatm is limited in scope, and a good 
deal ot it is irrelevant in being a comparison with the purely Madagascan Fossa. How- 
ever, they state that doiigolana is a whitish, not a grey, form, with a blackish line from 
the middle ot the back and black spots overlain with cimiamon, there being no mention 
ot their shape, size or arrangement. The tail is described as differing verv widely from 
gcnctta and "'singularly short", the number of rings unstated. 

The specimen. No. 1939. 1768, mentioned above as most closely representing sene- 
galaisis corresponds pretty well to this sketchy description and there seems little doubt 
that the two torms are really one and the same. 

Distribution. This being so, we have a range tor scncgaknsis troni Senegal to the 
Nile and beyond (Setzer, 1956) to the Red Sea, that is to say transcontinentally through 
the Sahel and Subdesert zones of vegetation. Actually, as will shortly be seen, it occurs 
also further south in tlie Sudan and Doka zones. Specimens exist in the British Museum 
from Aoudcras (Air — Subdesert) ; Fort Lamy (Chad — Sahel woodland) ; Zuarangu 
(Ghana — Sudan woodland) ; Famiso and Kabvvir (Nigeria — Sudan woodland) ; Zaria 
district and Wukrum Hills (Nigeria — Doka woodland). It is also probably this species 
which A. J. Hopson (private communication) observed to be trequent in creeper-hung 
trees in the Yobe Valley, near Yo, Lake Chad (Sahel). 

Description. The Senegal genet, as represented by these specimens (Plate 3), 
is a distinctly pale animal with a background colour ot whitish, cream or buftish, 
and spots that arc mostly light or medium brown but more rarely blackish-brown. 
They vary in size but are never large, rarely having a maximum diameter ot more 
than 20 mm and usually less. One unusual example from the Wukrum Hills has very 
small spots, bearing the same relationship of pattern to the other skins as the servaline 
does to the serval. They arc also very variable in shape but could be described in 
general as irregularly roimd or oval, occasionally oblong; and though here and there, 
chiefly over the rump, joined together, they are in general sufticiently widely separated 
for there to be as much or more ground-colour as spots showing. In this last, as in 



GENETTA SENEOALENSIS 



195 




Fig. 29. Cenetla seiiegatensis: skull, B.M. No. 1939. 1766, (J, X i 

Other ways, they differ from the small-spotted servalina in which the spots dominate 
the ground-colour. 

The pelage, although considerably shorter than in genctta, is nevertheless still long 
by comparison with other West African species, the bristles, in typical non-moulting 
specimens, measuring about 25 to 30 mm over the rump, the underfur about 12 to 
14 mm. There is a long-bristled spinal crest from just posterior of the shoulders to the 



196 THE CABNIVOHES OF WEST AFIUC:A 

root ot tlic t.ul. rhis IS almost al\va)s black but in some cases merely deep brown. 

The distinctive character of the tail unites all these skins and shows their affinity to 
i^ctictru. It is long-haired, the coarse, terete bristles measure trom 45 to ss mm near the 
root, the uiiderhir 13 to 15 mm. As in all this group, the annulation, though mostly 
quite clear, is not sharp-margined because of the overlap ot the long bristles. For this 
reason the basal two or three rings are difficult to determine; and this, in some measure, 
accounts for differing opinions regarding the number of rings. The most usual number 
of dark rings is 10, though in some cases there seem to be only 9. The tip of the tail is 
white but often has a tew black-tipped hairs mixed with it. The dark rings may be 
black or deep brown; the pale rings are always pure white below, but dorsally they 
generally have a good deal of light-brown mixed with them, especially medially, 
or are entirely golden-brown. The length of the tail seems to be irregular. Only 5 skins 
have field measurements; and in these, two tails measure slightly less than head & 
body, one appreciably less, and two appreciably more. 

The forelegs are pale above and blackish below. The hindteet are white above; 
the legs arc blackish below and on the outside and back ot the lower leg (tibia) but not 
usually entirely aroiuid the ankle as mgeiictta ajra. The usual genet white tacial markings 
are present — below the eye, around the rhinarium and between the eyes. The chin is 
always white medially, bordered by a broader or narrower black area running along 
the lower lips. 

From the limited figures available and trom the general appearance ot the skins this 
genet would seem to be smaller than the animals living in northeni Atrica and Europe 
((;cuiitd). The actual mean measurements ot British Museum specimens will be tound 
in the table on page 219. 

Skull (tig. 29). Seven skulls, of which only tive are mature, exist 111 London. Though 
the latter are pretty even in size they show, as throughout this genus, contusingly 
irregular characters, the various inconsistencies being distributed over all the specimens, 
irrespective of sex, and not confined to one or two exceptional examples. In two temales 
and one male there is no trace of a sagittal crest other than a posterior portion adjoining 
the supraoccipital crest; in one other oldish male there is a clear, but very low, crest 
across the whole length of the cranium; and in the third, oldish, male such a rudi- 
mentary crest is partly present. In tour ot the specimens p'^ has a large internal cusp; 
but from the fifth, a medium-aged tcmale, all trace of this is completely lacking. In 
this skull, also, in- is very much reduced in size. The development ot the postorbital 
processes varies trom long points to fairly short ones. One lower jaw has a large 
quadrangular, four-cusped iiij.: but in the remainder this tooth is triangular and thiee- 
cusped. 

B. servalina Group 

The animals included here are usually regarded as a single species covering tour 
described subspecies; but the two forms occurring in West Atrica seem to be sufficiently 
distinct to constitute two discrete though, as in the case of (jc/dVAi, very closely connected 
species. This will be dealt with more fully below in the taxonomic section. 



GENETTA SERVALINA 197 

Description. There is little room for doubt in tlie determination of the scrvaline 
genets. The characters, especially in comparison with genetta are very distinctive., 
The dorsal pelage is short compared witli that of typical genetta, though not markedly 
shorter than in seiiegalciisis; but it has none ot the harslmcss of the genetta group, being 
instead rather silky. This is because ot the quite different form of the bristle-hairs (20 
to 25 mm long) which, in this group, arc only expanded for the distal one-third of their 
length, tapering abruptly proximally to a long narrow pedicel not readily distinguish- 
able by eye from the underfur, than which it is not a great deal wider. The distal 
portion is of oval section, not terete. The luidcrfur is from 15 to 18 mm long. The 
ground colour in this group is, in general, redder than in genetta because the pale-grey- 
ended underfur that forms its basis is overlain by bristles that have wide subterminal 
bands of red-brown or even bright orange. This reddening always shows but varies 
in extent, being sometimes in young skins and at moult considerably reduced. The 
pattern is of very numerous, small-sized, dense black spots not always very clearly 
disposed in longitudinal lines; but where they can be coimted they form at least seven 
such on each side of the body. The spots are generally well under 20 mm diameter; 
and they are mostly separate though occasionally they may coalesce near the spine. 
There may or may not be a spinal crest. 

The tail is the other very characteristic feature of this group. It is subcylindrical, 
relatively short-haired and soft-furred, and very distinctly annulated from root to tip, 
the edges of the rings being fairly sharply cut. The bristle-hairs do not greatly exceed 
in length the abundant underfur. Whereas in genetta the tail broadly matches the body 
colour, in seri'alina it, as a rule, torms something of a contrast by reason of the fairly 
pure white of its light rings. The dark rings, which are generally black or very deep 
brown, are usually 10 to 12 in number, the tip being white, though as a rule not so 
purely white as the remainder ot the light rings. The alternate rings are subequal in 
breadth or, more commonly, the white narrower than the black. The tail is shorter 
than the head & body. The legs may be black or pale; the usual genet pale facial 
markings below and between the eyes and around the rhinarium are present but 
mostlv more obscure than in genetta. 

Skull. There are not enough mature skulls to make comparisons with the genetta 
group particularly useful. The characters are variable. In the six skulls available the 
sagittal crest is restricted to the extreme posterior segment in four and, across the main 
body of the braincasc, is incipient in two, a medium-aged male and a very old female. 
The internal cusp on p^ is large in two examples; and in the others ranges from vestigial 
through a mere bulge to a low tubercle. Both /))- and "12 vary a good deal in size, 
and the latter may be quadrate with four cusps or triangular with three. The postorbital 
processes may be long fine points or blunt and dctlccted. From such variable material 
it is scarcely possible to derive any useful diagnostic characten. The most constant 
feature is that the intertemporal breadth is narrower than the interorbital ; but in one 
case it is appreciably wider. In the genetta group examples it is always wider. 

Taxonomy. The group is usually regarded as a clear-cut one. Today it is universally 
held to be monospecific although in the past three discrete species were named. In this 



ipS THE CARNIVORES OF WEST AFRICA 

present accoiuit it is suggested that there arc three vaHd species, two ot thcin occurring 
within the area dealt with in this work. 

The group first became known when in 1H55 Puchcran described sen'aliiia from 
Gaboon. His brief Latin diagnosis was not in itself very precise; but since it makes it 
clear that the spots arc extremely numerous, the limbs almost completely black, and 
the tail long with broad black and narrower white rings the species is readily recog- 
nisable and has never been questioned. At the same time Pucheran named a second 
specimen from the same area aiihryana, the description being vague, the only apparent 
difference from scrvalina lying in the ground-colour. As tliis last is alwa\s variable the 
two names have for long rightly been regarded as synonymous. 

In 1902 Thomas described hcttoni trom Kenya as a discrete species on the score ot 
smaller size and annulate spots. This, which has proved to be a very common form in 
eastern Africa, is now regarded as nothing more than a race oi scrvalina. The annulation 
of the dorsal spots is not constant; but the markedly small size of the skull and the 
shortness of the rostrum and palate probably justify Thomas's original evaluation as a 
species. The same cannot be said ot Lonnberg's i\itcusa from the Congo, the author 
himself being doubtful ot its validity as a race of seri'aliiia. Ot more immediate concern 
to West Atnca is the third named race ot this species, cristata Hayman. Paler ground- 
colour and pale forelegs apart, there are certain features of this animal which appear 
to the present writer to entitle it to consideration as a discrete species. These arc the 
possession not only of a black spinal crest, absent from true scrvalina, but of an erectile 
nuchal crest as well. The distribution, so far as we at present know, lies between the 
Cross River and the Sanaga; whereas scrvalina itselt is tound, in West Africa, only to 
the south of the latter river, that is to say entirely extralimitally. 

The evidence that scrvalina occurs as far west as the River Niger, as stated m Schwarz 
(1930) and Rosevear (1953) is slender. No record has been traced to account for 
Schwarz's assertion; the green spot shown on map No. iSia in Rosevear relates to a 
rather poor skin collected by the present writer near Benin and of doubttul standing. 
Though from its general appearance, including the typical short-turred, wholly- 
annulated tail, it obviously belongs to this present group it is certainly neither scrvalina 
nor cristata. The ground-colour is a richer orange-browai than in these forms; the spots 
are appreciably larger and less numerous — so much so that it was originally thought 
to be a "niaailata" . This was later altered to scrvalina by R. W. Hayman. The forelegs 
are black, much as in scrvalina, but there is a very clear, black spinal crest on the lower 
back, not occurring in that form, though no nuchal crest as in cristata The tail has only 
nine black rings, the tip (the 9th) being black Whether this is a species or merely a 
race is open to question. It is here treated as a new species in its own right. 

The two, consequent. West African forms can be readily told apart by means ot the 
key on page 189. 

GENETTA CRISTATA Hayman Crested Genet 

Gcuctta seri'ttlina criitaia H.iynun, 1940, Trans, zool. Soc. Land. 24: 6Sf)-6S,s. Okoiyoiig, Mamfc Division, 
Cimcroun. Type 111 the British Museum, No. 39.323, ^; skin in good condition, skull good except for 



GENETTA CRISTATA 199 

the loss of four upper teeth. The subspecific name was given with reference to both the nuchal and 
spinal crests. 

Distribution and general. This genet (Plate 3) is known in the British Museum 
from 5 adult and 9 yoiuig or juvenile specimens, all from the Cross River — Cameroun 
forests. Only 3 of the adult specimens arc complete with both skin and skull. The 
majority of these come from various places in the Mamfe Division, collected by Sander- 
son (1940): Mamfe, Binjong, Bashauo, Olulu and Okoiyong; but there are others 
from the Rumpi Hills (Kuniba Division, Cameroun) and Oban (south-eastern Nigeria). 
In addition to these, Eisentraut (1963) records 2 specimens from Mount Cameroiui at 
Buea and Malendc. The whole area of known distribution, therefore, lies between the 
Cross River basin and the Cameroun Mountain, a heavily forested and, imtil recently, 
secluded region, about 100 miles long by the same wide, well north of the Sanaga 
River, south of which the species is replaced by servalina. 

Description. In the servalina group the ground-colour of the dorsal pelage is some- 
what variable from specimen to specimen ; but in so far as a generalisation can be made 
it is rather paler in aistata than in servalina itself. It is buffish-grey ; the orange- 
brown subterminal rings of the individual hairs adulterate this colour to a greater or 
lesser extent in different examples, sometimes scarcely at all except romid the anterior 
edges of the black spots. The latter vary a little in size and shape but most typically 
may be regarded as roimd, and 10 or 15 mm in diameter, though sometimes they tend 
to become rather longer than broad. They are, in general, independent but there is in 
some specimens a greater tendency to unite than in others; and almost always they 
torm short lines near the spinal crest in the region of the hindquarters. The bristle-hairs 
of the lower back are about 20 to 22 mm long; the underfur some 10 to 13 mm. 
In servalina the belly is dark, almost blackish-brown; the chin and throat are lighter 
but nevertheless darkish grey and, especially in the latter, washed with golden-brown ; 
cristata, on the other hand, is considerably paler below; medium buffy-brown on the 
belly but the chin and throat cold whitish-grey with little or no warm over-wash. 

The black spinal crest becomes noticeably long from the mid-back to the root of the 
tail; but it is, in a shortened form, really continuous with the nuchal crest. This latter 
is much shorter than the main spinal crest being, indeed, little longer than the surroimd- 
ing tur from which it stands out by reason of its ercctness instead of being adpressed 
and directed backwards. Though it is most conveniently referred to as the nuchal 
crest it, in fact, has its origin, in a very low form, on the front of the face from about 
between the eyes, whence it continues over the crown to the neck where, the fur being 
longer, it becomes far more obvioas. It is caused by the direction of the hairs, which 
converge towards the medial line. On the neck it involves the central black stripe and 
the brown-ringed contiguous hairs, and is thus of mixed colour, not wholly black like 
the spinal crest. It is of interest to note that rudimentary traces of the frontal crest 
arc to be found in typical servalina. 

The tail may have fewer black rings than servalina, the type having only 8 clear 
rings, thougli others, including all the juveniles, count up to 10; but the reckoning 
of the terminal rings is open to argument, there being a very obscure narrow darkening 



200 THE CARNMVORr.S OV WEST AIKICA 

that could or could not be regarded as a ring. The tail-tip is really white, but because 
of this indecisive blackening becomes greyed. The forefeet arc in their upper portion 
light-coloured and spotted similarly to tjic back, but become blackish towards the toes 
and on their outer and lower aspects. The hindlegs are somewhat similar but often 
carrv a distinct pale, greyisli, patch over the metatarsal region above. 

Skull. There is no particular distinction between aistata and sevvaliiia. From the 
very few examples available for comparison it is just possible the inner cusp of /(^ is 
better developed in the former than in the latter, as in the three available specimens 
it is large in t\\ o and represented by a prominent swelling in the tliird. All skulls have 
a posterior sagittal flange; but a crest across the lengtli of the braiiicase is rudimentary 
in two, and absent from two, an old male and a fairly old female. This old male is 
the only specimen in which the postorbital breadtii is greater than the interorbital. 
The transverse breadths of /ii- and ;»2 are rather greater than m scivaUna. The bulla is 
illustrated in fig. 28c. 

Habits. Sanderson (1940) characterised the habitat as scrub, low tangled vegetation 
and bare ground below trees. He considered the species as detmitely terrestrial, only 
one specimen having been seen in a low tree, and that returned to the groiuid to make 
its escape. 

GENETTA BINI spec, not: Benin Genet 

The erection of this species from one very incomplete specimen seems to be justified 
bv its apparent difference from other known forms, from its extreme western pro- 
venance, and the wide gap of 750 kilometres separating it from its nearest recorded 
neighbour, across the two fauiial barriers of the lower Niger and Cross Rivers. It is, 
in fact, described below from a, in some respects, poorish skin devoid of anv supporting 
skull. 

Description. The type, B.M. No. 50.315, o (Plate 4), was collected for the present 
writer in December 193S by R. II. Hide, a forest officer, in the Ohosu Forest Reserve 
some 65 kilometres north-west of Benin City, western Nigeria, altitude 250 metres. 
Although appreciably different in detail, the overall resemblance of the pelage and 
tail patterns to scrviiliiia make it quite certain that it belongs to that group. The ground- 
colour of the dorsal pelage is of a much richer red-brown than in the two other more 
commonly known western species, or in the more golden-brown of iiitciisa from eastern 
Congo. This is because the subterminal bands of the bristles are both broader and more 
deeply coloured than in the other species. The pelage length is much the same as m 
cristata. 

This new proposed form (I'late 4) diHers from sciraliini 111 the piossession of a long 
black spinal crest from the mid-back to the tail; and from crinata by the absence of a 
nuchal crest. The fur on the back of the neck, however, is directed forward and meets 
the backwardly directed hair of the frontal region in a line across the crown between 
the ears. Where the differently orientated hairs of the neck and back meet there would 
appear to be something in the nature of a whorl; but tlie only skin is so poorly made 
that the position here is a little obscure. 



GENETTA PARDINA 



The greatest difference between this new and the other species, however, Hes in 
the size, shape and number of the spots. They arc much more irregular, less round, 
in shape, and appreciably larger and fewer in number. Many of them are subquadrate 
or triangular, 20 mm or more across, those of the middle lines being larger than those 
near the spine or lower down. Between the shoulder and the hip there are not more 
than 7 or 8 in a row as compared with 10 to 12 in the hitherto known forms. A few 
merge over the rump and one or two are confluent with the spinal crest. Owing to 
the make-up of the skin and the virtual absence of the belly it is not possible to say 
precisely how many rows there arc, but it is probably 6. 

There are 9 dark rings on the tail including the black tip. The pale rings, which 
are only about half the width of the black ones, although white below, arc dorsally 
distinctly more coloured than in the other species of this group, especially the proximal 
ones, being greyish with a fair mixture of reddish and black hairs. The bristles of the 
proximal black rings are rather longer than normal and overlap the light rings more 
than is usual. The chin and throat are medium grey, fairly well spotted. Although 
most of the legs are present there is only one foot, the left forefoot. The upper arms 
are deep grey and spotted on top but black beneath like the whole of the remainder of 
leg and foot. The lower parts of the hindlcgs are blackish all roiuid. 

Nothing is recorded of this animal except that it lives in closed high forest, and the 
only specimen known smelt strongly of musk. The name Bini, pronounced bee-nee, 
is the name of the ethnic group inhabiting Benin and the surroiuiding district. 

C. pardina Group 

In all the confusion that exists in the genus Genctta the greatest, without question, 
concerns the three classic species, tigrina (1778), jAina (1811) and paidina (1832). The 
last is of closest concern to West Africa; and though it has in some measure, through 
the attribution of the same forms to one or the other, become involved with the first 
an endeavour must be made initially to sort out the question o( pardina alone. 

With the pardina group we reach a remarkably mixed assortment of forms, far and 
away the most variable in visible characters. Wliether the group embraces more than 
a single species is an open question. The animals lying at opposite ends of it are assuredly 
markedly different in appearance; yet there seems to be no clear-cut division between 
them. Great play has been made by taxonomists with spot size and colour; with the 
number of tail annulations and the relative widths of the light and dark rings; and with 
the colour of the feet. There is little doubt that these characters, and others less often 
taken into account, are subject to considerable variation, possibly to some slight 
degree local but for the most part individual; though from the largely disjoint study 
material in the British Museum — scarcely any two from the same locality, apart 
from questions of age and sex — it is hardly possible to do more than conjecture that 
this must be so. However, the 46 specimens collected by the American Museum 
Congo Expedition, by reason of their restricted provenance, cast a more certain light 
on the matter. J. A. Allen's detailed comments on variation of supposedly diagnostic 
characters as cxliibitcd by this scries have already been quoted on page 187. 



THE CARNIVORES OF WEST AFRICA 



Description. In ilic Uci: ot tlic \^■idc inconstancy thus revealed it sccnis difficult, 
it not impossible, to find dependable characters that might serve to defuic pauUiia. 
In truth the matter boils downi to a similarity of tail structure and markings throughout 
an otherwise capricious group, serving to differentiate the animals included here on 
the one har.d clearly (i-oni (jaicttci and on the otlier, less emphatically, trom scnudina. 
This tail is of the relatively short, backwardly-directed-haired, subcylindrical type 
charactcric o£ scri'aliiia, differing manitestly trom the ver)' long-, coarse-haired type 
ot du't^cnctla group in its extremely soft and silky nature. It differs from that oiscrvaliiia 
in being appreciably more melanistic, the pale rings being not only few er in number 
but, as a rule, of far less width than the black ones. Further, the terminal portion of the 
tail, sometimes only about 75 mm, sometimes very much more, is almost whollv 
black, particularly above, there being, not uncommonly, faint traces of white partial- 
rings on its underside. It is, in this species, more convenient to count the pale rings 
than the dark; but often there arc only halt-rings, or less, visible on the underside but 
not the top. The count varies, therefore, on the two aspects, dorsal and ventral, of the 
tail. Taking the latter, there are usually 6 to S pale rings or partial rings, but in excep- 
tional specimens there may be one less or more. Even in the basal portion the rnigs are 
often fir more obscure above than below. Their lesser number, the (as a rule) narro\\- 
ness and relative obscurity ot the pale rings, and the long black end to the tail serve 
as a distinction between even the small-spotted forms of this species and the scrvahiia 
group, in which the rings have a sharp clarit)', are subcqual in width, and continue, 
above and below, to the distal end. 

Taxonomy. The mix-up owcv paniiua stems partly trom Matschie (1902) and partly 
from Schwarz (1930), both, in an obscure situation, having long been regarded as the 
ultimate authorities; and what they have said, surmised or hinted at has, often blindly, 
been followed. Added to this is, at least as far as the British Museum is concerned, 
a plethora of specimens trom the rain-forest belt and a paucity trom the drier inland 
areas. Now, Isidore Geoffroy, clearly as the result of specific enquiries, described 
pardiua as emanating trom the interior of Senegal, that is to say from one of the more 
arid types of woodland. Sudan or Saliel, certainly not from the closed forest. There is a 
temptation today, and both Matschie and Schwarz succumbed to it, to criticise early 
notions ot West African geography and to amend reputed provenances to suit what 
seemed to them to be the tacts. It cannot be denied that this is sometimes justitied; 
but there is no reason whatsoever to suppose that Geotiroy was in this particular case 
misled. Senegal had tor long been intimately connected witli France and in Geoffrey's 
day was certainly closer, and a far more likely source of livhig animals, than any more 
heavily forested part of West Africa. 

Matschie showed himself to be in two nriiids regarding the vegetation and distri- 
bution proper to the species. On page 1135 of liis paper he connected pardiua with the 
coastal belt of Togo; but on page 1142 he gave as the range first north Cameroun, 
secondly Togo; and implied possible synonymy with pociisis, a forest form. Schwarz, 
too, who first identit'ied piudiua with tnacidata (a synonymy since necessarik reversed) 
was not very clear regarding the ecological background proper to the species. In a 
footnote on page 277 he said, in translation: "Gciictia uiaaiLua is p.itentK' the same av 



GENETTA PARDINA 203 

Genetta pardina I. Geoffroy, so that the second, more used, name must disappear as 
clearly a synonym. Matschie (1902) has already shown that maculata is no north African 
form as the original description says . . . (It) doubtless emanates from upper Guinea 
and, since it was imported live, can very well, like the type of G. pardina, stem from 
Senegal, which was at that time one of the most important export coimtries for African 
animals". In other words lie agreed to its relatively dry-zone origin. Yet in his main 
text, near the top of the same page, Schwarz quite unequivocally rated maculata (as he 
called the species) an upper Guinea closed-forest ("Waldgcbiet") animal, ranging from 
the Niger to the Gambia — thus vegetationally, and possibly distributionally, running 
cotuitcr to Gcoftroy's "interior of Senegal". J. A. Allen had obviously adopted an 
entirely different view of the distribution oi pardina in that he had ascribed the major 
part of his north-eastern Congo material to the species. Further, in another footnote, 
on the previous page, while allowing the species a certain degree of range in climate, 
Schwarz none the less still confmed it to the high-forest: "While among the forms of 
the moist forest G. maculata produces a small-spotted type 'pocnsis', the spots in the 
marginal areas oi the forest arc large and the groimd-colour pale — 'maculata' {'pardina'), 
'genetloidcs' ; intermediate types 'diibia', 'johnstoni' " . Attractive and reasonable as this 
hypothesis may seem, it will be seen shortly that Schwarz was quite wrong in con- 
necting spot size with humidity. 

Relatively recently an extremely interesting set of 10 flat skins, in only very fair 
condition and luifortimately without skulls, has been received bv the British Museum 
from Mr. J. D. Carter. They all emanate from a very restricted area some 30 kilo- 
metres long, in the Pctico-Zo locality ot the Wukrmn Hills, 25 kilometres north-east 
to 25 kilometres north-west of Lau, which lies on the upper Bcnue River in Nigeria 
at about ii°30 East. This area is in the Sudan woodland zone. The interest of this 
scries lies in the fact that it shows that no less than three quite distinct, though very 
similarly coloured, taxonomic forms, scncgalcnsis, pardina and tliierryi, live sympatri- 
cally; that eight of the skins which arc pardina display a wide rajige of size, shape and 
colour of spots, some of them being close enough to Geoffroy's description and illus- 
tration to make it clear that the type specimen o£ pardina may well have come from 
the Sudan zone and that Schwarz's limitation of the species distribution to the closed 
forest v»'as misleading. It must, 111 all fairness, be added that one exceptional skin. 
No. 39.132, from a small island in the mouth of the Volta River, near Ada, is not dis- 
similar from these Sudan zone specimens, though of a slightly colder tone, and quite 
distinct in its paler ground-colour from all southern Ghana forest material. But although 
so near the coast and thus expectcdly in the wet-forest zone, this area is, in fact, one of 
remarkably low rainfall, not greatly exceeding that of the much more inland Sudan 
zone, and the vegetation of a dry littoral kind, together with some mangroves. 

Because of the wide range of pattern and colouring in the pardina group the present 
writer is rather sceptical of the usefulness, or the feasibility without considerable further 
evidence, of drawing nomcnclatural distinctions between the various forms. The 
extremes are very distinct and could support independent names; and it thus seems 
unjustifiable to apply precisely the same term to the pale, creamy-butf, large-spotted 
animal ot the dry open country and to the relatively dark, greyish, small-spotted 



204 THE CARNIVORES OF WEST AFRICA 

inhabitant ot tlic dense rain-torcst. But, though there arc considerable laciuiac ni the 
study material, there is some reason to suppose that the forms arc wholly clinal, and 
in the attempt to differentiate everything one could be faced with devising an almost 
endless string of names, the limit of whose application was anything but clear-cut. 

In the face of this the question of what to do \\ ith puniiiui is a difficult one. The 
course adopted in tliis work is to regard it as a species-group, calling the pale open 
comitr)- forms pardiua. the large-spotted forest form <^aicttoiiIc.<, and the small-spotted 
iorcst form fiicnsis. This is more a matter of convenience than anything else and can 
at once be shown to be illogical since, while differentiating between the extreme 
forest forms, it tails to do the same for the dry-zone forms; and, as regards the forest 
forms themselves, whether a specimen is large-spotted or small-spotted must often 
remain purely a matter of personal opinion. But it has this merit: that it does not 
burden an obscure situation with a mass of new names that more abundant material 
and deeper continental-wide research would almost certainly show to be unnecessary 
and inept. 

This decision having been made it is necessarv to examine and discuss the standing ot 
certain named forms ui more detail in order to explain the difficulties which have arisen 
over some and the reasons tor the rejection of others from the West African list. 
The earliest of these is viaailata Gray, 1S30, which since Schwarz (1930) has been gener- 
ally accepted as synonymous with paniiua. Owing to doubts and contradictions in 
Gray's diagnosis the position is, in tact, not absolutely straightforward. I lis English 
description is of a grey-brown animal with a brown erectile crest from the shoulders 
to the tail, and spots in 3 rows on each side, the 1 innermost square, those next to the 
spine largest and nearest together, the lowest series roundish; below these are scattered 
black-browii spots. There would thus appear to be 3 clearly defined rows ot spots 
on each side with further spots, not disposed in lines, below them, a not unusual 
arrangement in the genets, giving rise to the common diagnosis "spots in 4 or s rows 
on each side". But Gray's Latin diagnosis gives the spots as in 6 rows with scattered 
round spots on the flajiks. hi view ot the English description this could logically be 
taken as meaning a total derived from two sides of 3 each. The accompanying illus- 
tration, however, clearly shows 6 well-defined rows of independent spots on a single 
side; and since Gray must presumably have seen the drawing before publication it 
might be taken that this is what he, in fact, meant — though it scarcely accords with 
his verbal description ot 3 series of spots with scattered spots on the sides ot the bell)'. 
Neither does the illustration bear out the written description ot the tail. This states 
that there arc 7 black bands increasing in breadth towards the end. The picture, how- 
ever, shows 8 together with a black tip making, in all, 9 black bands, which show no 
appreciable increase of width. 

Schwarz, synonymising maculata znA pardiua, drew attention to the fact that Matschie 
had already expressed the view that the former was no North African animal, as Gray 
had stated, and had noticed, also, the discrepancies between Gray's diagnosis and his 
illustration. Schwarz, ignoring the confusion over the spots and contming himself to the 
tail, expressed the view tjiat the written description "mav be regarded as authoritative 
as it speaks ot 7 dark tail-rings, which .igrecs with the upper (iumea genet. Apart 



GENETTA PARDINA 205 

from the inexactness of the drawing of the tail the picture shows a long-legged, long- 
faced genet that doubtless emanates from upper Guinea . . .". It will be seen that 
Schwarz accepted or rejected the illustration to suit his own theory and his own 
defmitions. He further supported his view by a pure guess that the country of origin of 
Gray's specimen w^as really Senegal. 

With doubt cast on the body pattern and on the tail there seems little enough in 
the type description that may be picked upon as usefully diagnostic apart from large, 
quadrangular, more or less separate spots. One thing is certainly lacking from both 
verbal and pictorial descriptions : any reference to the markedly melanistic tail with the 
extra-long black tip and pale half-rings which arc the essential characteristic of all 
parditia. The illustration is, too, of a distinctly more tapering and fmely-tipped tail 
than is general in that species. The true country of origin of Gray's specimen is, in fact, 
not without relevancy; tor his figure is, apart from an extra line of spots on the flank, 
not a bad representation o(^ genetta afra, which, if the animal came from North Africa 
as he said, it could well be. 

The name maailata has been rejected as unavailable because of prior occupation. 
Were this not so it could justifiably be abandoned on the grounds of obscurity of 
definition. This latter also renders its correct position in synonymy uncertain, but in 
order to avoid introducing unnecessary confusion it has herein been retained in its 
customary place under paidlua in the synonymic list which follows later on page 209. 

It is perhaps of general interest, as a commentary on Gray's contused diagnosis, 
to quote here a view of his reliability written not from the purely zoological angle 
but as a pointer to his character: "He published a steady stream of brief zoological 
papers with such haste that in a number of instances he changed his mind and published 
a second paper correcting his first; indeed, in some cases he soon published a third 
paper amending his second". (Paden, 1964: 13). 

The forms pociisis and gcncttoidcs will be commented on in their own appropriate 
sections later. The status of a few other names must be examined here. The first calling 
for attention is ficUiaiia Du Chaillu, i860, because from time to time West African 
specimens have been referred, directly or indirectly, to it or it has been stated to occur 
in that region — as, for example, in Rosevear (1953). Du Chaillu's description is not 
very helpful but makes it probable that his animal belonged, in fact, to the pardiim 
group. This is fairly clear from the tail, which was said to have 7 dark rings, the first 
incomplete below, the last indistinct; the last 5 inches of the tail brownish-black. 
Other points in the description tend to confirm this. The size of the spots is not given : 
but, indirectly, by mentioning that the two lateral rows on each side of the spinal 
band arc broken up into five or six smaller longitudinal spots he indicates that these 
must in fact be largish, as in genettoides. 

J. A. Allen (1924) treated ficldiana as a race ot pardiiia; but Sch\\ arz {1930) regarded 
it as a subspecies of tigrina, bemg followed 111 this by G. M. Allen (1939). In view of the 
character of the tail this latter classification seems improbable. Not much is revealed 
by skulls in Gciietia; but there is a Du Chaillu skull of an old animal in London (No. 
1645a, sex?), without a skin, that conforms pretty closely to the general riui of the 
pardiiia group except that it is on the whole rather below average size and has an 



:o6 Till" rAUMVdliKS (II- WIST AIRKA 



:i' 



uiubually narrow iiucrtcmporal auisuiction. However, it is matched m this lattc 
by one of the laigc-spottcd oencnoiilcs from Ghana (No. 46.397). Tlic standing of 
Ucld'uiiui is only incidental to West Africa but is regarded herein as jiertaimng to the 
Ihudiiia species-group, the name being iirobablv synonymous widi ocucttoidcs. 

Matschie's duhia, the exact provenance ot whicli in West Atrica is unknown, is also 
almost certainly fn'itdtoides. No set description ot diihiii was given by Matschic and 
diagnosis ot the form must be built up from his scattered key characters. The most 
significant ot these are: spots large, in 5 rows; tail short-haired, the hairs not more 
than 2s mm long at the root; the tail itselt very short, two-thirds of the head & bodv 
length (tliere is a misprint in the key), with only 6 pale rings, which arc at most as 
wide as the dark. Tails are. 111 the pardiiux group, normally rather shorter than the 
head & body length (about S'i per cent); but it is doubtful whether such an extremely 
short-tailed animal exists except as a treak or the result ot accident. The torin dulna. 
theretore, isjiere regarded as, in iact.frciicttoidcs. Schwarz, also, put the two in synonymy. 

Cabrera s liistilctrls. described trom Fernando Poo in 1921 because no trace ot pociisis 
Itselt had so tar come to light on the island, must next be considered. Schw'arz grouped 
this with rifJiiiia; but Cabrera himselt likened it in general to pardina, oi which he 
considered it to be an island representative. He described it as having few spots, large 
and annulatcd for the first two rows, which are more irregular and imperfect than 111 
pdidiihi. many of them formed trom two black spots enclosing an intermediate, ill- 
detmcd reddisji space; the base of the pelage infused with cinnamon, the flanks and 
tjiighs very bright; the feet grey. No mention is made ot the tail. There is no reason 
to suppose that Cabrera's estimate ot taxononuc relationship was wrong and that 
Schwarz was more justified in allocating the form to ti(iriiia. The form is, therefore, 
in this work s\nonynnscd whh f^ciicttoidcs; but it is possible that with tunher collecting 
It may be found to merit some more independent status within the pardina group. 

No form is attributed in this present work t<i tii^iiua; but since this species has so 
often been mentioned in previous paragraphs and has been held to occur in West 
Atrica It is necessary to consider the question here. G. tii^ruia constituted one ot Schwarz's 
six basic forms. Indeed, it was certainly one ot the major species since he divided it 
into tliree groups embracing 1 1 subspecies besides X other named forms sunk into these 
as synonyms. According to him its range was the entire jiigh-forest block except west 
of the Niger, and the neighbouring dry woodlands; the savajuiah north of the forest, 
including west ot t]ie Niger as tar as Senegal; also Sudan and Ethiopia; and East Africa 
to the coast and to the Cape. It was thus in Schwarz's view extremely widespread, 
both geographically and vegetationally. In view of the importance attaclied to this 
naiiK Its standing ,ind tlie ditticulties connected with it must be examined in sonte 
detail. 

The species iiuisl under the International Code be dated from Sclireber, 1778, though 
this author cleari\ states that the animal was unknown to him and that Ins description 
was copied from that of Vosinaer and the accompanying plate. Sclireber quoted as his 
source the French edition of Vosmaer (1771) though tlie description had originalK^ 
been published in I )iitc]i in the same \ear. Hut whether the French or the Dutch version 
be takiii Stlirelnr's re-\\riting i>l 11 in (ieriii.iii was iicH alwa\s \-er\' acctir.ite, .is when 



GENETTA PARDINA 207 

he said there was a black stripe troin head to tail whereas Vosmaer quite clearly des- 
cribed it as running from the middle of the back. It may be as well to add here that 
the plate accompanying Schreber's type diagnosis is, as far as outline is concerned, 
an exact copy of Vosmacr's or perhaps made from the same block; but the colouring, 
in the British Museum version at least, is of a very markedly darker brown than in 
any ot tour different copies of Vosmaer. 

Very few characters of any taxonomic value arc, in tact, given by either author. 
Schrcber stated that there were many irregular browii spots; this he derived purely 
from the illustration, Vosmaer having made no reference to them. This picture, which 
carries the subscription that it was drawn trom life, depicts them as extremely irregular 
in shape, that is to say neither roundish nor squarish but with far more jagged outlines 
than are tound in any existing genet skin. This is explicable m that, if the drawing 
was truly made from a living annnal, the tur may well have been in a rough, wind- 
blown state, not brushed smooth as in a prepared specimen. This would argue a long, 
loose pelage; but it is difticult to make the markings ot any existing specimen ot genet 
take on the size, appearance and positioning ot the blotches depicted by Vosmaer's artist. 

The tail was described as ringed with black and white and having a black or brownish- 
black tip; but the number of rings was not stated, nor does the illustration make this 
clear, the only two, very narrow, pale rings visible, near the distal end, run at a slope 
such as is foimd in no genet. The dark tip is relatively short, narrow and pointed. 
The overall impression given is that ot a long-haired structure, matching the similar 
inference respecting the dorsal pelage made above. Other, more minor, points cannot 
be touched on here. 

Clearly there is very little that is reliable in the way ot characters to be squeezed 
out of the original diagnosis. When we come to consider what has been made ot 
lif^rina since by Matschie (1902), Thomas (1906, 1908), Schwarz (1930) and Roberts 
(195 1) we get a pretty mixed picture, sometimes in direct conflict with the type des- 
cription and/or its illustration. Each has attributed to the species characters of his own 
devising. Matschie, for example, defmitel)' stated that the hairs at the tail root were 
at least 30 cm {sic) long; that, despite the obvious taper in the illustration, it was about 
as broad at the base as at the tip; that the last quarter of its length was black with only 
traces ot narrow half-rings below; that the tirst two dark tail rings, neither visible 
in the illustration nor described, were at least as wide as the succeeding pale rings. 
This may be tigrina Matschie but it certain!}' is not tigrina as described by Schreber. 
Again, Thomas & Schwann (1906) stated, directly and indirectly in their key, that the 
whole midersurtace and forelimbs \\ ere dark-brown or black — which trom Schreber 
and Vosmaer was patently untrue; and, in tact, Thomas & Wroughton (1908) 
modified this to forefeet black — the original description ("feet with a lot ot brown") 
and the illustration belying this misleading assertion. Schwarz endo\\ed tigrina with a 
number of new characteristics: short legs, short face; pear-shaped bullae, strong post- 
orbital processes and a distinct sagittal crest — all very inconstant characters in Gencltti 
skulls. He stated the pelage to be variable but never shaggy, with large, mostly brown- 
centred spots; and the tail to have 8 to 9 dark rings. Other uncertainties and contra- 
dictions in diagnosis concern the niiier cusp on /)■'. 



20S TIIF. CARNIVOKFS OT WEST AIRUA 

The characters which most authors concur in tor n^i;n'»ci and whicii do not conflict 
with Schreber's and Vosmacr's descriptions, and which also accord with specimens 
trom the reputed region of provenance. South Africa, are: that the pelage is longish, 
with large, mostly independent spots in about 4 longitudinal rows, and a dark, probably 
erectile stripe on the lower halt of the back; the tail with about cS dark rings and a dark 
tip; the legs and feet with a good deal ot dark colour on them, especially the back ot 
the hindlcg contiguous to the heel, the hindfoot having at least some grey on its upper 
side. However, in view of the doubt and conflict of interpretation that ti^rina has 
engendered a good case could be made for the abandonment of the name. But whatever 
attitude may be adopted towards this, it seems clear that, in so far as it is possible to 
gather, the animal intended by Vosmaer and Schreber is probably more akin to the 
long-haired ficnctta group than to the relatively short- and soft-furred pcvdliui; and 
that as regards specimens from eastern Nigeria and Cameroun, as well as Allen's 
eastern Congo series, Schwarzs hypothesis about distribution is entirely at fault, 
these animals being certainly pardiiia. 

The question ot iiihi'^iiicsii Pucheran, iHa, is otten involved with that ot rif^riiid, 
as to whether they are either one and the same, the former a race of the latter, or 
specifically discrete. Though ot no tirst-hand concern to West Africa, the following, 
which emerged incidentally from the wide investigations necessarily made, may be 
added here as an appendix to the above. Pucheran's curt Latin diagnosis would appear 
to be too vague to be of any real use whatsoever; whitish-grey washed with tawny; 
the spots of the back almost totally rusty; the tail furnished at the base with four rusty 
rings then four black. That is all. It is too obscure, 111 the light of now known colour 
variability, to render riihi{iiiio<ci certainly determinable. It seems almost pointless to 
continue trying to tie such ill defined names to specific forms. Schwarz looked on 
nihi(;iiiosa as a synonym of ti\^riua, though Thomas in his 1908 key had — with what 
lustitication is surely open to question — devised his own distinction between them; 
that m fi(in>/(i the spots were large and the forefeet black, whereas in riihigiiwsa they 
were of medium size and the forefeet pale. 

Skull. The general form of this in the pardiua group is ver)- much as in other genets. 
Possible points of distinction, such as the existence ot an inner cusp on p-^, or ot a sagittal 
crest, are as variable as in other groups. One dittereiuial character alone is practically 
constant; in fully adult skulls the postorbital constriction is, with few cxcep>tions, less, 
and sometimes markedly less, than the interorbital breadth, this difference being on 
the average greater than it is m the scwaliua group. No significant variation of external 
appearance can be detected in exceptional examples. On the average the excess of 
interorbital over postorbital breadth is ot the order of 2 mm. At its greatest it may 
reach 5 or even 6 mm, the latter width dropping to as little as 8-0 mm, the cranial 
constriction becoming a very marked feature of the skull. One skull in which this 
occurs. No. 46.397, is a quite typical large-spotted f^cncttoidcs from Oda (Ghana); 
the other. No. 1645a from Gaboon, has no skin. 

Of the two adult skulls 111 which the intertemporal breadth is greater than the 
interorbital, one has an excess of oiiK' or mm. In the other. No. 27.8. 12. i (Ghana), 
the excess is 41 mm; ,ind there .ire iiileresting ditterences 111 the teeth. />'. )»' and m- 



GENETTA PARDINA 209 

being smaller than normal, especially the last which is only half the usual size and lacking 
from one side. A similar state of affairs with regard to iifi exists also in a Mamfe skull ; 
but neither skin is in any wav out of the usual run o^ ^cnettoides save that the Ghana 
animal was in full moult. 

Two other skulls have unusual features but both arc without skins by which to check 
their standing. No. 61.42, from Musaia (Sierra Leone) is narrower across the upper 
camassials than any other specimen and also has an appreciably smaller p^ and m'. 
And No. 68.493 has remarkably long, backwardly directed postorbital processes, 
measuring 27-1 mm across in contrast to a mean of 19-5 mm. These points are men- 
tioned since they have at one time or another in single skulls been regarded as having 
some taxonomic significance. There is no clear reason to suppose that they have; 
the material available for study is quite inadequate to provide sound data; and they 
are connected with no readily detectable external differences. Other characters which 
liave been used taxonomically yield, as regards British Museum specimens, the follow- 
ing data. An iinier cusp on p^ is of irregular occurrence and is independent of age or 
sex. hi adult skulls of the pardina group it is present and large in 9 specimens, small in 3, 
and lacking from 4. As for a sagittal crest: it is present in 8 old skulls, 3 of them ^, 
5 of unknown sex; and it is absent from 4 medium-age skulls {zSS, 2??). There are 
no adult female skulls from which to derive data. Finally, the nature of the postorbital 
processes has been suggested as of possible taxonomic significance. However, they 
show no fLxed form, in length, pointedness or direction, varying from short, blunt 
and deflected to sharp and long-pointed irrespective of other characters. 

The three West African species included in this species-group may be separated by 
the key on page 189. 

GENETTA PARDINA I. Geotiroy Parduie Genet 

Viverra maciilala Gra\', 1S30, Spicilegia Zoologka .... 2: 9, pi. 9. Type localit)- given as Nortli Africa 
but in tlie opinions of Matscfiie (1902) and Schwarz (1930) this should really be West Africa. This 
was described from a living animal in the Tower of London and there docs not appear to be any 
preserved type. The specific name is the Latin word for spotted. Preoccupied b\' I'ii'crra inaailaia 
Kerr, 1792 (= Dasyiirops iiiaailattis Kerr). See p. 205 herein. 

Gcnctia pardina I. Geoffroy, 1832, Mag. de Zoo}. 2iid year, class I, pi. S and accompanying text. Interior 
of Senegal. The name is a diminutive of the Latin pardiis lepoard. 

Cciietta pamherina Hamilton Smith, 1842, in Jardhie's Naliiralisl's Library: 35 (13 of Mammalia): 171 
No type locality given. This, from the Greek panther leopard, would appear to be nothing more than 
a renaming, or misnaming, of pardina \. Geoffroy resulting from the French name also given b\' 
Geoffroy at the head of his description, G.(cncttc) panthcrine. 

Distribution. As imdcrstood in this present work, and as defined above, pardina 
occurs in the drier open woodlands from Cameroun to Senegal. How far east it ranges 
in these belts has not been determined; but the present writer has little doubt that 
East African forms from as far afield as Tanzania, and usually nomenclaturally associated 
with tigrina, belong, in fact, to the pardina group, if not pardina itself. West African 
specimens ot pardina exist in the British Museum from the Wukrum Hills (Northern 
Nigeria. Sudan woodland) and Fort Lamy (Chad, Sahel woodland). 



::to iin: (aumvores or wrsr ai kh a 

Description. The scries ot K flat. p.irtK' incomplete skins trom the Wukriuii Hilis 
has alrcidy been referred to. No two skins are precisely alike in their colouring and 
marking. Six have the ground-colour a sort ot creamy- or sandy-buff; the seventh is 
rather grcvcr; and the eighth somewhat more whitish — though, nevertheless, still 
palely washed with sandy tips. There is an erectile spinal crest trom the shoulders to 
the tail, black in most cases but rusty in two. On each side ot this are four well defmed 
lines of spots, ranging, in different skins, between almost entirely black to almost 
entirely red-brown, some being obscurely black-ringed with rust\ centres. The sizes 
and shapes of these markings arc highly variable. In some skuis thev are wholly indepen- 
dent, in some coalesced longitudinally into short lines; some are roughly quadrate, 
some more oval, some rounded; and in one skin they are almost Imear, the width ot 
the blotches being only some 7 or S mm, as contrasted with twice this or more in other 
skins. The dense soft undcrfur is about 17 mm long, the tine-shafted bristle-hairs 
about 24 mm. 

The tails arc relatively short-, sott-haircd and exhibit dorsally 5 to 7 pale rings and a 
black end portion 75 to 150 mm long, this terminal part interrupted laterally and/or 
below by 2, occasionally 3, partial white rings. The legs and feet have been mutilated; 
but are pale rather than black, diough the inside of the hindlcg appears to be dark, 
at least in some cases. 

The Fort Lamy skin differs in being of a more yellowish-brown colour and a less 
well-detincd pattern of whoUv brown spots. The fulvous colour, which extends also to 
the rings of the tail and to the undcrsurtacc of the skin itself, seems as if it might be due 
to drying over a smoky fire. 

Skull. There are no skulls. 

Habits. No special notes regarding habits accompany the skins. There is no reason 
to suppose that open-countrj' genets differ materially 111 these from their closed-forest 
kindred except in so far as they probably more trequendy make use of rock crevices 
tor shelter and breeding than high-forest forms do; and ground birds, francolins and 
guinea-fowl, doubtless play a more prominent role in their dietary. 



GENETTA GENETTOIDES Tcmminck Pel's Genet 



I 'nrrr.i •jcnctioiilcs Tcmminck, 1S53, Esipiiacs zoologiipus sin In cote tie Ciiific: Sy. River lioutry and 
Hiiuina, Gh.ina. According tu ]cntink (1887 and 1S92) there are four r\'pes in the llijksmuseum van 
Natmirlijkc Historic, Leydcn: 2 mounted ++ with skulls separate, tVom the River Boutry; and 2 
mounted animals ot unknown sexes from Elmina. Tcmminck coined this name from Gcnclta .ind the 
(ircek termination -o-cidcs implying resemblance, trom tides lorm. 

(jcncttci ficldiaiia Du Chaillu, i860, Pioc. BpsIoii Sec. nat. Hist., 7: 302. hiterior ot Ciabooii. This was 
named in honour of his "distinguished follow citizen", Cyrus W. Field. 

( H-nctta duhia Matschie, 1902, V'crh. des V intanationaleii Zoologcn-Concircssts zii Berlin, igoi: 1141. West 
Alric.i, locality unknown. The Latin word diibia h.is several meanings, of which doubtful was that 
most likely in Matschic's mind in reference more to the species' uncerM'ii standing (p. 1137) than to 
Its indefmite provenance. 

Cmclta insiilaris Cabrera, iy2i, Bolii R. Soc. ap. Hist, iinl., 21: 261. Rcbola, Island ot Fern.indo Poo. 
rhc specific name is from the Latin iiisiiln island. 



Plate 4 




Pel's Genet (Gvm-tla_icncmmi,<): Small-spotted (ienct (G.wiu, i>oo,„>): Uen.n Genet (Cauv,, him) 



GENETTA GENETTOIDES 



Distriburion. This form (Plate 4) iiiliabits the closed-forest and neighbouring Guinea 
woodland, plentifully from Gaboon to Sierra Leone and thence, less commonly, to 
Gambia. The localities from which specimens exist in the British Museum are too 
numerous to cite; but, closed forest apart, it may be worth mentioning that the most 
inland places which have yielded specimens are Katsina Ala (Northern Nigeria), 
Musaia (Sierra Leone) and Fouta Jallon (Guinea). These Guinea zone specimens may 
possibly come from high-forest remnants or extensions ("fringing forest") rather 
than from the more open woodland itself; but on this point there appears to be no 
positive information. 

Description. As typically described this form comes from the Ghana coastal forests. 
One such skin in the British Museum, No. 57.9.16.3, is, in fact, from its history, almost 
certainly one of Pel's original specimens and therefore a paratype of Temminck's 
genettoides. Temminck wrote at some length of this genet, regarding it as very closely 
related to genetta itself He gave no clear account of the dorsal markings, laying his 
greatest emphasis on the nature of the pelage. From tliis and from his description of 
the tail — not at all bushy, the hairs being half as long as in gaictta, the annulation 
irregidar, the black rings wider than the white, the terminal third black with brown 
half-rings below — it is abundantly clear that he was wrong in his estimation of relation- 
ship and that the form without doubt belongs to the pardina group. 

It has already been explained that pattern and colour are very variable and that 
there seems to be no clear-cut division between this form and pociisis. In the most 
typical gencttoides the spots are fairly large, separate, oval or quadrate, either wholly 
black or with varying amoimts of ginger-brown hairs at the centre, making them to a 
greater or less degree annulate. In size, those nearest the black spinal stripe are 30 to 
40 mm long and 15 to 25 mm wide. They arc in 4 rows, with, often, a small part 
of a fifth; and there are about 5 of these large blotches in the innermost row between 
the shoulder and the hip. The legs may be either pale or blackish; but neither this nor 
the colour and size of die spots is regarded as having any real taxonomic significance. 
The hindfect are dark above with a more or less pronounced pale streak running along 
the inner edge from about the base of the 2nd digit. 

Between this and pocnsis at the other end of the scale, with small-sized spots too 
numerous to count, there are all manner of intermediate stages of size, colour, shape 
and independence. Some are deep black and remarkably clear-cut; other brownish 
and confused. These variations show no connexion with localities. Two corresponding 
to the deep-black clear-cut category come, the one from Sapoba (Benin) west of the 
Niger, the other from Ikot Mbo (Calabar) well to the east; and from Ejura (Ghana) 
come fairly large-spotted and very small-spotted skins. The last, however, is pretty 
young, and it is possible that its markings would increase in size with age. 

The pelage ingcncttoidcs is of medium length and fairly soft, and consists of abundant, 
very fme, rather sinuous underfur 12 to 14 mm long together with finely stalked, 
terminally broadened and flattened bristle-hairs measuring 19 to 21 mm. Ground- 
colour is usually of a mediumly-dark grey washed with pale brovwi; but one veiy- 
large-amiulate-spotted skin from a small island in the mouth of the Volta River near 
Ada (Ghana, strand and mangrove vegetation, of very low rainfall) is a pale whitisii- 



;i2 Tllr. ( AUNIVOKTS OI- WPST ArillCA 

grey, not dissiiiiiljr lu toiio tioiu opcn-countrv panliua itsL-lt but witliout this hitter's 
warmer creani-buft tint. 

Alh of course, have the typical jhirdiihi group tail, short-haired, snaooth, subcyhu- 
drical, and silky in texture. Tliis last is because the bristle-hairs, which are only some 
19 to 2.Z mm iu lengdi, are slender and taper to a very fme base. The abundant iindertiir 
IS about 9 to II mm long and plays a prominent role in determining the te.xture and 
shape ot the structure. All this, wliile vastly different from the extremely long, coarse 
bristles oi i^cncliii, is not luihke savalina; but m. ^cmttoidcs, as in others of the pardina 
group, black is the dominant colour since the pale rings (mostly wliitish) are generally 
much narrower than the dark and only completely encircle the tail in the basal halt. 
The black end portion is usually at least 75 mm in length and often much more, with 
2 or 3 halt-rings (or shorter) below, or sometimes with no trace whatsoever of white. 

Skull. The bulla is shown in fig. 2Sb. Notliing appears to distinguish the larger- 
spotted (jci/c/A'/f/c^' trom the smaller-spotted pocnsis. No comparison, cither of form or 
of size, can be made with open-coiuitry pardina itself since no skulls of diis last are 
available for study. 

Taxonomy. Sutticient has already been said regarding this in the introductory 
remarks to the group. 

Habits. There are no collectors' tield notes turnishing any usetul information 
peculiar 10 f^cncttoidcs. One specimen was trapped in a farm, one shot at night; and the 
species is said to be tairly common in Sierra Leone. The first of these collectors' remarks 
runs coimter to Sanderson (1940) who said that tins species in Camerouii (under the 
name G. lifirina fieldiana) inhabited low trees m both high deciduous forest and ram 
forest, occasionally entering old secondary forest, but never recently cleared land, 
i.e. "cultivated land and tertiary growth". Common observation also casts doubt on 
what Sanderson wrote, since these genets are tairly rehably known to come into open 
cultivated compounds and firms for the purpose of stealing fowls. 

GENETTA POENSIS Waterhouse Small-spotted Genet 

Ot-ncttti piH'iisis W-itcrliousc, I1S3S, Proi. zool. Soi. Loud.: 59. Island ot Fcinando P(')o; but doubt was 
cast oil this provenance by Pocock (njoSb). Type in the British Museum, No. ss. 12. 24. 41:!, sex ?; 
skin only, in fairly good condition, but legs incomplete. 

Description. The doubtful validity of this species (Plate 4) but its descriptive 
usefulness have already been brieH\' discussed above. It remains to enter into this matter 
a little more fully. 

The species /)oi';).w.s has suttered niiire tips and downs than possibly any other form ot 
(jcnctta. Waterhouse himself thought that his newly proposed species was nearest to 
pardithi: and Matschie (1930), possibly acting on this opinion and seemingly never 
having seen the skni or read the description with much care, s ynonymised the two. 
Pocock (1908b) was astonished at this as "it would be hard to fmd two more dis- 
similar species in the genus". This is very true, as reference to Geoffroy's plate and 
diagnosis abundantly shows, and if these alone are taken as the sole criteria o{ pardina. 



GENETTA POENSIS 2I3 

The markings ofpoetisis (Plate 4) are widely different from the large, squarish, indepen- 
dent blotches of fully typical pardina. They are very abundant, narrow, of much the 
same width as the spinal stripe or even less {i.e. rarely more than about 10 mm), the 
two rows on either side of this being largely united into similar almost black bands 
but in fact comprising some 8 or more spots between the shoulder and hip. Besides 
these partial stripes there are about four more longitudinal rows of roundish or oblong, 
more or less independent spots on each side. 

Spots apart, one of the first striking differences between the type skin of pocnsis 
and other specimens of the pardina group is its golden-yellow groiuid-colour. Indeed, 
with this colour, the abundant narrow spots and the forelegs being black, the skin at 
first sight closely recalls scrvalina, but is at once distinguished by its typical pardina 
tail with its lengthy black distal portion and indisrinct or partial pale rings. The authen- 
ticity of tliis colour, however, has been called into question. Pocock (1908b), acting 
on a hint from Oldficld Thomas, thought that the yellowish tone might well be due to 
drying over a smoky fire. It is now not possible to be sure. The only West African 
specimen in the British Museum at aU resembling it in its warm colour is the animal 
herein named hitii: But there is a marked difference in that the reddish colour of the 
latter is bright and hvcly whereas that o( the pocnsis type is dull, pervasive and seems 
very likely to be adventitious. This is, in a way, confirmed by other specimens closely 
conforming to the pocnsis pattern, Nos. 1938.7.25.5 and 57.12. 5.1, in which the ground- 
colour agrees well with that of the ordinary run of high-forest genets of the pardina 
group. The same may be said of No. 39.686 from west of the upper Cavally River in 
north-east Liberia. But this specimen, which was figured by Pocock (1908), has some- 
what aberrant features. In it the rows of spots adjacent to the spine are more con- 
current than usual and also in places join the crest itself so that the medial dorsal pattern 
is rather darker and more confused than in other examples. The fur of the back of the 
neck is not in good condition but it appears that the linear pattern is there not so clearlv 
defined as in the r)'pe and Kumasi specimens. Added to this is the fact that the tail is 
almost completely melanistic, the pale annulations consisting solely of 5 rather obscure, 
basal half-rings. Moreover it is very short, recalling what Matschie must have meant for 
duhia; but it is by no meaiis certain that it is all there. 

Pocock's scepticism regarding the identity oH poensis with pardina — or with ^c;)f«- 
oidcs, which he would at that time have regarded as pardina — is excusable. Yet, 
in the light of more abundant material than was available to Pocock, it seems to the 
present writer that the two may well be demonstrated in the future to be the same, 
pocnsis being at the small-spotted end of a pattern range. Nevertheless, until that time, 
and in the present state of doubt, it seems worth while retaining the name even if 
only for its succinct convenience in description. It nmst be noted, however, that were 
it to become established that transition from large to small spots was in fact of this 
nature, or that spot size was simply a matter of individual idiosyncrasy within a single 
species, then the name pocnsis would have to replace gcnettoidcs, which it antedates. 

Pocock was also inclined to doubt the island provenance o( pocnsis. In this, in view 
of the early date of its collection and the imprecise conception of West African geo- 
graphy at the time, he may well be justified. At any rate, nothing siirdlar has come to 



2\i THE CAUMVOIiFS 1)1 V^' 1 S T AFRICA 

light smcc troni rcniando Poo. One ot the skins referred to above as most closely 
conforiniiig to the ty'pc came from Ghana, the other is, unfortunately, also an early 
one and labelled merely "West Africa". It will be seen from the clinal nature of the 
paniliui forms that it must often be dirticult to assign smaller-spotted specimens un- 
hesitatingly to a specific form. Apart from the type and the other skins just mentioned. 
It would seem that a few specimens from the closed forests of Ghana, Western Nigeria, 
and possibly Cameroun, might be regarded as pccnsis. 

Skull. There is no type skull; but that belonging to the skin which, apart from 
colour, very closely agrees with Waterhouse's diagnosis o£ pociisis, No. 1938.7.25.5, 
collected at about 64 kilometres north of Kumasi (Ghana), is far larger than any other 
of the pardind group, being between 7 and 8 per cent larger in many measurements but 
with a toodirow rather more than 20 per cent longer than average. It is tempting to 
regard this as validly diagnostic and as confirming the specific distinctness oi. pocmis; 
but it would be very ill-advised to jump to conclusions on the strength of this single 
specimen. This is the more so as the only other adult skin closely resembling pociisis 
(No. 57. 12.5. i) has a skull that differs quite appreciably in form and is of average size 
wnth an average toothrow. 

Habits. Nothing whatsoever has been recorded of habits specially pertaining to 
piU-iisis. 

D. thierryi group 

The relationship to other genets of the species included imder this heading is obscure; 
but, for various reasons, it seems to merit separation from the species-groups so far 
dealt with. Since there is only one such species known it caiuiot logically be considered 
as part of a species group, and it is so dealt with here solelv for the purpose of main- 
taining uniformity of treatment. 

GENETTA THIERRYI Matschic Hausa Genet 

Gfm'/M thicriyi Matschic, iy02, Will, ties V intcriiatioiialcii ZooL-'i;cii-Ccii}<;rcsscs zn Berlin, tgor. 1142. 
Hinterland of Togo from latitude 9'N. onwards. Specimen No. 19015 in the Iiistitut fiir Spcziellc 
Zoologie und Znologisches Museum der Humboldt-Univcrsitat zu BerHn is herein, below on page 
215, designated as lectotvpe. There appears to be no record ot exactly who Thierrj' was, in honour 
of whom this was named; but since he was an Obcrleum.mt it is probable that he was an army officer 
posted to Togo. 

Psiudi.-'gi.ncua villiersi Dckeyser, 1949, Bull. Mis. natn. Hist. imi. Aims, (2) 21 : 421-424. Mcssirah, Senegal. 
Type in the Museum National d'Histoire Naturelle, Pans, No. C.G.i9s2, No. 4 (I.F.A.N. 48.5.32 
(453)), stuffed skin and skull; 6 paratypes in the Institiit Francais d'Afrique Noire, Dakar. Dr. Villiers 
.ifter whom this was named is a well-known West African collector and zooloaist. 



m 



Taxonomy. It is necessary to discuss this fust before any pronouncement can be 
ade respecting distribution and other matters. So ftr as published Hterature reveals 
there does not appear to have been any critical re-exannnation of Matschie's material 
since his erection of the species at the beginning of this century. There is therefore no 
first-hand description additional to the rather inadequate accoiuit which can be put 
together from his kevs. The characters there given vield the following: the spots are 



GENETTA THIERRYI 215 

small and very narrow, sometimes several united into a long stripe; colour is not 
mentioned, but there are 4 complete rows with at least 9 spots in the first row, and 
those of the tliird row further apart than their length. Both fore and liindfcct are 
pale. The tail is short-haired, the hairs not longer than 25 mm at the base; some up- 
standing hairs at the root; there are 10 pale rings, the last few visible only on the 
underside; the light rings at most as broad as the dark ones. 

By the kindness of Dr. Rcnate Angcrmann of the histitut fiir Spczielle Zoologie und 
Zoologischcs Museum dcr Humboldt-Universitat zu Berlin it has now been possible 
to study some of Thierr)''s specimens used by Matschie. There is no type o( thicrryi, 
nor did Matschie ever label any of these skins with names. According to Dr. Angermann 
(i)j lilt.) there is "a series of 10, all of them prepared in the same manner, all without 
skulls and without precise dates or locality ("Togo")". Of the 3 skins examined in 
London 2 belong, without question, to the pardina group; the third. No. 19015, agrees, 
as near as can be told, with Matscluc's key characters for thicrryi and also with several 
skins in the British Museum from the interior of Ghana — that is, from the area neigh- 
bouring that from which Matschie described tliis species. The present writer, and 
R. W. Hayman (1935) previously, had independently come to the conclusion that these 
British Muscimi specimens represented what Matschie intended by thkrryi, an opinion 
which has been strengthened to virtual certainty by comparison with the Berlin 
material. Dr. Angcrmann has \vritten that the majority of the series of 10, not sent for 
study, agree with specimen No. 19015, which is here designated as lectotype. 

Before proceeding further with this question it must be said here that it does not 
appear to the present author that there is any good reason to bcHeve that villiersi 
Dekeyser is materially different from thierryi, or that Pseudogenetta Dekeyser, of which 
genus villiersi was made the type species, merits separation from Ccnetta even as a 
subgenus. This latter opinion was held also by Kuhn (i960). 

Distribution. On this basis, specimens o{ thicrryi exist in the British Museum from 
Senegal (Bakcl and an unspecified locality). Sierra Leone (Rokupr), Ghana (Ejura, 
and Mole Reserve) and Nigeria (Wukrum Hills, north of Lau on the Bcnue). These 
provide 1 1 skins and 7 adult or fairly adult skulls. The Berlin Thierry skins are labelled 
as coming from Borugu, Togo, which would appear, from an old German colonial 
map, to be the equivalent of Borgu, io°46N, o°34E, about 30 kilometres north of 
Sansanne Mangu, just in the Sudan woodland zone. It is not known on what grounds 
Matschie gave the distribution of the species as Dahomey and Togo westwards; these 
specimens show that it ranges through the open woodlands of the Guinea, Sudan 
and Sahel zones from at least the upper Senegal River to the western side of Northern 
Nigeria. Further, and of considerable taxonomic interest, the last British Museum 
specimen cited above demonstrates that three species, sau'g<iletisis, pardina and thiirryi, 
are sympatric within a restricted area just north of the Benue. The Sierra Leone specimen 
is interesting in that Rokupr, though nominally in the closed-forest zone, is today in 
the Invasive Guinea woodland zone where very little forest now remains, and the 
species therefore is presumably a fairly recent immigrant. F. de Beaufort (1965) also 
records specimens (under the name Psciidogenettn villiersi) from Guinea, Ivory Coast, 
Dahomey and Cameroun. 



2i6 nil ( AUNiVDUfs oi vvi;sr airica 

Description. Tlicsc specimens tiuiiish the tollowiDg description ot thkrryi (I'Kitc 3). 
Tliis is a disrinctly smaller species than the others, as the measurements given in the 
table on page 219 show. The general impression is that of a pale, small-spotted genet. 
The actual tone and the spot colour vary not only from place to place — the Saliel 
woodland specimens are lighter than those from the Guinea belt — but also within a 
single loeahty. The Berlin specimen. No. 19015, has appreciably brighter red spots 
than the general rim of British JVluseum material, though the rather obscure markings 
of the aberrant No. 19. 7.7.3710 referred to below are almost the same bright colour. 
Speaking generally, then, the groiuid-colour is buftish, to a greater or lesser degree 
speckled with deep brown or black hair-tips. One ot the most obvious difterences 
from all the forms hitherto dealt with is the absence of a prominent black, and generally 
obviously erectile, spinal stripe. A medial stripe nevertheless exists; but it is of precisely 
the same brown colour as the rest of the dorsal markings and is very frequently itself 
narrowly split longitudinally down the middle by a line of paler hairs. It is narrow, 
.It most 10 mm wide, and on either side of it are 4 rows of spots which are no wider, 
and otten narrower, than this. Those in the rows adjacent to the medial stripe may be 
united into ahnost conrinuous lines, or they may be ahnost entirely separate. Those 
of the next row are sometimes partly coalesced; but in the two outer rows they are 
always quite independent and often slightly broader and rounder than in the inner 
series. Over the back of die neck these rows are continued as much narrower hues or 
series of spots, sometimes very clear, but sometimes eventually becoming confused 
anteriorly. They are not always easy to count but, in broad terms, there are somcrfiing 
like 10 spots in the iimer row between shoulder and hip. The colour of the markings, 
while uniform throughout a given example, may be deep brown or ginger in different 
specimens from a single locaht)'. The pelage is fairly short, that is to say, the bristle- 
hairs about 15 to 20 mm long, the undcrfur about 12 mm. There are spots on the 
upper thigh but the rest of the hindleg and all the foreleg is pale buft, or even becoming 
whitish towards the feet. 

The tail is somcdiing between (jc/u'/Ai and parJiii.i. It has the long blackish terminal 
section with pale half-rings of the latter: but, though it is a very much narrower 
structure, it has, particularly near its base, something ot the longer-, backwardly- 
directed-haired appearance ot the former, and with its less clear-cut boundaries to the 
rings as well. Its bristles are about 25 to 28 mm long, die underfur 15 to 17 mm. There 
are 8 or 9 pale rings and half-rings, and occasionally some very slight indication of a 
tenth. Though all tails conform to this general pattern they do vary considerably in 
the clarity or obscurit)' of their annulation, in the length of their hair, especially basall)', 
and m the colour of the dark rings, winch varies trom blackish to i_)range-bro\vn. 

One of the two Bakel skins. No. 19. 7.7. 3710, while obviously lliicrryi, is exceptional. 
The medial pattern of spinal line and contiguous rows is rather confused; and die 
remainder of the spots are even smaller than normal. The basal portiiin of die tail is 
very long-haired, almost of a (jc/;cWii character. 

Skull. Apart from its markedly smaller size this corresponds closely to the general 
pattern of Cciictta. There is no old male skull amongst the British Museum material 
and this may account for any but the most vestigial indie.ition of a sagittal crest, except. 



GENETTA JOIINSTONI 217 

of course, in the extreme posterior of the cranium where there is, as throughout the 
genus, a short, sharp flange forming a T with the pronounced supraoccipital crest. 
Without exception ui the 6 skulls examined, the intertemporal constriction is some 
2 mm or more broader than the interorbital breadth. There is a clear, and usually 
large, internal cusp on p^. The molars, both upper and lower, are smaller than usual, 
II fi being particularly reduced; and 1112 is for the most part subtriangular with only 
rarely more than 3 cusps. 

Habits. Nothing particular is known of the habits o( thienyi. However, since it is 
an inhabitant of the open woodlands where for the most part the trees are of low 
stature and their crowns do not afford a great deal of visual protection, it is possible 
that tliis species more commonly breeds in holes in the ground or amongst rocks 
than the closed-forest forms do. 

Subgenus PARAGENETTA Kuhn, i960 
KiJin's Genets 

Some of the main characters which serve to distinguish the animal covered by this 
subgenus from those included in Gcnctta saisu striclo have been given briefly in the key 
on page 1 89 : and as there is only a single species at present known in Pariii^cnclta all 
further information relative to the subgenus can be gathered from the account of 
G. johnswiii which follows. 

GENETTA JOHNSTONI Pocock Johnston's Genet 

Geiwtta johiiiUmi Pocock, 190S, Proc. zool. Soc. Loud, for 1907: 1041, pi. 54, f. i & 2. About 25 to 30 
kilometres west of ttie Putu Mountains, eastern Liberia. Type in the Britisfi Museum, No. 8.8.23.1, 
sex?; flat skin only, in good condition apart from lacking head and forepaws. This was named in 
honour of Sir Harpi' Johnston, a well-known explorer and collector, who wrote a detailed account 
of Liberia. 

Gcnella [Paragcnetla) lelinianni Kulm, i960, Saiigetierl:. Mitt. 8: 154-160, t. 1-12. Kpeaple, Liberia. Type 
in the Museum Alexander Koenig, Bonn, No. 57.11, sex unknown; adult skull only. This was named 
after Dr. Ernst von Lehmami of the Museum Alexander Koenig. 

General. This species has had an erratic liistory. It was first described from "flat, 
native-prepared headless skins", with no skulls, collected about 1907 by Mr. Leonard 
Leighton. There were originally five specimens (Pocock, 1908b: 1038) but only the 
type and one other exist in the British Museum. The former bears a Zoological Society 
of London label but was registered, in 1908, as having been presented to the Museum 
by the collector. The second did not come to the Museum until 1930, when it was 
registered as presented by the Zoological Society but bears a label in Pocock's hand- 
writing as die paratype figured by him alongisde die r)'pe in his original description. 

The new species was thought to have nothing special about it apart from its slightly 
diflcrcnt pattern of spots and tail. Pocock, indeed, thought that it might, with more 
complete data, prove to be notliing more than a subspecies of pardina. No further 
material, however, was forthcoming, and there the matter rested until over half .1 



IC 



;i8 THE CARNIVORES OF WEST AFRICA 

century later a genet skull from Liberia, without a skin, came to the notice of Kuhn 
The cranial, and especially dental, characters of this were sufficiently different from tl 
general run of genets to merit, in his opinion, not only the erection of a new species, 
klwiiiiwi, but its assignment as well to a new subgenus of its own. No connexion with 
Pocock's johnslciii was suspected until, at last, a complete specimen of skin and skull, 
demonstrating the svnonymy of the two, was collected by Kulin near Tappita (Liberia) 
in the early 1960s. The British Museum still has no skull assignable to this species. 

Distribution. Gcnctta [Paraficmtia) johnstoni is now known by specimens from the 
following Liberian locahties: west of the Putu Mountains, Kpeaple, Bo, Tappita, 
forest south of Freemantown (Kuhn, 1965). These arc all in the western part of the 
countrv-, well inland from the coast in dense forest; but it is not known whether the 
species is restricted to this locality or ranges further east and west. It does not occupy 
the area to the exclusion of other species since Pocock (1908b) rccoids pocnsis as forming 
part of the collection obtained there by Leighton. There is no information as to the 
degree of commoness o{ johnstoni. The small number of specimens known might seem 
to argue rarirv^; but this region is relatively poorly explored zoologically, and the 
fact that the original collector obtained no less than five skins indicates that such a 
deduction might well be erroneous. 

Description. The type and paratype are of slightly different appearance, the latter 
being a shade paler, its spots appreciably less concurrent. The foUowing description is 
basically that of the type. The pelage is dense, soft and not very long (bristles 20 mm, 
underfur i •; mm). The ground-colour is yellowish-brown. There is a black, moderately 
long-haired and almost certainly erectile spinal crest, some 10 to 12 mm wide, from 
well short of the shoulders to the root of the tail. On cither side of this are four fairly 
clear rows of spots of approximately the same width as the spinal stripe and short 
confused portions of a smaller-spotted fifth and sixth. These spots are blackish-brown 
more or less heavily speckled with red-brown, so that in the paratype the latter colour 
dominates, the contrast between the spots and spinal line bemg considerably more 
marked than in the type itself In this latter the furst two rows on either side of the 
black medial stripe are almost wholly coalesced into complete lines. Because of this it 
is not really possible to count the number of spots in the furst row accurately; but there 
would appear to be basically something in the nature of 8 or 9 between the shoulder 
and hip. These series continue forward over the back of the neck in the usual genet 
fashion, either (paratype) as independent spots or (type) united into fme lines. The belly 
IS whitish. The thighs are heavily spotted; but in the type the remainder of the hindlegs 
and feet are deep blackish-brown. The sides of the neck are also well spotted; and 
what remains of the forelegs shows them to be so too; but in the type the markings 
arc obscured by a blackish sufliusion similar to, but not so intense as, that of the hind- 
legs. 

The tail resembles that of the servalina group more than any other. It is long, bub- 
cylindrical, of a soft and furry nature, that is to say densely haired, the hairs not markedly 
long, with the underfur playing as prominent a role as the not very much longer 
bristles — in this respect completely different from the (icnctia group. The amiulation, 
also, recalls that of frrrnlina in its relative clarit)- and its continuation, though less 



CENETTA — MEASUKEMENTS 



219 



Table 10: Numerical data for species of Ccnctia 





^t'uctta 










jolmsloni 




{? "H 


snicgalciisis 


(^encttoidi-s 


ciistala 


thk-nyi 


(from 
Kuhn) 


Vegetation 


Sahel? 


Doka— 
Subdcsert 


Forest and 
Guinea 


Forest 


Guinea, 
Sudan, 
Sahel 


Forest 


Number in mean 


I 


5 


9 


4 


6 


2 


Condylobasal length 


92-9 


84-8 


93-4 


93-4 


75-1 


93-0 


Basilar length 


(85-5) 


79-0 


85-9 


86-0 


69-5 


— 


Palatilar length 


43-0 


38-1 


42-5 


427 


32-7 


— 


Zygomatic breadtli 


46-3 


43-5 


46-7 


45-6 


38-5 


44-2 


Upper cheekteeth breadth 


27-S 


26-4 


28-0 


27-8 


24-0 


24-4 


Interorbital breadth 


12-3 


12-3 


13-6 


13-3 


10-7 


13-4 


Postorbital constriction 


13-6 


13-5 


II-5 


12-6 


12-8 


13-0 


Braincase breadth 


29-3 


28-8 


30-8 


30-9 


28-5 


— 


Toothrow (c — hi-) 


36-3 


33-8 


3 5 '5 


36-7 


28-3 


— 


p^ length 


8-6 


7-9 


8-6 


8-6 


rt-9 


6-6 


M|l breadth 


S-6 


7-2 


7-8 


7-7 


6-5 


5-6 


iifi breadth 


4-9 


4-^ 


4-5 


4-5 


2-9 


2-3 


(Ml length 


7-5 


7-0 


7-C, 


7-5 


(-,■3 


5-8 


i«2 length 


3-8 


3-7 


4-1 


3-9 


3-1 


3-1 


Head & bodv 


(490) 


438 


503 


540 


390 


— 


Tail 


(300?) 


445 


427 


425 


385 


— 


Hindfoot 




78 


80 


90 


73 


— 


Ear 


— 


44 


40 


46 


38 


— 


RATIOS (per cent) 














Tail/head & bod\- 


(60?) 


103 


85 


79 


99 


— 


Zygom. br./cond)lob. 1. 


50 


51 


50 


49 


51 


48 


Braincase/condylob. I. 


32 


34 


33 


33 


38 


— 


Braincase/zygom. br. 


63 


66 


66 


68 


74 


— 


Palatilar 1,/condylob. 1. 


46 


45 


46 


46 


44 


— 


Interorb./postorb. 


91 


91 


iiS 


105 


84 


103 


pi/c— "1- 


^3-7 


^3-4 


23-4 


24-2 


24-5 


— 


mijmi + '"3 


197 


1 89 


185 


192 


203 


187 



clearly than 111 that group, practically to the tip. Though in the distal portion the last 
two or three pale rings are certainly more obscure than proximally, they are neverthe- 
less clearly observable, above as well as below, as contrasted with the almost complete 
terminal dorsal blackening of the paydiiui group. In the two study specimens available 
it is difficult to know how much, if any, of the tail tip is missing. However, the tip 
would appear to be fundamentally one of the pale scries; but in the type it is very 
heavily dusted with brown though still distinguishable from the penultimate, dark, 
ring. In the paratype, on the other hand, only a very small area of white is detectable, 
and the tip would be classified as black. There are 8 black rings, sometimes equal to, 
sometimes broader than the white. 

Skull. In the absence of any skull from the British Museum the following is founded 
on Kuhn (i960). The species, and hence subgenus, differs from Gciu-thi by the smalLncss 



THE CARN'IVORES OF WEST AFRICA 



ot the tcctli, onlv the incisors being of equivalent size. Tlie upper incisors are soniewlrat 
tilted backwards, but the lower ones are more horizontal in the mandible than in 
Gok'tfii, imparting the appearance of many Insecrivores. Also the lower canines are 
more forwardly inclined in the jaw, though bent upwards in the middle. Because ot 
this the chin is ver)- flat. The upper canines show no sign of the fme longitudinal 
fiuTows generally present on the outer face in Gmctta ; they arc more slender and more 
backwardly curved. The first premolar is just as high as in Gaictta, only somewhat 
narrower. All the remainijig premolars are much lower and narrower than in Gcnctta; 
the imier cusp on /'^ is only lunted at. The fourth upper premolar is altogether much 
smaller; nfl is tiny. 

The zygomatic arch is narrower than in Gcnctta; the supraorbital processes are 
strong; and a sagittal crest is absent except posteriorly. 

Taxonomy. From the nature of the dorsal pattern one would be tempted, as 
I'ocock was, to relate this species to the pardina group; from the tail the comrexion 
would seem to be rather more with that oi scrvalina. But the teeth clearly show it to 
stand apart from both of these and probably, also, justify the retention of the subgenus. 

Habits. Nothing at all is recorded of these. 

Table lo shows mean measurements derived from such material as exists in the 
British Museum, ox{johnstoni) from recently published figures. 

Genus POIANA Gray, 1865 
African Linsangs 

Poiana Gray, 1865, Proc. zool. Soc. Lond. (for 1S64); 520. Type species Gcitctia rkhardsmii Thomson. 
This name is generally thought to have been derived from tlic second halt of the name ot the type 
locality, Fernando Poo, often in English spelt Po; recently de Beaufort (1965) has suggested derivation 
iVom the vernacular name oyaii; but Gray had coined the name 33 years betore G. L. Bates's Benito 
River specimens first recording this naine had come to notice. 

General. The true linsangs are Asiatic, the name itself being a vernacular one for 
these animals used in parts of the East Indies. Linsain^ was also used (Miiller, 1838) 
as the scientific name of these oriental forms but has now been replaced in part by the 
carher Prionodon Horsfield, 1824, and in part by Pardictis Thomas, 1925. There is a 
degree of superficial resemblance, external and cranial betrwcen these last animals, 
from Nepal and Tonkin, and those from Africa; and despite the lack of any existing 
link it has often been thought that the relationship between these widely separated 
forms is close. This is exanuned in greater detail later; but it may be said here that 
there are certain clear differences which caused Simpson (1945) to place the Asiatic 
Linsangs in a separate tribe, the Priouodontini, the Afi.-ican forms remaining widr 
Gcnctta, Naiulinia, Civcttictis and others of extralimital concern in the Viverrini. It can 
be argued from this and the fact that the name is strictly speaking East Indian that it is 
inappropriate to call the African animals linsangs; and, indeed, the French use the name 
poianc. But it would be ditflcult now to substitute this or any other specially coined 
English name for the generally accepted term African linsangs, which has the merit ot 
indicating that other possibly related forms exist elsewhere. 



POIANA 



Distribution. African linsangs occur only in tropical Africa within the forest belt, 
on the West Coast between Liberia and Gaboon, thence extending inland to the 
eastern half of the Congolese Republic from its extreme north-eastern corner (J. A. 
Allen, 1924) to about as far as 4° South (Sclioutcden, 1948). They are known also 
from the island of Fernando Poo. Whether the distribution is continuous throughout 
this range is quite another matter. To judge from museum material linsangs arc very 
rare and probably extremely local in occurrence. This may not, of course, reflect the 
true position on the gromid; chance or an especial cumiing m the avoidance of traps 
or of exposure may accoimt for the paucity of known specimens. Nevertheless, this 
seems unlikely and these animals can be pretrv' safely reckoned as some of the rarest 
in Africa. The recorded places of collection are listed below in the two specific accoimts. 

Description. African linsangs are in appearance ver)' like small genets, to which, 
in fact, they arc nearly related; but their long bodies, clothed with spotted pelage, 
are yet more slender and lithe than in Gcnctra, their legs even shorter, and their ringed 
tails relatively longer. The head is small and the face pointed, the prominent ears 
upstanding and oval and with a fairly large bursa, the posterior flap arising above beliind 
the pimia. The two West African species differ in their coloration but the general 
impression is that of a paler or darker brownish animal hberally marked with black 
spots which are rounded or oval and always small or very small. These are not quite 
so clearly disposed in regular longitudinal lines as in the genets, but about 4 or 5 rows 
can be detected on each side of the body. There may or may not be a narrow spinal 
line. The back of the neck is marked with three, or sometimes foiu% black lines or 
linear series of spots, which posteriorly are continuous with the dorsal markings. 

The dorsal pelage is soft and ver)' short but dense, comprising abundant fuie underfur 
into which are mixed not very much longer bristle-hairs, flattish and shghtly expanded 
distally but with long slender stalks scarcely distinguishable, even imder magnification, 
from the fine underfur. The underparts of the body are whitish or creamy, without 
any markings. 

The short legs are spotted and terminate m 5-toed tect which are clad, and especially 
on the sides and below, with dense, very short hair giving them a velvety touch and 
appearance. The sharp claws are not quite completely retractile. The soles of the feet 
carry the usual naked pads, the subdigital ones ver)' small and rounded: the central 
pad on the forefoot consists of four anterior and lateral pads together with a partially 
divided posterior pad, the space enclosed by these being entirely bare or only ver)' 
shghtly hairy. On the hindfoot the central pad is composed of four distinct pads; but 
this foot is chiefly remarkable for its medial, narrow, naked pad, di\dded anteriorly, 
and extendhig nearly to the heel. It is the presence of this pad that is one of the chief 
differences between the African linsangs and the wholly hairy metatarsus of the oriental 
ones. 

The tail in Poiiiiiu, as in Gciictta, is a very distinctive structure. It is narrow, more 
truly cylindrical from root to tip than that of any genet, and verv^ long. It is, indeed, 
ften stated, as a fu-mly diagnostic character, to be longer than the head & bod)-; 
but this may not always be so. The measurements given by G. L. Bates, an experienced 
and extremely painstaking collector, show the tail to be the longer in only half of the 



o: 



222 Till-. CARMVOltl-S OI WLST AIUICA 

t> sp(.\'iiiicus ot riihdiJsoni collected b)' hini and nicasui'cd in the held, the mean length 
ot all 6 specimens working out at 95 per cent of the mean head tV body measurement. 
It is true that this does not accord with the figures given by J. A. Allen (1924) where 
in all of 4 specimens, 3 of diem |uveiiilcs, the tail exceeds the head &: body. The greatest 
excess of tail over body, 52 mm, m the Bates specimens occurs in his youngest animal, 
a subadult: it is also a subadult that exhibits the largest difference, 53 mm, in Allen's 
Congo series. The tail is clad with dense, more or less erect, iinderfur and only ver)' 
slightly longer, soft, bristle-hairs. There are from to to 14 dark rings according to 
species, somewhat narrower than the intervening pale ones, which mav sometimes by 
faintly divided bv very narrow accessory dark rings. 

There appears to be no pubhshed account of the detmite e.visteiice or absence of 
scent glands; and no Uve or spirit material has ever existed in London from which this 
question could be determined. But Pocock, who, was intensely interested in this 
matter, said in a footnote (1933 : 970) that he h.id found I'vidence on made-up skins of 
their existence in Poiivui. 

Skull (tig. 30). Since no skulls ot lfii;liioiti exist in the British Museum the tollowmg 
description applies more especially to richardsoiii. Apart from its considerably smaller 
size the Pciiiiid skull differs most obviously from that oiGviutt.i 111 its lesser development 
ot the posterior region. Although an occipital crest does exist it is relatively incon- 
spicuous, often httle more than a ridge, and there is no very marked excavation of the 
posterior braincase to rise again, as 111 Go/cffj, to elevated occipital flanges. Nor is 
there any short posterior sagittal crest, as throughout that genus. It may be mentioned 
here that in these matters the oriental Unsangs differ from the African, their skulls 
more closely agreeing with those of the genets. The braincase is smooth and ovoid; 
the postorbital processes may be sharply pointed, though never lengthy, or the\- 
maybe poorly developed and blunt; and in 4 of the 7 London skulls this region is 
fenestrated. The postorbital constriction is wider than the interorbital breadth, the 
difference averaging rather less than 2 mm. The rostrum is narrow and pointed. The 
zygomatic arch is strong and in most cases has a will-developed, pointed, jugal process; 
but this appears to be a question of age since this section of the circumorbital ring is 
more or less absent from the subadult skull. The bullae are tairly large, the posterior 
portion far exceeding the anterior part, the two being conspicuously divided by a waist; 
the meatus is large. The post-dental palate is relatively longer and narrower than in 
(jciuriti. and the notches that separate it from the mam palate shallower and rather less 
obvious than in that genus. 

The dentition normally differs from that ot Gciictrii m having one less upper molar, 
j-y-jrs — 3N; but III- may be occasionally present, though minute, as in B.M. No. 
I. II. 21. 7, a medium-aged female; iii^ varies a good deal in the London specimens but 
is on the whole more reduced than in Gaictia. In the iiiandibL-, which is relativeK 
weak and flat, ;»l> is minute and seems, like its upper counterpart, to be on the way out. 
The canines ot both jaws have niinute longitudinal furrows along their outer faces, 
.1 character the\- sliare with Gcmtui aiiil Wiinlliil,:. and one so unusual th.it it must 
assuredly indie.ilc phylogenetic atfinitv. 



Habits. Exceedingly little is known of the linsangs iii life. Their apparent rarir\' 
and purely local occurrence, their secretive and almost certainly nocturnal habits 
have prevented their way of life from being observed save by a hmited number of 
African hunters who, in the way of things, have never recorded what they know. 
African linsangs, at least, seem never to have been kept in captivity, either in zoos or, 
more locally, as domestic pets. Further, it is possible that more specimens have not 
come to light because of some special value traditionally attached to the pelt. Over a 




Fig. 30. Poiaiia richardsoiii: skull, B.M. No. yS. 3. 19.11, q, x i 



century ago Allen & Thomson (1848) recorded: "This is a rather scarce animal at 
Fernando Po, and as the skin is considered one of the most s.icred and valuable amulets 
or charms of the Edeeyalis, they are unwilling to part with it. The small specimen in 
the British Museum, presented by Dr. Thomson, was skinned from the mouth, and is a 
proof of the mgenuity of that singular people". The special significance of these rare 
skins noted by these authors in Fernando Poo seems to be reflected also in Liberia 
where, according to Dr. Hans-Jiirg Kuhn in observations referred to below, the 



^;.l Tlin CARNIVORKS OF WEST ATRICA 

pi'lts arc usfd to make mcclicuic bags, that is to say containers tor materials regarded as 
possessmg exceptional properties aiid worth. However, the mystical value of the 
Pciiin.i skin doubtless varies with different individuals or may have suffered a decline, 
since Seimimd, responsible for the greater part of British Museum collecting in Fer- 
nando Poo, recorded that a t.iil and fragments of a pelt were "got m exchange for 
rum '. 

Habit notes boil down at present to two collectors. Hates c\ Kuhn. The former 
(190s), who characterized Poiaiui richardsoni in Cameroun as a rather rare little beast, 
said that it was foimd only in the forest, sleeping in the daytime on thick tangled vines, 
and walking (? waking) onl)- when disturbed. A female brought to him in October 
had miUc in two teats; and he was told that these animals produce two yoimg at a 
birth. More recently, Kuhn, in an unpublished commimication quoted in Walker 
(1964), as a result of intensive mammal collection and study in Liberia, has given the 
most abundant information so far available. In this he states that the nests are round, 
at least z metres and usually more from the ground, and ot green material, several 
animals sleeping there for a tew days and then moving on to construct a fresh nest. 
Further, although these animals have been recorded as using abandoned squirrel dreys 
as shelters, he was informed by reliable Airican hunters that the reverse was the case, 
squirrels taking over abandoned linsang nests. Kuhn says that the diet includes cola- 
luits, insects, young birds and plant material. Doubtless it also covers small rodents 
.md possibly reptiles as in the genets. This is die sum total of our present knowledge. 

Taxonomy. It has often been assumed that the African Imsangs, Poiana, are closely 
related to the Asiatic Imsangs of the genera Piionodon Horsfield and Pcirdictis Thomas 
in spite ot being ver)' widely geographically separated with no known connecting 
forms. Both groups are tropical rain-forest dwellers; and Misoime (1963 and 1965) 
has shown that the existence ot any forest bridge between Africa and Asia later than the 
Oligocene is improbable. The likelihood of either group's being an offshoot ot the 
other, or both the relatively recent descendants of some common linsang stock is 
remote; and the affinity between the African and Far Eastern genera is, in fact, con- 
siderably more distant than often supposed. Poidita is without doubt more closely 
akin to Gciutta than to Piioiwdoii and Pardictis. This is in some measure expressed in 
Simpson (1945) by the allocation ot the two groups of linsangs to different tribes, 
the Viverrmi and the I'rionodontim. 

Resemblances which on the surface appear striking are on closer exannnation seen 
to be not so ex.ict, and are assuredly fortuitous, the result rather of convergence than ot 
artinirv'. The position cannot be more than sunnnarily glanced at here. Any corres- 
pondence between the skulls is no more than frequently occurs ui this family; and, 
111 tact, the supraoecipital region and the zygoma are of diverse forms in the t\vo 
groups. Loss of the posterior molars, giving rise to similarity of dental foriiuilae, is ot 
common e\olutionary occurrenee and of no overriding significance. Against this 
numerical consonance, the forms of /ii and ot iiio ;ive distinctly different. Li the Asiatic 
Imsangs b>>th of these teeth are somewhat more complex, p^, when not too worn, 
Iviiig cle.irK tncuspidate; while ///a, in them, is fir more laterally compressed and more 
sharply, luieveiily and line.illy cuspid. ite tli.in the lallier sqii.it qiiadi.ilc or ti langiilai 



POIANA 22.S 

tooth of Poiaiiii, bearing in side aspect a passable resemblance to a small premolar. 
Finally, and quite significantly in a rare though slight character, the canines of the 
oriental species show no trace of the fane furrows on their external faces so typical of 
Poiana and Gcnetta. These cranial and dental disconformities are supported by external 
ones of pattern, pelage and feet, that make it clear that close aft'uiity between the 
African and Asiatic linsangs is more apparent than real. As Pocock (1908) indicated, 
a character that would go far towards clinching the question of relationship, but 
about which no information was available, would be the presence or absence of perineal 
scent glands in Poiana since it is known tliat they are lacking from the Asiatic linsangs. 
A quarter of a centmy after pointing this out Pocock (1933: 970 f.n.) had convinced 
himself, by the re-examination of the same dried skins as were previously available to 
him, that there was, in fact, evidence in them of the existence of such organs; and he 
therefore concluded that the affinities of Poiana quite defmitely lay with Gemrta and 
not with Prionodon. 

Three forms of Poiana have been described: richardsoiii, ochracca and ki^htoni, the 
two last as subspecies ot the first. No study skulls of these two exist, at least m London; 
but from external characters it seems probable that whereas ochracea is merely a colour 
variant of richardsoni, leigjitoni merits specific status. Only two of the three forms 
occur in West Africa, ochracca Thomas & Wroughton, 1907, being described from the 
Aruwimi River, eastern Congo. The two relevant species may be differentiated in the 
follo^^^ng way. 

KEY TO THE SPECIES OF POIANA 
(previous key page 166) 

Dorsal ground-colour dull reddish-browii; belly off-white; no continuous dark 

spijial stripe; rmgs of the tail parallel-sided richardsoni [patic 22s) 

Dorsal ground-colour bright bufi^'; belly pure white; a narrow, more or less 
continuous dark spinal line present; dark rings of the tail rather chevron- 
shaped. . leightoni (panic 227) 

POIANA RICHARDSONI (Thomson) Richardson's Linsang 

Genelta ricliardsoiiii Thomson, 1842, Ami. Mag. nat. Hist, (i) 10: 204. Fernando Poo. Type in the British 
Museum, No. 42. 10.18. 1, sex unknown; unmounted young skin, in poor condition; no skull. This 
was named in honour of Dr. John Richardson, Inspector of the Naval Hospital at Haslar. Thomson, 
whose name was consistently misspelt Thompson by Gray in synonymies, was one of the medical 
ofiicers to the 1841 expedition to explore the Niger and co-author of the published account. A note 
in Gray's handwriting in the British Museum Register for 1842 indicates that it was he who actually 
prepared the description "for Mr. Thompson". 

Distribution and general. This is the better recorded of the two African linsangs, 
the British Museum possessing 8 skins, one of which is juvenile, one fragmentar)% 
and 7 skulls. These all come from the Bight of Biafra, that is to say more specificallv. 



226 THE CAIiXlVDRVS Ol- WFST AFRICA 

the island ot Fernando I'oo, Benito Jdwr m Spanish Ciuinca, and Bityc in lower 
Camcroim, G. L. Bates being responsible for all the 6 mainland, and only perfect 
adult, specimens. This area, of course lies wholly within the rain-forest belt. The 
skin referred to by Pocock (1908b) as being labelled from Sierra Leone has iiot been 
traced, nor any evidence fouird of its existence. However, one of the skulls. No. iof<4a, 
was said to have come from Sierra Leone, the relevant note being in Gray's own 
handwriting on the page so numbered of his draft catalogue but imder an 1854 Register 
niunber that does not, in fact, exist. But, though purchased from a third party, it was 
also stated to have been one of Eraser's collection — and, thus, with little doubt from 
Fernando Poo. It bears a label to this last effect; and Gray himself must have come to 
this decision since he quoted this skull m Gcrrard (1S62) and illustrated it, reversed left 
to right. (1865a and 1869), making no mention at all of Sierra Leone as one oi^ Pciivui's 
possible habitats. It does, hideed, seem highly luilikely that richardsoni docs, or ever 
did, exist in Sierra Leone. PeiTct iv Acllen (1956) obtained a specimen from Foulassi, 
Cameroun, which is not fir distant from Bitye; and Auerbach (1913) recorded one 
from Yaoimde, also Cameroun but somewhat further north. 

Description. Richardson's linsang (Plate 5) appears to be the larger of the two West 
African species — though no measurements of Icigbtoni seem ever to have been pub- 
lished; but it is still smaller than any known genet, to which animals it bears some 
considerable superficial resemblance. Fiowcvcr, it has a yet more slender body; and it 
is very readily recognisable by the complete absence of any continuous spinal stripe. 
Its small, roimd or oval, fully independent black spots, and its very long, absolutely 
c\lindrical, amiulated tail carrying 12 to 14 dark, parallel-sided rings. 

The pelage of Richardson's linsang is short and soft, consisting of abundant, relatively 
straight underfur, 10 to 11 mm long, with which are mixed comparatively few, only 
slightly longer (13 to 15 mm) bristle-hairs. These are expanded subterminally into a 
narrow, flattish blade but taper proxinially to a very long stalk almost as fuie as the 
underfur. The base of all the pelage is medium grey, the visible pattern of the back 
being produced by either brown or black distal portions of the hairs. The ground- 
colour of this dorsal pattern is a sort of dull orangey-brown on which is superimposed a 
pattern of black spots. Down the spine is a discontinuous series of narrow, and hence 
rather linear, spots, on cither side of which are about 4 rows of roundish or oval spots, 
well separated from each other. The back of the neck carries three, sometimes four, 
more or less parallel longitudinal black marks; the medial one (or two) narrow, some- 
times a discontinuous series of spots, and on cither side of tliis a far bolder, continuous 
band. The undcrparts arc creamy or off-white, though the exact colour in hfe is difficult 
to determine from old and hence rather soiled specimens. 

The head is small, the ears big and rounded, the upper lips pale. The proximal parts 
of both fore and hind limbs are spotted; the feet much the same colour as the back or a 
little more sepia. The outside edge of the hindf'oot, in all but one specimen, is deep 
blackish-brown. The long tail is remarkable in its narrow, almost perfectly cylindrical 
nature and the regularit)- and clarity of its parallel-sided annulation from root to tip. 
The colour of the light rings above is precisely that of the ground-colour of the dorsum ; 
but below IS rather paler. The dark rings are deep-brown rather than black, that is to 





Richardson's Linsang {Poiaiia richardsom); Two-spotted Palm Civet [Kaiidinia Iniioiata) 



r o I A N A 227 

say, noticeably paler than the dorsal niaculation. hi a number of specmiens, but not 
all, a few, nuich narrower, relatively indistinct intermediate dark rings occur in a 
few of the pale annulations. The composition of the more or less upstanding tail fur 
is very similar to that of the back; abundant underfur 10 to it mm long, and bristle- 
hairs 13 to 14 mm long; but widely scattered guard-hairs about 20 mm long, also 
occur. 

Skull (fig. 30). This has already been sufficiently described in the introduction to 
the genus. 

Habits. There is nothing to add to what has been given above under the generic 
heading. 

POIANA LEIGHTONI Pocock Leighton's Linsang 

Poiaiia ricliardsciii libcrkiisis Pocock, 190S, Proc. zool. Soc. Loud, (for 1907): 1043. This name is rejected as 
a /ii/'ji(5 calami in favour of that which next follows, Icightoui Pocock, as explained thereunder. 

Poiana richardsoni Uifililoni Pocock, 1908, Proc. zool. Soc. Loud, (for 1907): 1043-1045, pi. 54, f 3. North- 
east Liberia, 24 to 32 kilometres west of the Putu Mountains, situated west of the Duboc and Cavally 
Rivers. Type in the British Museum, No. S.8.23.2, sex?; a headless flat skin, lacking also the forefeet, 
but otherwise in fair condition; no skull. This name is accepted in spite of the apparent page priority 
oi liheriaisis, which was an obvious lapsus calami. It has generally been assumed that the lapsus must 
be in respect of the name Icightoui used in the key two pages later than lihcriciisis. But that Pocock's 
real intention had been to give the animal the former of these names is clearly shown both by a printed 
amendment slip m the Proc. zool. Soc. and the fact that tor his own private separates of his paper he 
had the name amended on page 1043 of the text to leightoni before the reprinting. Pocock's pen, 
therefore, obviously slipped in the inadvertent use of lihcriciisis, not vice versa. This name was given 
in compliment to Mr. Leonard Leighton of the Liberian Rubber Company whose small collection 
of mammals contained the r\pe. 

Distribution. Although Pocock named this as merely a local race of richardsoni it 
would seem, even from the very incomplete material m the British Museum that it is 
probably a discrete species, geographically widely separated from the Bight of Biafra 
animal. Pocock (1908b; 1038) mentions that Leighton obtained 6 skins; but only 2 
ever came to the British Museum, the type in 1908, a second from the Zoological Society 
of London, via Pocock, in 1939. There are no skulls. Kuhn (1965) gives the following 
additional places in Liberia from which he personally has obtained specimens; Biple, 
Bongle, Deaple, Duotown, Igua, Siamoiirovia, Tappita. on the road between Bia and 
Zwedru. All these are in north-eastern Liberia, rather further north than the type 
localiry, and west of the upper River Cess. The vegetation is rain-forest. F. de Beaufort 
(1965) says the species also occurs in the Ivory Coast at Gagnoa. Like richardsoni, there- 
fore, h-ightoni would appear to be a rare and very localized animal. 

Description. Leighton's linsang is appreciably paler than Richardson's, the dorsal 
groiuid-colour being a sort of golden-buff. As in the latter species the niaculation is 
black, though appreciably more intense; and the spots are of similarly small size but 
often of more angular outline. There is a definite, more or less continuous, narrow 
black spinal stripe. The dorsal pelage is very short, vers' soft and springy, the abundant 
imderfur measuring about 10 mm, the bristle-hairs about 12 mm but exceptionally 
reaching 15 mm. The head and neck being missing from both British Museum speci- 



228 Till, <AIIM\01U:S (IP WISI MUICA 

iiK'Us u IS not possible to describe iheii) ; but what rein.iins nt the legs and tect seems 
to be quite similar to those of richardsoni. The belly and insides of tlie liindlegs are 
pure white, a clear distinction from the other species. 

The tail is even softer than in riclumlsoiii, clothed with very dense imderfur about 
7 to 9 mm long together with flat ended, long-stalked brisdc-haiis lo to ii mm long. 
Tliere are also scattered, terete guard-hairs up to 15 mm m length. The structure is 
readily distinguishable from that of richanisoiii by its aimulation. There are only from 
10 to 12 dark rings, wliicli are dense black or ver}' deep brown; and dresc instead of 
being parallel-margined are more chevron-shaped, with a forwardly directed point. 
The tails of specimens so far known show no sign of intermediate dark rings. There 
are no measurements: but the skins as they stand arc of a somewhat smaller animal 
than richinihoiii, though they may possibly be only those of yoimg adults. 

Skull. No skulls exist in London, nor does there appear to be any published account. 

Habits. Nothing whatsoever is known of these. 

The following table gives the mean measurements of British Museum material of 
Poitimi richardsoni; it is not possible to give comparative figures for Ici'^htcni. 

Tabic 11: Numerical tl.it.i lor Poiiiiin r'uhtirilsoni 

I'uhardioui 

Vegetation Forest 
Number in mean 7 

Coudylobasal lengtli 67-6 

Basilar Icngtli 61 -y 

Palatilar length 29-9 

Zygom.atic breadth 34-5 

Upper cheekteeth breadth I9'4 

hiterorhital breadth lo-o 

Richardsoni 

I'ostorhital constriction 11 '6 

Brainc.Tse breadth 24' i 

Tootlirow (1 — »|l) 2S'i 

/)• length 0-3 

ihI breadth 4-6 

i»l length 5-4 

i»2 length 1-5 

Head &• body 384 

Tail 365 

fhndfoot 59 

Ear 34 

l^ATIOS (per cent) 

Tail/head & body 95 

Zygoin. br./condylob. I. 51 

Braincase/condylob. 1. 36 

Braincasc/zygom. br. 70 

I'alatilar l./condylob. 1. 44 

hitcrorb./postorb. 86 

p'^jc — m^ 25-0 

m\jm2 ; »I3 360 



PARADOXURINAE ZZp 

Subfamily PARADOXURINAE Gill, 1872 
African and Asian Palm Civets, Binturong etc. 

Taxonomy. Somctliing has already been said, on page 163, of the disputed classifica- 
tion of the Paradoxurinae. Pocock (1929) considered that the African animal now to 
be dealt with merited separation into a full family of its own, the Nandiniidae, and 
thus be clearly more widely detached from the oriental palm civets with which it had 
customarily been associated, the latter remaining as the subfamily Paradoxurinae of 
the Viverridae. There is certainly a great deal to be said for Pocock's view of the taxon- 
omic distincmess of the African and Asiatic groups. Resemblance between them, as in 
the parallel case of the linsangs, is with very little doubt largely fortuitous, being rather 
a matter of convergence than an indication of close phylogenetic aftuiit)-. 

Such resemblance, in the present instance, concerns not so much an overall similarity 
of external appearance as the nature of the soles of the feet and a general correspondence 
of skull shape. The character of the feet was well investigated and analysed by Pocock 
(1915b); and the considerable differences that he found, together with those of other 
external features such as vibrissae, rhinaria, ear pinnae, and, more importantly, the 
form and siting of the perineal scent glands, led him with little hesitation to the con- 
clusion of phylogenetic divergence now under discussion. He might have added that 
superficial resemblance of build in the skulls of the two groups is no more than runs 
through a wide area of these small carnivores; but that there are, in fact, important 
distinctions, notably in respect of the bullae and of dental form, that lend support to 
the conclusions he arrived at from external features alone. 

With little doubt, then, the African and Asiatic palm civets arc phylogenetically 
more separate than has been commonly accepted; but whether that division should be 
drawn at family level is another matter. Simpson, while indicating some degree of 
taxonomic distinction, rated it as no higher than tribal. This is probably insufficient; 
but Simpson's classification is nevertheless retained in this present work since, in the 
absence of a major revision, mere tinkering on a minor scale with currently accepted 
forms is liable to add more to confusion than it does to clarity. 

The subfamily Paradoxurinae is divided by Simpson into three tribes: the Para- 
doxurini Simpson, 1945, and the Arctogahdiini Simpson, 1945, both Asiatic; and the 
purely African Nandiniini Simpson, 1945, with which alone we are here concerned. 

General. In view of the probable disunion of the African and Asiatic animals at 
present covered by this subfamily, and the fact that there is only a single genus in 
Africa, nothing further is said here regarding the general attributes of the Paradoxurinae 
as a group since all matters relevant to this present work concerning distribution, form, 
skull, habits and so forth are contained in the following account. The name of the 
subfamily is formed from that of the Asiatic palm civets, musangs or toddy cats, 
Pdradoxunis F. Cuvier, 1821, derived from the Greek words pariidcxtis strange, and 
oiira tail — a name given mistakenly without any substantial justification for its ety- 
mology. 



230 THI (AUNUOIUN (11 \V I s I A I l< I ( A 

Cicnus NANDINIA Gray, i<S43 
African Palm Civets 

Wwilina Gray, 1843, List oj tlif S/'faHu/is 11/ Maiiiiiuiliti in the . . . Bririili Miisiiiiir. 54; and 1S65, I'roc 
C(V'/. 5i'c LiVhl for 1S64: 520-530. Tvpe species IVivmj hinolala Gray. 

As this is a nionospccihc genus all relevant characteristics are to be found in the 
succeeding account of the Two-spotted Palm Civet. 

NANDINIA BINOTATA (Ciray) Two-spotted Palm Civet 

I'li'tr/.i biiwlala Gray, 1S30, Spicilfgla ZcH'/iujiVii . . . :y. Ashanti (Ghana). This was described from a 
specimen in the Nctlierlands Museum, Lcyden; but the type no longer seems to exist since it is not 
listed by Jemiuk (18S7 & 1892). The attribution ot this name to Rcinwardt, often given by authors 
(e'.jj. G. M. Allen, 1939), appears to be erroneous and stems from Gray (1 843) where he hmisell accredited 
the name to that author. However, Keinwardt does not appear ever to have published it; and in later 
works Gray {1865 & 1S69) indicated that Rcinwardt had only suggested the name in manuscript. 
Incidentally, Gray more than once muddled his List of Specimens (1843) with his Catalogue (1869), 
the tirst of the two works cited in the previous sentence providing one such confusing example. The 
specitic name is tormed from the Latin hi- two, and iiolola marked, with reicrence to the conspicuous 
pale shoulder spots. 

I'tiradoMiriis hamiUonii Gray, 1832, Proc. zooL Soc. Loiul.: 67; and 1835, lUiislrations oj Indian Zoology, 
2, pi. 10. In both cases the type locality is indicated, wrongly, as India; this was subsequently amended, 
in two old British Museum registers and in Gray (1843), to Fernando Poo. This revised provenance 
must also be reg.irded with considerable suspicion; it was m all likelihood due solely to Edward Cross 
of the Surrey Zoological Gardens, where the type animal had been displayed. It is well known that 
the knowledge ot West African geography possessed by commercial importeis in the early I9di 
century was cNtremely sketchy and most trequently conveniently crystallised itself into Fernando 
Poo, at which island nearly every ship called. Thomas (1904) rejected Nanilinia from the island's tauna; 
Cabrera (190S) was doubtful; and no specimen seems to have authentically been forthcoming since 
the type. Type in the British Museum, No. Soa, q ; skin in good condition except for an incomplete 
tail; skull with the back and floor of the braincasc missing. The species was called after Dr. Hamilton. 

General. The common name two-spotted palm civet tor this animal is pretty 
firmly established in spite of several objections that can be lodged against its use. Li 
the first place it could radier more sensibly be termed, as it sometimes is, the twin- 
spotted palm civet since the pelage has, in tact, a multitude t)t spots but only one 
matched pair ot distinctively coloured ones, situated on the shoulders. It is also, less 
commonly, known as the tree civet; and this has possibly more to be said m its tavour 
than the others as these animals spend a good deal ot their lives m trees of all hinds, 
palms being probably relatively infrequent. It is interesting to see how the name p. 
civet came to be applied to Nandinia. It was not touiided on any knowledge ot th 
animal's habits, these being to this day largely unknown; but, as recorded in th 
introduction to this present subfamily, the Atricaii species now under discussion was 
considered to be very^ closely akin to certain Oriental viverrids, chiefly ot the genus 
Parddi'Minis. The best known species of this last Asiatic genus, P. hciiiMpliioditiis (Pallas), 
was called hx Anglo-Indians the palm civet or todd\' cat because it was reputed to 



aim 
e 



NANDINIA 231 

climb the toddy palm [Phoenix iih'cstris) to steal the toddy ("wine") from the receptacles 
into which it was being tapped. Because of its supposed close affinity, Natidinia was 
regarded as the exact Ethiopian counterpart of the Indian animal, whose common 
name, with suitable qualification, consequently became transferred to it; though 
there is no evidence, so far as the present writer knows, cither factual or in hearsay- 
folklore, that Nandinid chmbs palms to get at the palm-wine. 

Distribution. In general appearance the African palm civet (Plate 5), with its 
spotted body and long ringed tail resembles a heavily built dark brown-coloured 
genet. It would seem from its abundance in collections that it is probably the common- 
est of the African viverrids. It is widely distributed throughout the whole of the 
rain-forest block and some of the contiguous Guinea or Guinea-type woodland zone 
where there is fringing forest or forest renuiants — for this is essentially an animal of 
fairly high trees and the shade of their dense crowns, not of exposure to siuishinc in 
low-growing open-country species. The southern limit of its range is roughly a line 
drawn from 15° South on the west coast to 20° South on the eastern side of the con- 
tinent; thence it spreads north to extreme south-western Sudan, taking in Uganda 
and south-west Kenya, on the one side and, on the other, Sierra Leone, and probably 
further west to Portuguese Guinea, though the specimen on which this last range is 
gromided was in fact a captive one (Monard, 1940). Nividiiiia is commonly stated to 
occur also in Fernando Poo. Something has been said on this subject above in the 
synonymy. The claim appears to be based on two specimens alone; the type, already 
dealt with, and another British Museum skin. No. 55.12.24.413. Tlus latter formed 
part of a parcel of over a thousand specimens purchased m 1855 from the Zoological 
Society of London, and its reputed provenance is open to precisely the same doubt 
as given above in respect of the type o( luiinihoiiii. The case for the occurrence of the 
palm civet on Fernando Poo is slender. 

Description. The dorsal fur of Nandinia is dense, of moderate length, soft to the 
touch if quite clean, but slightly harsh if dirty. The overall colouring is variable in 
different specimens, some being appreciably darker than others; but it may be charac- 
terized in general tcruis as of a medium bro\vn hue with blackish spots. The fur is 
deep sepia based throughout, the visible background colour and that of the dark 
maculation residing solely in the extreme distal portions of the hairs. The dominant 
component of the pelage is long, fme, dense, wzwj underfur, measuring about 12 to 
15 mm, though this varies somewhat with different skins and the state of moult. This 
has short golden-brown tips. Amongst this dense imderfur arc set, much more widely 
dispersed, rather longer bristle-hairs, some 20 to 23 mm, ver\' slender throughout 
most of their length but, be)'ond the reach of the luiderfur, expanded into a stouter, 
terete or very shghtly flat-sectioned terminal portion. The pointed tips of these are 
black; but the majority have a subterminal band of golden-brown, which often gives 
way proximally to a narrow creamy zone. This accounts for the rather nondescript 
slightly reddish-browii flecked with yellow of the back. The dark spots arc the outcome 
of the lack of any subterminal zone, the entire distal half of the hair being black. In 
the two pale spots that normally occur, one on each shoulder, the reverse is the case, 
there being no black tips. There are also much longer black guard-hairs scattered 



rill ( Ai(M\ (11(1 '. cii w IS r Ai im A 



>,> 



thioiighout till.- diMj.il pchigc. liic black, occmsikiuHv dci.'p-bro\vn, spots torniiiig 
the dorsal pattern arc ot fairly small size and rather irregularly disposed, not more or 
less clearly arranged in longitudinal series as they are in most Giitctiii. Broadly speaking 
they are roughly siibcirciilar, about to mm in diameter, or less, mostly independent of 
one another, though in a few cases tending to coalesce transversely. But in some speci- 
mens they arc much smaller. There is no black spinal stripe or crest, although occasion- 
ally in some skins there is a combination of spots that gives some impression of a 
medial stripe. 

The flanks are almost completely unspotted; the belly is yellowish and clearly, 
though not sharply, divided from die flanks. The two yellow or creamy spots on the 
shoulders are somewhat oval and at their maximum are about 25 mm long and vers- 
plain to see; but they vary and are often indistinct or sometimes virtually lacking. 
Forward ot these lie the markings of the neck. These are considerably variable but 
basically consist of a medial black line reaching to the crown of the head between the 
ears. This is flanked on either side by another, more or less parallel, black line reaching 
to the base of the ear. There may be a row of dots, sometimes very faint, between die 
medial and outer band; or tlris last may itself by broken up into spots: or some or all 
t these elements may be entirely lacking. However, in most, though not all, West 
African specimens the three main black lines are clearly present. The face is greyish- 
brown, without distinct markings. The ears are veiw rounded, low 111 height but broad 
at the base; and a bursa is always present, its semicircular posterior flap arising, bodi 
top and bottom, behind the piima. 

The feet are 5-toed, each digit armed with a very sharp, well-curved claw like a 
cat's in shape and similarly retractile. They are slightly webbed between the basal 
parts of the toes. The soles are very characteristic, apart from the naked, v.-ell-developcd 
pads almost completely densely but shortly hairy. Li the forefoot the palmar and the 
carpal pads are united; a central depression boimded by the four main palmar and two 
carpal sections naked but of granular appearance. The ist di2;it of this foot is joined to 
Its section of the palmar pad by a naked strip; and there are four small triangular 
naked areas just anterior to the other sections. In the hmdfoot the plantar and tarsal 
pads are similarly joined, but the latter unite into a single, large pad, very wrinkled 
or ridged posteriorly and reaching almost to the heel. There are similai small naked 
areas forward of the central pad and also joining it to rile 1st digit. 

The tail is somewhat longer than the head ib\ body, bushy and woolly throughout 
Its length but usually distinctly broader proxiniallv than distally. It is darker above 
than below through the presence of long-black-tipped hairs, which in some specimens 
impart a wholK blackish appearance to the terniiiial four niches. Its covering consists, 
like that of the bodv, of abundant, long, dense underfur, about 30 mm in length: 
and of sparser bristle-hairs, normally measurmg about 40 mm but which here and 
there attain almost 60 mm. The tail may be looselv described as ringed; but it is very 
irregularly so, nothing like the genets or Imsang, the "rings" being unevenly spaced 
and mostly only half-rings across the dorsal side, or sometimes little more than broad 
patches. The scent glands have been fully described by Focock (1915b): those of the 
male are situated anterior to the penis, diose of die female in front of the vulva. 



NANDINIA 233 

Skull (figs. 3 I and. 32). This is considerably larger than that oi Gciniui but otherwise 
superhciallv ver)- similar. The braiiicasc is long, ovoid, and bounded anteriorly by a 
marked intertemporal constriction wliich is narrower than the interorbital breadth. 
The supraorbital processes arc long and sharp. In old males and very old females there 
is a pronounced sagittal crest, v^^hich posteriorly jouis a broad, flange-Uke supraoccipital 
crest. The zygomatic arch is strong, almost semicircularly curved, the jugal process 
slight. 

The most notable pecuharity of the Niindinici skull lies in the bulla, the posterior 
part of which is generally held to be entirely cartilaginous and is lacking from normally 
prepared specimens. A detailed description of this auditory region is given by Hough 
(1948), who finds it, despite the peculiarity now under discussion, otherwise quite 
rv'pically viverrine. She accepts the generally held view that the entire posterior portion 
of the bulla (the entot)'mpanic) is unossificd; Van Valen (1963), however, disagrees 
with this, holding tliat "the dorsomedial side of the entorv'oipanic is commonly, 
perhaps always, ossified in mature and nearly mature individuals, contrary to the 
usual statement, although normally there is a cartilaginous region between the tympanic 
and ossihed cntorympanic where they approach each other". Be this as it may, for 
practical purposes of recognition almost ever)' prepared Naudiiiiii skull has the main 
bulbous portion of the tympanic bulla, so conspicuous a feature of other carnivorous 
skulls, lacking. Li this it is unique amongst the Hving (but not fossil) carnivora, and 
exceptional as regards the mammalia as a whole. The anterior, wholly ossified, portion 
around the meatus is often missing too, but for the entirely different reason that it has 
been swept away through lack of its normal mechanical support provided by the bony 
posterior region. The lack of a posterior portion of the bullae is accompamed by a 
second peculiarity of the Nandinia skull. In other Feloidea, in which the bullae are 
fully ossified, the paroccipital processes are closely applied to their posterior faces, 
over which they spread to a greater or less extent; in XiVidiniii there is no ossified 
bulbous portion to which they could become attached and they stand isolated as con- 
spicuous components of the posterior part of the skull, long and pointed, and distinctly 
more canoid than feloid in appearance. 

The dental formula is ^-ia-i ~ 4-°" '^^^ upper mcisors are set m a straight or slightly 
curved, compact row, the outer ones being somewhat stouter than the iimer ones. 
The upper canines arc much straighter than usual, grooved on the outer face as in 
Gcnetta but more obviously so; and also on the inner face, a character that is obscure in, 
or lacking from, the genets. There is nothing remarkable about the rest of the check- 
teeth except that tifi is very small, peg-like; and, in fact, considering the much greater 
size of the animal the whole series is small in comparison wth Gawltn. 

The lower jaw is strongly built, wth deep rami. The incisors arc bifid; the canines 
are more curved tlian the upper ones and are similarly grooved on both outer and inner 
faces, the latter sometimes rather indistincdy. The posterior molar is relatively small, 
but nevertheless much better developed than the upper one. 

Habits. In spite ot its being both widespread and numerous not much is known of 
the life ot the two-spotted palm civet. Something has already been said, in the opening 



-34 



THE CARNIVOKKS OF WnST AFRICA 



paragraph dcaliiii; witli this animal, ot its conjectured attinity with the Asian I'did- 
iL'.xurus, in respect ot habits as nuich as of phylogeny. Though Wvuiiniii does chnib 
pahn trees, and has been shot in diem, there is at present no significant evidence that, 
despite its common Enghsh name, its habits are in any exckisive or predominant way 
connected widi them, and particularly in so far as thieving palm-wine is concerned. 
In an area such as West Africa where wine-tapping is so abiuidantly practised such a 
habit, if it existed would be a matter of common everyday knowledge. This is not so. 




Fic. 31. Naudinia liinotiila: skull, B.M. No. 48.814, 



I ; Literal 



Further, ex-Anglo-Lidiani, relyuig on parallelism with Asiatic species, have in the past 
saddled NaiiJiiiid with the character of being "a noted fowl thief", certainly a more 
credible failing in a viverrid than wine-bibbing. In combination of these two reputed 
habits, the present writer was warned many years ago that it was foolhardy to preserve 
oil-palms standing in one's compoiuid since there might always be a palm civet lurking 
therein, ready to descend at night to rob the fowl-house. Whether this danger is real 
is open to some doubt; for this small carnivore is, in truth, largely vegetarian. G. L. 
Bates (1905), who has recorded more of this animal in its natural surroiuidings than 
anyone else, wrote: "... there is no doubt that the usual food of the Nandine is 
vegetable. It never catches chickens, as do other J'ivnjiddc" . And T. S.Jones has the 
impression that ripe palm fruits form a significant part of the diet. On the other hand, 
Thorneycroft (195S), writing in Malawi implied that the oiJy time die palm civet was, 
as a rule, encountered, was "in connection with raids on the hen house". He also sup- 
posed a pair he saw in the bush to be interested in a flock of guinea-fowl. And R. W. 
Hayinan has noted {in lift.) that 6 adult specimens taken by him in the Ituri forest in 
1930 were all lured by baits of oifal into traps set on the ground between die buttresses 
ot giant trees. 



NANDINIA 



235 



The fruit-eating habit is well confirmed by those who have kept Xiindinia in cap- 
tivity'. Ball (1955) for example, who reaixd a yoiuig specimen in Nigeria, found it to 
eat banana at an early age; and this remained its favourite fruit though it would eat, 
also, pawpaw and avocado. This animal was, in addition, given crickets and other 
insects, fat moths being very welcome; but as it was diftkult to obtain a sufficiency of 




Fig. 32. Naiiditiia binoiaia: skull, B.M. No. 48.814, sc.\ ?, x i; p,iliital & dorsal views 



such items the young palm Livet was eventually provided with a small lizard each day. 
Such food had to be quite fresh, for even if it was only a little stale it would be refused. 
This animal also accepted scraps from the table of a varied kind such as might come 
from ordinary human mc.ils, including fish and pieces of meat; and it was very fond 
of sweetened condensed milk and of chocolate. It liardlv ever drank water. 



236 THE CARKIVOHf.S oy WEST AIRHA 

All this IS doubtk'ss a t.iir indication ot tlic sent ot dietary m die wild: a good deal 
ot hiiit tempered with insects, small rodents, birds or anything of like kind offering 
Itself easily. Bates, in fact, records raids by a two-spotted palm civet, repeated over a 
series ot nights, on the shreds of flesh still attached to the skeleton of a chimpanzee 
hung up to dr\''. He also indicates that the fruits of the umbrella tree, Mnsaiii^ii cccro- 
jHoiiks R. Br. and ot the climbing, yellow-flowered cucurbit Cof^iiiatixiii podolacna 
Baill. are recognised to be such favourite foods of this animal that they are used by 
local hunters to bait traps set with the object of capturing Kauduiia. Sanderson (1940) 
foimd the stomachs of his many Mamfe specimens "nivanably" crammed with vege- 
table remains — that is to say plantains when caught near farmed land. 

The two-spotted palm civet is nocturnal. Under natural conditions it wakes at about 
sunset and goes to bed soon after dawn, being, presumably, active during most of the 
night. It may theretore sometimes be seen 111 the half-light at either end ot the day; 
or It may be picked up by the red glow of its eyes in a torch in the darkness of the 
forest at night. Since this animal is both arboreal and terrestrial such glimpses may be 
either on the ground or on the lower branches of trees. The sight ot a pair of glowing 
eyes peering down through the blackness can be rather eerie. These nocturnal habits 
are basically retained in captivity but may be partially overcome in response to the 
offer of food. Climbing for this vivcrrid, with its extremely sharp cat-iike claws, is a 
matter of considerable ease; indeed Bates, in connexion with the nightly visits to the 
chimpanzee skeleton mentioned above, makes it clear that the animal concerned in 
this must have walked some distance clinging upside-down to the underside of the 
ridge-pole of a hut. Its sure-footedness is also indicated by 13all (1955) who relates 
how his domesticated animal used to spring onto the curtains, scramble to the top 
and walk along a curtain-rod only 13 mm in diameter. Li climbing vertically up the 
bole of a tree the claws may well receive assistance from the large, rough-surfaced pad 
under the hindfoot. It always climbs down heaci first, Taylor (1970) giving a series of 
pictures of the successive moves. 

Probably the majority of its foraging, however, is carried out on the ground; and this 
is one reason why it is fairly readUy trapped. When on the ground it either prowls in 
feline fishion or proceeds at a trot, though not very tast. Thorneycroft (1958) records 
an interesting incident concerning the palm civet's capacity tor leaping. He observed 
one high up a tree, frightened by the barking of dogs at the foot, to sail gracefully — 
"almost float" — to the ground 111 an effort to escape, descending at "an angle greater 
than half a right angle", legs and tail fully stretched out. The landing on the bare 
ground was perfect, on all four feet, at a considerable distance from the tree. This 
performance was rapidly repeated from a neighbouring tree; but an attempt to reach 
a third tree was foiled by the dogs. Taylor (1970) rates Xdiuliiiia as the most efficient 
African vivcrrid at lumping. 

During daylight hours the two-spotted palm civet curls up tightly and sleeps. Bates 
said that it did this in thick tangles of vines in the tree-tops. This is a little surprising 
in that climbers do not very commonly form tangles such as would naturally provide 
suitable beds for quite heavy animals; especially in the upper strata of trees, at which 
level lianes have generally achieved a good deal of independence. Tree-tops, also. 



NANUINIA 237 

would sccni to imply a good deal of sunlight, whereas Wmdiniii, in common wirii most 
daytime sleepers, prefers deep shade. Tangles certainly exist; but at ftirly low levels 
or in partially destroyed forest to which simlight has been admitted. It would seem 
probable that for their daytime sleep these animals would more commonly seek the 
shelter of holes or well-shaded crooks where large branches join the bole. There seems 
to be no record at all of an actual breeding nest. This must almost certainly be a hole; 
but whether up a tree or in the ground it is impossible to say. According to Walker 
(1964) the period of gestation is about 64 days, the litter size being 2 or 3. Juvemlc 
specimens exist in the British Museum captured in February, March, May, June and 
August, thus indicating that breeding may take place in either the wet or dry season. 

Two-spotted palm civets, if obtained very young, have shown themselves to make 
charming, friendly and trusting pets, though suspicious of and aggressive towards 
strangers. Given the free range of a house they have been said to preserve it free of 
rats and snakes. Ball (1955) found his yoimg animal to enjoy games, even fairly rough 
ones, with a dog, but not with a cat. This author also records that though his palm 
civet had the habit of marking out certain areas with its scent glands it never at any 
time had, itself, any offensive odour. Its fur, which was rather rough and bristly, 
always had a pleasant, warm, musty smell. Bates heard two palm civets calling to each 
other in the evening in the forest with a kind of faint kittenish mewing; Ball's pet 
when pleased uttered a sort o[ cwiick-cwuck sound mingled with a purr; but when 
amioyed gave a shrill, ratthng warning note. This animal was also responsive to human 
utterances, recognising the sound of its master's voice, and answering to its name and 
coming to a whistle. 

Taxonomy. Three races have been described from other parts of Africa. These are 
dependent on variations of colour or pattern. Some degree of variation is common ; 
(. A. Allen (1924) writing of nearly 75 north-east Congo specimens said that there was 
amongst them "a wide range of individual variation in coloration"; and the British 
Museum West African material, of about 50 skins, displays appreciable variety of both 
colour and markings. It is therefore ditficult to know how far races founded on these 
characters may be justified. At least one Sierra Leone skin has a more intense colour 
that recalls Cabrera & Luxton's ititcnsa from the southern Congo ; but it is not quite the 
same, and the pattern is certainly less marked. And though nuchal striation becomes 
obscure in certain other West African specimens there is none exhibiting the com- 
plete absence of any trace of dark markings in this area that is the mam characteristic 
of Thomas's (lerrarJi. J. A. Allen foimd this to be the case also with liis Congo series; 
and since, moreover, there are in London several skins from East Africa with this 
distinguisliing feature it seems probable tiiat i^crrardi is something more than a mere 
individual extreme. Allen considered the third named race, arborca Heller, "to be 
surprisingly different ". It would therefore seem that these races arc justified; and, that 
being so, the West African two-spotted palm civet is correctly designated Xandinia 
hinotatii hinolatd (Gray). If colour alone is regarded as a valid subspecific distinction it is 
always difficult to know where to stop; but it is possible that the appreciably paler 
skins of the Cross River basin (south-eastern Nigeria and Cumeroim) may merit a 
distinguishing name. 



;;,S Till'. CARNIVORES Of WEST AFRICA 
I .iWe 12: Niiniciic.tl ilatj lur .\'im</ii/iVi hiuohiia 

West Africa 
general 

Vegetation Forest 

Number in mean 17 

Condylobasal length 96^2 

Basilar length 91-0 

Palatilar length 42-6 

Zygomatic breadth 53-4. 

Upper cheekteeth breadth 29-8 

iutcrorbital breadth l8'0 

Postorbital constriction 14-0 

Braincase breadth 32-2 

Toothrow (r — m-) 35-9 

;i^ length 7-6 

m^ breadth 5'4 

»i- breadth 2-0 

(Ml length 7*2 

ni2 length 3-0 

Head & body 484 



Tail 



5S4 



Hindfoot 85 

Ear 37 

RATIOS (per cent) 

Tail/head & body 115 

Zygom. br./condylob. I. 55 

Bramcase/condylob. 1. 33 

Braincase/z\'gom. br. 60 

Palatilar l./condylob. 1. 44 

Interorb./postorb. 128 

p^jc — ufl 21 -2 

iiijIiiio -f- i»3 240 

Tabic 12 shows the average size ot 17 tuliy adult West Ahican specimens in the 
British Museum. 



HERPESTINAE 239 

Subfamily HERPESTINAE Gill, 1872 
Mongooses 

Distribution. The last of the three West African subfamilies of the Viverridae, 
the Herpestinae, has, like the other two, an Asiatic as well as an African distribution; 
and, like the Viverrinae, just fmds its way, as something of a rarity, into southern 
Europe. This is the largest of the three subfamilies, in respect of the number of different 
genera, of species, and almost certainly of individuals. It is in Africa that the group 
reaches its maximum development. As contrasted with 2 genera in Asia there are 
something in the nature of 16 in Africa, the majority, it is true, monospecific. Distribu- 
tion is mostly south of the Sahara, much of it almost entirely tropical, though some 
species range to South Africa which, moreover, has a few genera pecuhar to itself 
In this present account West Africa is credited with 9 genera covering 10 species. 

General character. Nevertheless, all members of the group, whatever their size 
or colour, have much more in common as regards both appearance and habits than 
is often the case with other subfamilies; something, neither exact of defmition nor 
con^pletely general, which nevertheless enables them all to be readily recognised as 
mongooses (Plates 6, 7, 8 and 9). Tliis is, for the most part, a long subcylindrical body 
carried close to the ground on short legs and clad as a ride in a rather coarse, long and 
bristly pelage, nearly always to a greater or less degree speckled; a sharpish face with 
low rounded ears set well down on the sides of the head; and a tapermg tail that is 
nearly always loosely haired, sometimes shaggy, and most often only a httle shorter 
than head & body. The ears, which differ from those of other West African viverrines 
in never having a bursa, may be completely closed by tightly folding together. 

Mongooses, certainly in West Africa, fall into two clearly distinct size classes: small, 
having a head & body length of under 400 mm and a weight usually well below 
2 kg {MuHi^os, Crossarchus, Gakrella) ; and large, with a head Si body length of about 
500 mm or much more, the adult body weight being from 3 to 5 kg [Herpestes, Atilix, 
Ichneumui, Galeriscus, Xciwgale). Lihcriictis, whose bodily characteristics are unknowai, 
can nevertheless, from its skull size, be certainly reckoned as belonging to the latter 
class, though a little smaller than the others. 

The pelage in the Herpestinae is composed almost invariably of longer or shorter, 
fine dense underfur and abundant, flattish-scctioned, much longer, annulated bristle- 
hairs; but Mungos is an exception in that the underfur is virtually lacking. The tail, 
which is of somewhat variable shape but usually distijictly tapering from root to tip, 
is clad in the same way but the bristle-hairs arc often much longer and sometimes give 
the structure a shaggy appearance. Mostly it is uniformly coloured throughout its 
length, generally similarly to the dorsum; but Ichncumia and Galeriscus form notable 
exceptions to this, while in Hcrpcstcs and Gdlcrclhi there is a sharply contrasting terminal 
tuft of jet-black or red. All the fur is erectile in anger or alarm. 

The rather pointed face ends in a naked rliinarium wliich, in different genera, takes 
on somewhat different shapes, involving the naked area itself, the nostrils, and the 
infranarial depth .and character of the upper lip. I'ocock (1916c), who figured several 



240 IHL CAKMVOIIhS Ul WIST A 1 R I C A 

L;;cnci.i and laid sonic raxunoniic signitR'ancc upon the dirtciciiccs he cited, used hvc 
or recently dead zoo animals as die basis of his observati(nis; but the relevant points 
are not often clearly to be seen in preserved specimens. 

The shortness of leg m the mongooses has already been mentioned; but there are 
several other noteworthy features attaching to the limbs in this subtainily. The niajorirv' 
ot species are sub-digitigrade, but Crossdirhiii has by comparison a noticeably more 
clumsy, tiat-footcd stance and gait. The feet themselves are in most genera fairly 
powerful, but in GakrcUa are of slender build. Mostly there are s digits, but Giikriicns 
has only 4 on each toot, while in Gcilcrclla the pollex and hallux are much reduced 
and sometimes lacking. In all cases except that ot the marsh mongoose (Aiiliix) digits 
11 to V are joined by webs, never vcrf pronovuiced and sometimes only basal. The 
claws are mostly strong, and on the forefoot long, m some cases (ci;. .\///);<,'c\*-, CrosSiir- 
ilnis) very noticeably so; but in GalcrclLj diey are quite different, short, slender, curved 
.md relatively sharp: 111 fact, much more adapted to climbing than in most. Finally, 
the soles, particidarly those of the hindfeet, are diverse in respect of the amount ot 
turr}- covering they carry, being either entirely hain.' [hhmiunia, G,iUrisais, AV/ioiyd/c), 
hairy in the posterior part only {Crossiiirhm, GiilcrclLi), or entirely naked {Mtiii(;os, 
Herpcfics). AtiLix is mostly naked but seems to be vanable. 

Like other viverrids the mongooses are provided with scent-glands probably tor 
die purposes of recognition and territorial marking though possibly not defence; 
but little actually defuiite has so tar been positively dctcrmmed regarding such uses. 
There are, in the Herpestmae, always a pair situated cither side of the rectum, the 
external orifices of their ducts appearing either diametrically lateral to the anus or 
slightly above. 15oth these orifices and the anus itself are surrounded by an upstanding, 
fairly thick, elliptical or subcircular wall forming a sac which is capable of closure, 
not by the contraction ot a circular sphincter muscle but by the juxtaposition of the 
upper and lower sections of the rim in a hp-like action. The form of the glands, the 
siting ot their openings, together with the shape and nature of the sac, though con- 
torming to this broad general pattern are, so far as known from limited investigation, 
variable m detail from genus to genus. A somewhat malodorous liquid seeps from the 
external glandular orifices into the sac and accords to the animal a faint odour, normally 
not highly objectionable; but whether a more energetic and active expulsion ot fluid, 
.IS III the polecat, is practised under stress bv mongooses in general is not clear, though 
one such case is referred to on page 30('>. 

All these matters are dealt with in some detail and with illustrations in Pocock's 
paper on the external characters ot die mongooses fig 1 fie), though by no means all 
West Atncaii species are covered. 

Skulls and Deutitiojl. As might be expected, mongoose skulls more closeK' 
resemble those ot the iiearK' related genets (Viverrmae) than those of an\ other West 
Afnciii carnivores. There are. .ipart from size, certain clear differences, which can be 
seen b\ a comp.irison ot tig. Z'j w itli tig. 36. In dorsal aspect the rostrum m the ITerpes- 
tiiiae, with the notable exception ot Lilwrliilis, is bro.ider; .md the also very much 
bro.ider frontal area combines widi the complete or almost complete orbital ring to 
gi\e tile structure .1 quite difhreiit .ippe.ir.uue in this region. In the veiitr,,! .ispect the 



IIEnrRSTINAF. 



241 



much longer post-dciual palate of the mongooses is at once obvious; and the bullae 
instead of being long arc for the most part much more highly and acutely domed. 
Mostly the anterior portion of the bulla is very much smaller than the inflated posterior 
chamber, with the single exception o{ GalcrclLi in which the two are subequal. 

The dentition in the Mungotinae is remarkable for its variery. There may be 36, 
38 or 40 teeth due to diversity- in the number of premolars, which may be ^ in Munoos 
and Crossiircluis; j in Gi^lcrcUa; or j ui Ichncwniii, Giilcrisais, Xeiwgalc and Liberiiais. 
Both Herpcstes and Atihx are capricious in this respect, the former having either 3 or 4 
lower premolars, the latter exhibiting the same variation both above and below. 
Occasionally skulls occur m these two genera with differing numbers of premolars 
on either side. The form of the cheekteeth is also widely variable m the different genera, 
being sometimes fairly typically carnivorous with sharp-edged flcsh-cutring carnassials 
(Xaiogale, Atilax), sometimes sharply-cusped insectivorous {Afiiii(;os, Crossarchus), 
and in one case, Galeriscus, fraiikly crushing. In some species the dentition is relatively 
reduced in size, and this is particularly so in Libcriictis. 




Fig. 33. Herpestinae: illustrating the two contrasting positions of the cheekteeth: 
a. p"^ well anterior to the root of the zygoma {Idmeumia albicaiida, B.M. No. 25. 5. 12. 13) ; 
b. posterior comer of p"* about level with the zygoma (Xeiwgale imso, B.M. No. T0.6.1.14) 



There are nvo distinct rs'pes of herpestinc skulls m accordance with the positioning 
of the cheekteeth in relation to the zygoma (fig. 33 a & b). There are two reference 
pomts: firstly the extreme posterior comer of the upper camassial; secondly the 
point, mostly fairly obvious though not exactly dcfmed, at which the lower curve of 
the zygomatic arch arises from the main body of the maxilla, just above the toothrow. 
In one category of skulls {Hcrpestcs, XfiW(;(ih\ GcilcrellT, Arihix), the posterior comer of 
/!■' is situated at this point and, in consequence, /n^ lies wholly posterior to it, its outer 
face forming a sharp angle with the outer face of the camassial. In the other category 
{khncwnia, Gakriscia, CrosSiirclnis, Miingos, Libcriictis), the comer of ;)^ and all or at 
least part of m^ lie anterior to the reference point, mostly very clearly so, but in some 
Mii/Jijo.'; skulls the root of the zygoma is rather obscure. In this gi^oup the outer faces 
of />•* and ml form oiJy a shallow angle with each other, or even a gentle curve. 

Habits. Compared with other groups the habits of the mongooses have in some 
respects been relatively well recorded, though often more as regards their everyday 
behaviour as captive animals than as concerns the more secretive aspects of their lives 



::.\i Tin ( arm\i)ui:s or wisi aiiika 

m their natural environments. This is because, by reason oi' some ot their activities 
and the exceptional readiness with which many of them take to domestication, they 
have excited man's sympathetic interest over many centuries. Of all wild African 
mammals there can be little doubt that mongooses make the most acceptable pets, 
presenting much less difficulty of control and management than monkeys and offering 
in return a certain indcfmable charm together with unending interest, amusement, 
often surprise and sometimes apparent affection. The subject must, of course, be taken 
in hand while still ver\' young; and, with this proviso, probably nearly all species 
can be tumed into quite lovable, diverting, playful companions — and even useful 
additions to the household since there is no doubt that one of these creatures helps to 
keep the premises free of luidesirable pests, be they rodents, reptiles or cockroaches. 
However, not all West African species have yet been tried out in this respect. Of those 
that have, probably the kusimanse {Civssardnis) would lead as favourite — from the 
point of view of size as well as ardency of spirit. Accounts of this and other species 
will be found below in the appropriate sections. 

Of all the activities or reputed activities tif mongooses that for which they are most 
widcl)- famous is snake-killing. There is a good deal of misapprehension and misin- 
tormation about this. It is often sweepingly beheved that all mongooses make a regular 
practice of dehberately seeking out and destroying snakes, the ancient Eg)'ptians going 
so far as to say that Hcrpcstcs habitually entered into long preparation tor the fray by 
encasmg itself in layers of hardened mud as an impenetrable protection against the 
tangs of its opponent. The information relative to snake-killing is by no means so 
positive as generally supposed. Possibly several species will kill a snake if they should 
happen in the course of foraging to come across one and it was not too large; but it is 
extremely doubtful whether any would purposefully seek out snakes in general as a 
matter ot sheer innate enmit)-, as popular renown would have it. Certainly as far as 
the mongooses dealt with in this present accoimt are concerned Herpi'stes ichneumon 
has since ancient times had a reputation, though seemingly unconfirmed by modern 
observation, for an exceptional devotion to killing snakes; Mimgos mtinf^o and Ichiiciimia 
dlhkandd have both been rehably reported as doing so and eating their prey; but for the 
other species evidence is either lacking or merely hearsay. Indeed, Cansdale is quoted 
m Hinton & Dium (1967) as observing that his tame kusimanse was afraid of even a 
ilead snake ; and recorded ( 1 946) the same thing of a marsh mongoose. 

There can be no doubt, however, that where a fight between a mongoose and a 
snake does take place the former is aided to victory by two tilings: its great speed and 
agihty which enable it to spring upon its enemy, make an effective bite and leap clear 
before its opponent can strike; and the deceptive nature of its long-bristled coat in 
full erection, giving a false impression of size and misleading the reptile mto striking 
at what proves to be notliing more solid than hair. It has been experimentally shown 
that some mongooses are markedly more resistant to the effects of snake venom than 
other animals are, though they nevertheless succumb to a sufficiently large dose. 

A peculiar activit}^ that is more certainly common to part, if not all, of the sub- 
family is the practice of smashing objects by throwing them against something hard — 
usualK' interpreted as being basically intended for cracking open eggs, but on good 



HEUPKSTINAE 243 

evidence applied to other things, sonic familiar as foods, some strange, and some frankly 
useless. Tliis operation is carried out in two ways. The more spectacular is hurling the 
object, with the forcpaws, horizontally backwards through the hindlcgs against a rock 
or wall; the other is to stand upright, hold the object in the forcpaws against the chest 
and then cast it with surprising force vertically down to the ground. Both methods 
are very effective ; and if they do not immediately achieve the desired end are repeated 
until they do. West African species that have so far been observed to act in one or the 
other of these ways are Ichiieuiniti albiaiudd, Mnn^cs inni{i;o and Atilax paliidinostis. 

As for more general habits, mongooses may be diurnal or nocturnal, sohtary or 
gregarious. They are, by and large, pretty omnivorous, taking flesh when they can 
get it, killing and eating small mammals, birds and their eggs, snakes, lizards, probably 
frogs, crabs, sometimes fish, snails and, seemingly always as an essential mainstay, 
insects, particularly orthoptera, beetles and their grubs. Some, at least, greedily eat 
fruit; and in captivity more exotic foods. Not a great deal has been recorded of shelters 
and breeding nests; but it would seem that all species normally use convenient holes 
in the ground, in fallen logs, amongst tree roots or gaps between boulders. Some live 
communally in so-called warrens but consisting usually of a single chamber though 
with two or three entrances. It has sometimes been asserted that mongooses are generally 
fairly closely associated with water. While this is clearly so in the case of some, such as 
Atilax and possibly Hcrpcstcs, it is not so obvious with other genera; and without 
question these animals are sometimes encountered at some remove from the nearest 
river. 

Little has been recorded of the various steps and behaviour patterns in raising a 
family. Courtship play and coupling have been seen, as noted later, in the case ot 
Muiiqos imingo; and some rare and very interesting though brief observations on the 
period of gestation and frequency' of pregnancy are given in the accoimt o£ Crossdrchns. 
Litter size can, to all intents and purposes, only be guessed at: 2 to 3 would seem to 
be the commonest, but some species have been credited \%dth 4 or even 5. 

One other peculiarity may be mentioned in this general sketch of the subfamily. 
All mongooses, as already described, are furnished with anal scent glands, one of the 
chief uses of which is for the demarcation of territor)\ This activity is carried out m 
two different ways: one is by squatting down and dragging the circumanal sac across 
the ground; the other by the much more imusual method ot standing erect, upside 
down on the forepaws — called in gymnastics making a "handstand" — and in this 
position pressing the scent pouch against trees, rocks or other objects well above 
ground level. The purpose of this is not altogether clear since the warning patch is well 
above the nose level of any mongoose of similar species that should happen to trespass 
on the territor)'. The scent, of course, diffuses and would doubtless be picked up at a 
distance by a keenly attuned nose; but the advantage gained or the end served by 
dehberately making it less obvious is obscure. 

Mongooses have little direct economic impact upon mankind. Their sometimes 
ornamental skins occasionally provide simple pouches, scabbards and so forth in \'illage 
communities; and their flesh similarly helps to reduce animal protein defxcieno,- in 
areas where the more conventional butchers' meat is scarce or irregular in supply. 



2-11 I III ( A l!M\ OKI S Ol \VI S r AI KICA 

Indirectly tlicy may be regarded as in some me.isiire benetlcial in lielpmg to keep in 
check a varied list of creatures many ot which can be looked upon as undesirable; 
but. on the other hand, it must not be overlooked that they also take birds such as 
trancolms and guinea-fowl that are themselves of human food value; and that in 
killing lizards and frogs they are disposing of potential insect destroyers. They them- 
selves arc preyed upon by the wild cats and dogs, hyaenas, pythons and hawks. Certainly 
they fill an important role in the complex network of nature; but the sum total of 
their effect on man himself one way or another, is probably not vcrf great. Mongooses 
of various kinds, mostly under very confused nomenclature, are credited (Stiles c\ 
Maker, 1935) with a long list of parasites, both external and internal. 

Taxonomy. Diiker's (19.S7) unorthodox belief, founded on observation of living 
animals rather than study material, that the Herpcstinae are in fact more nearly related 
to the Mustelidae dian to the Viverridae has been referred to earlier on page 163. 

The nomenclatural position in the Herpestniae is exceedingly involved. Had anyone 
deliberately set out for malign reasons to devise a situation so complex as to be almost 
beyond comprehension he could scarcely have bettered the labyrinthme confusion 
that by chance came about 111 diis group. The binominal system of Latinised nomen- 
clature was devised to overcome the imprecision of vernacular names; but m the 
1 lerpestinac the absurd position arose wherein in using the name Muliiics miiiiiio it 
became advisable, or even necessary, to make clear, m the vernacular, whether reference 
was intended to the common mongoose of hidia or to the only distantly related 
banded mongoose of Africa. This is only one of several traps and obscurities. Fortunately 
1. A. Allen eventually went into the matter with extreme thoroughness (1919b and, 
more fully, 1924) and by reasoned argument brought order out of chaos. In consequence 
there is little need tor more than brief mention ot this past complexit)' here; and all 
systematists must tor long be deeply grateful to Allen tor his painstakuig unravelment 
ot so tangled a skein. Those to whom the matter is one of detailed interest must study 
his accounts, where they will tuid a score of pages crannned with the history and 
pitfills ot lierpestine nomenclature. 

.Mien's tuidmgs have been prett)' generally accepted; but though, thanks to his 
labours, the situarion has been cleared up as far as modern writmgs are concerned the 
tact must always remain that the study of early literature is rendered highly misleading 
.ind ditticult of comprehension without a good deal ot knowledge of this complex 
history. So many scientitic names, both generic and specitie, and so many animals ot 
no present interest to West Africa are involved in the muddle that no good purpose 
would be served by their mention here; but the general warning may be given to the 
West African student who attempts to derive intonnation from natural histories or 
scientific literature published well into this century to be cautious in interpretation ot 
Miiiioo<. Hcrpvslcs. Cro.Viiriliiis. iimn^o, linciatiis and iclimiiiihvi, in various combinations 
with each other or with other genera such as the previously all-embracing t Vi'crn;. 

hi this general note the main questions of present interest are diose ot the name 
and the status of the mongooses as a taxonomic unit, hi Simpson's classification (1945) 
they stand as a subfamily, the Herpestinae, of the Viverridae; but some authors, notably 
I'ocock, had previousK' regarded them as constituting a fimilv in its own right. 



HERPESTINAE 245 

Pocock (1916c and 1919), examining the question almost purely from the standpoint of 
external characters, felt strongly that this was their correct taxonomic rank; and since 
at the time he believed Herpistcs to be preoccupied and Miingos comcquently the vahd 
name for '"the typical mongooses" he called the family the Mungotidae. J. A. Allen 
(1919b) showed this to be an error and the family, or subfamily, name to be more 
logically and correctly derived from Hcrpcstcs. More important than mere nomenclature 
is the question of the taxonomic rank to be assigned to the group. There is a good deal 
in favour of a reassessment of the Vivcrridac and its component subfamilies. The 
mongooses exhibit morphological differences, external, cranial and dental, that in 
sum present a strong case for independent family recognition. But this is a matter 
essentially beyond the limited scope of this present work. It involves questions far 
wider than the mongooses, indeed the whole present conception of the Fissipeda. To 
deal with such fundamental matters piecemeal in a regional work is probably to lead 
more to confusion than clarity, and the major lines of Simpson's classification are 
therefore adhered to for the purposes of the present account. 

As regards the full generic use herein of Gakrclhi and Xcnogiile, the arguments will 
be dealt with later under their appropriate heads; but it may be said in this general 
account that, despite a broad overall resemblance of the skulls of these two genera and 
that oi Hcrpcstcs, the present writer fully supports Allen's opinion that there is a sufficient 
degree of difference, external and cranial, to warrant regarding these three as generically 
distinct. It must be added that this view is not shared by G. Petter (1969) from the 
consideration of tooth structure alone. Similarly, Oldfield Thomas's conclusion that 
Gakriscus is generically separable from Bdcogalc seems, on several counts detailed later, 
to be quite justified. His reservation oi GalcrcUa for cchracca alone, with the consequent 
erection of Myonas to cover the remaining species in this complex is, however, not 
accepted, the arguments in favour of such action being at present not fully convmcing. 

Thus, in this work these 9 genera are considered to be valid for West Africa : Mimgcs, 
Hcrpcstcs, Crossiirchiis, Atihx, Ichiietimin, GalcrcUa, Gcilcriscus, Xenogalc, Libcriictis. 
These may be separated by the following keys : 

KEYS TO THE GENERA OF HERPESTINAE 
(previous key page 163) 

A. Cranial characters 

I. Condylobasal length of the mature skull about 92-96 mm; the rostrum long 
and narrow (length of skull anterior to the postorbital processes very 
nearly equal to that posterior) ; all the cheekteeth very small for the size of 
skull (e.g. breadth of/)'' is less than 6 mm) . . Liberiictis {fiigc 336) 

Mature skull usually either appreciably longer or markedly shorter than the 
above; rostrum short, blunt and mostly broad (length anterior to the 
postorbital processes nearly always appreciably less than that posterior); 
cheekteeth relatively much larger ....... 2 



246 THE CARNIVORES OF WEST AFRICA 

2. The posterior outer corner ot p* lies very' close to the point where the outer 

posterior margin ot the maxillary root ot the zygomatic arch arises, and 
all or nearly all of in^ conseqently lies posterior to this point; the outer 
face of »(i forms a sharp angle with that of /)'* (fig. 33b) ... 3 
The posterior outer corner of/)'' lies well anterior to the point defuied above, 
and all, or a great part, of »ii is consequently also forward ot it; the outer 
face of /»! forms a very obtuse angle with that of p^ (tig. 33a) . . 6 

3. Condylobasal length of the mature skull about 100 mm or more; the anterior 

chamber of the bulla flattish and much smaller than the inflated posterior 
portion ............ 4 

Condylobasal length ot the mature skull under 70 mm; the anterior chamber 
ot the bulla inflated and more comparable in size with the posterior 
portion ......... Galerclla (pin;c 307) 

4. Postdental palate about as broad as long; the interorbital breadth over 19 mm 

and the cheekteeth light in build, the occlusal outline ot the narrow 
upper carnassial being scalene, its anterior breadth appreciably less than 
the external length ....... Xctiogalc (/niijc 329) 

Postdental palate longer than broad; it the interorbital breadth is as much as 
19 mm then the cheekteeth are strongly bmlt, the occlusal outline ot the 
stout upper carnassial being more nearly equilateral, die anterior breadth 
usually well over 80 per cent of the external length .... 5 

5. Width of the rostrum at the canine alveoli nearly always less than 20 mm, 

never much more; greatest breadth across the outsides of the cheekteeth 
less than 33 mm; breadth ot />■* less than 7 mm; length of /;j2 less than 
5-5 mm . ....... Herpcstes {pa(;c iCifi) 

Width of the rostrum at the canines 22 mm or more; breadth across the 
cheekteeth over 33 mm; breadth of;)'* over 7 mm; length ot ;»■.; over 
5-5 mm. ....... Atilax (p,n^c 2gi) 

6. Condylobasal length ot mature skull well over 100 mm; premolars ;j . .7 
Condylobasal length of mature skull well under 100 mm; premolars y . 8 

7. Posterior upper cheekteeth about as broad as long, subquadrate in outline; 

upper canines straight, tall, above average size, laterally compressed and 
with peculiar, sharp, knite-like anterior and posterior edges; postdental 
palate broad (about 14 mm) ; interorbital breadth considerably greater than 
the postorbital constriction .... Galcriscus (paiic ill) 

Posterior upper cheekteeth generally markedly wider than long; upper canines 
curved and of normal subconical form; postdental palate narrow (mostly 
well under 14 mm); niterorbital breadth usually rather narrower than the 
postorbital constriction, occasionally a little wider Ichneumia (/'. ;(,'(' 300) 

8. Occlusal surface of upper molars subtriangular in outline, the lingual portion 

clearly converging proximally to a sharp apex; skull narrow, the zygo- 
matic and mastoid breadths under jo per cent and 40 per cent respectively 
of the condylobasal length .... Crossarclius {paoc 279) 



HERPESTINAB 247 

Lingual portion of the upper molars elongate and more or less parallel-sided, 
the proximal margin rounded; skull broader, with zygomatic and mastoid 
breadths over 50 per cent and 40 per cent respectively of the condylobasal 
length Mungos {page 248) 



B. External characters 

The external features oi Lihcriictis are at present quite unknown. 

1. Both fore and hindfeet with only 4 digits and soles completely hairy up to 

the main pad; a very large mongoose with black legs contrasting with the 
paler, usually grey, dorsum .... Galeriscus [page 121) 

Feet with 5 digits (occasionally Galerella, small mongooses, have the ist digit 

much reduced or lacking) ......... 2 

2. Digits II to V entirely unwebbed; hind soles usually quite naked, but some- 

times partially hairy; a large, most often entirely blackish-browai mon- 
goose ......... Atilax [page 291) 

Digits II to V with small webs comiecting at least the basal joints. . . 3 

3. Pelage dark blackish-brown, slightly grizzled on the head and neck; size large 

(head &: body about 520 mm); the long nose projecting well beyond the 
Ups; hind-sole completely hairy . . . Xenogale [page iig) 

Pelage of a lighter colour ......... 4 

4. Entire forelegs and hindfeet almost black, contrasting strongly with the rest 

of the long, loose pelage; the long-haired tail mostly unicolorous, white or 
black, but sometimes coarsely parti-coloured; hind-sole almost com- 
pletely hairy ....... Ichneumia [page 300) 

Not like this 5 

5. Size large, head &: body 500 mm or more; tail broad at the base, much 

narrower distally with a conspicuous, contrasting, long black tuft; soles 
naked ........ Herpestes [page 266) 

Size considerably smaller ......... 6 

6. Pelage short, close-lying, of fme texture and fmely speckled; tail terminating 

with a conspicuous black or reddish tuft; head & body of small size 
(about 350 mm) and very slender; claws very short; sole of the hindfoot 
naked for about three-quarters of its length. Galerella [page 307) 

Pelage longer, coarse; animals of small size but stocky build; claws of the 

forefeet very long .......... 7 

7. Pelage with abundant long miderfur; one or two whorls on the back of the 

neck; nose very long; sole of the hindfoot hairy on the proximal quarter 

Crossarchus [page 279) 
Pelage virtually without imderfur; no nuchal whorl; nose of normal length; 

sole of the hindfoot naked to the heel . . . Mungos [page 2^%) 



248 THE CARNIVORES OF WEST AFRICA 

Genus MUNGOS E. Geoffroy & G. Cuvier, 1795 
African Mungos 

Mimgcs E. Geoffroy & G. Cuvicr, 1795, Mugasin EiuyclopiJi.pic, Z: 184, 187. Type species I'ii'crra mungo 
Gmelin. This is a fonnalisation of an Indian vernacular (Marathi) name, maiigus. 

Aricia Gray, 1864, Proc. rcci/. Sec. Lond.: 565. Type species Hcrpates tacnianotus A. Smith. Derivation of 
this name is not certain but is most probably trom Shakespeare's light and airy spirit Ariel, given 
with reference to the swilt and lively movement exhibited on occasion by small mongooses. 

.MuMjjiU Gray, 1864, Proc. zool. Soc. Land.: 575. Type species, by virtual tautonymy, Hapeslcs mungo 
Desmarest(= Mungos (asciatus Gray). 

Taxonomy. The fact that a vernacular name from India is officially, and properly, 
applied to a wholly African genus is but one slight indication of the muddle that has 
existed in the nomenclature of the mongooses. However, this point of etymology is of 
minor importance compared with the chaos in scientific naming due to misimder- 
standings regarding the proper application of the names Muuiios and Hcrpcsics. It is 
not necessary to recount the tuU details here; those interested should consult J. A. Allen 
(1919b or 1924). Nevertheless, the following brief summary is given as a caution to 
those who might otherwise be misled by the study of bygone papers or books. 

For a reason which today is not altogether clear Thomas, for whom the name in 
its specific form had been "so utterly barbarous" that it could be ignored in favour 
of a later one (1882: 90 fn.), suddenly (1907) substituted the use ot Mhiii^os for Hcrpcstes 
which he had up to a short time previously (1906) employed in connexion with the 
identical species, ijriin//.?. This use of Mungos for species previously known as Hcrpcstes 
was continued by Wroughton (1907). At this time the specific name tnun<;o, imder the 
genus Hcrpcstes, was that widely accepted as pertaining to the common hidian mon- 
goose (e.g. Blanford, iSSS); and Wroughton, expressly stating that this was the oldest 
specific name for tliis species, and without comment replacing Hcrpcstes by Mtiiigos, 
thus introduced the combination Mungos niungo for this Lidian animal. Four years 
later J. A. Allen (1919b) showed that Mungos was "untenable as a genus name for any 
Indian mongoose" and should, in fict, be apphed to the banded mongoose of Africa, 
to which, further, the specific name niungo was correctly applicable. There was there- 
fore a sudden and sweeping change of meaning of the term Mungos mungo from the 
common Indian mongoose to the banded mongoose of Africa. This latter apphcation 
is accepted today without question; but without a knowledge of this history some 
references in literature are liable to misinterpretation. It is, for example, used in the 
now outdated sense of the common bidian mongoose in Pocock's important papers 
and keys (i9l6e and 1919). 

Only one other question of taxonomy arises in connexion with this genus : that of 
whether it properly embraces Crosscirchus as well, either completely synonymously 
or as a subgenus. In this present work the two are regarded as quite distinct, a matter 
that is more fully dealt with imder Crossarchus. 

Distribution and general. The genus Mungos is wholly African and, as at present 
understood, comprises two species, nningo and givnhiiuius, the latter purely West 
African, the former widely distributed, in many somewhat different local forms, 



MUNGOS 249 

throughout the continent south of the Sahara. Neither species enters the closed forest 
as such though they may occur within the nominal forest belt marked on maps where 
the original vegetation has been greatly destroyed or is mere open coastal scrub. 
The genus is well represented in museum collections and may be reckoned as moderately 
common in the field, gamhianus in West Africa, mioi^c elsewhere in the continent. 
The latter species, however, seems from the paucit)' of specimens collected, to be rare 
in the territory covered by this present work, its centre of distribution lying, apparently, 
far to the south-east. 

Descriprion. The two species differ markedly from each other in superficial appear- 
ance, the one being conspicuously cross-banded, the other an irregular mixture of 
colours. Apart from this, they are very similar in general form, being both of smalhsh 
size and stout, with short legs and a short tapering tail. The contour dorsal pelage 
consists of annulated bristle-hairs; in mungos the various corresponding colour zones 
fall together thus giving rise to a marked and very regular transverse pattern of alter- 
nating light and dark bands; whereas m fiamhianus they are randomly dispersed and 
consequently result in a diffuse speckling. Both species in form, and one [gamhianus) 
in general colouration as well, closely recall Crossarchns, and it is for this reason that 
the latter has by many authors been identified with Miingos. But, the marked pattern 
of nmngo aside, there are clear differences. Externally, the most readily observed of 
these arc, in Muiigos, an almost complete lack of underfur; absence of an extra-long, 
overshot snout; and a hindfoot whose sole is naked to the heel. Further, more detailed, 
external description will be found under the two separate species. 

Skull (fig. 34). This belongs to the category of small-sized skulls having a condylo- 
basal length of less than 75 mm. Two other West African genera fall in this category, 
GalercUa and Crossarchns. From the former, Mungos can be immediately distinguished 
by the more forward positioning in its jaw of the upper carnassial; in Mungos the 
posterior outer comer of this tooth, /)'', is situated anterior to the posterior root of the 
maxillarv' process, with all or a great part of in^ also lying in front of this point. 
Differentiation from Crossarchns is a more difficult matter; there is little that is absolute 
to go on, it being almost entirely a question of comparative lengths or proportions. 
The difficulty is added to by an almost complete lack of West African imtngo skulls 
in the British Museum, data in respect of this species consequently having to be taken 
largely from extralimital material, which amongst itself is pretty variable. The slight 
differences can best be appreciated by a comparison of fig. 34 with fig. 37. The Mungos 
skull is mostly somewhat broader; of 11 adult specimens measured only one had a 
zygomatic breadth shghtly under 53 per cent of the condylobasal length; whereas 
Crossarchns rarely exceeds 52 per cent and is almost always less. But the most obvious 
difference in the dorsal view is the long narrow rostrum of Crossarchns in contrast to 
the noticeably shorter, broader and blunter structure of Mungos. This discrepancy of 
length is, in skulls from which the suture has not been obliterated, almost always 
reflected in the mid-line length of the nasals: 16 mm or more in Crossarchns, 14 mm or 
less in Mungos~hut there arc some extralimital exceptions to the latter. Also clear is 
that the postdental palate, where it becomes parallel-sided, is in M. gamhianus and the 



250 



THE CARNIVORES OF WEST AFRICA 




Fig. 34. Mun^os gamhianus: skull, B.M. No. 36.10.30.12, ,^, X .j' 



MUNGOS 251 

type of M. caurinus always conspicuously shorter than broad, though in other M. mungo 
there are exceptions; in Crossarchus the two measurements arc about equal. 

The general appearance of the Munt^os skull varies slightly. The postorbital con- 
striction may be equal to or markedly more or less than the interorbital breadth. The 
postorbital processes are mostly short and sharp, never, even in the oldest skulls, joining 
up with the jugal processes to form complete orbital rings. A sagittal crest may be 
well-developed but in most available West African specimens it is either poorly 
represented or completely lacking, including well mature or oldish skulls. A supra- 
occipital crest is clearly present even in very yoimg skulls and becomes broad and 
flange-like in fully adult animals. The depth of the bones forming the zygoma is, 
for mongooses, relatively shallow. The anterior portion of the bulla is smaller than the 
well-inflated posterior chamber and carries on its ventral aspect a marked transverse 
depression and fissure covering the greater part of its breadth; and this makes the 
auditory meatus a flat oval. 

In Muiigos there are, apparently with complete constancy, only 3 premolars on each 
side, above and below. The outer cusps of the upper camassial are low and without 
any marked cutting function; its inner heel, as with those of the two molars, 
terminates in an almost equally high sharp cusp, and the whole cheek dentition appears 
adapted more to an insect diet than to anything else, tn^ is a fairly big tooth, generally 
somewhat wider transversely than the camassial; nfi is commonly wider across than 
the lingual section of m^ but in exceptional cases is much reduced or may be lacking 
from one or both sides. In the lower jaw the camassial is a tooth of little or no more 
importance than any of the others; in young animals it carries three small, subcqual 
cusps, the two lingual ones closely juxtaposed. These cusps soon become obliterated 
with wear, as do those of »i2 also. In both jaws the canines are, for the Camivora, 
relatively short. 

Habits. What is known of these will be found below under the two separate species. 
No general generic account can be given; for while there exists a number of observa- 
tions relating to the continentally-spread mungo next to nothing is recorded of the 
purely West African gamhianus. 

MUNGOS MUNGO (Gmelin) Banded Mongoose 

Viverra mwifio Gmelin, 1788, in Linnaeus' Sysft-ma Naturae, 13th edition, I: 84. Type locality' given as 
Asia, but fixed by Ogilby (1835), Prcc. zool. Soc. Land. : loi, as Gambia; though Thomas (1882) beheved 
it to be more probably the eastern part of the Cape Province of South Africa, and Roberts (1929) 
wrote that "By common consent the type localit)' is taken to be Natal . . .". The name is derived from 
the Marathi vernacular mangiis, being simply another form of the generic name. 

Herpcstcs fasciatiis Desmarest, 1823, Dicnotmaire des Sciences iiaiurellcs . . ., 29: 58. Type locality given as 
India, but as this was merely a renaming o{ immgo, above, the same considerations apply. The specific 
name is a Latin word meaning enveloped with bands, given in reference to the pelage pattern. 

Herpeslcs gothnch Heuglin & Fitzinger, 1866, Sher. Akad. IViss. IVieii, 54 sect, i: 560. Type locality 
Kordofan. The specific name is the Arabic term for this animal. Possibly usable subspecifically for 
certain West African specimens. 

Crossarchus iMoti Thomas & Wroughton, 1907, Ami. Mag. nat. Hist. (7) 19: 374. Type locality Bomu, 
north-east Nigeria. This was named after its collector P. Amaur)' Talbot, at that time a member of 
Boyd Alexander's "Niger to the Nile" expedition. Available subspecifically. 



2J2 THE CARNMVORES OF WEST AFRICA 

Crossarchus fasdalus mandjarum Schwarz, lyij, }b. imssai<. Vcr. Naltirb. 68: 63-^4. Type locality Bate, 
near the Uham River, Cameroun. This race was called after the Mandjia tribe dwelling around the 
headwaters of the Shari River. Available subspccifically. 

Mimgos aiurmis Thomas, 1926, Ann, Mag. nat. Hist. (9) 17: 1S2-183. Type locality Gunnal, Portuguese 
Guinea. The name is from the Latin cmmis, the north-west wmd, because of the situation of the speci- 
men's source relative to the main distribution ot the genus. Available subspccifically. 

Distribution. Like several other mongooses — of the genera Hcrpcstcs, Atihix, 
Ichncwniit and GidtnlLj — Miin^^os nuini^o has a wide distribution over the continent 
south of the Sahara ranging from Senegal to almost 30°S. It is an inhabitant of the 
open grass-woodlands, never of the closed forest, and seems to be centred more on 
south-eastern Africa than anywhere else. At any rate, although fairly common in 
that region it is apparently rare in West Africa, whence only 4 skins exist in the British 
Museum. These are from Gumial in Portuguese Guinea (2), liornu in north-eastern 
Nigeria, and Fort Laniy (Chad). But the species is reputed to have a wider distribution 
in West AfHca than this would imply, for specimens were brought to England from 
Gambia in 1835, which caused Ogilby to regard that as the real type locahty; and 
Schwarz (191. <;) described the race numdjimiin from Bate near the Uham River in the 
Central African Republic (about 6''4iN, I7°02E). These extremes embrace a wide 
variety of open-woodland vegetation: Guinea, Doka, Sudan and Sahel. 

Description. With its prommently and uniquely cross-banded pelage this mon- 
goose cannot be mistaken for any other in West Africa (Plate 6). This is one of the 
smaller mongooses, the head & body length being of the order of 380 mm, the tail 
about 230 mm, and the weight i to 1-5 kg. It is therefore just slightly more bulky 
than Crossdrcluis. The pelage is fairly long and slightly rough in appearance in life 
though possibly not quite so much so as m the kusimanse. Li prepared skins it generally 
hes very flat, and this is because it lacks the usual cushioning support of dense underfur, 
this category of hair being almost totally absent. The bristle-hairs are of flat section, 
about 24 to 31 mm long, and are markedly ringed light and dark. There are, almost 
without exception, four regions of aimulation. The long basal section may be either 
pure white or this colour interrupted by one or two indefmite blackish areas. The 
three distal zones are always the same: a deep, blackish-brown ring, an off-white or 
yellowish ring, and a black tip, usually longish. But there is no common length for 
each of these zones, and especially the basal one, even in a single specimen; for since 
the hairs arise at different points of the skin such irregularity is necessary for the corres- 
ponding zones to fall coincidentally from dispersed points of origin and so produce a 
total pelage effect of alternating light and dark transverse bands. 

This cross-banded pattern applies only to the top of the back posterior to the 
shoulders: it fides out on the flanks; and the shoulders, nape and head are an ordinary 
random "pepper and salt" mixture. The hairs of the underparts arc plain or very 
inconspicuously aimulated, and may be very sparse, leaving much of the belly almost 
naked. The hairs of the throat are directed forwards and outwards; those of the chin 
are directed backwards, so there is a pronounced line where the two meet; as there is, 
also, along the side of the neck, backwards from the lowest point of the ear, where 
the outAvardly directed hairs of the throat oppose the downwardly pointing neck pelage. 



Plate 6 




Kiisim;,mc {Crosscrclnii ohsam,>): C.inibMi, 



Mongoose (A/,,,,,,...,.,,,,/,,,,,,,,,); Talbot's IJandcd Monaoos. 



(.\/i(;/i>().< /»(,/,(.,i ,,,//„,„ 



M U N G O S 253 

The head is broad and the muzzle fairly blunt. The largish, rounded ears are positioned 
well to the side of the head. The tail, which is in broad terms about three-quarters of 
the head & body length, is long-haired but not bushy, of a nondescript speckled 
colouring like the anterior portion of the body pelage, but always with a blackish 
terminal zone. It is evenly tapered from base to tip, very much in the manner of 
Crcssiirchus. The legs are speckled ; but the backs of the forefeet and the digital area of 
the hindfeet arc generally, but not always, dark, even blackish. The digits are webbed 
between the proximal joints; the claws are very long, particularly those of the forefeet, 
deep basally and fairly well curved; and the sole of the hindfoot is naked to the heel. 

The scent glands have been described in some detail by both Chatin (1874) and 
Pocock (i9i6e), the former dealing with the male, the latter the female. The sac 
surroimding the anus and the orifices of the glands is subcircular m shape and very 
large compared with most species. The glandular system itself is more complex both 
in external appearance and in structure than in other mongooses; for, besides the usual 
simple pair of organs with single orifices cither side of the anus, there are subsidiary 
glands and ducts. The floor of the deep-wallcd sac is not a plain surface but is complexly 
wrinkled in the male, and in the female comprises integumental folds or depressions. 
Summarising Pocock's description of these latter, there are two pairs of depressions, 
an upper and a lower, above the anus, and a lateral pair situated one each side of the 
rectal orifice and more distant from it. Each of these six depressions contains a small 
glandular pit with several secreting pores, none of which, however, represents the true 
anal glands found in other species. The orifices of these, the normal glands, are to be 
found on each side of the sac close to the inner margins of the two lateral folds. These 
true anal glands arc a pair of large muscular sacs containing a dark-coloured, strongly 
smelling, oily fluid which escapes into the sac. The smaller, subsidiary glands are very 
different, of a more rudimentary nature in the female; but in the male, according to 
Chatin's description, appear to have become more highly developed and to approxi- 
mate to the true anal glands both in their structure and in the nature and amount of 
their secretion. 

Skull. A general description of the Mungos skull has already been given above; 
here is not a great deal to add to this in respect of the species. There is, in fact, in the 
British Museum only one adult West African skull of tnungo from which to derive 
data for comparison with five adult gtuiihianns. Recourse is therefore necessary to 
extralimital material of the former, and of this there is an abundance. Hayman (1936) 
first drew attention to significant differences between the skulls of the two species. 
The most unusual of these concerns the carotid foramen, which in the species now 
under discussion is a largish round opening in the basisphenoid, but which in gnm- 
hiimus is obscured, as described imdcr that species. 

The inwii;o skull appears to be slightly broader than that o[ (^dinhidiius its zygomatic 
breadth averaging about 56 per cent of the condylobasal length as compared with 
only some 53 or 54 per cent in the other species. The posterior projection of the ptery- 
goid process is narrow and club-shaped, terminating in a small knob. In the great 
majority of extrahmital imiiifio skulls measured the interorbital breadth is greater than 
the postorbital constriction, whereas in the available gambianus material it is almost 



254 T'"- CARNIVORES OF WI.ST AIRUA 

alw.iys less. The most obvious distinction between the two species lies m the teeth, 
which in ijii/ii/x'./iiii.'.' are strikingly smaller. This is ni some measure brought out in the 
table ot measurements on page 265. 

Habits. Because it is not imcommon and ticcurs ui large and noisy parties in relatively 
open coimtry the banded mongoose has possibly more frequently attracted attention 
to itself and been observed than any other African species. Moreover, it has often 
been kept as a pet or zoo specimen. There are consequently a good many notes on 
certain aspects of its life and behaviour — though not a single one emanating from 
West Africa. The most complete and up-to-date field accounts are those given by 
C. D. Simpson (1964 and 1966) and Neal (1970). These relate to the animal in the wild 
in the southern and eastern part of the continent and to captive or semi-captive indivi- 
duals but can with little doubt be taken as applying equally well to the forms occurring 
in the territory covered by this present work. The notes ^\■hich follow herein are 
indebted in large measure to the patient work of both these observers. 

The banded mongoose is probably wholly diurnal in its activities. Like the kusimanse 
It limits in companies which as a rule number between half-a-dozen and a score. 
Occasional packs of 30 or 35 have been reliably reported; but now and again claims 
to have seen very much larger congregations, even up to 100, have been made. If 
these estimates were accurate they concerned very exceptional gatherings. It is, in any 
case, generally no easy matter, except at the nocturnal shelter itself, to count even a 
moderate-sized band or merely to estimate their numbers with close accuracy since 
at any given moment a large proportion of the part}' is on the move and often in and 
out of the undergrowth. The members of the unit keep fairly close together but not 
m tightly packed formation unless they are rravelhng purposefully from one given 
point to another, say a new nest or fresh feeding grounds; or if danger threatens, when 
they bunch together or fall into line, head to tail. The younger members of the group 
may from time to time lag behind the rest, delayed by play and a lack of that fervent 
application to the business of food-finding commonly exhibited by their elders. 
Realising their separation they hasten forward and often actually run ahead of the 
main body. Whether, in point of fact, a single individual ultimately determines the 
general direction taken by the parry is unknown; but since it appears that some sort 
of dominance is recognised between different members of the community it seems at 
least probable that there is a tacitly admitted leader whose movements are instinctively 
followed. 

Throughout these foraging expeditions a fnrly continuous twittering is kept up, 
partly the outcome of the excitement of the hunt, and pardy as a means of keeping 
together by the creation of a communal spirit. Pocock (1916c) thought that it was one 
of the functions of the anal glands in the gregarious mongooses to assist in holding a 
group together by scent. This may be true in the more static sense of recognising and 
clinging to known friends rather than risking the society of strangers ; but while on 
the move it seems certain that sight and hearing, both highly developed senses in this 
species, play the greatest part in maintaining close contact between all members of the 
company. This, at any rate, was the opinion of Simpson. Nevertheless, should a few 
get cut off from the main body for any length of time, by being frightened into taking 



MUNGOS 255 

cover, Ncal observed them to rejoin the main parry by pursuing exactly the same 
intricate route as that taken by it, a procedure that could only be accomphshcd by 
scent. The combined voices of a party of banded mongooses on the forage do, in fact, 
carry a considerable distance. When it is necessary to sound an alarm or to demand the 
full concentration of the pack the twittering takes on a shriller, more urgent note. 

Foraging for food in the banded mongoose is no leisurely, part-time affair as it 
has the appearance of being in so many carnivores. It is a highly active, continuous 
performance, cverv' fully adult member of the unit displaying not only an astonishing 
voracity but a dedicated eagerness and keen inquisitivcness as well. The party fans out 
slightly and ever)- root, grass tuft, fallen log, rock, hole or crevice within the ambience 
of each individual is determinedly explored, surface litter raked apart and such stones 
as are readily movable turned over in the search for insects, their larvae, spiders, worms, 
snails, lizards, mice, eggs, fledglings and other small creatures that may help to satisfy 
seemingly insatiable appetites. Travel is therefore slow. An acute sense of smell enables 
insect grubs, caterpillars and chn,'salids to be sniffed out though well hidden below the 
soil or in tussocks. These are then rapidly dug out with the long claws of the forefeet. 
Droppings of large animals are of particular interest and eagerly broken apart since 
they arc often the source of a plentiful supply of insects, especially dung-beetles and 
sometimes masses of dipterous larvae. Because of their richness Neal found dung trails 
to be an important factor in determining the direction of the himt; and it is possible that 
the paucity of such animals as elephants, buffalos and other large game in West Africa 
accounts in some measure for the relative scarcit}' there of the banded mongoose. 
Mostly, each individual hunts purely on its own behalf, quickly gobbling down what it 
discovers before predatory interference from others can take place. But occasionally 
larger prey is the object of communal attack; Simpson records having once wimessed 
a pack of about 10 of these mongooses kill a large sand snake — Pstiniiiiophis sihilmis, 
a species that occurs also in West Africa. Li such an attack the mongooses fluff out their 
fur, dash in for a quick bite at the snake's body and leap clear before it can strike. 

Simpson's statement that prey such as mice are killed by shaking, dog fashion, 
is interesting as running contrary to the habit of the in many respects similar kusimanse, 
in which Ewer (1968) observed that shaking was not carried out. Larger prey that 
cannot be demolished at a gulp by one individual is pounced upon by the pack in 
general, each member that can get at it ripping off as big a chunk as it can and carrying 
it off to consume at leisure apart from its companions. There seems no doubt, however, 
that insects are the basic food, and vast niunbers must be eaten in the course of a day. 
In captivity a wide range of human foods is accepted; but in sharp contrast with the 
kusimanse, fruit was found to be not very acceptable to captive specimens of muu(;o 
though it is said that, as in a number of other small carnivores, fallen fruits form part 
of the natural diet. If water is available these animals drink, by lapping, once or twice a 
day; but though the species has often been held to favour the vicinity of water, com- 
panies have been observed in prett)' arid territory where bodily water requirements 
must be satisfied by extraction from food. It is possible that in such localities wild 
fruits play a more important role. 

No one has yet succeeded in following a pack continuously throughout the day, 



2<,f> Till; t AIlNIVdUF.S OF WEST AFRICA 

tliough both Simpson and Ncal have observed dieni for long periods, the latter with 
binoculars from vantage points. It is not certainly kirown, therefore, whether or not 
hmiting is, in fact, unremitting throughout the entire daylight hours. For any animal 
to be persistently and so vigorously active as banded mongooses from dawn to dusk 
would be a severe tax of energ)', and it seems at least possible that they, like so many 
other animals, large and small, nocturnal as well as diurnal, suspend their pursuit of 
food for a period, or periods, of rest. Be this as it may, at the end of the day, at least, 
the entire company seeks the shelter of a common warren. This commmial resting 
place may occasionally, m suitable territory, be more or less on the surface amongst 
boulders, tree roots or other convenient cover; but more often it is subterranean, 
having as its basis a burrow originally dug out by some larger fossorial animal such as 
an aardvark, or, more commonly, a hole beneath a deserted termite mound. This last 
certainly seems to be the favoured site, particularly if it is on a slope affording a clear 
view roimd, has partial shrubby cover and a not too distant water supply. Entrances of 
convenient size are dug or enlarged with the long-clawed front paws which cast the 
excavated earth backwards between the hindlegs much in the common hcrpestine 
manner of "egg-throwing". 

Once the general appearance of these warrens has been grasped it becomes, according 
to Neal, fairly easy to pick them out. If one has been in occupation for any length of 
time it can be readily identified by the piles of droppings on its summit and scattered 
for some distance around. There are generally two or three separate entrances and 
several feet of internal tiumels leading to a central chamber. In the deserted warren 
excavated by Neal this chamber measured some i -o by 1-5 metres and had a mid-height 
of 0-5 metre. There were also two side tunnels leading to small chambers; and it is 
likely that this special accommodation is for the use of breeding mothers. 

It would appear that more than one commmiity may occupy a single warren; for 
Shortridge (1934) mentions "an unusually large colony of at least thirry individuals" 
that daily divided into several troops which hunted independently. A warren may, 
by and large, be occupied for a long period of months; though not necessarily con- 
tinuously, for the pack may, possibly when food runs short in the immediate vicrmty, 
temporarily desert it for other warrens within its hunting territory. Simpson observed 
one pack to use three different warrens located within an area of more than a square 
mile. Besides these more or less permanent nocturnal shelters there may be within a 
territory other temporary "bolt holes" to which a foraging company can retreat in the 
face of danger; and it is at least possible that such established refuges, above or below 
ground, may be used for a recuperative phase of inactivity during the heat of the day. 
Two general points concerning warrens may be mentioned here. The first is that a 
good deal of noisy activity, clearly audible from above, takes place below ground; 
and the second, that Simpson found the entrances to abound in fleas. 

The extent of a hiuitmg range has not been closely investigated and there is nothing 
to indicate the sort of area covered beyond Simpson's incidental observation just 
quoted. Nor is there much information concerning scent-marking of boundaries. 
Neal makes no reference to this at all; but Simpson recorded that there is little activity 
of this kind during food foraging but that it was common at drinking sites. The methods 



MUNGOS 257 

of registering scent signals follow the usual pattern : pressing the anal pouch against 
convenient trees, logs or rocks, sometimes at ground level, sometimes by raising the 
hindquarters well into the air; and, besides this dabbing, marking by dragging the anal 
region across the ground. 

According to Snnpson (1966) the birth and nursnig of the young takes place "in a 
grass-hned chamber deep in the warren". Whether this statement is derived from actual 
excavation of the site or is deduced from the fact that a tame female offered several 
alternative possibilities for a natal chamber always chose a dark box accessible through 
a tumiel in a mound of soil is not clear. Sexual maturity occurs certainly in females 
at the age of 9 or 10 months, possibly a httle later in males, hi heat, a good deal of the 
more active, dehberate courtship seems to be carried out by the female, lying on her 
back and wrestling with the male, or flattening herself against his back and rubbing 
her vulva through the fur. Both sexes, however, become playful, jumping around, 
pouncing on each other, or nudging with the head; and during this the male also 
"scents" the female, though seemingly less designedly than in her case. His anal glands 
become enlarged and exude quantities of a white secretion with which the female 
unavoidably becomes covered during this precoitional romping. Finally, the two 
animals generally chase each other roiuid and round in ever narrowing circles, or the 
male may circle the female, his tail held high. Neal observed mounting to be repeated 
three or four times, with a short chase between each; and on one occasion a second 
male, which had been watching throughout, mounted soon after the first male had 
separated. 

The period of gestation is in the nature of 2 months, its precise length being as yet 
undetermined. It is not possible to derive information regarding a preferred breeding 
season in West Africa from the limited material available ; but it seems probable (C. D. 
Simpson, 1966) that, like the kusimanse, these mongooses are capable of breeding more 
or less conrinuously. From 3 to 5 young seems to be the general number at a birth; 
but a litter of 6 has been recorded. They are born blind, with a dark skin and a fine 
transparent pelage which, however, faintly reveals black cross-banding on shoulders 
and rump. The eyes open about the loth day. At between 2 and 3 weeks the pelage 
begins to become more apparent; and it starts to assume adult appearance after 6 weeks, 
though it does not attain its mature coloration imtil the age of 3 months. The young 
appear to be held in common; for in one community Neal observed that when the 
pack left on the daily forage one female remained behind to suckle and look after the 
eight young. And on return of the parry to the warren in the cvcnuig all the females 
niu-sed any of the young without distinction. 

The weight at birth is about 20 g but this increases very rapidly, though the figures 
published by Simpson (1964) exhibit an extremely wide variation between different 
individuals. One young male achieved what may be regarded as a fully adult weight 
of 1-75 kg as early as about 5 months; an adult male, on the other hand, is given as 
only 155 kg; but, an adult female reached the exceptionally high weight of 225 kg. 
As regards longevity, one banded mongoose is known to have lived in captivity to the 
age of II years; and another of 8| years. 

Several miscellaneous matters of behaviour must be briefly glanced at. In connexion 



258 



THE CARNIVORES OF WEST AFRICA 



With feeding it is interesting to note that the backward egg-hurhng procedure common 
to several members of the siibtamily is well confirmed in inmijio. It has been observed 
and illustrated by C. D. Simpson (1964 and 1966), Davis (1966) and Kiiiloch (1964). 
The force with which this is carried out is surprisingly powerful; nevertheless, an egg 
docs not always crack sufficiently at the first attempt, in which case the action is deter- 
minedly repeated until success is achieved. This procedure is, however, not reserved 
solely for eggs but is employed with other desirable but encased foodstuffs, some of them 
necessarily strange to the mongoose; and, moreover, as KinJoch has shown, is applied 
to a variety of quite unprofitable articles, including even a bimch of keys. One obser- 
vation involving this behaviour, made by Davis in the Bronx Zoo, is of especial 
interest in as much as it concerns large millipedes so commonly occurring in Africa. 
A banded mongoose, discovering after repeated attempts that its teeth were imablc to 
pierce a cc-iiled up specimen of one of these, fmally hurled it dirough its hiudlegs against 
the wall twice, thereby cracking the hard chitmous coat sufficiently for its teeth to 
effect penetration. This achieved, the millipede was then readily consumed. Neal, 
in fact, found these myriapods to constitute a very important item in the dietary in 
Uganda, imini^o droppings invariably showing clear and abundant remains of the 
chitmous shell. 

Lrke others of its kind this mongoose is almost wholly terrestrial and probably 
normally has little urge to climb for any distance up vertical tree trunks. Its long, almost 
straight claws are scarcely adapted to this; and more especially to getting down again. 
Kinloch describes a few rather inept attempts at tree climbing made by his pet /;i/»njp, 
in which, it is true, the upper branches of acacias were attained but descent thence 
found too nerve-racking to be imdertaken. Simpson, too, in his opening paragraph 
(1964) makes some mention of tree chmbing; but this may refer to a pack of banded 
mongooses that, escaping from hunting dogs, scrambled to the top branches of a tree 
that had been pushed over by elephants. The sloping trunk and interlacing branches 
of this would, of course, considerably lessen the difficulty, bodi of getting up and of 
getting down. Wliere there arc positive footholds such as offered by the wire netting 
of a cage it is known that iiningo can reach a good height from the ground. Apart from 
an ability to scramble over low obstructions it can leap onto them provided they are 
not more than 500 mm or so high. Though banded mongooses often live near water 
Simpson says that they never seem to take to it of their own accord and are, in any 
case, poor swimmers; on the other hand, Dalton (1961a) records how a pet of this species 
leapt into a flooded river in order to rejoin her on the other bank and successfully 
swam across. 

Play is an important factor 111 the lives ot all young carnivores, and tins fact is con- 
stantly illustrated by juvenile and subadult banded mongooses, both in the vicinity of 
the burrow and while actually on the move durmg forays. Play follows the common 
pattern of mock-fighting, chest to chest wrestling, tail-chasLng and scampering after 
each odier. Simpson found, however, that play does not occupy much of the adults' 
time. These, basking in the early morning sun or resting around the wan^en after the 
day's hunt, go in for a good deal of grooming of dieir own or other adidts' fur. At 
diese times, too, but especially in the morning after rising, they like to stretch their 



MUNGOS 259 

limbs out quite straight tore and aft, lying with their bellies pressed to the soil. Play 
and other intercourse within the group, however, whatever it may appear on the 
surface is not necessarily on a socially equal basis; for domination of one animal over 
another is established, in the observation of Simpson, by putting the foreleg across the 
shoulders of the inferior animal and seizing this latter by the back of the head lightly 
in the jaws. Such ascendency by no means always belongs to the male. Simpson noted 
female dominance in two different groups, in one of which a female established her 
position over two males of the same litter at the age of about three months. 

All who have studied banded mongooses in the wild or had dealings with them at 
closer quarters have been impressed by the obvious acuteness of the senses of sight, 
hearing and smell. These animals, though recklessly courageous in the face of close 
danger, are at all other times highly suspicious of possible attack and timorous to the 
point of fleeing for cover at the least imagined threat. This caution and an obsessive 
curiosity that is often vital to the discovery of food hidden 111 soil or vegetation lead to a 
constant exercise of these highly developed senses. Both at the warren and frequently 
during the course of the daily excursion one, or sometimes nearly all, of the group will 
sit upright on the haunches, or if necessary stretch up fully erect on the hindfeet, and peer 
round, intently watching and listening. High sounds seem to make more impression 
than low; and there has been no suggestion that crackling is a terrifying noise as it is 
to the kusimanse. Simpson noted that pet animals could pick him out visually at some 
distance, and that lizards cUmbing trees or insects flying away were followed by sight. 
Neal found that this constant vigilance and acuteness of perception rendered close 
observation of a pack a matter of difficulty since a stationary Land Rover or other 
strange object at about 40 metres aroused nervous suspicion, and the shghtest movement 
was detected at 20 to 25 metres, at which distance, also, the click of a camera attracted 
attention. 

A display of wariness can be seen on the occasion of leaving or return to the home. 
The banded mongoose is not so early a riser as other diurnal animals, mostly making 
its first appearance an hour or so after sunrise rather than in the grey light of dawn. 
Before actual emergence from the warren the immediate vicinity is carefully scrutmised 
from the mouth of an ennance hole, and if all looks clear one or two adults come out 
and, if it is an old termitarium, moimt to the top of the nest and stand or sit erect 
scanning the surrounding territory. Then, if no danger threatens, there is a general 
exit followed, as a rule, by a period of play, siui-basking, yawning, stretching, grooming 
and defaecation before moving off^on the day's forage. The return home is commonly 
effected a little before night closes in; though should the afternoon be heavily overcast 
the pack, possibly deceived by the poor light, may arrive at the home territory con- 
siderably earlier. The returning party never enters the warren direct but pauses to rest 
and relax at a little distance from it. Thence, approach to and actual entry into the 
nest is carried out in gradual and cautious stages, puncuated by the usual play and 
grooming activities, irntil the final bolt down an entrance tminel as night falls. 

Banded mongooses have a range of different calls and notes. That most commonly 
heard is the rather bird-like twittering of the whole pack when on the hunt. This 
may change when danger is suspected to a strident note of alarm; but if a single in- 



260 THE CARNIVORES OE WEST AFRICA 

dividual is scared or threatened it spits radier like a cat and growls, and if this proves 
ineffective it alters to a staccato chatter ot rage — or, rather, ot blind fury, for at these 
moments one of these mongooses will hurl itself without reserve in attack at whatsoever 
has aroused its anger. At the other end of the scale, in moments of pleasure the note 
uttered is something of a low whine or even a kind of soft purr. 

The species appears to have few natural enemies apart from birds of prey, which it 
quite obviously holds in considerable fear; for a hawk may swoop swiftly from the 
sky, seize an mdividual and be safely away without the possibihty of counter-attack; 
whereas the combined tooth-power and lightning agihty of 20 or 30 enraged mongooses 
must be a considerable deterrent to almost any ground predator. Nevertheless it is 
said that the larger carmvores such as lions and leopards will pursue and attack a mun^o 
hunting party. As some sort of defence against assault from the sky, of which their 
keen eyes give them early warning, these mongooses instinctively bimch together m 
tight formation; and this may possibly have the beneficial effect of deceivuig hawks 
and eagles, making them more hesitant of attacking an apparently large body than 
they would be of attempting the comparatively simple task of picking up isolated 
individuals. 

The majority of mongooses, obtained sufficiently young, make excellent and 
fascinating pets, and Mungos mungo is no exception. These animals, however, cannot 
be kept confmed to small cages but must be given more or less free range. This being 
the case it shoidd be fully realised before taking on the responsibihry for such a pet 
that, while the effort is usually very rewarding, a great deal of time and trouble are 
unavoidably involved m feeding, play and general watchfulness, together with good- 
humoured patience in the face of inevitable accidents to possessions due to an insatiable 
curiosity, and restraint sometimes in response to slight personal injury, the outcome 
of an incalculable uncertainty of animal temper. The banded mongoose is at heart a 
very friendly creature, with its owti kind, with humans, and often with dogs; m 
domestication it becomes trusting and, at least apparently, affectionate. It is by nature a 
cleanly animal, easily house-trained; and its bodily smell is, as a general rule, shght 
and moffcnsive. Its feeding is a matter ot simplicity since it avidly accepts a very wide 
range of foods. All this has been vividly brought out by Kinloch (1964)- 

Taxonomy. A species such as imin{;o with a pelage composed predominantly of 
light and dark ringed hairs must inevitably display a range of colour variation, within 
a smgle population as well as, iir the wider aspect, in response to differing ecological 
conditions. No less than 16 subspecies are, mdeed, hstcd in G. M. Allen's African 
Checklist (1939) ; and although, in the type descriptions of these, general body and tooth 
sizes receive occasional reference and rarer emphasis, the great majority of the forms 
are, m fact, described almost purely on the basis of colour, and in almost every case 
from one or two specimens only. 

Without question, clear differences of colouring exist between specimens; but even 
now, after many years of collecting, the amomit of material available for study is for 
the most part quite insufficient to judge reliably the constancy of colour in any given 
locality, and hence the real validity of presumed races. Thomas, erecting CMrinus on 
the basis of two differing specimens, himself thought that "probably it will be foimd to 



MUNGOS 261 

grade hereafter into others of the banded forms . . .". Certainly as far as West Africa 
is concerned it is futile to pretend to the accuracy of deterniuiation implied by a tliird 
name. Four specimens alone exist in the British Museum from the region. Two of 
these, from a single locality and both ascribed by Thomas to caurinus, one being the 
type, differ very distiiictly from each other in coloration. The other two, from well 
separated locahties but both ascribed to talhoti, one the type, also differ from one 
another, possibly due to age, possibly not. Until a great deal more material has been 
gathered together for study it seems pointless to go into fmer division than the species 
and imply a degree of understanding that does not exist. 

Moreover, for practical purposes in a work such as this present it is not possible to 
go any further. The plain fact is that with only a couple of sharply different forms to 
differentiate or ^vith comparative specimens actually in the hand such expressions as 
"paler" or "darker", "pinker" or "buffer" may well be full of meaning; but with a 
range of several dehcate nuances it is an almost impossible task to convey absolute 
colour to a student possessed of a single skin. Very few users of this present work 
have access to standard colour charts; and even with these at hand the task is still fraught 
with difficulty since the selection of colours in them is too limited to find an exact 
match for the slight nauances involved. Li any case the colours themselves vary quite 
appreciably in different parts of the pelage. 

However, since certain names have been connected with West Africa the foUowing 
notes are given for what they are worth. At the outset it must be pointed out that 
there is uncertainty regarding the precise appearance of even the nominate race since 
three different African localities have been assumed to be the type locahty. If Ogilby 
was right in thinking tliis to be Gambia — and there is no positive reason for supposing 
his guess to be less accurate than those of Thomas or Roberts — then caurinus might 
fall into the category of a synonym. 

Mungos mungo gothneh (Heughn & Fitzinger) Kordofan Banded Mongoose 

This was described from Kordofan (Sudan) in the Sahel zone of vegetation but has 
sometimes, with possible justification, been assumed to range westwards through this 
same belt at least to the eastern side of Lake Chad. If it does this there is no reason why 
it should not range yet further. The only clue to its appearance given m the type des- 
cription, and that almost useless, is that the white colour of the imdcrside is less well 
developed than in zebra, an Ethiopian species. Measurements of a few Sudanese speci- 
mens assumed, from their locahty of origin, to be gothneh are given in the table on 
page 265. 

Mungos mungo talboti (Thomas & Wroughton) Talbot's Banded Mongoose 

The brief distinguisliing features of this race (Plate 6) from Bornu (north-east Nigeria) 
were given as medium size and very pale coloration — though Schwarz later wrote 
that it was not always so pale as Thomas & Wroughton had said. What specimens 
Schwarz had as a foundation for tliis statement, and on what grounds those specimens, 
differing as they must have done from the t^'pc, could be held to be tallwii is not clear. 
A second specimen in the British Museum from Fort Lamy, and determined two years 
later as this race, is, in truth, somewhat darker; but it is a juvenile and not strictly 
s 



THE CARNlVOliliS OV WEST AFUICA 



comparable, liirthcr, the two may, thougli not certainly, have conic h'om dirterent 
vcL!;ctation zones; Fort Laniy lies hi the Sahel, Aduia raddiana, type of vegetation, 
but Bornii is a district covermg a large area which incluties both Sudan and Sahel 
woodland. Comparison was made in the type description with sonudiais, the general 
appearance being characterised as even paler than in that race; and udhoti was said also 
to have a black tip to its tail, a feature that certainly does not show very clearly today 
in the type, the end of the tail being missing. 

Mungos imingo niandjariini (Schwarz) Schwarz's Banded Mongoose 

Bate, whence the type specimen of this race came, is in the Central African Republic 
(6'4iN, 17"02E) and in a much moister vegetation zone, Doka woodland, than is 
usual for this species. Possibly for this reason its coloration is duller: Schwarz described 
the form as ver}' like i.tlhoil but at once distinguishable by its darker, more yellow or 
browiiish ground colour, especially on tlie head and neck, and feet that arc quite 
bkick. He also ascribed to this race two odier specimens trom i .So km further south, 
from the source of the Pama River, which flows into the Ubangi and is hence just 
extralimital to this work and in die (niiiiea woodland zone. 

Muiigos mtingo caurinus Thomas North-west Banded Mongoose 

Thomas described this race as "rather small"; the skull of the t)pe, an old animal, 
m fact exlubits little diflerence of measurement from Sudan specimens presumed to be 
gotlinch. The general colour was said by him to be "nearly as dark as in typical M. iimngi.\ 
quite unhkethat of the pale M. lalboti of Bornu". hi the r\'pc the colour is certainly 
relatively dark; but Thomas omitted to say that the specimen was in moult and that 
in consequence tlie bands are obscure or entirely lacking from the posterior halt of the 
back. A second specimen, yoiuig, taken at the same place (CTiinnal, Portuguese Guinea) 
six weeks carher is very different; and if it is compared with the type q{ taihoti it will 
be seen that though the flanks oi taihoti are paler the backs match exactly. The material 
available is quite inadequate to form the basis of any expression of opinion on the validity 
of this proposed race, the type locahty of which lies in the Guinea woodland. 

MUNGOS GAMBIANUS (Ogilby) Gambian Mongoose 

Hcrpistcs ^anihiamis Ogilby, 1835, Pioc. zool. Sot: Loud.: 102. Gambia. Type m tlic liritisli Museum 
No. 55. 12. 24.22ft, 9; skin in fair condition but probably much faded since it was once mounted and 
on exhibition, and with h.ilfthe tail missing; skull much damaged. 

Distribution and general. Although there has been some confusion regarding the 
cenus of this mongoose, it having tor a long time been held to be Crosfardius — and, 
mdeed, so classifie'd in G. M. Allen's Checklist (1939)— there seems never to have 
been the slightest question respecting the species. Quite difterent m general appearance 
from its strikingly cross-banded near relative imiu\i,\ it does, in fact, bear considerable 
external resemblance to the more distant C.rossarchus ohsciinn, with which it has in 
consequence quite often been confused both in the field and in the museum (Plate 6). 
The essentia! distinctions between the two are brought out m the section devoted to 
this latter species and need not be further entered into here. 



MUNGOS 263 

This is a small-sized mongoose apparently entirely of West Africa, its known range 
being from Gambia to western Nigeria. It occurs almost exclusively in the Guinea 
woodland zone just inland of the high forest but may penetrate into the rather similar 
Doka belt. However, it has also been obtained from Bonthe island on the coast of 
Sierra Leone, the vegetation of which, apart from fringes of mangroves, consists of 
sand ridges and sparse grass. How the species got there is an interesting speculation; 
but its presence there implies that it may also be discovered in other stretches of coastal 
scrub such as the Accra plains, Ghana. G. S. Child (private commiuiication) records 
that packs are not i:ifrequcntly to be seen in the dry-season throughout the Borgu 
Game Reserve area in western Nigeria. 

The actual places from wliich the 1 1 British Museum specimens came are : Gambia, 
unspecified (2); Sierra Leone, Bonthe and Dumbaia; Ghana, Ejura (4) and Bole; 
Togo, Kete Krachi; Nigeria, near Shcpeteri, Ogun Forest Reserve. OiJy 6 of these 
are adult. 

Description. The overall colour o( qiwihianus varies somewhat with the area from 
which the specimen comes ; but the broad general appearance is that of a smalhsh, 
dark brown, heavily speckled mongoose, with a head & body length of some 350 mm 
and a tapering tail measuring about three-fifths of this. The face, though not exactly 
blunt, lacks the conspicuously long protruding snout so characterstic of the rather 
similar-looking kusimanse. 

The pelage is fairly long and fairly harsh. On inspection it will be found, like the 
other species of this genus, iniingo, to lack any obvious sign of underfur below the 
contour coat of bristle-hairs. These latter are flattish, but not quite so much as in other 
genera. They vary considerably in length as regards extreme examples, from 30 to 
52 mm; but the more representative range is from about 35 to 45 mm. The annulation 
of these bristles is distinctive and serves as one of the points of difference between 
this mongoose and the kusimanse, as explained in the description of that species, hi the 
proximal half there may be four zones of coloration, but these are not always all 
present. Exceptionally there is a basal white portion measuring 3 to 5 mm; the next 
distal zone, present in most specimens, is black and averages 5 to 7 mm; nearly every 
specimen exhibits the next zone, which is white and averages 5 mm ; and this is very 
often, but not always, followed by about i mm of deep yellow, dividing the proximal 
region from the three terminal, and constant, rings. First of these is a black zone the 
length of which varies between different specimens, in a Ghana example averaging 
II mm, in a Nigerian one 18 mm; there is then a golden-yellow subterminal zone 
averaging from 5 to 7 mm ; followed by a fmc black tip which averages 5 mm, is rarely 
less than 4 mm, and may reach 10 mm. Because of the much longer subterminal yellow 
zone the overall pelage colour in (^ambiaims is almost always rather lighter than in 
C. obscurus, the ticking having a considerably coarser appearance than the fme puctula- 
tion of the coat in this latter. Unlike the nearly related timngo there is not the least 
suggestion of the corresponding colour zones falhng together in the pelage and there- 
fore no hmt of any regular transverse pattern. 

The fur on the belly is very scanty and, for the most part, imicolorous red. On the 
throat it is imicolorous white, or yellowish, and directed outwards and upwards 



264 



THE CARNIVORES OF WEST AFRICA 



from a longitudinal nuxli.il parting. Where its direction conies into opposition with 
the shorter, posteriorly directed hair of the neck it curves backwards and forms a more 
or less distinct white line. The short fur with which it comes iitto contact is blackish; 
so a characteristic giiiiihiiiniis pattern is formed along the side of the neck, from the ear 
to the foreleg, of two longitudinal black and white lines, but much clearer in some 
specimens than in others. The rounded ears, set well down on the sides of the head, 
arc clad with short, speckled hair, both inside and on their backs. Since the nose is not 
elongated the depth of the upper lip from the rhinarium to the mouth opening is short, 
quite different from the kusimanse. The arms and legs are ticked with yellow in the 
manner of the back, but the backs of the forefeet and at least the digital area of the 
hindfeet are black. The toes are slightly webbed; the sole ot the hindfoot naked to the 
heel. The evenly tapering tail is mostly speckled but has a longer or shorter black 
termmal region. No one appears to have investigated the form of the anal scent glands 
and pouch. 




Fig. 35. MiiwiiL^s i^anthiaitjts: huWa, ■ 2 

Skull (fig. 34). This follows the general MtDii^os pattern as described above under 
the generic head. There are, however, certain points of difference from nmngo. The 
most interesting of these, to which Hayman (1936) first drew attention, is the existence 
in i^ainhiivnis of a projection of bone from the antero-intemal angle of the bulla which 
at least partially covers and obscures the relatively small carotid foramen in the basi- 
sphenoid, with which bone it eventually becomes fused around the latero-anterior 
rim of the foramen (fig. 35). This structure appeared to Hayman to be unique in the 
carnivores; but a similar bony projection is, m fact, to be found from time to rime in 
other mongoose skulls, especially in }hriHsiis ichneumon, as for example m B.M. No. 
95.9.4.6 and B.M. No. 33.3.3.5. It can be seen also, though less well developed, in the 
type o[ Crossarchus iinsor\;ci, B.M. No. 10.4.8.7. Judging from the slender material at 
present available, five adult specimens, this is a somewhat narrower skull than that ot 
niun^o, the 2r\-gomatic breadth averaging about 53 or 54 per cent of the condylobasal 
length. The mterorbital breadth is less than the postorbital constriction ui four of the 
five adult skulls, in the fifth the two measurements being equal. Whereas in numfio 
the postdcntal palate is generally slightly broader than long, in '^amhianu'. the difference 
IS more marked, the length being only about halt the breadth. The posterior part of the 



MUNGOS 



265 



The most obvious diffl^rencc bcm^m the two species, however, lies m the teeth 

sTm ",f:r k;;i'rTr/""f ^'^^ smalJ very considerably less in bulk for .n Zt 

imilar sized skull. The visual sharpness of this distinction ,s only partially conveyed bv 

the measurements given 111 the Table below. ^ tonveyea by 

noI^tS;^" 'VTu' '""^ ft[P°'" "°''^^g '' ^''''''' of tl^^^^^ not a single field 
record a in^Tf ''v?' 7^ 'f'''°' '''^ '^^ independent observation havmg been 

XtneaTthe ti"ot T t''^'' "" "^Z" '^ J°""'^ ^P^"'"™^ ^o- Bonthe wa^ 
young near the end of June. Li a personal communication he states that he collected 
another young specimen at the end of September; and young or very yot^^ ones 

t^rw^TheTh'Ih " ^.^-t'^-"gJ--n' and Februa'ry. if is ther^o^e nS even 
known whether this species shares m any way the communal habit of ,,,,.,^0 and the 



Table 13 : Numerical data for species o£ Mimgos 



Vegetation 



Number in mean 
Condylobasal length 
Basilar length 
Palatilar length 
Zygomatic breadth 
Upper cheekteeth breadth 
Nasals, length 
Interorbiul breadth 
Postorbital constriction 
Braincase breadth 

Toothrow (c — 111-) 

p* length 

("1 breadth 

m^ breadth 

"11 length 

"12 length 

Head & body 
Tail 

Hindfoot 
Ear 



mimgo imwgo 

?gotlmch caiirimis 

(ex Sudan) (Type) 



gambiamis 



RATIOS (per cent) 
Tail/head & body 
Zygom. br./condylob. 
Braincase/condylob. I. 
Braincase/zygom. br. 
Palatilar l./condylob. I. 
Interorb./postorb. 



Sahel 


Guinea 


Guinea and 

Coastal 

scrub 


4- 


I 


5 


67-0 


66-2 


69-9 


63-3 


61 -2 


65-0 


33-7 


32-6 


35-1 


37-3 


39-5 


37-4 


23-1 


22-8 


2I"7 


i8-3 


— 


i8-8 


14'0 


14-0 


13-4 


I2-I 


13-8 


14-2 


26-9 


27-4 


28-3 


22-8 


23-1 


22-9 


5-3 


4-4 


4-3 


6-5 


S-6 


5-2 


45 


— 


4-0 


4-7 


4-1 


4-2 


4-4 


— 


3-6 


— 


38s 


350 


•^ 


23s 


210 


— 


72 


69 


' — 


26 


23 


— 


61 


60 


5<S 


60 


54 


40 


41 


40 


72 


69 


76 


50 


49 


50 


116 


lOI 


94 


23-2 


i9-r 


l8-8 



266 TUF CARNIVORES OF WEST AFRICA 

similar kusimansc. It can only be assumed from the small size of the teeth that the 
food is weak and soft, probably almost entirely insects. 

Taxonomy. There is no doubt, as Hayman (193C') pointed out, t\ut (^diiihiiuius is a 
full)' independent species and not merely a West African race of /)ii/»/i;(i. No subspecific 
division ot it has yet been suggested; and though it is true that minor differences of 
coloration can be foiuid amongst the available material these arc certainly too in- 
significant to justify any such distinction. 

Genus HERPESTES lUiger, 181 1 
Typical Mongooses 

Ichncuiihvt L.iccpcde, 1799, TahUan ilcs divisions, sous-ilivisiotis, ordrcs c ( (jcdri's dcs Mammijlrcs . . .: 7. Not 
ol Linnaeus, 1758. This n.ime is Greek and means the tracker, from ichncno to hunt, with reference to 
the reputed hahit of hunting out and consuming crocodiles' eggs. 

Hcrpnlcs Illigcr, iSi [, Prodniiinis systcmatis Mammaliinn ct Avium . . .: 135; this misprinted spelling was 
corrected to Hcrpcslcs in a list of Errata on the tmal page (302, but actually unnumbered). Type species 
Vivcrra ichiicwnon Linnaeus, as designated by J. Anderson, 1878, Anatoniital ami zoolofiital resfiirclies . . . 
of tlic two vxfH'ditioiis to Western Ynnnan . . ., I: 171. The name, as stated by Illiger, means creeper, 
from the Greek hcrpcin, and was presumably given with reference to the animal's stealthy habit. 

There are several other, extraliniital, synonyms. 

Taxonomy. Hirpcstcs is a genus of very disputed scope. Ilhgcr, in first naming the 
genus, gave no country of origin nor any type species — though he did cite the name 
ichneumon, which, as Vivcrra ichneumon Linnaeus from Africa, in fict, very much 
later (1878) became the type species by designation. Meanwhile, it was not long (1823) 
before the name became generically attached to a species from India, quickly followed 
by others from the far east. Thus the position soon became established in which the 
genus was widely regarded as satisfactorily covering all the Asiatic as well as several 
African species, a situation that holds today — though at one time or another attempts 
have been made by different authors to assign the eastern species to other genera (c.i;. 
Mciiiiiiista Horsfield, 1822; Urva Hodgson, 1837; C(ilo(;ak Gray, 1865; Calictis Gray, 
1865). Palaeontologically, Hcrpcslcs is, indeed, regarded as having originated outside 
Africa and to have spread from its source over much of the oriental region and dirough- 
out a great part of Africa, dividing itself mensurally in the course of evolution into 
large species and small species. 

The question of the generic identity of the Asiatic species is necessarily of relatively 
minor interest to the area now being covered; there are much more urgent problems 
concerning the genus in Africa. In that continent, apart from the type species Hcrpcstcs 
idiiicHiiwii, basically from Egypt but, including its southern representative dijjcr, 
spread from the Mediterranean to the Cape, there have been ascribed to the genus 
one other large species, named by dc Winton H. iuiso, and a host of smaller animals 
belonging to a multitude of forms, of which siiiif^tiinciis, ^^rdcilis and ochrMctis may be 
mentioned as of prime importance to the present study. 

To deal with these small mongooses first: all were assigned generically by their 
original authors to Hcrpcstcs as, at least outwardly, nearly resembling ichneumon though 
on a smaller scale. Gray (1865a), however, split up the large assemblage of species. 



IIERPESTES 267 

both African and Asiatic, by that time ascribed to Hcrpcstes into a number of smaller 
genera erected by him for this purpose. H. s<in(iuineus and H. (;racilis were placed in 
C.j/oijii/c, and H. ochramis in Gakrclla. The former genus comprised Indian as well as 
African species; but eventually, largely through the work of Thomas (1882), it became 
restricted to oriental mongooses. Gray's genera did not, in fact, fmd much favour; 
tlie African species continued to be associated with Hcrpcstes (or Mungos, which as 
part of the general nomcnclatural mix-up, was then the more commonly accepted 
name for this group) until J. A. Allen (1924) made out a good case for reviving Gakrclla 
generically to cover all the small African forms. But this name, though widely recog- 
nised and understood by African mammalogists, has been little used in literature. The 
reasons are several. Four years after Allen's paper Thomas (1928) drew attention to it, 
only to express the opinion that Gakrclla was, in fact, apphcable solely to the north- 
east African ochracca, which had been named type species. He consequently erected 
Myonax to accommodate the numerous remaining pan-African forms to which Allen 
had shown the name Hcrpcstes to be inappropriate. Myonax then became the fashionable 
name for the majority of these small African mongooses imtil a few years later Schwarz 
(1935) expressed the opinion that there was no justification for it since there were no 
generically valid morphological distinctions between ochracca and the others, thus 
restoring Gakrclla to the scope that J. A. Allen (1924) had accorded it. However, 
G. M. Allen (1939) in his Checklist did not accept this view and Myonax consequently, 
though somewhat latently, remained cuiTent during the period of relative nomcn- 
clatural inactivity brought about by the second world war. Simpson's Classification 
(1945) pubhshed at the close of tliis period, at once to become the universally accepted 
criterion of nomenclature, sank both Gakrclla and Myonax in Hcrpcstes; and this 
widespread and numerous group of small mongooses thus imderwent its fifth change of 
generic name, Mnni^os — Hcrpcstes — Gakrclla — Myonax — Hcrpcstes, in 20 years. The 
current practice is, following Simpson, to use, at choice, Gakrclla or Myonax sub- 
generically. 

The argument for not generically separating the Asiatic and African mongooses 
under consideration here, as briefly expressed in Ellerman, Morrison-Scott &' Hayman 
(1953: 119 £n.), is that "it should be borne in mind that although the small African 
mongooses appear very distinct from H. ichneumon (the type of the genus) there are 
many more small species of Hcrpcstes in tropical Asia". However, the present author 
accepts J. A. Allen's contention that the small mongooses under discussion merit full 
generic separation as Gakrclla from the large Egyptian mongoose, Hcrpcstes. The 
characters taken into consideration in this and in the rejection oi Myonax will be detailed 
later in the account of the first of these genera. It is now necessary to glance at the other 
large species named by de Win ton Hcrpcstes naso. 

In his preliminary notes on the carnivora collected in the north-eastern Congo J. A. 
Allen (1919b) foimd it necessary to erect a new genus, Xciun;ale, to cover what he 
thought to be a hitherto undescribed forest mongoose, which he named microdon. 
Hayman (in Sanderson, 1940) satisfactorily showed that this species was identical with 
de Winton's long-established Htrpcstcs naso; the question remains simply one of genus 
— is Xenogalc validly separable from Hcrpcstes at generic level? In erecting his new 



268 THE CARNIVORES OF WEST AFRICA 

genus Allen adduced a number ot differential characters, both external and cranial, 
that he regarded as sufficient to separate AVmnj.i/i' from Hcrpcstcs, khncwnia and Atilax, 
to each of which the new mongoose bore some measure ot resemblance. These charac- 
ters will be entered into later in the appropriate place; meanwhile, here it must be 
said that Hayman, in the work just cited, rejected Allen's arguments on the grounds 
that though Xciio^iilc did exJiibit well-marked, differences from the African Hcrpcsics 
ichneumon these were, in tact, no greater than the divergence of this last from certain 
Asiatic species commonly allocated to Hcrpestcs. This, of course, raises the basic question 
of whether these oriental mongooses arc properly included in Hcrpestcs or whether, 
like AV/ii'ijii/c, they would be better split off as Gray and others long ago contended 
should be the case. It is true that a decision on this point of taxonomy must affect 
West African nomenclature; but the matter is too wide to be decisively probed in 
connexion with this regional investigation. It can only be said here that the present 
wTriter considers Allen's proposal to be correct, and that it is justifiable as well as con- 
venient to regard Xciiof^iik as generically independent. 

In this work, therefore, instead of Hcrpcsics being taken as covering icluiciiiiioii. luiso 
and the san(;uinciis-<<mcilis-ochriKeus-piih'ciHlcntus-riitlamuclii-ctc. complex, either integ- 
rally or divided into the subgenera Galcrclla and Xcno'^alc, it is regarded as apphcable 
solely to the Egyptian mongoose, H. ichiictimon. 

In view of this reduction in scope, questions of general generic description, habits 
and so forth are reserved for the account of the single species which follows. 

HERPESTES ICHNEUMON (Liimaeus) Ichneumon or Egyptian Mongoose 

IVi'iTni ichneumon Linu.icus, 175S, Syslcnin Nnlnuic, lOtli cd., I: 43. Egypt, 011 the hanks ot the Nile. The 
specific name is Greek .and means tracker, given because this mongoose was since ancient days reputed 
to hunt out crocodiles' nests to rob them of eggs. 

Ichneumon phamon Laccpcde, 1799, Tahk'an tics ilii'iiiions, sons-tlirisions, odrcs ft (^cnrc^ iies Mamnnjl-rcs . . . : 7. 
This name associates the animal with the Pharaohs of Egypt. 

Ichneumon aciiypti Tiedemarm, 1808, Zoologie . . ., I: 364. 

Ichneumon major E. Geoffroy, 1812, Description des Mammijercs ijui se tronvcnt en Egyple, 2: 139 (footnote). 
Egypt. The specific name is the Latin for larger and was given because the mongoose described (based 
on La Grande Mangoustc) was thought to be bigger than others in the group dealt with by Geoffroy 
under the general heading of ichneumons. 

Herpcsles ichneumon occidenlalis Monard, 1940, ArcJios Mus. Bocjge, II: 193-194. Mansoa and Cacheu, 
Portuguese Guinea. The name ocadcntalis is Latin for western. Valid as a race. 

Hcrpestcs ichnctnnon aithos subsp. nov. Described on page 276 of this present work. Type locality Newton, 
Sierra Leone. Type specimen in the British Museum No. 53.130, '+' of medium age; skin complete and 
in good condition, skull with cranium badly damaged and partly missing. One meaning of dif/ins 
(Greek) is red-hrov™. 

Herpcstes ichnctnnon mesos subsp. nov. Described on page 278 ot this present work. Type locality Anara 
Forest Reserve, near Kaduna, northern Nigeria. Type specimen in the British Museum, No. 50.320, 
J of medium age; skin fiir, hindfect and part ot the tail missing; skull structurally complete but with 
1 5 teeth missing. One sense of the Greek word mesos is in the middle ot two extremes. 

Distribittion and general. This is almost certainly the oldest-known mongoose 
to v.'cstcrn civilization. It figures in ancient Greek and Latin literature; but long before 
that it was depicted on Egyptian tombs and temples nearly 3000 years B.C. It was 



HERPESTES 269 

certainly revered in Egypt and was commonly mummified. The so-called Egyptian 
mongoose is, in fact, distributed over the length and breadth of Africa from the Mediter- 
ranean to the Cape, though it is not found everywhere. Basically it is an open-country 
ammal but forms arc also known, though less frequently, in the dense forest. It is, 
moreover, the only mongoose in the European fauna, occurring in Spain and Portugal. 
It is also found in Palestine but, curiously enough, docs not seem to have received 
mention in ancient Jewish writings as it did in those of other neighbouring pre- 
christian civilizations. 

One of the names it was known by in early days was Pharaoh's rat; but so far from 
being a nuisance it was widely revered in Egypt, most probably primarily because of 
its pest-destroymg value, being kept m temples and the houses of the rich largely for 
its snake-kilhng propensity, and being honoured in priestly circles for its reputed 
enmity to the crocodile. Throughout the long and doctrinally varied history of Egypt 
this mongoose was, in fact, at different times or places credited with a number of 
mysterious or momentous attributes which cannot be entered into here. Doubtless 
having its origin in Egyptian practice, the ichneumon seems also to have been kept in 
some well-to-do Roman households; but this was without religious significance and 
was most likely little more than a matter of possessing a rather unusual domestic pet, 
though some idea of pest discouragement may have played its part too. 

Description. The ichneumon (Plate 7) is one of the larger mongooses, having an 
adult weight of 3 to 4 kg, a long, relatively slender head &: body of some 550 to 
600 mm together with a notable tail measuring roughly nine-tenths as much. This is a 
very short-legged animal considering the size of the body, standuig only some 150 to 
200 mm at the shoulder, hi fact the long pelage, especially that of the hindquarters, 
often reaches to near the ground, hiding much of the legs. As in most mongooses, 
the face is long and pointed, the muzzle terminating in a naked black rhinarium; the 
romided ears are short but wide and are set on the sides of the head. The higWy speckled 
pelage is coarse, consisting of long wiry bristle-hairs, annulated with several rings of 
alternating light and dark colours, and fairly long, abundant, fine, richly coloured 
imderfur. The bristle-hairs are long throughout all the pelage right up to the nape but 
are especially so over the rump, where they sometimes attain a length of 65-75 mni. 
The imderfur measures 1 5 to 22 mm. The fur of the head, face, chin and neck is similarly 
coloured and speckled but very much shorter and close-lying. The area around the 
eye is bare. Two widely distinct colour forms of pelage occur in West Africa, together 
with a third intermediate form, all described in detail later. 

The tail is of characteristic appearance. It is very broad basally being there clad with 
very long hair which obscures its junction with the similarly long-furred body. Froni 
this stout root it tapers to about its nud-length whence it continues to almost the 
end as narrowly subcyhndrical and terminates in an intensely black, long-bristled tuft. 
The short legs and the feet are fairly powerful, the latter all with five long digits, 
webbed in their basal half and armed with long, slightly curved claws. The soles of 
both fore and hindfeet are naked. There are anal scent glands, openmg not within the 
recmm but into an external oval sac which surrounds and encloses their orifices and 
that of the anus itself. These have been described and figured by Charin (1874). 



2V0 



Till ( AliNlVORES OF WEST ATRICA 




Fi<;. yi. //(T/ii>7c< iiliih-tiiiioii: skull, li.M. Nn. 9.11.2.10, sex ?, x i 



HERPESTES 27I 

Skull (fig. 36). The skull is long and rather narrow, terminating anteriorly in a very 
short, bliint but narrow rostrum. The ovoid braiiicase carries, with few exceptions, 
a well-developed sagittal crest in males and females alike, posteriorly joining the 
broad, flange-like supraoccipital crest in a T. The frontal region is fairly broad and, 
for the most part, iirflated to a slightly higher level than the cranium, quite smooth 
and falling away in an even curve to the rostrum, and also laterally to the orbits. 
The ratio of the intcrorbital breadth to the postorbital constriction is pretty widely 
variable being sometimes lO per cent less, sometimes lO per cent more. The post- 
orbital processes nearly always meet and fuse with the jugal processes to form com- 
plete circumorbital rings. The nasal sutures also fuse and disappear fairly early. The 
zygomata are strong and broadened through most of their length. The postdcntal 
palate is clearly longer than it is wide; the anterior part of the bulla small, the posterior 
chamber very inflated, high. 

The incisors, top and bottom, are in a compact straight row, the outer ones larger 
than the inner. The upper canines are slightly curved, the lower ones rather more so. 
There arc always 4 upper premolars on each side; and there are mostly 4 in the lower 
jaw; but of 11 skulls examined 3 mandibles have only 3 premolars on each side, and 
one has 4 and 3. The anterior premolar is always very much smaller than the rest, 
a simple peg. The posterior outer corner of the upper camassial lies near the outer 
posterior root of the maxillary process, m^ is of moderate size, short but wide, its bulk 
about that of the buccal section of/)"*, its outer face forming a sharp angle with that of 
this latter tooth. The outer and anterior cusps of the lower camassial (;»i) are markedly 
larger than the iimer posterior one. i)fi is small, its transverse width about that of the 
lingual portion of /h^, its bulk clearly less. 

Habits. In spite of the fact that the alleged habits of this animal have been written of 
spoken of, and handed down almost as folklore for many htmdreds of years the truth 
IS that very little that is up-to-date and reliably factual has ever been recorded. The 
ichneumon has a long-established and almost unassailable reputation as the prime 
enemy of both crocodiles and poisonous snakes, having, besides courage, such dedica- 
tion to the destruction of the latter that it was reputed in ancient times habitually and 
deliberately to prepare itself for the fray by covering itself in successive layers of mud, 
drying out each in turn, until it was enclosed in a thick fang-proof armour of hardened 
clay. With this protection it was able to seize its opponent in its jaws; when, instead 
of biting it to death, it fliuig it into the Nile to drown. As for its other foe, not only 
did it, according to the ancients, destroy crocodile eggs but it also regularly and courage- 
ously leapt do^vn the open throats of these animals, which commonly sleep with their 
mouths agape, and ate its way through the internal organs, making an exit through a 
hole gnawed in the belly wall. The crocodile was presumably supposed to remain 
placidly on dry land imtil this intestinal timnelling had been brought to a conclusion 
and a safe emergence achieved. Even modern accounts of this mongoose are invested 
with some air of fantasy, as when Brehna (i88o) describes a family on the himt, the 
male in the lead, closely followed by his mate, and she again by the young ones all in 
line close behind one another "as diough the whole chain of animals were only a 
single being, somewhat rescmbhng a remarkable long snake". 



27- llll ( ARNINUKES Ol WKST AlKICA 

Olio may suppose that such ancient and persistent stones, together with tlie Greek 
name "tracker", must have some foiuidation in fict however slender. Also that there 
should be some significant basis to religious veneration, although this last would seem 
to be discounted by the sacred ibis, which has little to it beyond its rather striking 
colciration yet was embalmed in countless numbers. The crocodile stoiy may well 
be a simple enhancement of egg destruction as an important factor in. the control of 
these greatly feared reptiles. Nevertheless, it has been suggested that even this last 
reputed practice may be pure invention since nest-robbing would be difficult if not 
impossible with the female crocodile constantly on the watch in nearby undergrowth. 
But that Nile monitors (I 'araiius niloiiais) acting in consort can rapidly create havoc in a 
cache of crocodile's eggs is well-established, and the marsh mongoose has been said 
to be a great destroyer of these too (page 297); so there seems no good reason why 
I Icrpcstcs should not behave similarly. However, no record has been traced of such 
depredation having been reliably observed. 

In regard to snakc-killing one must distinguish berween habit and ahiliry. While 
ichneumons may kill and consume snakes met with in the ordinary course of foragmg, 
it is very unlikely that plain inborn enmity would lead them purposeful!)' and regularly 
to seek out these reptiles. The ancient Egyptians knew, and doubtless honoured, the 
ichneumon's capacity' to attack and overcome so dangerous and feared an opponent 
as the asp or cobra; and possibly the spectacle of such combats was from time to time 
staged in or aroimd temples, where these mongooses were habitally kept as sacred 
animals. Yet there seems to be no record of anyone ever having witnessed a fight 
between ichneumon and snake in the wild; or, for that matter, any recent account of a 
staged one. Hinton & Dunn (1967) in writing a general work on mongooses and all 
that was kno\\ai of them obviously combed the literature pretty closely; but they 
make no reference to any fight, even a put-up one, between a snake and an Egyptian 
mongoose. Even such verbal accoimts from travellers as may from time to time come 
one's way may be open to some question since there has in the past been a good deal of 
traffic between India and Egypt; those who claim to have seen such set fights arc rarely 
competent to tell one kind of mongoose from another, and the Indian mongooses 
arc not unlike the iclincumon m appearance. Flower (1932), indeed, recorded that 
Itinerant conjurers were often to be seen in Egypt with the small hidian mongooses 
as part ot their stock in trade. 

It is, in fact. Flower who gives the most positive, though very brief account oi the 
habits oi Hcrpcsics, at least in so far as Egypt is concerned. He says that it is diumal and 
crepuscular, never to be seen in the early morning but on the move equally in brilliant 
noon-day sun and m die dusk ot evening. It must be noted that other observers state 
It to be positively nocturnal or possibly nocturnal as well as diurnal; and T. S. Jones 
(personal commmiication) says that in Sierra Leone it appears to be largely nocturnal. 
Flower characterises it as hunting aroimd in leisurely fashion; and it seems to be little 
disturbed by the presence of people or vehicular traffic since it has been known to cross 
Cairo streets, force motor-cars to slow down, and to have been run over by a tram. 
It is as diough the ages have passed it by and it still expects the protection and respect 
that were its heritage in ancient Egypt. 



IIERPESTES 273 

Hcrpfsics is, indeed, very unsuspicious and easily trapped. It takes readily to water 
and swims well, being sometimes caught by accident in fish-traps when it is attempting 
to rob thcni (Pitman, 1954). It is, in fact, commonly, but not exclusively, riparian, 
dwelling and hunting amongst reeds and other dense imdergrowth; at the same time 
it is by no means averse to woodlands, open countr)' at a distance from water, and 
even to montane grassland (Ansell, 1965). Wlien living on river banks it is said to 
feed on fish and possibly crabs and frogs; but away from such a locality it takes small 
mammals, birds, insects and reptiles. It has a doubtful reputation as a fowl thief. T. S. 
[ones states that in Sierra Leone it will certainly attack chickens when it gets any chance; 
and it has been said to play havoc in a fowl house and to be detested for this reason 
throughout modern Egypt. Flower (1932), on the other hand definitely states that the 
peasants assured him that the iclmeumon never steals chickens — though a fairly well 
authenticated record of the destruction of seven geese was once sent to him by a medical 
officer. Flower's assertion is the more interesting in running directly counter to views 
expressed by Zeuner (1963) that Hcrpesles had declined from the position of a sacred 
and much respected animal in ancient Egypt to that of a widely hated pest in later 
times due to the introduction of the domestic fowl from Lidia and its eventual spread 
amongst all classes, with consequent bitter resentment of depredations amongst both 
birds and eggs which had come to occupy a highly important place in the Egyptian 
domestic economy. 

Little, then, that is rcbable or not controversial has been recorded of the daily and 
yearly hfe of the Egyptian mongoose, and it is to be regretted that no systematic field 
study of it has so far been undertaken, as it has for some of the larger carnivores. The 
fact is that, while such predators as lions, cheetahs, hyaenas, jackals or himting-dogs 
are fairly readily observable and capable of being followed, the ichneumon, though 
widespread over the continent and not uncommon, besides being of less immediate 
economic moment presents considerably more difficulty. It spends much of its time 
in rather concealed conditions of dense ground vegetation and can, as a rule, be caught 
sight of only for brief moments as it hurries from one cover to another. Hoogstraal 
(1964), however, states that sometimes these mongooses sit, play or move slowly in 
easy sight of humans. The gait is normally a quick, purposeful trot. 

For the rest, it is known that Hcrpcstcs shelters, and doubtless breeds, in burrows, 
which may be of its own making or taken over from some other excavator; but gaps 
in rocks, holes under tree roots and similar convenient cavities arc also made use of. 
So far as limited observations indicate there seems to be no fixed breeding season; 
there are no West African juveniles in the British Museum from which deductions 
might be drawn, but one Sierra Leone specimen taken on 2nd November contained 
two foetuses. Beyond this, defmite figures for the number of young 111 a litter are 
almost non-existent; but though these animals probably lead much of their hves in 
solitary fashion they have been said when raising young to go about in family parties 
of about half-a-dozen. Like other vivcrrids, the ichneumon can emit an offensive 
odour when alarmed or otherwise excited; and like so many other mongooses, when 
caught young, it takes readily and kindly to domestication and makes an admirable 
pet. In view of the fact that it has been known in this capacirs' for five thousand years 



274 T'"- < ARMVORtb Ol WEST AFRICA 

It IS remarkable that more has not been recorded of its behaviour, its nature, hkes, 
dislikes, eating and sleephig habits, postures, breedmg, period of gestation, litter size, 
parental care, voice and so forth. However, there is one detailed account (Diicker, 
i960) of mating-play, copulation and parturition as exhibited by a young pair, of a 
single litter boni in captivity. Play of sexual significance started between these two 
when they were only about 15 months old; but real sexual readiness, evinced by a 
swelling and redness of the vulva, did not come about for another 6 months. Con- 
siderable wooing play took place thenceforward together with repeatedly imsuccessful 
attempts at coition. The male frequently uttered an 00-00-00 sound, which was answered 
by the female. At moments of great agitation this cry became something like hc-hc- 
hc-hc, rising in speed and intensity to a pitch of excitement. When copulation was 
actually achieved the act took some 4 to 5 minutes, during which the male repeatedly 
nudged the female in the region of the throat with wide open muzzle. Two young 
were eventually born at a half-hour interval ; but owing to the long-dra\vn-out nature 
of the mating period it was not possible to tell the actual period of gestation. From the 
dates indicated it might he an)n,vhere between, say, 7 and 11 weeks. 

After the birth the parents were both very excited, ruiming and climbing about the 
cage: but it was some time before they took any notice of the cries of their young. 
At lengdi, after about a couple of hours, the mother carried the babies, separately and 
held by the middle of the body, into the sleeping box, where they simply lay on the 
floor while the parents sat snuggled up closely against one another taking no interest 
in their oftspring. The only occasion on which they seemed to be aware of a parental 
dur\' was when the cage was approached and they growled and erected their fur. 
Next morning there was no sign of the babies, which must have been eaten by their 
parents — a common event in nature, and more so under the unnatural stresses of 
captivit)-. Ten days after parturition the female was again m season and the pair strongly 
sexually interested in one another. This would seem to confirm that breeding is not 
merely an annual event related to a particular time of the year. Also, it might be 
inferred that two is a common litter size — there is the London record of two foetuses, 
these two parents were themselves twins, and they had two offspring. No further 
records were possible with this pair since the female suddenly died. It may here be 
noted that a specimen has been known to have lived in captivity to the age of about 
124' years [Int. Zoo Yr. Bh., 2). 

Taxonomy. The broadest question is whether there is, in fact, more than one species 
in the genus; that is whether the separation of the southern African animals at this level, 
as differ, from the northern ones, once the practice, is valid. It seems fairly unanimously 
held today that the species ichneumon satisfactorily and correctly covers all mongooses 
of diis genus from Spain to the Cape; and this view is held in this present work. 

With so wide a distribution it is obvious that local races must almost certainly exist; 
but the problem here is not so easy of solution. With a pelage of the composition of 
Hcrpcstcs, consisting of a top coat of long amiulated bristle-hairs in which black and 
white play a dominant role there is, as in other similar genera, particular scope for an 
almost infinite variet)' of effect due to slight differences of ring width or nuances in 
the tone of the two main colours or the transitional zones between them. Some of 



HERPESTES 275 

these may be simply idios^ticratic; others may, in truth, be pecuHar to all the inhabi- 
tants of a locality or set of conditions, though the plain fact is, as so commonly the 
case, that existing study material is often too meagre to cstabhsh on a firm basis whether 
such divergencies are sufficiently constant over a given area to justify nomenclatural 
recognition. 

Ten or eleven races oi ichncuimin are, in fact, currently recognised, only one of them, 
occidentaUs Monard from Portuguese Guinea, dcfmitely West African, though it has 
been suggested that parvidens Lonnberg from the lower Congo possibly reaches Lake 
Chad. One incontrovertible fact exists, however: there are m West Africa two very 
distinct colour extremes, the one pale grey, the other dark red-browTi. The former is 
not ver)- dissimilar from specimens from the type area, Eg)'pt — which, not unusually, 
differ somewhat amongst themselves. But in the West African examples the hairs of 
the dorsal pelage have rather wider and whiter rings; the miderfur is orange, whereas 
in Eygptian material it may sometimes be redder, sometimes drabber; and the belly 
is better clothed, with longer, paler ("oft-white") hairs. Menard's description oi ccci- 
daitiilis, though in respect of mdividual components detailed, fails to convey the overall 
appearance of the coat. His type has not been seen; but by the courtesy of Dr. Villy 
Acllen and Dr. J. L. Perret of Geneva it has been possible to make a direct comparison 
of a small but adequate sample of the type pelage. From this and Monard's characterisa- 
tion of the underfur as orange it seems sufficiently clear that the three pale West African 
Herpesics in the British Museum are indeed occidciittilis. These are all from the Guinea 
or Doka woodland. 

The five dark form specimens, of very marked contrast to those just dealt with, are, 
with one possible exception gone into later, animals of the closed forest zone between 
western Nigeria and Sierra Leone. There is nothing m the large London collection 
like these dark reddish-brown skms except an aberrant one from Spain, and they 
would thus certainly seem to merit a descriptive name of their own. It might be deduced 
from tins that Hcrpcstes ichneumon is pale grey in arid country becoming deeper and 
redder as the atmospheric humidity increases; but there is an intermediate form which 
comes not from an intermediate ecological zone but, so far as can be told from the verv 
limited material available of each of these forms, from somewhat further inland than 
the pale form, that is the drier Doka and Sudan zones. These skins partake rather 
more of the reddish forest colour than they do of the grey; but the pale rings are broader 
and whiter. Since they do not fit into either category they, too, though less obviously, 
seem to justif,- a distinctive name. 

The subspeciation described below hinges entirely upon coat colour; there are 
two other specimens of West African ichneumon in the British Museum, one from 
Gambia (no localit)') the other from Oda (Ghana), which since they consist of skulls 
only cannot be particularised. No satisfactor)' evidence for the existence of Lonnberg's 
parvidens in West Africa has been foimd in this present investigation. 

Herpestes ichneumon occidentaUs Monard Western Ichneumon 

Distriburion. Monard had two specimens before him when he named tliis race. 



1-6 THE C:AIINIV0RES op west AFRICA 

.111 adult tciiKilc h'oiii Mansoa and .1 male frmn Cachcu, both 111 Portuguese Guinea and 
both, accorduig to the author, in dense tropical forest. The following are the particulars 
and description ot the three London specimens believed to represent this race. They 
come from Thics (Senegal), Cape St. Mary (Gambia), and Kita Sudan (Upper Guinea) — 
which seems most likely to be Kita in the former French Sudan, now Mali. The first 
two places are situated in Guinea woodland, the last in Doka. 

Description. The overall impression ot the dorsal pelage of these is pale grey flecked 
with chocolate (Plate 7). The furly dense, fme, not very long underfur, on the average 
about I s mm though individual hairs may be longer, is, as Monard described, a beautiful 
orange colour. This is almost completely concealed, in repose, by the very long (at the 
lower back 60 to 65 mm) bristle-hairs. These arc ringed black and white, the basal 
and two other zones white, separated by two major dark zones and ending with a 
rather short black tip; that is to say 6 alternating rings. The "black" is seen on close 
examination to be really a very intense red-brown; and there are some rings in which 
this true nature of the pigment is more apparent than in others; and there is always a 
brief transitional zone to the white where it becomes progressively less dense and hence 
pale red. The coat, therefore, always displays a fnr amoimt of reddish colour, mostly 
deep, which, as the corresponding zones of the hairs come to lie somewhat together, 
tend to form short, very irregular, transverse lines or chevrons. 

On the belly the imdcrfur becomes very pale buff, and the still very long bristles 
cream, with one, or sometimes two, rather pale, inconspicuous dark rings. The hairs 
of die head and fice have the same sort of colouring as the back but are quite short, 
with correspondingly short annulation. The same can be said of the legs, where, 
however, the dark clement is more prominent; so that, while the dorsal pelage may 
be said to be grey speckled with dark, the legs arc dark stippled with white. The tail 
is similar to the back m the broad basal half beyond which it becomes much more 
yellowdsh-brown up to the long pure black terminal tuft. 

Skull (fig. y>). It is not possible to draw useful conclusions from the single skull 
available for study. There are in tliis only two marked differences from the remaining 
West African material, and these may well prove merely idiosyncratic : the zygomatic 
breadth is appreciably greater than in the others, amoimting to 55 per cent of the 
condylobasal length as compared with rather less dian 50 per cent; and the breadth of 
nfi is less than in any other specimen. 

Herpestes ichneumon aithos mhip. iiov. Forest Ichneumon 

Distribution and general. The contrast between this and the last is very marked 
(Plate 7). Five specimens exist in the British Museum; these come from Newton (Sierra 
Leone); Wulade, Kissi country (northern Liberia), 2 specimens; Idumiye Ono, 4 miles 
north of Iseluku, west of Asaba on the River Niger, Benin Province, Nigeria (these 
are spelt Idumuje Uno and Issale Uku on modern maps); and Senegal, no specified 
localitv'. The first four of these lie within the closed high-forest belt, though Newton 
has become degraded by fire and cultivation into grass with patches of secondaiy bush 
near water. The last would seem to be possibly an exception, though, in the absence of 



Plate 7 




Forest Icliiicuiiicn {Hcrpcftcs idwaimon aiilios): Western Iclineunioii (HiT/Jcsfex Ulwaumv! oaichiialis): 
M^irsli Mongoose {Alikx yw/i/i/iHusiis) 



HERPESTES 277 

an exact locality and the fact that in 1863, when the specimen was registered, the 
limits implied by "Senegal" were not precise, one cannot be at all sure of the vegetation. 
There is another ground for not blindly accepting at its face value this accrediting of 
an apparently high forest mongoose to predominantly dry open woodland country 
in the vicinage of the Senegal River. The collector was the late 19th century con- 
troversial author Winwood Reade. It is a fact that on this visit to the then very httle 
known and not very well comprehended western side of Africa he went to Gaboon, 
Angola, Casamance, Gambia and Senegal, in that order. He might easily have said, 
speaking generally, when visiting the British Museum that he had just returned from 
Senegal and had brought some specimens for identification, thus conveying a wrong 
impression; whereas the specimens may qiute wcU have been obtained at various pouits 
of his tour; this one, for example, in the forests of Gaboon, whither he had gone 
deliberately to follow the footsteps of the then famous collector Du Chaillu. 

The Newton (Sierra Leone) specimen, collector T. S. Jones, B.M. No. 53.130, 
a mediumly-agcd $, gravid with 2 foetuses, is nominated as type of this new race 
because, although the skull is badly broken, the skin is complete; whereas the only 
perfect skull in the material available has the skin incomplete. 

Description. The overall impression of the dorsal pelage is a deep reddish-brown 
abimdantly ticked with yellow. The underfur is of an appreciably deeper rufous shade 
than the orange o( occidcnhilis; it is about 15 to 18 mm long, and though in a large 
measure hidden by the bristle-hairs it does to some extent show through and help to 
account for the reddish impression of the pelage. The bristle-hairs are not much more 
than 50 mm long, and their annulation is quite distinct from that oi occidentalis; for 
whereas in this latter race there are, including the black tip, about 6 alternating rmgs 
in which the light and dark are not very different in length, in this newly proposed 
form there are about 9 or 10 rmgs, the dark ones very much predominant, the pale 
ones being mostly reduced to widths of 3 or 4 mm. The dark annidations are to all 
intents and purposes black; but the two narrow distal pale ones, the only ones which 
show beyond the underfur and constitute the ticking element in the pelage, are yellow 
or off-white. Because of the vastly different widths of the aimulation in the two races 
the actual physical appearance of the pale tickmg irrespective of questions of colour, 
is very distinct in occicienUilis and aithcs, being in the former long and longitudinally 
almost continuous, but in the latter very short, very much interrupted and numerically 
very abimdant. There is a fuller note on annulation at the end of this generic section. 

The belly fur, though in all rather paler, is very similar to that of the back, being 
annulated yellow and blackish-brown. It is paler because the pale rings are far wider 
than they are dorsally. The neck and top of the head are fairly similar to the back but 
ticked more fmely; but whereas in ocddait,ilis the ticking continues forward almost 
to the rhinarium, in this new race it stops short about the level of the eyes, the upper 
part of the nose being clad with plain, d.irk blackish-browar bristle-hairs. The ticking 
on the chin, throat and underside of the neck is whiter than that of the back. The 
upper parts of the legs are ticked hke the back but the feet are deep red-bro\vn. The tail 
resembles the back throughout nearly all its length but is for a short distance before the 
black tuft very slightly redder. 



278 THE CARNIVOliES OF WEST AFRICA 

Skull. The measurements given ni the table on page 279 sliow that the zygomatic 
breadth is less than in occidciinilis, amounting to 49 per cent of the condylobasal length. 
The interorbital breadth is less than the postorbital constriction. Li the type skull 
there are only 3 premolars on each side ot the lower jaw; in one other skull there are 
4 on one side, 3 on the other; the third skull has tlie normal 4 on each side. 

Paratypes. Other specimens exannned are No. 7.12.17.1 (skull broken), and No. 
7.12. 17.2, both Liberia; No. 63.5.9.7 (no skull), Senegal; and No. 13.11. 6.1 (skull 
perfect), western Nigeria. Although in these the shade of yellow in die ticking, and the 
width ot the pale annulations do not absolutely correspond with diose given above 
for the tvpe, the difterences are too slight to be ot aiiv account. 

Herpcstes ichneumon niesos .<;i(/i.';^). noi\ Nigerian Iclineumon 

Distribution and general. Three specimens in the British Museum collection are 
assignable to this: from Aiiara Forest Reserve, near Kaduna, northern Nigeria; from 
Kode area, 24 kilometres north-east of Lau (on the Benue), northern Nigeria; and from 
the Northern Territories, C.hana. The first of these places lies 111 the l^oka zone, the 
second in the Sudan zone; the third specimen, having no clue to die precise locality, 
might have come from any of the Guinea, Doka or Sudan zones. 

The Anara specimen, collector D. R. Rosevear, B.M. No. 50.320, a mediuni-aged 
to old S> IS taken as the rj'pe of this new race. The skin is poorly stutted, the hmdteet 
and part ot the tail missing; the skull in itself is complete but has several teeth missing. 

Description. There is no doubt of the intermediate nature of this proposed new 
race; m overall appearance it is appreciably redder than oaidcntiilis but a markedly 
lighter-coloured animal than aitlios. Like other features, the underfur in this proposed 
race is intermediate between the bright orange o( ociiiiciitalis and the rich red o( aithos. 
It is about 13 mm long. As in occiiliiitdlis, the bristle-hairs have about 6 annulations. 
Their overall length is somewhat shorter, about 50 to 60 mm; the pale rings, besides 
being shorter, are buti and the dark ones are a less intense red-brown. A comparison 
of this ammlation with that of the other races is made below at the end ot the account 
of this genus. 

The imdertur on the belly has lost most ot the red of the upper side, becoming pale 
brown; the long bristles are butf and their dark rings relatively pale. The colouring of 
the head and face is very much that ot the back but the hairs and their annulations very 
short; the specklmg continues forward dorsally very nearly to the rhinanuni, there 
being only a few millimetres of luispeckled hairs in this region, markedly less than in 
aitlios. The forelegs are speckled, of a slightly darker tone than the back, the feet being 
a deep brown; the upper part of the hmdlegs in the r\'pe is pale buff\'-browii with very 
little obvious annulation; the proximal part of the feet is darkish with short speckled 
hairs similar to the head; their end part is missing from die type, but in another specimen 
is plain dark browni. The very long-haired tail is similar in appearance to the back; 
its terminal portion is lacking from the type and one other specimen (No. 56.250); 
but in the third specimen (No. 22. 12. 2.1) it becomes very rufous in the narrow part; 
and the terminal tuft is not the intense black of the other races but a little reddish 
throughout and more particularly so distally. 



HERPESTES 



279 



The two paratypcs exhibit minor, but not significant, differences from the type as 
described above. 

Skull. This has no very distinctive features discernible in the two sets of measure- 
ments available. It is of average width, the zygomatic breadth amounts to precisely 
50 per cent of the condylobasal length. The braincasc is a little narrower than in the 
other two races; the interorbital breadth is in one case (the t)-pe) larger, in the other 

Table 14: Hair annulation in Herpcstes khneunmt subspecies 



occiJentalis (7) 


Basal region 
Dark Pale Dark 


Middle region 

Pale Dark Pale Dark 

17 II 7 13 


Terminal region 
Pale Red Black 

7-5 


= 60 mm 


mesos (12) 
aithos (5) 


2 3 6 


13 10 5 14 
5 10 3 14 


5 - 10 
3 2 5 


= 57 mm 
= 53 mm 



Table 15: Numerical data for Herpestes ichneumon 





occidenlalis 


mesos. 


mesos. 


aithos. 


aithos. 






Type 


means 


Type 


means 


Vegetation 


Guinea and 
Doka 


Doka 


Doka and 
Sudan 


Forest 


Forest 


Number in mean 


I 


I 


2 


I 


3 


Condylobasal length 


99-4 


96-4 


98-4 


100-4 


I0I-3 


Basilar length 


92-6 


88-7 


95-4 


— 


94-1 


Palatilar length 


56-0 


52-5 


54-2 


— 


55-2 


Zygomatic breadth 


54-8 


48-1 


49-2 


50-6 


49-8 


Upper cheekteeth breadth 


32-6 


29-2 


29-7 


3I-I 


31-4 


Nasals, length 






24-2 




23-5 


Interorbital constriction 


18-4 


17-0 


17-4 


(l8-2) 


i8-i 


Postorbital breadth 


17-9 


i6-l 


17-3 


20-5 


20-0 


Braincase breadth 


36-1 


32-8 


32-9 


35-0 


35-3 


Toothrow [c — m^) 


36-3 


34-6 


35-4 


35-2 


36-7 


p* length 


9-7 


9-5 


9-5 


9-5 


9-5 


ml breadth 


9-3 


8-6 


8-6 


8-8 


91 


m- breadth 


4-4 


— 


5-3 


4-7 


5-3 


mi length 


8-7 


8-4 


8-5 


8-1 


8-6 


m2 length 


5-0 


4-3 


4-2 


4-6 


4-7 


Head & body 


— 


600 


600 


545 


545 


Tail 


— 


300 


300 


492 


492 


Hindfoot 


— 


— 




95 


95 


Ear 


— 


26 


26 


36 


36 


RATIOS (per cent) 












Tail/head & body 


— 


50 


50 


90 


90 


Zygom. br./condylob. 1. 


55 


50 


50 


49 


49 


Braincasc/condylob. 1. 


36 


33 


33 


34 


35 


Braincase/zygom. br. 


66 


68 


67 


69 


71 


Palatilar l./condylob. 1. 


56 


55 


55 




55 


Interorb./postorb. 


103 


105 


100 


89 


90 


p*lc—m~ 


26-7 


37-4 


26'9 


27-0 


25-9 



280 THE CAKNIVOUKS OF WEST AFRICA 

less than the postorbital constriction. The type specimen h.is 4, tlie otlier skull 3, 
premolars on each side ot the lower jaw. The breadth ot /»' and the leiiuith ot 1112 
both seem to be less than in the other subspecies. 

General comments, h must be repeated that no duect comparison ot the specimen 
(No. 9.1 1. 2. 10) here assumed to be McMiard's caiJiinulis with his r\'pe has been possible: 
if it should be shown to differ materially it would require a new subspecific name. 

The table given on p. 279 shows a comparison of the lengths ot the various annulations 
of the bristle-hairs in the three races, as illustrated by the mean values derived from a 
number of diflerent hairs from each of a number ot ditlerent specimens. The ring- 
margins are never clear-cut, the dark zone fading throughout a longer or shorter 
distance (mosdy of the order of 0-5 to i mm) into the light zone, tliis transitional region 
exliibiting a pale red-brown colour. Precise measurement is therefore never possible; 
hut, with this proviso and the tact that there are always occasional exceptions, the 
ring-widths arc pretty constant m all the specimens examined and for the most part 
do not vary from the mean lengths shown below by more than 2 mm, usually less. 

Three groups of zones may be recognised. There is a terminal region measuring 
10 to 15 mm consisting usually ot two zones, the tmely tapering black tip and a pro- 
ximal pale ring; but in dithos the transitional area between these two is of sutticient 
width and importance to merit separate recognition. Proximal to this terminal region 
is a second group of 4 annulations of the utmost importance in determining the appear- 
ance of the coat; for whereas it will be seen, in the table, that the width ot the dark 
rings remains almost constant throughout all three races that of the pale rings diminishes 
very greatly, mi til in iilihos it is a half, or evcMi a third, ot what it measures in occidcniiilis. 
It will be appreciated diat, ignoring colour, this results in a marked ditterence of 
appearance, hi addition to the two sets of rings just detailed, iiithos has a further, long 
basal region of 3 (and in one hair specimen 4) alternating dark and pale bands. The 
length of this extra region is more than compensated for by the extreme shortening 
of the pale zones of the adjacent region, to which attention has just been directed, 
so that the overall length of the hair is, in fact, less than those with only 6 alternating 
light and dark bands. 

Genus CROSSARCHUS F. Cuvier, 1825 
Lesser Long-nosed Mongooses 

Miiiii^os E. GcofFroy & G. Ciivicr, 1795, A/iyiisiH Ei)cycl^j\iiiqii,\ 2: 1S4, 1X7. Type species I 'ivcrra iiiiiiif;o 
Gnielin; in part. 

Cwssarihus F. Cuvier, 1.H25, in E. GeortVoy & F. Cuvier, Hiiloirc Niiltiifllc i1t:< Mitiiiiniji-m . . ., 3, pt. 47, 
Le M.inguc: 3. Type species Crossarclun obsainis F. Cuvier. West Atrici. This n.uiic was coined from 
the Greek words crossolps fringed, and arclios anus, with reference to the appearance ot the circumanal 
glandular sac, the wrinkled folds of which bear a fancied rescnibl.ince to the old-fashioned pleated 
collar known as a ruff. 

Distribution. The genus Cros<archui, as at present constituted, embraces three 
species inhabiting mainly the closed forest block but sometimes reaching into the 



CROSSARCHUS 28l 

adjacent woodland zones by means of forest remnants or riverain extensions. Only 
one of these species, obscurus, is found in West Africa, where it is known from Sierra 
Leone to lower Cameroon at about 3''N. The other two species are alcxandri from the 
Central African Republic and northern Congo; and anscr^ei from north-western 
Angola. Where these mongooses occur they may be reckoned as moderately common 
animals since they mostly associate together in fairly large companies. 

General description. The range of size of the three species is wide, from the small 
ii»5or(;ci with a head & body length of 320 mm and a skull of 58 mm to the relatively 
large alcxaudri which measures 450 mm and 81 mm; but even this latter is, of course, 
small in comparison with the really large mongooses such as Hcrpcstcs, Xawgale and 
Gakrisais. In between the two extremes of size in Crossiuchus stands the very common 
West African obscurus. All forms arc, most typically, darkish-brown animals ticked to a 
greater or less extent with golden-yellow, pale yellow, or even white; but ansorgei 
and an occasional obsainis specimen exhibit considerably less ticking than normal. 
The rather harsh pelage is composed of long underfur and considerably longer bristle- 
hairs; and there is, almost without exception, a whorl or whorls on the back of the 
neck. The snout is exceptionally long. The tail is shorter than head & body and tapers 
evenly from root to tip; and though it is clad with long bristle-hairs it is not, when in 
repose, in any way bushy. The legs are short; the 5-toed feet armed with very long 
claws and webbed between the proximal joints. The soles of the hindfeet are mostly 
naked except for the last quarter near the heel. The rectum is flanked by a pair of scent 
glands, their external orifices either side of the anus, all being surrounded by a thick- 
walled circumanal sac. While both obsainis and akxandri conform well to this general 
description, ausor\;ci is in several respects somewhat exceptional. 

Skull. This is long and narrow especially as regards the rostrum, though this last is 
less marked in aiiforff-i than in the others. Crossarchus belongs to the category of mon- 
gooses in which the upper carnassial is set well forward of the root of the maxillary 
process, the cheekteeth being sharply cuspidate but not particularly sectorial. There 
are three premolars on each side, above and below. These are the essential generic 
points; the remainder of the skull characters may be gathered from the description of 
those oi^ obsainis below. 

Taxonomy. The tspe species obsainis was from the start assigned to Crossiuchus, 
which F. Cuvier simultaneously especially erected to accommodate it; but though this 
genus has to a large extent managed to maintain its independence there has been some 
tendency to identify it with Mun^os. Thomas, in his 1882 review of the mongooses 
considered obsainis to be congeneric with fasciams, zebra and qnmbiiiiuis. These are 
now regarded as belonging to Muiii<os Geoffroy & Cuvier; nevertheless he ascribed 
them all to Crossarchus supposing this name to antedate Munj^os, which he wTongly 
attributed to Gray, 1865. The complex muddle over the names Mungos and Hcrpcstcs 
has been referred to earlier in this work; it has been very fuUy dealt with and clarified 
by J. A. Allen (1924) and need not be gone into here. But since it has sometimes recently 
become the practice to treat the Crossarchus species as very closely allied to the banded 
mongoose, AIuik^os immgo, and to deal with Crossarchus cither as completely SAiiony- 
mou5 with Mungos Geoffrey & Cuvier (Hill & Carter, 1941) or as merely a subgenus 



282 THE CARNIVORP.S OF WEST AERICA 

of it (Ellcrnian, Morrison-Scott ^ Haynian, 1953) it may be profitable to quote here 
Allen's opinion, which at the same time succinctly incorporates the views of others 
on this matter: "The restoration o( Muiii^os to its proper place in nomenclature need 
not in the least disturb the stability of Crossimhiis F. Cuvicr (1835), which has, by 
monon'py, Crossiirchin ohsciinis F. Cuvier as its gcnor^-pe, for which and later described 
aUicd forms it should be retained. As thus restricted Crossiiirhiis forms a group very 
different from the banded mongooses for which Miiii(>cs is available and to which it 
should be restricted. Gray showed good judgment in separating the two groups 
gcnerically. Attention has recently been called to the generic distinctness of these 
groups by Pocock, he adopting for die banded mongooses Gray's unavailable name 
.4nVl<!. He also calls attention to the fict that the inclusion of the two groups luider 
Crossarchus was due to erroneous information concerning the structure of the anal 
glands. Before meeting with Pocock's paper I had become strongly impD?sscd with 
their incongruity and their evident generic distinctness". 

The two genera, especially as exemplified by C. cihsciiins and M. (^iunhiauiis, share 
something of a common size and appearance; and it is difficult, as a glance at the keys 
given herein on pp. 245 and 247 will show, to separate them, either externally or 
cranially, on commonly accepted morphological characters. But this is not the whole 
story; and, like Allen and Pocock, the present author has little hesitation m regarding 
them suft'iciently different animals as to justify separate generic assignment. They arc 
therefore so dealt with in this work. 

As far as internal division of Crossanliiis is concerned, oidy the three species men- 
tioned in the opening paragraph have so far been described. C. dlcxaiiilri closely resem- 
bles ohscums, differing from it chiefly by its markedly greater size, and it is possible 
that the two may be conspecific. However, the present author regards the size dis- 
tinction as more than racial and the two are accepted herein as specifically distinct. 
C. ansor<!^ci, on the other hand, besides being considerably smaller, is aberrant in several 
respects, notably the shormess of the rostrum; and it seems at least possible that it is, 
in fact, not a true Crossarchus. 

CROSSARCHUS OBSCURUS F. Cuvier Kusimanse 

Crossarchus ohsiuriis F. Cuvicr, 1S25, in E. Gcoffroy & F. Cuvicr, Histoirc Naiurcllc civs Maiiiiiiijcrfs . . ., 
3, pt. 47, Lc Mangue: 3. Type species Crossarchus ohsciirus F. Cuvier, West Africi. The specific name 
is the Latin adjective meaning dark or dusky, given witli reference to the pelage colour. 

Distribution and general. The name kusimanse, pronounced m three syllables 
.and sometimes spelt cusimanse, is of obscure origin. It is said to be a vernacular name 
but even its approximate source appears to be unrecorded; and it does not seem possible 
to trace its use back beyond "Wood's Natural History published m 1861. It has also 
been used by Germ,in writers; but never by the French, who originally coined the 
term mangue, now confmed to the genus Miiin;os and replaced, for the animal at 
present imder discussion, by a simplified form of the scientific name, crossarche. The 
kusimanse has sometimes been referred to also as the Long-nosed mongoose; but if 



CROSSARCHUS 283 

such an expression is used for Crossarchiis it should, be quahficd as the lesser long-nosed 
mongoose since the very much larger Xeuof^ak tiaso is similarly characterised by the 
unusual length of its snout. 

The kusiniansc is wholly a West African animal, ranging from Sierra Leone to 
lower Cameroun. The list of places from which it has been recorded is too lengthy 
to quote, and the fact is that it may be expected almost anywhere in the forest away 
from populous or much disturbed areas. It has been taken at an altitude of about 
600 metres on the Cameroun Moiuitain and about looo metres on Moiuit Bintamane 
(Sierra Leone). It is found solely in the high forest, chiefly in the main block but 
occasionally in extensions of this vegetation into the contiguous grass-woodland zone; 
but it appears never to stray from these outhers into the grassland itself. Because it is 
diurnal it is rather more frequently encountered than most of the other mongooses; 
and when it is seen it is in some numbers since this species associates and hunts in 
companies. It is to be recognised in the field partly by this habit and partly by its small 
size, plump low-set body and dark family speckled coat, its noticeably long pink nose 
and its short tapering tail. It might from its size and superficial appearance be confused 
with the Gambian mongoose {Mwigos gambianus); but this latter, apart from having a 
shorter muzzle, is a species of the Guinea grass-woodland and the tw'O are therefore 
unlikely to be foiuid in the same sort of place. Puchased skins, however, are another 
matter and recourse must be had to a close examination of the pelage. 

Description. The kusimanse (Plate 6) is nearly the smallest of West African mon- 
gooses and has a head & body length of about 340 mm, a tail of about 170 mm and 
an adult body weight of from i to 1-5 kg. Before proceeding to the following detailed 
description it may be well to draw attention to the fact that some of the chief distinctions 
between Crofsarclius and Mun(;os rest in the character, composition and annulation 
pattern of the pelage; and this applies to all known species of each genus, extralimital 
as much as West African. It is therefore as well to get a clear grasp of the points involved. 
As a matter of interest, it is not difficult to distinguish between the two genera by a 
single bristle-hair. 

The fur o( Cwssimlnis has mostly a rough, slightly bristly appearance, or very much 
so when erected, as it commonly is, in excitement. The most usual overall general 
impression is dark brown ticked to a very varying degree with yellowish. However, 
the basic colour ranges, less commonly, between something almost black, in specimens 
in which the ticking is reduced to a minimum, to a heavily speckled lighter brown in 
young animals, or even a very red-brown in juveniles. The pelage is composed of 
densely abundant, fairly long, slightly wavy, very fme imderfur, and plentiful though 
much more widely scattered and very much lengthier bristle-hairs. The existence of 
dense underfur is, with a specimen in the hand, the most ready, and completely certain, 
distinction between Crcssarclws and Mwii:;cs. It is some 17 to 19 mm long and carries 
alternating bands of dark and light colour, the basal one, occup-sing nearly half the 
length is dark, ranging m different specimens from chocolate to nearly black. Distally 
of this, and the v-isually most obvious zone when the fur is turned back is a grey, white 
or yellowish ring, 4 or 5 mm wide; and this is succeeded by a narrow black ring of 
about 3 mm and a golden tip of 2 mm, often obscure. The bristle-hairs are commonly 



284 



THE CARNIVORES OV WEST AFRICA 




Fig. 37. Crossarchus ohsctinis: skull, B.M. No. 48.841, sex ?, X 



CROSSARCHUS 285 

some 37 or 38 nun long, though m specimens from some areas they may reach only 
28 to 30 mm. They have a pale base followed by a long black zone, some 25 to 30 mm 
and thus occupying the greater part of the hair; distally of this is a narrow subterminal 
gold ring of 2 to 3 mm and a short black tip of about 2 mm. It is the lengths of these 
three terminal zones that, almost without exception, enable Crosstinlius to be differen- 
tiated from Muni;os, in which the median black is much shorter and the distal gold 
ring and slender black tip both appreciably longer. Juveniles are chiefly clad in dense 
imderfur; bristle-hairs may be present but are relatively inconspicuous and do not, 
at first, exceed the imderfur in length. 

The underside is similar to the back but mostly somewhat more sparsely and shorter 
haired. One of the most characteristic features of the Crossarchus pelage is the presence 
of a pair of large whorls on the back of the neck a little posterior of the ears and crown; 
but though these are for the most part very conspicuous they are not always so and 
may, exceptionally, apparently be wholly lacking. The hair on the throat is directed 
forwards and there is therefore another whorl below, parting this from the chest. 
The low, rounded ears are situated very much on the side of the head, widely separated 
medially. The head and face are covered with very short, very fmely speckled, medium- 
brown hair but there is a barish patch anterior to the eye. This latter organ is bright 
and keen. But the most noticeable feature of the head is the lengthy, fleshy nose with 
its pink rhinarium, a long way overshooting the lower jaw. 

The tail is short, being as a rule not much more than half the length of head & body. 
It is markedly tapering from a fairly broad base, clothed with long bristles and underfur 
in a similar maimer to the back, but not bushy except when the hairs are deliberately 
erected in anger. It often, at least in museum specimens, seems to be broken; when 
it is complete the extreme tip is dull reddish, there never being a black terminal zone 
as in Mungos. The legs and feet are dark-brown or blackish, the soles of the hindfeet 
partly hairy near the heels; the toes are webbed between their proximal joints; the 
claws, especially of the forefeet, arc very long, differing from those of MHHi[os m being 
rather more slenderly built, that is to say laterally more compressed and basally not 
so deep, hi stance and rim the animal is very flat-footed and this makes the already short 
legs seem even shorter in contrast to other mongooses v/hich are clearly digitigrade. 
There are two scent glands situated one on each side of the rectum; their orifices and 
the anus are surroimded by a sac remarkable for the parallel ridges of skin connecting 
its upper margin to the anus (Pocock, I9i6e). It was the appearance of these that gave 
rise to the generic name as explained above in the synonymy. 

Skull (fig. 37). All known West African skulls have a condylobasal length of well 
under 75 mm. This is a narrower skull than that of Mungos, the zygomatic breadth 
being almost always (11 out of 13) less than 53 per cent of the condylobasal length. 
The braincase is in every way much narrower and less voluminous than that oi Munoos, 
much more so visually than the measurements in the tables on pages 265 and 290 
would suggest (compare figs. 3 4 and 37). In 11 out of 1 7 specimens the mastoid breadth 
IS imdcr 41 per cent of the condylobasal length, the remaining 6 being slightly over; 
whereas all Mungcs skulls measured are over 41 per cent, mostly, except m g^ambianm, 
appreciably so. Two features which influence the overall appearance of the skull are 



286 Till; CARNIVCHiF.S Of WEST AFRICA 

curiously variable: tirstly. in 9 spccmicns the internrbital breadth is greater than the 
postorhital constriction, and m S it is less, the range being from S<; to 159 per cent. 
Secondly, a sagittal crest is absent from 9 skulls, rudimentary or very poorly developed 
in 7, and prominent in 4. This, providing the sexing has been correctly done in the 
tield, appears to h,ave little to do with age or sex: only adult skulls are included, the 
best developed is in an oldish female, and one old male is devoid of any sign. A well 
developed supraoccipital crest is present in all adult skulls. The postorbital processes, 
although quite well developed, are, m comparison with those of other mongooses, 
nearly always relatively short and tor the most part pretr\- widely separated from the 
short jugal processes. The zygoma is, tor this tamily, a rather weak structure. The 
anterior portion ot the bulla is always smaller than the posterior chamber; but its 
degree of development and the amoimt of inflation of the latter are both variable, 
so that their size relationship is pretty inconstant. However, the anterior chamber is 
rarely so deeply transversely depressed as it is in Ahiinios. The postdcntal palate where 
it becomes parallel-sided is, for the most part, about as long as it is broad, but may in a 
few cases be either slightly longer or slightly shorter. 

CrcsSiinhiis belongs to that section of mongooses in which the posterior outer corner 
of the upper carnassial together with the whole of m'^ are situated anterior to the 
posterior root of the maxillary process. The upper incisors are compact; the four inner 
ones form an almost straight transverse row but the two outer, rather larger ones are 
set back and are more lateral than frontal in their positioning. The canines of both 
jaws are, for cariiivora, relatively short. The premolars are, without exception in 
;o skulls, y All the cheekteeth are sharply and fliirly evenly cusped, including the 
inner heels of/)^ to /»-. That on p^ is nearly always conspicuous and sharp; and although 
a similar cusp often exists in A/i/j/(;iv it is usually not quite so clearly developed, iifl is a 
well-developed tooth, appreciably bigger than the lingual portion of ;;|i. In the lower 
jaw the forward half of i;;' carries three cusps, one buccal, slightly the largest, and two 
lingual, not quite so closely juxtaposed as in Muii(;os; 1112 is also abundantly and sharply 
cusped. To judge from British Museum material both these molars retain their cuspida- 
tion to an advanced age. The dentition as a whole has little sectorial capacity but is 
well adapted to an insect diet. 

Habits. Croisarchiii is one of those animals the behaviour ot which has been moder- 
•itely well observed in the status of a pet but of whose life in nature little has been 
recorded. Accounts of the species have been published by Haig (1931. under the 
pseudonym ■■Bushman"), Naimdortf (1936) and Ewer (196S); these mostly concern 
captive specimens, the first two each relating to one specitic animal kept about the 
house, the last regarding items of observed kusimanse behaviour in deeper perspective 
as elements of a general study of mammalian ethology. To these can be added a few 
disconnected notes made by various tield naturalists. 

Crofsairhui inhabits the rain-forest, ranging at a jog-trot through the tairly dense 
herbaceous inidergrowth. It has been said to prefer the banks ot streams; but though it 
IS known to occur in such conditions and to consume water-loving animals there is at 
present little factual evidence to indicate any particular preference for such sites. Opinion 



CROSSARCHUS 287 

is almost imaninious that the kusimansc is in nature a purely diurnal mongoose. T. S. 
Jones (personal communication), for one, observed it to be so in the forest on Bintamane 
Mountain, Sierra Leone; and Haig (193 1) found his tame specimen to sleep more or 
less peacefully from dusk to dawn. Yet Sanderson (1940) states positively that "at 
night the animals often visit cleared land and farms in search of food". Whether this 
contradictory observation was the outcome of night-trapping of a fair porportion of 
his 9 specimens is not clear. On the other hand, Durrell (1958), who worked in niuch 
the same area (upper Cameroim), writes of Crossarchtis as not uncommonly seen by 
day. But this latter author also records watching a solitary animal fishing for crabs 
whereas the kusimanse is renowned for its custom of hunting m companies of a dozen 
or twenty or sometimes many more. This latter is without doubt the general habit of 
the species. It is presumed that these parties, comprising old and yoiuig specimens of 
both sexes, consist of one, two or three family imits, probably, to judge from the 
breeding note given later, made up of parents and the surviving members of two or 
three litters still holding together. Whether these assemblages are fortuitous and 
tcmporaiy or have some greater or less degree of permanence is quite unknown. That 
the kusimanse is a highly sociable and apparently affectionate animal is testified to 
by all who have experienced these charming pets; and this sociabihty is not confined, 
in captivity, to the limited circle of those who regularly supply food and shelter but 
embraces a wide sphere of strangers, dogs, cats and other animals. It may, nevertheless, 
in excitement or greed but without deliberate malice, occasionally bite the hand that 
feeds it, without, however, effecting a great deal of damage. 

In nature, a large part of the food intake probably consists of insects. The teeth are 
well adapted to such a diet. Those who have had the task of feeding these mongooses 
and at the same time keeping them in good health know that they avidly, and appar- 
ently necessarily, consume quantities of grasshoppers and crickets. T. S.Jones (personal 
commiuiication) has observed that they are particidarly fond of the grubs of mason- 
wasps, breaking open the mud nests with a great deal of excitement. But, of course, 
other things are taken. Naundorff mentions worms and snails; and it is also known that 
small river crabs are commonly eaten. Sanderson (1940), in fact, found the stomachs 
which he examined to contain nothing else but crabs and insects. Durrell (1958) observed 
a kusimanse to secure crabs by flinging them with the forepaws from shallow water 
to the bank; and he also recoimts attempts to capture frogs in a similar manner — 
unsuccessful because immediately on landing they leapt away. As a domestic pet 
CrosSiircliiis willingly accepts prepared foods such as cooked meat or chicken; and Haig 
found fruit salad, comprising largely orange and banana, to be greedily taken. Sweet 
biscuit, so distant from any food likely to occur in nature, was also readily eaten; and 
cooked egg was a titbit. In this last comiexion it is interesting to note that Naundorli 's 
kusimanse cxliibited the widely used mongoose shell-breaking technique, using a ball, 
flinging it backwards towards a wall with its forepaws, at the same rime bouncing its 
hindlegs oft the ground to clear them out of the ball's line of flight. This appears to 
be the only record of this not uncommon mongoose procedure in CrosSiircluis. 

Besides crabs, insects and other small creatures already mentioned it would seem that 
in nature the kusimanse must, at least from time to time, take more typical carnivore 



'S 



2SS 1111 (AHNIVOIUS Ol WIS! MUICA 

tennis. Ewer records one stalking .i small bird, using a cat-like crouching, slinking 
attitude of approach. In the same work Civss,iiihtis is described as killing a mouse by a 
simple neck-bite without the addition of any dog-like shaking of the prey. It may be 
noted here that these animals drink readily in captivity. 

Generallv. a meal is not eaten straight from a dish in the manner of a dog or cat 
but individual pieces are flicked out with the paws onto the floor and taken from 
dierc. The forefeet, indeed, appear to be very sensitive and much-used structures as in 
some other mongooses (cf. Atihix, page 298). Nautidorft foimd her kusimanse to use 
diem much as hands. This animal ceaselessly handled evei"ything in its path, turning 
objects over if they were at all moveable, including all manner ot things m the house. 
This was doubtless part of the himt tor insects, spiders and similar creatures commonly 
hiding beneath stones or rotting wood. An intense curiosity is, indeed, one of the 
characteristics of this species which, during its periods of activity, endlessly explores 
cracks, crevices and holes of all kinds. Ewer (1968) says that they dig with their long 
claws and use the snout to assist in moving the soil aside; and also that they arc good 
climbers. Others, too, have asserted that they can climb trees when alarmed; but it 
appears to the present writer that such statements may well give a wrong impression. 
Crossarclnis can, and readily will, nm up sloping objects whether they be tree trmiks 
or human legs; but to scale a vertical bole like a cat with its claws is not within its 
capability. Nevertheless, if R. W. Hayman"s experience with the very similar eastern 
Congo species dlcxaiuirl is a guide these mongooses, given fivourahle circumstances, 
may be able to achieve considerable heights in trees, using epiphytic figs and closely 
clinging lianes as aids to climbing. This observer saw three specimens of ahwaiidri, 
111 north-east Congo, bolt up a big tree enrwined by a Flciis up the stems of which the 
mongooses climbed. Two eventually got away but one attained about 12 metres before 
it was shot trying to hide by squeezing itselt between the stout stem of the epiphyte 
and the tree trunk — a situation in which it in fact got lodged (personal communication). 
Li the ordinary way, to climb any but low upright obstructions offering no positive 
footholds or other means ot support seems a matter of difliculty; and diis, too, was 
Haig's experience. 

The senses of both sight and smell are excellent. Haig's kusimanse could spot a hawk 
in die sky, no matter how distant. Naiiiidorft, too, found acuitv of vision to be remark- 
able even at long distances; and there was no difliculty in ciistinguishiiig between 
persons, though possibly smell played a part in diis latter. Without doubt, information 
conveyed by the nose plays a big role in the life ot the kusimanse; and Naundorft 
observed that the correct friendly smell was important in producing assured relaxation. 
Nevertheless, if her pet animal lagged behind when out walking it found its way to 
her more by the sound of her voice than by her scent. Hearing, too, is well developed. 
Captive animals answer readily to their names, not only figuratively by coming when 
called but sometimes literally, also, by means ot a griuU. Naundorft indicates that 
other words, or perhaps tones, were distinguished and elicited response. 

As tor their own voices, kusimanses utter a good deal of bird-like twittering when 
exploring tor food, and a party of them on the hunt in the torest can produce quite a 
volume ot sound. In anger thev growl. At the same time they double their size by 



CHOSSARCHUS 289 

erecting the whole of the pelage until they take on the appearance of a very angry 
fluffy ball. They are for the most part very courageous and will attack strange dogs 
without hesitation if they deem it necessary, despite the vast discrepancy of size. 
Normally, however, they arc very friendly little creatures and delight ni play, either 
with human beings or such other animals as will tolerate them. They often invite 
friendly scratcliing of the head, legs, armpits or belly. There is a curious exception to 
their normal fearlessness; they become extremely alarmed at any loud rustling or 
crackling soimd and at once retreat to cover. Both Ewer and Haig refer to this. The 
latter when travelhng with his mongoose in relevant zones found it to be comiectcd 
with the soimd and smell of grassland bushiires; and one might reasonably suppose 
this to be a logical general explanation but for the fact that Crossarchiis is, in nature, 
a rain-forest dweller where no such fires occur. Haig also observed his kusimansc to 
react in alarm to any overhead shadow which caused it to think that a hawk was 
hovering in the sky. 

This mongoose is a clean animal, according to Haig readily house-trained. Naundorff" 
never observed hers to take any steps to clean its fur yet it was always free of vermin. 
From time to time it suffered from an inconvenient smell owing to exudation from the 
anal glands. Li nature these glands are used for the demarcation of territory, Crossimhiis 
being one of several mongooses that for some unknown reason scent-mark objects 
well above normal nose level by standing on their hands and pressing the anal pouch 
against objects thus brought within reach. Ewer also mentions a simple dragging of 
the anus, in the female, across the floor. 

There appear to be no published records relatmg to breeding in this species. The 
following information supplied by T. S. Jones is therefore of considerable interest. 
A very young pair was taken into captivity in June 1959; 10 months later, on nth 
April i960, the female gave birth to 4 young; 73 days after this, on 23rd June i960, 
she again produced 4 young; and she was once more pregnant, and estimated at about 
3 weeks from term, when 48 days later, on loth August i960, she was accidentally 
killed. This shows that in the kusimanse breeding can start at a prett)' early age; that 
the period of gestation may be about 10 weeks; that litters follow in rapid succession, 
at least 3 a year being possible; and that 4 yoimg at a birth seems a likely average. 
Of the state at birth nothing is known; but the very young are clothed with imderfur 
alone, and their faces are not particularly long. It is only at the age of about 6 weeks 
(Haig, 193 1) that the nose starts to assume its fmal remarkable length and that the bristle- 
hairs lengthen and begin to dominate the pelage. If breeding in the wild is as con- 
tinuous as it would seem to be in captivity' there can be little choice of season. Certainly, 
British Museum specimens give no clear indication of a preferred time of the year: 
juveniles have been taken in West Africa during April and August, and quite young 
specimens in December and January. As regards length of life, a single published 
record relates to a kusimanse known to have lived in the London Zoo for about 
4i years. 

No one has ever described the natural nocturnal shelter or the breeding home of 
CrosSiirchiis. In captivity the species always likes, when resting, to be covered with a 
cloth or to hide itself in straw; and as it has shown itself to be intensely interested in 



290 



THE CARNIVORES OF WEST AIRICA 



Tjblc i6: Numerical data tor Crosiarilim oh.vurus 





ohsi'unis, 


cbsatrus. 




means 


range 


Vegetation 


Forest 




Number in mean 


19 




Coudylobasal length 


70-1 


64-5-74-0 


Basilar length 


64-1 


58-I-68-0 


Palatilar length 


37-7 


34-3-39-4 


Zygomatic breadth 


36'0 


33-9-3S-S 


Upper cheekteeth breadth 


22-0 


20-O-23-3 


Nasals, length 


I9-X 


I7-6-22-0 


Interorbital breadth 


14-2 


I3-I-I5-9 


Postorbital constriction 


13-3 


I1-5-I5-8 


Braincase breadth 


27-3 


25-8-28-7 


Toothrow (f — III-) 


24-1 


22-9-25-6 


p* length 


4-9 


4-4-5-5 


H|l breadth 


5-7 


5-I-6-2 


MI- breadth 


4-5 


3 •7-5-2 


III! length 


5-0 


4-6-5-3 


1112 length 


4-3 


3-5-4'9 


Head & body 


337 


316-369 


Tad 


1 66 


150-190 


Hindfoot 


64 


60-67 


Ear 


23 


21-26 


RATIOS (per cent) 






Tail/head and body 


49 




Zygom. br./condylob. 1. 


51 




Braincase/condylob. 1. 


38 




Braincase/zygom. hr. 


76 




Palatilar l./condylob. 1. 


53 




Interorb./postorb. 


107 




p*lc-iifl 


20-3 





holes of all kinds it must be assuiiicd th.it parturition and family raising, at least, take 
place 111 a hole m the groiuid. The comnuuial life of the kusinianse does, indeed, offer 
several interesting problems for iield study in West Africa. Wliat is the composition of, 
and what degree of stability attaches to the himting parties? Aix they single family 
units; or colonies comprising an aggregation of such imits; and if the latter, do they 
meet fortuitously and spasmodically or is there a deliberate daily reunion? Do the 
companies range the forest at random, fuiishing at dusk far from where they set out at 
dawn; or are they centred on a permanent home, covering only a circumscribed 
territory and returning to it nightly? It might be considered most probable that with 
individual females producing one litter after another and with an association of females, 
therefore, almost continuously in labour or raising yoiuig, there must be some per- 
manent focal point where they can safely remain. If this is so, it at once raises the 
question whether such a rendezvous is a single communal nest or a series of neigh- 
bouring, or even juxtaposed, homes in the style of a warren. No such common home 



ATILAX 291 

or community has, however, been recorded, as it has in the case o( Alnni^os and Lyaion 
(pages 256 and 88), though such a congregation must surely by now have attracted 
attention to itself. Alternatively, a gravid female, when her time arrives, might be left 
by a nomadic pack to fend for herself and, if successful, form the nucleus of a new 
party by the interbreeding of her yoimg or the crossing of the mother again by one 
of her offspring. Such a system might well lead more often to failure than to success; 
but in a fast and frequently breeding society the wastage must necessarily, in point of 
fact, be high. Durrell's lone crab-fisher, mentioned earlier, may well have been such a 
mother whose litter had fallen victims to predators. 

Taxonomy. There arc very clear differences of colour and of the degree of speckling 
between specimens; but such distinctions do not, from present material, seem to be 
in any settled way connected with localities or other factors; and, in fact, very con- 
siderable variation exists within a single region as may be seen amongst skins from 
Cameromi. The drawing of racial distinction is therefore considered neither possible 
nor desirable upon present evidence. 



ATILAX F. Cuvier 
Marsh Mongooses 

Aliltix F. Cuvier, 1826, in E. Gcoffroy & F. Cuvier, Histoire Naturelle des Mammifcrcs . . ., part 54, text 
on the "Vansire": 2. Type species the "Vansirc" of F. Cuvier Hcrpcstcs palndinosns G. Cuvier. The 
position is complex and is referred to below in the taxononiic section. The name is derived from the 
Greek thyhx a pouch, with the prefix a denoting without, given in the mistaken supposition that this 
mongoose lacked the usual anal pocket enclosing scent glands. 

Athylax Blainville, 1S37, Aniils Sci. nat. 8: 272. Correction of the spelling of the above in accordance 
with its etymology. 

Distribution and general, hi spite of early confusion and doubt regarding its 
correct naming and identity, dealt with below in the taxonomic section, Atilax is 
today one of the few mongooses about whose generic independence and identit)' 
there is no dispute. The marsh mongoose, frequently also called the water mongoose is, 
as these names imply, possibly more closely associated with water than is any other 
species. Taylor (1970), indeed, states it to be the only viverrid known to be a good 
swimmer, and this is so for the region covered by this present work; but there are, 
of course, elsewhere in Africa the aquatic civet [Oshoniictis), and in Asia the otter-civet 
(Cyijoijiilc). Yet it belies all expectation that might logically arise from this fact in 
being the only one of its kind to lack any sign of webbing between the toes — a character 
which, thus, constitutes an infallible external means of recognition. Atilax is widely 
spread in Africa south of the Sahara, from Senegal or Portuguese Guinea in the west, 
and Ethiopia in the east, southwards to the Cape. It is, of course, not of equal dis- 
tribution throughout this extensive range, but it is almost certainly to be fomid where 
there is adequate permanent water and vegetation of sufficient density to afford secure 
cover; and in localities in which these conditions arc well met it may even be regarded 
as common. 



292 THE CARNIVORES OE WEST AfRICA 

Dirtcrcnt species have troni time to time been named, but today the genus is luuver- 
sally held to be monospecific. This being the case, generic description, cither of morph- 
ology or habits, is luicalled for, all matters relevant to these being contained in the 
account o( piilniiinosiis which follows. 

Taxonomy. Reference has already been made above to the fact that the early 
history of the name .-lf/7,i.v and the precise animal to which it was intended to refer 
was subject to some doubt. Fortunately, J. A. AJlen (1924) cleared the matter up and 
there is no point m going over the whole ground again and making more than a 
passing reference to it here. The question was of sufficient complexity to take Allen 
three or four pages of closely packed fact and reasoned argument. It was part and parcel 
of the vexed question of the identity of an animal known in early literature as the 
"Vansire" and variously described and referred to a complex mixture of different 
teclmical names by a number of early naturalists, including Bufton, Daubenton, 
Schreber, Erxlebcn, Flacourt, the two Cuviers, the two Gcolfroys and others. The 
animal was alternatively held to inhabit Madagascar or various parts of the African 
continent: the names involved were I'lVcrw, Mustchi, Ciilidiais, Galidia, Hanii^alidia, 
SdhiHoia, MdWiiiistci, Hcrpcslcs, Atihix, sjiilcri;, paludiiiosiis, ch\^ai:s, ra».\//c, iHWi^-shin-, 
uriiiairix and possibly more. Allen boiled all this down to die virtual certainty that 
the Vansire of Button i5\' Daubenton (1765) was not that of F. Cuvier (1826), and that 
in the process of describing that animal the last audior proposed the genus Aiilix which, 
though in point of fact ineptly named from a supposed character that had no validity, 
was sufficiently described m another, the uniquely unwcbbed toes, as to render its 
identit)' quite luimistakable. Allen's conclusions can be thankfully accepted and 
there seems no virtue in ever raking over this troubled soil again. It would be a fortunate 
thing for mammalian taxonomy if all early classic genera and species could be so 
satisfactorily and conclusively pin-pointed. 

Li the matter of species no serious challenge has been made, or long sustained, 
to the monospecific nature of the genus though Gray (1865a) proposed rohnstus from 
the White Nile and was followed in this by J. A. Allen (1924) who at the same time 
described maavdoii from the eastern Congo. These, being extralimital, arc of no 
particular concern to this work; but they are today usually regarded as nothing more 
than two of the ten local races so far described. 

ATILAX PALUDINOSUS (G. Cuvier) JMarsh Mongoose 

Miistcia \;aleia SchrcLicr, 1776, Die Sini^ciliicrc in AblnldHn(;at uath lia Wilin .'. ., pi. 135; 1778, text 3: 
493. Madag.iscar. Regarded as unidentifiable (J. A. Allen, 1924). The name jjn/tm is Latin for helmet, 
probably given with leferencc to the skull. 

A/iii(i/a voaii(;-shire Zimmennann, 1777, S/'hi/hcii zoolc^iac giographicae, etc.: 487. Zimmermann, 1777, 
has been ruled an unavailable work {Bull. :ivl. Noiii., 1950, 4: 547). This specific name is a translitera- 
tion of the reputed vernacular name. 

HirpeitespaluJinosHsG. Cuvier, 1829, Le Ri}<iic Animal . . . cd. 2, 1: 15S, Cape of Good Hope. The specific 
name is derived trom the Latin pains, pahidis a marsh, and refers to the habitat. 

Maii(;usta urinatri.\ A. Smith, 1829, Zocl.J. Li'iii/. 4: 437. South Africa. This specific name is the teminine 
form of the Latin uriiiator a diver, in reference to the animal's habit while foraging in the water. 

Hcrpeshs atilax Wagner, 1841, m Schreber, Die Saugetliiere . . . Supplement, 2; 305. South Africa. Deriva- 
tion ot this name is given under the genus. 



ATILAX 293 

Alilax vaiisire F. Cuvier, 1842, in E. Geoffroy & F. Cuvier, Hisloire NauireUc dcs Mannniferes, etc. Table 

gencrale, 3. This is the reputed vernacular name of the animal whose exact identity gave rise to so 

much confusion as explained above. 
Aihylax paluJosiis Gray, 1865, Proc. zool. Soc. Lond. for 1864: 557. The Latin adjective pahidosiis means 

marshy and refers to the annnal's habitat. Cape of Good Hope. 
Hcrpesh's pliito Temminck, 1853, Esqiiissfs zoologiqiies siir la cote de Giii'ik': 95-96. Dabocrom and River 

Boutr)', botli Ghana. Pluto was the Greek god of the underworld; the name was thus intended to 

indicate the entirely sombre, blackish appearance of the animal. 
Atilax paludinosus giiitieeiis i s Monard, 1940, Arches Mus. Bocage, II: 198-201. Catio, Portuguese Guinea 

Distribution. There are more West African specimens of this mongoose in the 
British Museum than of almost any other species, and from a wider range of locahries. 
This is probably a good measure of its relative abundance on the ground not only in 
West Africa but as regards the continent as a whole. The Marsh Mongoose is, in fact, 
a well-kiiown species from a wide variety of locahties from 12 or 14 degrees north 
southwards to the Cape. 

Li so far as the tcrntory covered by this book is concerned there is an early specimen 
(1863) labelled as coming from Senegal, without any specified place, and this may 
possibly be correct or it may be nothing more than a loose gencrahzation, not un- 
common at that period, for somewhere on the western coast of Africa. However, 
the species is defmitely recorded from as far west as Portuguese Guinea. Thence it is 
known from ever)- West African country to Cameroon, and there seems httle point 
in recording all the names of the many places at which it has been taken. All in the 
British Museum are from the closed-forest belt or the contiguous Guinea woodland; 
but this is not to say that this mongoose does not occur in the drier zones of vegetation, 
for extralimitall)' the species commonly inhabits Acacia country-. The basic criterion 
for the habitat would seem to be the existence of adequate permanent water suitable 
for di\Tng, swimming and fishing, bordered by sufEciently dense vegetation to afford 
safe cover. It does not matter whether that vegetation is of the forest, shrub and herb, 
type or close grass and reeds with no backing but fairly open country behind it. The 
most commonly used English name, marsh mongoose, must not be taken too literally 
but only as implying the sort of waterside habitat that may or may not be subject to 
some degree of inimdation. The requisite conditions are undoubtedly of most frequent 
occurrence, in West Africa, in the rain-forest and Guinea woodland, in which forest 
river fringes commonly exist. There are in the former zone also extensive areas of 
freshwater swamp and of mangrove swamp, in the latter of which, despite the im- 
promising brackish nature of the water, the marsh mongoose is known to be common. 
There seems basically little reason why this animal should not be encountered over a 
much wider range of countr)' than that from which it is at present recorded since 
rivers or lakes with bordering vegetation are common features of all but the most arid 
zones — and even here sometimes there are dense marshes. However, A. J. Hopson, 
an interested observer, states (in a private communication) that in 7 years on Lake 
Chad he neither saw it nor heard of any record of its occurrence there. There is, how- 
ever, one important factor that has had a great influence on the actual present day 
distribution of this mongoose. Riverine vegetation is, for one reason or another, 
u 



294 '"' cAi!Nivoiu;s or wist aiuica 

pi'culi.irly pioiic to dL'structioii, and especially tor die pturpose ot cultivation. The 
enonnoiis increase \n luinian population and the concomitant agriculture that has 
taken place in the past three-quarters ot a century has resulted, in West Africa, in the 
complete disappearance of vast areas ot potential shelter for AliLix. and it can be said 
with considerable certainty that the occurrence of this genus in the drier woodland 
7cines is today very much more restricted than it was tifty years ago. 

Description. Atilax (l*late 7) is one of the group of larger mongooses, ranking with 
Ilciihiiti, Ithnciiiiiiii. Xfiioi^iilv and Galcrifciis, though, at least 111 West Africa, a little 
less in size than any ot them, especially the last. It is usually described as ,1 wholly 
"black" mongoose, that is to say of a very intense brown; but though this is its most 
connnon appearance dicre is, in fact, a considerable degree ot colour variation and die 
pelage sometimes assumes a rich reddish-brown; or it may be a medium-brown and 
heavily speckled. In its normal dark form the animal can verv easily be confused with 
AV»(\'ii/c' ihiso, a ditiiculty dealt wuh under that species. 

The pelage ot AtiLix is long, glossy and tairly harsh. It is composed ot dense, sinuous, 
tine, moderately long (about 15 mm) miderfur and bristle-hairs, 32 to 40 mm long, 
tapering to a narrow base and of flat oval section. These, though abiuidant, do not 
always conceal the imderfur. They have a longish pale-brown basal region, but any 
banding on them is irregular m situation and often obscure, hi some specimens the 
bristle-hairs may be wholly dark, almost black, v,-liile 111 others there are indistinct wliite 
rings, narrow and inconspicuous. Some, on the other hand, arc by contrast notable 
for a series of three very clear, pale, widely separated rings, which impart a "ticked" 
etfect to die whole dorsal pelage. These are all narrow, diat is not more than about 
2 mm broad, the two lower ones v/hitish-yellow, the subterminal one orange. It will 
be seen, dien, diat it is not practicable to give any general description that might not 
be misleading ot a coat that shows so many degrees of variation, except to say that the 
impression is mostly of a very dark animal. The underfur is similar in colouring to die 
back but a shade paler; the hairs on die throat are directed forwards. 

The head o( Atihix is wide, the muzzle not very long, die rhinarium broad and black 
with a furrow down its front face that is continued below as a naked strip parting the 
upper lips. The crov/11 and clieeks arc usually fuiely speckled; and there is mostly a 
relatively bare area around the eye. The ears are of the usual low, broad, rounded mon- 
goose form, set well on the side of the head below the crown and rather obscured by a 
thick growth ot hair in tront of diem. The tail, which is roughly two-thirds of the 
jiead & body length, tapers evenly from a broad base, not clearly differentiated from 
die body, to a narrow tip; and it is long-haired throughout but there is no terminal 
tuft. The short legs and feet are of much the same tone as the back or a little darker; 
the feet themselves all s-toed, with strong but not exceptionally long claws, and 
notable as being, alone of all mongooses, completely unwebbed. The soles are wholly 
naked around the pads, and in die hindfoot this nakedness is continued posteriorly to 
the heel. As explained previously, the name Alilax was given 111 the belief that this genus 
was imfurnished with any glandular anal sac; but this is not so, the folds of skin 
surrounding and enclosing the scent glands and anus being, in point of fict, well 
developed. The glands are situated one cMi each side of the rectal orifice. 



ATILAX 295 

Skull (tig. 38). This is amongst the broadest of West African mongooses, the zygo- 
matic breadth amounting on the average to some 58 per cent of the condylobasal 
length. The braincase is narrow ovoid with a well-developed, flange-like supraoccipital 
crest. There is always a short posterior section of sagittal crest joining this m a T; 
but development of die former over the main body of the cranium is variable. It is 
mostly better in the males than females: in existing British Museum material it is 
pronoiuiced in 3 and low in 6 males, slight or very slight in 5 fenialcs and rudimentarv' 
111 4. hi only one case out of 21 measurements is the interorbital breadth very slightly 
less than the postorbital constriction; it is mostly markedly greater, varying in extent 
from about 5 to 50 per cent more. The postorbital processes are long and sharp, and 
not infrequently nearly, and occasionally completely, meet the jugal processes to 
form an orbital ring. The frontal region is broad and slightly inflated; the rostrum short, 
blunt and broad; and in fully mature skulls the nasal sutures almost entirely disappear by 
fusion. The zygomata are strong, sometimes broadened in the middle section, sometimes 
not. The postdental palate is obviously longer than it is broad mostly from about 30 to 
50 per cent but occasionally rather less. The anterior chamber of the bulla is small and 
flat; the posterior chamber very much larger, well inflated, rather pear-shaped. 

Both upper and lower incisors arc set in compact, practically straight transverse 
rows, the outer ones, especially in the upper series, being appreciably larger than the 
inner. All the upper ones, in unworn teeth, are clearly, though mmutely, cuspcd on 
their posterior faces. The upper canines are rather compressed from side to side and 
are furnished anteriorly and posteriorly with sharp knife edges; but they are neither 
so large nor so straight as those o{ Giilcrisciis. The posterior outer corner of/)'* is situated 
very close to the point where the posterior outer corner of the maxillary root of the 
zygoma arises; and in consequence the whole of i»i lies behind this point. 

One of the standard characters generally given for Atilax is that the premolars 
arc y, and this is so in nearly all southern and eastern African specimens. But in 26 
British Museum West African, mostly fully adult, skulls it is true in only 9 cases; 
in 6 other skulls these teeth arc 4; in 5 they are j; in one other --,; and in a juvenile 5. 
The remaining 4 cases arc more complex, the lower jaw having a differing number of 
premolars on either side: they arc j^j, jj2, jn 'I'^d jij- It is true that the "extra" pre- 
molars arc always very small; but they are little or no smaller than the corresponding 
teeth in genera characterised as habitually possessing 4. Wliile it is clear that the anterior 
premolars in Aiilax are on the way out it is obviously misleading in West Africa to 
state that the cheekteeth are j^, the exceptions being, seemingly, appreciably com- 
moner than the rule, m^ is a large tooth, nearly as large as /)''; iii- is small but not ver\- 
small; that is, it is transversely somewhat wider than the lingual section of »|i and is 
subequal to this in bulk. There is, however, a very occasional exception to this, the 
tooth being of reduced size, hi the mandible the caiuncs are much more curved than 
in the upper jaw; mi has, in its anterior half, three pyramidal cusps, equilaterally 
set and all very nearly the same height, not with the external one clearly the highest 
as in several other genera. 



296 



Tlir. CARNIVOIUS or WEST AIRICA 




Fig. 3S. Anla\ paliuihnmir. ■•kiill, B.M. No. T3.4.1.1, ^^ x 



ATILAX 297 

Habits. Somcthmg has already been said of the water-hauiiting habits ot this mon- 
goose : it dives and swims almost as well as an otter and will travel below the surface 
for some distance. Indeed, it is said that when alarmed it will remain for a long period 
completely submerged except for its nostrils. Certainly when disturbed on land it 
makes off towards the nearest water and, if still pursued, plunges in. Taylor (1970) 
describes it as swimming by lateral undulations of the body with the back partly 
exposed. He observed a specimen to remain submerged for 15 seconds, and states 
that the fur becomes completely wet. Yet, in spite of all this, observers have from 
time to time reported Atihix from dry^ country at some distance from the nearest river. 
Little or nothing that is defmite is known of the breeding habits of this mongoose; 
some say that it makes a nest amongst the reeds, others that it raises a family in a 
burrow in the banks of a stream. There are said to be two in a litter; but hunters in the 
Cross River forests insisted to Gerald Durrell that it consisted of only a single baby 
(personal communication). There is nothing from which any particular breeding season 
can be inferred in West Africa. 

Except when actually bringmg up a family the marsh mongoose lives smgly or 
sometimes in pairs. Because of the hidden life it leads amongst dense waterside vegeta- 
tion htde is knowm of its daOy habits, and because it is so little seen it has the reputation 
of being entirely nocturnal, hi actual fact, although it may be active on bright moonlit 
nights, it IS for the most part, as Lombard (1958) has pointed out, crepuscular rather 
than nocturnal and almost certainly carries out the majority of its foraging and feeding 
in the early morning and late evening. It is sometimes to be seen trotting along a path 
or a motor road at midday. 

The food consists of frogs (these have been taken from the stomach of a Nigerian 
specimen), reptiles such as lizards, small rodents, a large number of insects of various 
kinds, fresh-water mussels and fish — though Lombard doubts whether it has the 
ability to capture large specimens of these last and thinks it probably confines its 
attention to smaller fry in shallow water. Crabs, too, are always cited as one of the 
staple articles of diet; but whether Atilax can satisfactorily deal with any but the 
smaller, softer-shelled varieties does not appear to have been adequately shown. To all 
this can be added an at least occasional diet of small birds and eggs; the nests of reed- 
haunting species such as rails and others must almost certainly be raided as opportunity 
offers. Stevenson-Hamilton (1947) rates this mongoose with the monitor lizard as the 
deadliest enemies of the crocodile, diggmg out and avidly consuming its eggs. 

Several kinds of mongoose have been observed to crack eggs or similar shelled 
objects by hurtling them with the forefeet forcibly backwards between the arched 
hindlegs against a rock or other resistcnt body. Whether Atihx is one of these docs not 
seem to have been recorded; but that it takes objects in its forepaws, stretches up and 
then flings them to the ground with great force is well authenticated. Steinbacher 
(1939 and 1951) describes how two separate captive specimens dealt with food problems 
in this \\-ay, the one with nuts, the other with snails that had withdrawn into their 
shells. Snails that were fully exposed were bitten into directly. It is of interest to note 
that nuts meant nothing to this second mongoose which made no attempt to open them 
as the other specimen had. Lombard (1958) observed AtiLix to break mussels open 



298 Tin; carni vdkls oi- wlst mkica 

using tlic same method. He .ilso recorded tliat it an egg could not be bitten by the 
teeth It would be cracked on the ground, and a one-meli hole opened up througli which 
the hquid egg could be licked. 

The same author, who gives the most complete account ot the habits ot the marsh 
mongoose that has so tar been published, describes how it uses its highly sensitive 
forcpaws to feel about in the mud for river mussels; and also, on land, continually 
pokes them under rocks or into holes and crevices in the hope of tinding insects. He 
says, too, that its very keen sense of smell enables it to locate articles of food, even 
subterraneous worms. Fmallv, he provides a rare record of a pet marsh mongoose 
having of its own free will on two separate occasions attacked and killed snakes. 
It did not eat them. 

One other thing must be mentioned m coimexion with teeding. AtiLix is one ot the 
many viverrincs with a reputation as a fowl thief and for this reason, as well as for its 
own value as protein food, it is much hunted, h may take eggs it they are readily 
available, as they so often are in Africa, but the fowl stealing, in a water-haunting animal 
which lives mostly on small prey, is probably exaggerated. However, it is the subject 
of a widely spread fable that must be mentioned here, though at present there seems 
to be no record ot it m West Africa, hi the vicinity of a fowl Aiilax is said to stand still 
with its hindquarters towards the bird and to open the sac that surromids and normally 
conceals the anus and the rwo laterally sited scent glands. The fowl, seeing the bright 
colours thus exposed, becomes curious and, thinking them to be those ot some ripe 
and attractive fruit, advances to peck at the spot — whereupon the mongoose turns and 
springs upon its foolish prey. Pitman (1931 and 1954) m commenting on this belief is 
inclined to think that a tale so widespread in Africa might have some foimdation in 
fact. The story is the more intriguing in the light of the literal implication of the generic 
name AtiLix and offers an amusing comparison between the powers of observation of 
African field naturalists and European study mammalogists. 

Like several others ot its kind the marsh mongoose respiMids well to captivity and 
becomes very friendly — some specimens, at least, trusting enougli to invite head 
scratching from complete strangers. When pleased it makes a purring sound, somethmg 
after the style of a cat; when alarmed or displeased it growls. Having scent glands it 
has, at least occasionally, a musky odour; but it opens up the circumanal pouch and 
actively emits a malodorous fluid only as a last resort in danger. Normally these glands 
arc used for recognition purposes, imparting a characteristic smell to the droppings, 
and being used more dehberately for demarcation purposes by being pressed against 
key boundary marks. Hediger (1949), Fiedler (1957) and Hinton & Dumi (1967) all 
say that .-lfi7.!.v carries out this delinntation of territory, if necessary, on high objects 
by performing a "handstand"', thus bringing die anal pouch into a position well above 
the normal body level; but the taxonomic determination ot the East African mongoose 
forming the basis of this observation is m some doubt. 

In a cage, AliLix has shown itself capable of climbing several teet vertic.illy up wire 
netting. Such a surface, of course, affords positive footholds difficult to equal in nature; 
nevertheless, it seems certain diat low mud walls or trees with sloping trunks could be 
successfully tackled if necessarv. The normal gait on land of the marsh mongoose is a 



ATILAX 299 

Steady trot vvidi the tail stretched out behind, soinetiines a Uttlc curled upwards. One 
of these animals is recorded {Int. Zoo Yr. Bk. 2) as having hved iii years in caprivity. 
Taxonomy. A number of local races has been named. This is to be expected in an 
animal which exhibits such variation of external appearance as Atilax does; but whether 
such variations can be constantly related to localities or to definable ecological back- 
groimds is very much open to question. Two races have been set up for West Africa, 
phito Temminck and ^uineensis Monard. Temminck's name was given to a blackish 
form from Ghana; and such animals certainly occur at various places in West Africa; 
but there are as many lighter, browner specimens current in the same areas. Monard 
described {^iiiturnsis from Portuguese Guinea as being smaller than the typical form, 
with a less massive skull and vsith 4 premolars above and below. Measurements of 
West African material in the British Museum, including specimens from Portuguese 
Guinea, arc not very different from those given for typical South African paliidinosiis 

Tabic 17: Numerical data (or Atilax pahidiiiosus 





pahiditiosiis. 


paludiiwsus. 




means 


range 


Vegetation 


Mostly 


Forest 




and Guinea 


Number in mean 


20 




Condylobasal length 


100-6 


96-o-:o6-5 


Basilar length 


91-3 


87-4-98-8 


Palatilar length 


55-^ 


SI-2-60-7 


Zygomatic breadth 


58-0 


52-5-^8-6 


Upper cheekteeth breadth 


35-5 


33-3-38-3 


Nasals, length 


25-2 


21 ■9-27-8 


Interorbita breadth 


19-9 


I7-3-24-0 


Postorbital constriction 


i6-8 


13-0-19-2 


Braincase breadth 


37-7 


35-2-40-3 


Toothrow (f — III-) 


36-2 


31-8-40-1 


p* length 


9-8 


8-3-11-2 


i»l breadth 


10-5 


9-2-11-8 


iii2 breadth 


6-5 


5-0-7-6 


mi length 


9-5 


8-4-10-8 


012 length 


6-3 


5-2-7-2 


Head & body 


501 


442-553 


Tail 


322 


250-355 


Hindfoot 


96 


84-102 


Ear 


33 


28-40 


RATIOS (per cent) 






Tail/head & body 


62 


57-71 


Zygom. br./condylob. 1. 


58 




Braincase/condylob. 1. 


37 




Braincase/zygom. br. 


65 




Palatilar l./condylob. 1. 


55 




Interorb./postorb. 


118 





p4/c. 



<*lc—n 



300 TH1-: CAHNIVOUFS OF WLST AIUICA 

by Roberts (lysi): -iiid the presence ot /)' and /m. though coninioner in West Atric.i 
than elsewhere, as shown above .md in I layman (1935). is of irregular occurrence. 
Before naming this race Monard siibmitted his data to Professor Bourdelle of Paris 
who, on Monard's own admission, seemed very dubious of the existence of any valid 
distinction between the Portuguese Ckiinea animals and r\'pical material. 

The present writer is very doubtful that local races can be usefully named and 
distinguished; and certainly not before considerably more abundant data dian at present 
available have been gathered. 



Genus ICHNEUMIA I. (ieotfroy. 1X37 
\X/]iite-tailed Mongooses 

Lasiopus I. Geoftroy, 1S35, in Gervais' Rvsiiiiii- As Lf^oiis lic Mainiiialogic . . . piojesitrs an Museum de 
Paris I: 37. Not Lasiopus Dejean, 1833, Colcoptcr.i. Type species Hcrpeslcs aWicaudtis G. Ciivicr. This 
name is derived from the Greek lasics hairy and pous foot, from the fnrry nature of the soles. 

liluicuinia I. GeolTroy, 1837, .41111/1' Sci. iiat., Zool. (2) 8: 251. A new name to replace Lrii/ii^iis, preoccupied. 
Type species Hcrpcstcs alhicaudus G. Cuvier. Tlie name, trom the Greek, ichneuo to hunt or track, 
refers to the animal's habits. 

Distribution. This is one of the commonest of African mongooses, ranging through 
the grass-woodlands, but not the forest, soudi of the Sahara, from Portuguese Guinea 
in the west and Sudan and Somaliland in the east to parts of the Cape Province 111 
South Africa. It also occurs across the Red Sea in southern Arabia. The white-tailed 
mongoose is one of the largest and, when it lives up to us English name, certainly 
one of the most striking in appearance: but this unfortunately it does not always do, 
in many West African specimens the tail being black. The combination of size, white 
tail and black legs could lead to confusion with the Black-footed Mongoose (Galcri- 
sciis): but. apart trom the fact of the ranges of the two species being quite different, 
the animal now about to be dealt with 111 diis section has a far bushier tail and long, 
loose fur (see Plate 8). 

Taxonoiny. Iduuiiiuia is one of tlie tew African mongooses about whose identit)', 
naming and taxonomy there has been a minimum of argument. Its striking appearance 
and clear-cut characters have trom die start virtually prevented confusion with others, 
though in common with most mongooses it was at one period regarded as a species of 
the all-embracing genus Hcrpcsrcs — which, as explained elsewhere (page 248), itself 
tor a time became confused with Miin{;os. Thomas in his 1883 arrangement of the 
African mongooses set Ichnciiinia at the level of a subgenus q{ Hcrpeslcs; but Pocock 
(i9i6e) recognised it as a full genus, a status that has since remained iniquestioned. 
Since Hcrpcstcs belongs to the group of mongooses with cheekteeth of a predominantly 
sectorial character set well back in the jaw, and Ichncuiiiia to diat with teeth of an 
obviously more crushing rj-pe set much further forward, it is doubtful whether the 
two genera are, in fact, particularly closely related. 

As for the composition of the genus, it is universally regarded toda)' as monospecific 
though .1 number o[ independent species have from time to tune been erected. These 



ICHNEUMIA 301 



mostly date from early days; but as recently as 1924 J. A. Allen, dcalmg with eastern 
Congo mongooses, referred them to leiiaira rather than the now generally accepted 
alhicaudd as it did not appear to him probable, on geographic as well as other grounds, 
that the two could be specifically identical, hi this present work only alhicauda is regarded 
as valid. 



ICHNEUMIA ALBICAUDA (G. Cuvier) White-tailed Mongoose 

Hvrpeslcs alhicmidus G. Ciivicr, 1829, Le Rc'^ne Animal, ed. 2, I: 158. Senegal. The specific name is from 

the Latin alba white, and caiida tail. 
Herpesics leumrus Hemprich & Ehrenberg, iS3}{jidc a MS note of Sherbom's), Sytiibolae Physicae, sen 

hones et Dcsaiplioiies Maiiiiiialiuii: . . . decas 2, folios /(, 1, fe, Pi. 12. Dongola, Sudan. This name was 

derived from the Greek leucos white and oiira tail. 
Hcrpestes locnipo Tenmiinck, 1 853, Esqiiisses zoologiqiics sur la cole de Guine. "All the coast of Guinea" — 

almost certainly Ghana. The name was said to be a vernacular one meaning man-eater because this 

mongoose was held to rob graves and devour the corpses. 
hlmeumia iiigricatida Pucheran, 1855, i?ei'iif Mag. Zool. (2) 7: 394. Senegal. The name is compounded 

from the Latin nigra black and caiida tail. 

Distriburion. This large mongoose (Plate 8) of striking appearance and a weight 
of about 5 kg is spread over a good deal of Africa south of the Sahara but avoids the 
high forest and the zones of intense aridit)'. It may, however, occasionally be found 
at places which are nominally in the forest belt but which have, in fact, been persis- 
tently cleared and arc now open farmlands or invaded by Guinea woodland vege- 
tation. 

Dekcyser (1955) gives the range as extending as fir west as Senegal; and Monard 
(1940) records specimens from four places in Portuguese Guinea. Thence, eastwards. 
It may almost certainly be expected anywhere in the Guinea, Doka and Sudan wood- 
lands. The most inland specimen in the British Museum is from Maiduguri (north- 
eastern Nigeria); but a reliable observer stated a few years since that this animal was 
very commonly caught and brought in by dogs in Sokoto station (north-western 
Nigeria, 14° N.); and A. }. Hopson {in lilt.) found it to be common amongst salt bushes 
(5i!/i'ii(/(>n( pcrsica) near the shores of Lake Chad and the adjacent dunes; and it was 
often seen at night on the outskirts of lakeside villages; but was never observed at any 
great remove from the lake itself. Angus Buchanan obtained three specimens at Faniiso 
(Kano, Sudan woodland) but no others occur in his comprehensive collections from 
further inland in the more arid Sahcl and Subdescrt zones. Altogether, 21 skins and 
15 skulls exist in the British Museum from Sierra Leone, Ghana and Nigeria. The 
white-tailed mongoose may therefore be regarded as being, in the right localiries, 
fairly plentiful. 

Description. Icliiiciiiiiiii ranks approximately in size with Hcrpestes, AViiinjdlc, 
.d//7.;.v and Giikrisciis, that is to say it has a head & body length of at least 500 mm and 
a weight of some 4 to 5 kg; but it is not likely to be confused in the field with any of 
these, Hcrpestes having a very sharply tapering tail with a contrastmg black tuft, 
Xenogiile and .4fi7(j.v being mostly very dark animals, and Galeriais, despite its super- 



30a llli; CARNlVOUliS Ol whj.t aikica 

ticially very similar bl.ick legs and whitf tail, is contiiic'tl to the dense tdiests of the 
Cross River. Ichncunm is, in tact, a very easily recognisable mongoose, with a loose- 
furred buft or grcyish-biift coat more or less heavily splashed with black, wholly black 
legs, and an abinidantly long-haired tail which in its most typical form is brilliantly 
white (Plate S). 1 lowever, in West Africa it is at least as commonly black, there being 
in the British Museum collection 7 white to 12 black tails, and 2 half and half; but 
whichever it is, it and the black legs and teet bodi contrast sharply with the basically 
buft body. 

The lengthy dorsal pelage is, in texture, loose, fairly sott and springy. It is composed 
chiefly of dense, relatively long, slightly wav)', tine underfur and extremely long, 
oval-sectioned bristle-hairs. The underfur measures 20 to 25 mm and is buftish, pale- 
brown or grey-brown and plays a considerable part 111 determining the overall colour 
impression. The bristle-hairs are always vei"y long, reaching their ma.ximuni on the 
lower back; the actual length varies a good deal from specimen to specimen, in some 
measurnig about 50 mm or even a little less, in others attaining as much as 65 mm. 
They also var)' considerably in their colournig. There is usually a white base, which 
may be from 3 to 9 mm long, and then the remainder ot the hair may be either all 
black or may comprise one or two white zones alternatnig with pale or very pale 
brow-n. The white is never very conspicuous as such in the general impression of the 
coat; but the pale areas of the bristle-hairs, together with the tuiderfur, make up the 
basic pelage colour, die length and number of dark terminal zones determining the 
degree of superimposed black. These two elements, wlule producing coats that are 
recogmsably similar, account tor pale animals or tiark animals. On the underside the 
fur is sparser and paler with many tewer bristle-hairs and, generally, much less black. 
The fur of the throat is directed forwards and meets that of the chin at the angle of the 
jaw m a transverse wave. In the very young the coat is always pale and contains very 
fev,- black-ended bnstle-hairs, which increase in number with maturity. 

The head is ot moderate breaddi, the muzzle furly pointed; both crown and face 
are paler dian the rest of the upper pelage, speckled; and there is a bare area round the 
eve. The ears are low in height, roimded and very wide; the outer portion situated 
very much on the side of the head but the inner margins reaching much further up the 
crown than in most other African mongooses. Their forward aspect is rather obscured 
by a screen of long hairs rising m front ot them. The median groove ot the rhinarium 
is conrinued downwards as a narrow bare band parting the upper lip. 

The hindfeet and all of the forelegs and feet are practically black, but differ from 
the similar ones of Gakriscus in having 5 toes in place of 4. These are webbed to the 
subdigital pads. The hindfeet arc appreciably longer than those of other mongooses 
of similar size and Icliiiciiiiiid tends to stand higher oft the ground; the soles are com- 
pletely, or almost completely, hairy up to the pads. The very long-haired, tapering 
and ver}' shaggy tail is somewhat shorter than the head &: body; but different collectors' 
figures give widely differing degrees of this, var)'ing from 69 to 99 per cent. This is 
probably partlv due to some measurements having been taken to the last joint, some 
to die tip of the fur. But irrespective of diis there does seem to be a fiirly wide variation 
of ratio troin specimen to specimen; and this has something to do widi the degree of 



Whitc-tailcd Mongoose {Ichnvtitnut alhkauda)'. Black-tooted Mongoose {Giileri^ais nii^ripes) 



ICHNEUMIA 303 

maturity; Dalton (1961) recording that tail length increased enormously with advancmg 
age. The tail, as already noted above, may be either white or black or half-and-half, 
black prcdoniinatnig in West Africa, though relatively rare elsewhere. Maclnnes 
(1952) records being mformed that both black and wliite tails may occur in a single 
litter; but no positive evidence is given. There is the usual herpestine anal sac surround- 
ing and, when closed, concealing the anus and the two laterally sited orifices of the 
scent glands. 

Skull (tig. 39). The West African material in the British Museum is very inadequate. 
Although there are 15 skulls only 8 are reasonably mature, and of these 3 are broken. 
If the skulls of the other large mongooses, Herpestes, Xenot^ale, Atilax and Galeriscus are 
compared with that oi Ichncwnia it is seen that, although the distance from the glenoid 
tossa to the condyles is much the same in every case, the length of skull anterior to this 
IS, with the exception of Galeriscus, appreciably greater in Iclinciiiiu<i. The postorbital 
constriction, too, is markedly broader in this than in other genera and the braincase 
consequently rather less ovoid. There is no great difference in measurement between 
the interorbital breadth and the postorbital constriction though the former is in the 
majority of cases, but not invariably, slightly the smaller. The whole frontal region 
appears broad and inflated. The supraoccipital crest is as m other genera a pronounced 
flange; the sagittal crest is mostly not very highly developed, but old females as well 
as old males occasionally produce a deep knife-edge. In fully mature skulls, but not 
young adults, the orbital ring is complete. The palate, though not wholly vaulted as in 
Giilcrisciis, is in most cases not entirely flat as in Hcrpcslcs, Xcnci^ale and Atilax but has a 
broad, very shallow, medial, longitudinal depression, reaching approximately from 
/)" to fii^. Li these three genera also, there is very little extension of the palate immedia- 
tely posterior to in- ; but in Ichncuinia there is a shelf of bone sufficiently large to accom- 
modate a further molar of fair size. The central postdental palate is somewhat longer 
than broad. The anterior chamber of the bulla is small, the posterior one highly in- 
flated. 

The toothrow is relatively somewhat longer than in other genera. The posterior 
outer corner of/)'' is situated well anterior to the posterior root of the maxillary process, 
in^ correspondingly lying wholly anterior to that point (fig. 33a). There are always 
4 premolars each side in both upper and lower jaws; and the anterior ones, although 
always small, are somewhat bigger than in other genera in which they are present. 
)»i is a large tooth, equal in bulk to the upper carnassial; and iifi is also large. The cusps 
arc low, even in young teeth, and the occlusal surfaces of the posterior cheekteeth 
are adapted more to crushing than to cutting. The upper canines are subconical and 
shghtly curved. The upper incisors are in a compact, very shghtly curved row; all, 
unless well worn, are cusped on the posterior face, I'l and i'^ minutely so, i'^, which is 
larger than the others, more pronoiuicedly so. The lower canines arc short, strong 
and fairly abruptly curved. 1112 is only slightly smaller than im. and on both of these 
teeth the cusps arc low and soon obscured; but when unworn those of the anterior 
portion of iin are subcqual in size, the posterior lingual one being very slightly the 
smallest. 

Habits. Although this is such a common and widespread mongoose very little that 



304 



Tin; c;ahm\()ki;s oi \%isr airica 




fic. i'j. /(/i»ci/H/i.i rt//'iuiKi/,i: sUill, H.M. No. 12.4.3.3, ;, ■ I 



ICHNEUMIA 305 

IS precise has been recorded of its habits in the wild. It is always said to be shyer than 
most and reserved even in captivity — though this last is not always true, as will be 
seen later. Ichncwnia is a large and conspicuous animal and one might suppose it, 
in consequence, to be well observed. It is certainly common enough in museum collec- 
tions, but these specimens must be largely the outcome of trapping since no one seems 
to have recorded encounters with it in the field, apart from occasional views in the 
headlamps of cars. The fact is that, besides being solitary, it is pretty strictly nocturnal, 
spending the daylight hours hidden deep in holes or in dense undergrowth, the holes 
being those excavated or constructed by others — aardvarks, porcupines, termites 
and so forth. There arc, nevertheless, exceptions to this: T. S.Jones (personal com- 
munication) has noted that one regularly crossed a garden in Makcni, Sierra Leone, 
every evening at about 5.0 p.m., bemg eventually killed by a gardener. Some writers 
connect Ichneumia with the vicinity of water; but it certainly occurs at more distant 
sites; and even when it does live in the neighbourhood of water it is not so aquatic as 
Atihix, though Taylor (1970) says that it does sometimes swim. However, Dalton 
(1961b), writing of a tame specimen taken to a stream, said that it made its feet wet 
but made no attempt to enter the water further. 

There is no account of the breeding of this species, the period of gestation or the 
number of young; it can only be assumed that parturition and initial care of the off- 
spring take place m a hole in the groiuid; and there are slight indications that two is 
probably a common litter size. The only two juveniles with relevant data in the 
British Museum suggest the dry-season as the favoured breedmg period in West Africa, 
one having been taken on the 5th January, the other m February. 

The white-tailed mongoose has the reputation not only of beuig shy but also of 
being imfriendly. However, if captured very yoimg it can become a very charming 
pet. Dalton (1961b), who gives what appears to be the only full accoimt of this animal 
in the home, describes it as of sweet and gentle disposition, never biting or scratching 
cither in temper or in play. It romped round joyfully, with dogs or played with human 
beings, even with strangers. This animal was persistently nocturnal, retiring to rest 
in the morning, sleeping soimdly throughout the day despite disturbance or unfavour- 
able conditions, emerging in the evening to feed. It seemed equally satisfied with any 
kind of table food; it killed and ate rats, and consumed enormous quantities of moths 
and other insects that flocked to lamps at night. When allowed free range it would 
also claw out beetles; and at intervals it would stand up on its hindlegs, peering round 
for danger and chattering excitedly. Like many other nocturnal animals this one was 
not continuously active throughout the night but returned at intervals from its foraging 
to take a short rest. One of the most interesting observations made by Dalton is that 
the white-tailed mongoose, or this specimen at any rate, employs the common mungo- 
tine trick of breaking eggs by hurtling them backwards through the hindlegs — in 
this case against the leg of a table. The eggs were not proffered as food but discovered 
and stolen. This animal came to an early end through being found in someone's fowl- 
run, whither it had followed some wild specimens of its kmd; but a white-tailed 
mongoose is known to have lived in captivity for nearly 12! years {Int. Zoo Yr. Bk., 2). 

In spite of the reputation oi Ichncwnia for shyness it would seem that it is not much 



306 THF. CARN'IVORF.S OP WtST AFRICA 

.itV.ud lit luini.ii! proximity: it lias alrcidy bcL-n niciitioiu'd above that tins niongoosi.' 
used to be i-oninionly caught by dogs in Sokoto station (Nigeria): and Pitman ([954) 
records that m Uganda it frequents towns in large numbers. This observer further 
(1931) gives a remarkable account of having watched a white-t.iiled mongoose, every 
hair on its head, body and tail erect, perform a dance in the bright moonlight in front 
of a fowl-run with the purpose of arousing the curiosity- of the birds. Any which, 
in order to get a better view of this strange sight, thrust Us head through the wire 
netting had it at once bitten oft". This performance went on fir several nights until 
the raider was trapped. Pitman's own curiosity had originally been roused by the 
discovery of mangled corpses inside the rim without any sign ot an intruder. 

Roosevelt (1910) tells another peculiar story of this species, namely that it com- 
monly climbs trees, steals honey from the hives set therein, and kills hyraxes "following 
them everywhere among the trcctops". It would seem improbable that a mongoose 
with ill-adapted feet and claws, let alone one ot the size ot Ichncuiuia, would be able to 
climb tree stems with any facility or be much at ease chasing hyraxes along branches: 
and It would thus appear likely that this account of its habits nuist be a case of nnstaken 
identity. Roosevelt also recounts another astonishing story ot a white-tailed mongoose 
starting to eat a small puff-adder at the middle of its back before even attempting to 
kill it. The snake consequently turned and buried its tangs twice in the mongoose, 
which nevertheless took no notice, ate it all up, and showed no signs of having received 
any damage. That Ichncuiuia does, in fict, eat snakes is confirmed by an observation 
111 Shortridge (1934) that a small cobra, rodents, large beetles and termites have all 
been taken from stomachs. 

Finally, it mav be added that in self-defence this mongoose can eject a very foul- 
smelling liquid which, fide Pitman, it is impossible to wash off, the odour persisting 
for a very long time; and T. S. Jones (private communication) was told by hunters 
in Sierra Leone that the recent passage of one of these mongooses can always be detected 
bv the powerful smell. 

Taxonomy. It has already been stated diat only one species is today recognised 
as existing throughout Africa. It remains to examine the matter of races. Two are 
named for West Africa; locfui'o Temminck and uif^ricaudd Pucheran; and G. M. Allen 
(1939) lists five others from elsewhere. These have all been described on colour or the 
size and character of the teeth. Whereas there is some slight evidence that drier zones 
arc iidiabited by averagely somewhat paler animals, and I'icc vena, the degrees of 
ditference are slight and the practicability of drawing defined lines between forms non- 
existent. It has yet to be convincingly shown that described differences of size, corporal 
and cranial, or of dental cuspidation are constant to an area and do not fall within the 
range of variation normal to the nominate form. 

The present wiiter holds the view that no sufficiently convincuig case has been 
made out for the validity of any of the proposed subspecies. There is a wide degree 
of variation of colour in any area (see, for example, Moiiard, 1940): and factors such 
as moult and age also enter into the question. The data at present available arc quite 
inadequate to constitute the basis of any practically useful subspecific naming. As 
regards the West African races 111 particular, Teinminck's description of /(H;i//h' fails 



fiALF.RELLA 



307 



to draw any clear distinction between it and alhiaiuda itself; while Pucheran's lu^iricauda 
rests on the black tail which is known to be a commonly occurring variant. 



Tabic 18: Nunicric.ll d.iU for /r/i 



Vcgct.ition 

Number in mean 
Condylobasal length 
Basilar length 
Palatilar length 
Zygomatic breadth 
Upper cheekteeth breadth 
Nasals, length 
Interorbital breadth 
Postorbital constriction 
Braincase breadth 
Toothrow (f — m-) 
p* length 
i»l breadth 
III- breadth 
i/il length 
iii-y length 

Head & body 
Tail 

Hindfoot 
Ear 

RATIOS (per cent) 
Tail/head & body 
Zygom. br./condylob. 1. 
Braincase/condylob. 1. 
Braincase/zygom. br. 
Palatilar l./condylob. 1. 
Interorb./postorb. 
p*lc—iifi 



Iclnicttmid ah 


buattda 


albicauda, 


albicaiiiiiif 


means 


range 


Guinea — 




Sudan 




10 




107-2 


98-6-II5-7 


99-7 


9I-2-I07-3 


62-3 


j6-5-66-S 


54-6 


50-3-58-4 


33-9 


3i-3-36-r> 


26-5 


2I-9-29-3 


21-2 


I8-6-23-7 


22-1 


20-3-24-0 


3S-3 


3 1 •7-37-3 


40-2 


3 8 •4-44-0 


8-1 


7-7-8-9 


9-0 


8-I-I0-0 


7-6 


6-5-8-2 


8-4 


7-7-9-0 


7-4 


6-5-X-2 


482 


423-545 


361 


330-395 


107 


99-121 


40 


33-50 


75 




SI 




33 




6S 




58 




96 




20T 





Genus GALERELLA Gray, 1865 
Slender Mongooses 

Galerella Gray, 1865, Proc. zool. Soc. Lond. for 1864: 564. Type species Hcrpestes ochracciis Gray from 
Ethiopia. This name was obviously intended as a diminutive of the Greek gale weasel, a basic com- 
ponent of a variety of combinations applied to a number of vivcrrids and mustelids, and in this present 
case especially apt for the small, slender, weasel-like mongoose for which Gray devised this particular 
variant. 

Myonax Thomas, 1928, Ann. Mag. nat. Hist. (10) 2: 408. Type species Herpesles gracilis Riippell from 
Eritrea. The name is a combination of the Greek words mys, iiiyos mouse, and ana.x king, implying. 



lOS IHE CAKNIVOKLS Or Wl-Sr AFIUCA 

111 Tlioiii.ii\ oun cxpLm.ition, "the King or I'yiaiit ot the R.its and Mice", leterriiig tci the reputed 
ahilitv of thi<^ mongoose as a domestic pet to rid a house of rats and other vermin. 

Taxonomy. Soiiictlimg has .ilrcady been said (pages 24') and 2(^(>) ot the question ot 
wliether Giihrclla is synonynious with Ihrpcstcs or merits independent status: but the 
matter must now be entered into rather more fully, before proceeding further, in order 
to make the position adopted clear and to etiect detinition of the genus as nnderstood 
in this present work. 

Giihrclld was erected by Gray tor an Ethiopian species, ochiiiu\i, which he had named 
some years earlier and at that time assigned to Hcrpcstcs. The proposed genus, however, 
never came into use during the next 60 years luitil it was revived by J. A. Allen (1924); 
and no new form has ever been directly ascribed to it as a genus except when 111 ii;35 
Schwarz, following Allen's lead, used it in connexion with two subspecies of Siim^uincci. 
There had, indeed, come mto existence only one new species that might have been 
assigned to it, Heugliu's riificauJa, 1S77, between its creation by Gray and its virtual 
killing by Thomas in his 1S82 review of the mongooses. Li this youthful paper Thomas 
made a sweeping condemnation of Gray's proposals in which, he wrote, "such a large 
number of untenable genera are formed, and so many bad species are made . . ."'. 
He retransferred all of Gray's 1865 genera back to Hcrpcstcs and set a nomenclatural 
pattern that lasted through several decades. In his more mature years, however, he came 
(1929), as a result of J. A. Allen's 1924 paper, to see that, so far as the animals here at 
present under discussion were concerned, he had made a mistake and that Allen was 
"imquestionably right in maintaining that the small African Mongooses should be 
considered as genericallv distinct from die larger Egyptian species, the t)'pe of the 
genus Hcrpcstcs . . .' . 

There was, indeed, no tmge ot doubt or ambigiuty about Alien's views through the 
use ot minced words: "hi general features Hcrpcstcs and GalcrcUa are about as diverse 
as two genera can well be and be referable to the same subfamily . . . ". He supported 
this exceptionally luicomproinising opinion by contrasting the two genera in respect 
of overall size and form, of the tail, the pelage, limbs, digits, claws and soles, as well 
of cranial characters and dental formula. To sweep this aside by such a comment as 
". . . it should be borne in mind that although the small African mongooses appear 
very distinct from H. iciiiiciiiitoii (the type of die genus) there are many more small 
species of Hcrpcstcs in Tropical Asia" (Ellcrman, Morrison-Scott & Haymaii, 1953: 
119 fn.) seems unfiirly to diminish Allen's argument, in which size was only one, 
the least important, character adduced. 

Apart from the vast disparity of size and the altogether more slender structure of 
Giilcrclla the points cited by Allen as divorcnig this genus from Hcrpcstcs are that it 
has a pelage of a very diti'crent character; a narrow distichous tail; short limbs with 
relatively small and weak feet in which the pollex and hallux are reduced: and soles 
furred for mucli of the proximal half Li the skull, GalcrcUa ditlers in its uniquely 
inflated anterior chamber to the bulla; ui the shape and siting of the postorbital con- 
striction; and in the constant possession of only 3 lower premolars on each side, whereas 
Hcrpcstcs most commonly has 4. The present author finds that tliese difterential charac- 



GALERELLA 309 

ters are valid and in sum lift the matter beyond the compromise course now often 
adopted {e.g. Simpson, 1945; Ellerman, Morrison-Scott &: Hayman, 1953) of making 
Gahrella a subgenus of Hcrpcstcs. 

Li spite of Thomas's eventual broad agreement with Allen regarding the generic 
independence from Hcrpcstcs of the small African mongooses, however, he at once 
introduced into the controversy a second, cross argument in that he held that GalcrcUa 
was applicable only to its type species, ochracca. This last, he maintained, differed in 
its skull and in the nature of its feet from the remaining numerous forms, for which 
he erected a new genus, Myonax. The skull in ochracca was said to be relatively shortened, 
and Its bullae enlarged. The first is to some extent true; the second not so obvious. 
But Thomas placed most reliance on the feet, wliich in Galcrclla, m his restricted 
sense, he foimd "quite peculiar, very slender, narrowed throughout, with the naked 
area extended backwards in a prominent line to the heel, the corresponding region 
in the ordmary mongooses bemg hairy . . .". hi addition, the hallux is very much 
reduced or entirely absent. Schwarz (1935) disagreed that anything more than a specific 
difference was involved; and Ellerman, Morrison-Scott & Hayman (1953) adopted 
the same attitude. Nevertheless, Thomas did, in fact, have something more of a case 
than these contrary views might indicate. The skull differences he cited exist, though 
sometimes in no very marked form; and close examination shows that without doubt 
the hallux is either entirely lacking in ochracca or at least far more reduced than in the 
other species ; and the nakedness of the sole does extend, in most cases, further towards 
the heel. He could also have drawn attention to a difference in pelage character. But 
whether these points add up to a generic or even subgencric distinction is open to 
doubt; and in this present work Myonax has been rejected as requiring more con- 
clusive data than at present available. 

Turning now from generic to specific level, one of the basic problems as far so 
West Africa is concerned is to determine whether the red-tail-tufted, pale-pelagcd 
sanguinca and the black-tail-tufted gracilis are conspecific or not. This is but part of the 
far wider problem embracing the whole vast range of described forms from other areas 
of the continent, the whole matter boiling douii to whether all these small-sized 
speckled mongooses belong to a single species complex spread over the whole of 
tropical and southern Africa or whether vahd rpecific differences exist within a super- 
ficially similar group. 

hi so far as this present work is concerned two apparently distinct groups exist 
difi"erentiated most obviously by the colour of the terminal pencil of the tail. These are 
either identical with or related to saiigiiinca andgracilis, two species described by Riippell, 
the former with a red tail from Kordofan, the latter black-tuftcd from Eritrea. 
Wroughton (1907: 115-116) examined the question of specific distinction bervveen 
these two and came to the conclusion that there was none; and since that time it has 
been customary to sink the latter in the former, which had page priorit)' in Riippell's 
work. Thomas (1917) was inclined to disagree; and in 1929 quite dcfmitely did so 
when he deliberately named (^r.jc//!.!;, not saiigiiinca, as the type species of his new genus 
Myonax. Wroughton had based much of his conclusion on a specimen, B.M. No. 
6.10.2.9 from Erkowit near Suakin, which he regarded as intermediate between 
v 



310 THi; CARNIVOHI'.S OF WEST AHilCA 

i^raiilis and Siim^iiiiifa, the tail tiitt being "lialt choct)latc-browii and half black". The 
colour contrast is by no means so crystal clear as this description nnght lead one to 
visualise, and the tuft, in flict. has little or nothing of the intense jet-black that charac- 
terises the '"(jriifife" forms. Thomas thought it to be nothing more than "merely one 
of the ordinary i;ni('i7i.'.' t)'pe with a more or less bleached tail-tip". Wroughton formed 
his opinion on the basis, to all intents and purposes, of a single presumed aberrant, 
intermediate specimen. Thomas had never seen "examples of the true Kordofan 
SiiiUJiiiiuiis", that is to say topotypical material; or indeed, from the known history of the 
collections, any red-tutted specimen other than the one he then had before hmi and 
was in the process of naming phocnkiinis. The opinions expressed by either author 
regarding the validit)' or otherwise of the two species can therefore scarcely be taken 
as cariying any greater weight than tentative guesses. 

The truth is that the material available is still inadequate to form the basis ot a really 
sound conclusion. There are no topotypical examples ot (;ni(i7/V trom Massawa; but 
there are three skins, with only two skulls, from near Suakm, some 400 km to the north 
and, fide Keay ct al. (1959), in precisely the same vegetational belt. Coastal Subdesert. 
It is of interest to note that these differ quite appreciably from each other in their 
general pelage coloration, both above and below; and this is not merely a matter of 
tone but of rechiess or greyness, demonstrating the degree of variation that may occur 
even within a restricted locality'. None more than approximately resembles Riippell's 
rv'pc illustration. One is that already reterred to as possessing a parti-coloured tail; 
the other two are also of interest in diis connexion. Although by the standards ot 
early 19th century mammalogical description all three tail tips could well be classed as 
"black" none is, in flict, of quite the same pure jet-black that characterises so many 
of tills genus. All have in some degree something of an exceedingly dark chestnut-red, 
in one case so intense as to pass for black except in critical comparison widi the real 
thing. The skulls have almost precisely the same mean measurements as saw^iiiucd 
from Darfur and are otherwise indistinguishable except for a rather narrower post- 
dental palate. 

The present study material is far too meagre to support any tirm decision regarding 
the taxonomic significance of red tails and black tails; but since there is no morpho- 
logical distinction between specimens possessing the one or the other Wroughton's 
view that qnuilis is synonymous with Siiii'i^iiiiicd is adopted in this work. To what 
degree the vegetational background plays a part is not clear. Red tail tips are unknown 
from moister forest habitats and in West Africa occur from the Subdesert to the 
Doka zone. On the other hand, in West Africa the black-tipped forms are foiuid in the 
forest belt except for Ciuui which came from Cape Verde (Guinea woodland). Extra- 
limitally, black tailed Gahrclhi are commonly inhabitants of drier vegetational types. 
The position of Gii/i/rl/if, therefore, adopted in this present work is: 

Gdlcrclla santiuinca saiiquima (Riippell), iS}S- Extralimital. 

,, ,, );it7(!/;/(rii (Martin), 1836. Forest. 

,, ,, caiui (Wroughton), 1907. Guinea woodland. 

„ ,, pliociiicunis (Thomas), 1912. Doka woodland. 

,, ,, sahai'dc (Thomas), 1925. Subdesert. 



GALERELLA 311 

There is no present evidence that G. s. iiiiistcla Schwarz, 1935, from lower Cameroun 
occurs within the Hniits chosen for this work; but it might eventually turn up from the 
Cross River area. 

Distribution and general. The genus is spread over nearly the whole of Africa 
south of the Sahara except for the extreme south-west. Li the territory dealt with in 
this present work specimens are known from the extreme west at Cape Verde, Sierra 
Leone, Liberia, Ivory Coast, Ghana, western and northern Nigeria, and Air. There 
does not appear to be any preserved specimen or published record of the species from 
eastern Nigeria and that portion of Cameroun that lies north of the Sanaga River; 
but, in so fir as the latter is concerned, Gerald Durrcll (personal commmiication) 
obtained specimens at Bafut and Wum, roughly 20 and 50 km north of Bamenda. 
Over much of its range the genus can be said to be common; and this is true of the 
West African high forest form, mclanura, but not, apparently, of the other races within 
the region. These are all distinctive Httle creatures that could scarcely be confused with 
any other mongoose or small mammal occurring in West Africa, except possibly, 
at a distance, a ground-squirrel. 

Description. The GalcreUn mongooses (Plate 9) are all small, very slenderly built 
animals, weighing from 0-5 to 0-75 kg, with fmely speckled pelage, sharp faces and 
longish tails. Head & body length is of the order of 300 mm, more or less. Field 
measurements are notoriously undependable, and in one relevant case, the type of 
Siilumic, those for the head & body and the tail appear from the dried skin to have been 
reversed by a usually extremely reliable collector (Buchanan). Considerable dis- 
crepancy arises in long-haired genera from whether the tail measurement has been 
made to the last joint or to the tips of the hairs; but so far as can be gathered from 
label data the tail in GiilcrclLi is for the most part somewhat shorter than the head & 
body, reaching some 85 to 95 per cent; though in a few cases it seems to be genuinely 
longer. 

The pelage consists of long, very fme, dense or fairly dense underfur and abmidant 
annulated, flat-sectioned bristle-hairs that do not so markedly exceed this in length 
as they do in many other species. The whole texture of the coat is fme, silky and sleek; 
and in this it may be said to differ from all other West Africa mongooses. The luiderfur 
is of variable length in all specimens, being composed of longer and shorter hairs 
ranging from about 6 to 12 mm. It is curiously variable in colour from specimen to 
specimen of the same race. The annulation of the bristle-hairs at first seems confusingly 
different in the different forms; but on analysis a fairly consistent generic pattern 
emerges. There are specific differences of ring-width and colour, and in a few cases a 
complication is introduced by the division of the two pale zones by a subsidiary darkish 
band; nevertheless, fundamentally the bristle-hair pattern is made up in the following 
way. There is a pale basal region whose range of mean lengths is from 8 to ii| mm; 
distally of this lies a narrow blackish ring mostly 2 to 5 mm wide; and this is followed 
by a pale (white, yellow, orange) zone of ver)' variable width ranging from about 
I J to 9J mm; fmally there is a browai or black tip of at least 2 mm but when imbroken 
or unworn reaching 5 mm, and very fmely drawn out. Very occasionally this tip 
engulfs the subtenninal pale zone, joining the next dark ring to form a long, wholly 



312 THE CARNIVOKES OF WEST AIRICA 

blackish distal halt to the hair. The total bristlc-hair length averages from 17 to 24 mm. 
The overall effect is of a fuiely speckled coat. In some specimens the different colour 
bands lie randomly; in others they fall to some extent m groups, giving rise to a some- 
what irregular, rather narrow cross-banding, in no way comparable to the bold pattern 
o( Miiii'^os imiii^io. This is by no means an invariable feature ot every specimen of the 
same form; where it occurs, the speckling takes on a rather coarser appearance than the 
fme pimctulation produced by colour bands that are randomly disposed. 

Li many forms the imderside is imicolorous and fiirly sharply divided from the 
flanks; but this is not always so, especially as regards the common West African race 
tiuliiinir,!. Near each axilla there is a whorl of hair, and from here over the anterior 
part of the chest the hair is directed forwards; but on the throat, though the lie ot the 
hair is abnormal in not being evenly directed rearwards, exact detail of arrangement 
appears to vary from form to form or specimen to specimen. Generally, here, the fur 
is directed outwards from a longitudinal medial parting, but further detail in this 
region is rather confused, there being sometimes a distinct gular whorl, and sometimes 
part of the hair is directed forwards and meets the opposing lur of the chin in a transverse 
ridge. 

The head in ddcnlLi is small, the tace pointed, and the upper lip beneath the naked 
rhinarium very narrow. The rounded ears are very broad but low and set well on the 
sides of the head. The legs are short and speckled 111 the manner of the back; the narrow 
feet either similar or pale and more or less unicolorous. Both pollex and hallux are 
ver\' much reduced, the remaining digits slightly webbed basally, the claws short and 
curved. The sole ot the hindtoot is hair)' in the posterior portion near the heel. The 
tail though clad with long bristle-hairs is not bushy but the hairs are arranged somewhat 
distichously, though this is less evident in some specimens, or forms, than others, 
hi general it can be characterised as narrow, tapering from a moderate base to a tuie 
tip, the terminal two or three inches being unicolorous, either jet-black or rufous. 
The main portion of the tail is ot the same colour and speckling as the back on top but 
sometimes more or less imicolorous, paler, below or with a rutous medial longitudinal 
stripe. There is the usual herpestine subcircular pouch surrounding the anus and external 
orifices of the scent glands; but no one appears to have studied these latter in detail. 
The females carry two abdominal pairs of mammae. 

Skull (fig. 40). Galcrcllii skulls are the smallest of the mongooses occurring m West 
Africa the condylobasal length being always well under 70 mm. The long ovoid 
braincase terminates posteriorly in a broad flange-like supraoccipital crest; but in the 
vast majority of cases there is nothing more than a rudimentary or very slight sagittal 
crest. In only one of some fifry skulls examined, a mediumly-aged male, could this 
crest be characterised as well-developed into an erect flange of appreciable depth. 
The small extreme posterior section adjacent to the supraoccipital crest is nearly 
always present except in very young specimens. The relationship ot the measurements 
of the interorbital breadth and the postorbital constriction is variable; the former is 
roughly constant at about 11 mm, more or less; but the latter ranges from less than 
10 nnn to nearly 14 mm even in the same group of approximately comparable ages, 
and the ratio of the former to the latter trom 77 to 120 per cent. 13ut whatever its 



GALERELLA 



313 




Fig. 40. Calerella saiiguitica mclamira: skull, B.M. No. 35-I-30.53. 3, x I 



314 THE CARNIVORES OF WEST ATRICA 

relative size one of tlie cliiet features ot the Giilcnlln skull distinguishing it from 
Hcrpcstcs and Xciio<^iilc is the taet that this postorbital constriction lies immediately 
behind or even a little forward of the back margin of the circumorbital ring, not 
several millimetres posterior to it as in the others (tigs. 40, 36 and 44). In nearly every 
adult skull the postorbital processes unite, or almost unite, with the jugal processes 
to form ossified complete circumorbital rings. The frontal region is broad and rounded; 
the rostrum short and blunt. Fusion and complete disappearance of the nasal sutures 
takes place early. The zygomata are fairly strong, the zygomatic breadth about average, 
that is to say rouglily half the condylobasal length. The postdcntal palate is ratlicr 
narrower than it is long, sometimes more so than others, and occasionally markedly 




Fig. 41. GalcrcUa saii^^ititwa iiwhiiinyti: bulla, -' 2 

vvaisted. The bullae in this genus arc unique amongst West African mongooses in 
that the anterior chamber is much more inflated than m the other genera and begins to 
be comparable in size to the posterior one, though less globular in shape (fig. 41). 

The premolars are very constantly y Only three skulls of many examined had the 
upper cheekteeth with the first premolar lacking from one side alone. The anterior 
upper premolar is always an extremely small tooth; all the remaining cheekteeth arc, 
when unworn, very sharply cusped, the dentition having the appearance of being 
largely insectivorous. Nevertheless, the carnassials are relatively amongst the largest 
in the subfamily, the upper ones occupying 27 per cent of the length of the toothrow, 
and the lower one having a large sharp outer blade, the posterior outer cusp being 
the largest and the inner posterior the smallest of the three on the anterior portion of the 
tooth, in^ is a fairly large tooth, but iii- is much reduced and of less bulk than the 
lingual portion of »|1. The tightly packed incisors are in an almost straight transverse 
row; the canines are of moderate size. 

Habits. Although mongooses of this genus are widespread and pretty common 
over a good deal of die continent south of the Sahara not much has been written of the 
details of their way of life. There may be some variation of habit between different 
forms; but because of the confusion which has long existed in respect of siibgeneric 
classification and nomenclature this is a matter which must necessarily for the present 
remain obscure. The only course is to assemble on a generic basis the little that has 



GALERELLA 315 

been recorded, irrespective of the specific or racial names with which the observations 
were originally connected. As usual, there is almost nothing in the way of field notes 
from West Africa itself. Most recent accounts, in fact, all stem from a single source, 
the notes assembled by Shortridge (1934) concerning the genus in southern Africa; 
but Taylor (1970) has included the genus in his observations on locomotion made in 
East Africa. Except for a few odd scraps of information no other accounts of habits 
have been traced. 

The shortage is due in part to the fact that few people appear to have kept Galcrclk 
as a pet; and seemingly no one who has done so has felt the urge to describe in writing 
the animal's nature and habits as revealed imdcr these conditions. Opinions, indeed, 
diifcr regarding the possibility or desirability of keeping the slender mongoose as a 
house companion. Some have held that like others of its kind it is readily tamed (Astley 
Maberly) and makes a splendid pet keeping the home free of vermin and cockroaches; 
others have found it to take much less readily to captivity than most and to be largely 
untameable. Cansdale (1946), indeed, writing of the West African form mclanura, 
could not understand why anyone should ever wish to keep it about the house since 
it was foul-smelling and in point of fact so objectionable that it was one of the few 
animals of the forest that nobody would eat. Illustrating the saying that one man's 
meat is another man's poison there is a note in Hinton & Diuin (1967: 10), under the 
name "Dwarf Mongoose", that according to Dr. S. Toye the hunters of Northern 
Nigeria eat it despite its unpleasant odour. 

GalerclLi is almost entirely dmrnal, hunting singly or in pairs at almost any hour of 
the day, though possibly more in the early morning than during the heat of the after- 
noon. It has been said to come out sometimes on moonlight nights. As it runs, not 
trots, (Taylor, 1970) along in search of prey it holds its tail out behind with the terminal 
half curved upwards; and since its size and sometimes its coloration are not dissimilar 
it may, especially when scurrying across a road, be mistaken for a ground squirrel, 
the nature and carriage of the tail, however, helping to distinguish it. It is widely 
agreed that this mongoose differs sharply from others in the relative facility with 
which it can climb trees, and even descend head-foremost, though it lacks the expert 
ability of the cats and squirrels. Its short, curved claws and light weight help it in this 
activity. Taylor (1970), however, while fmding that this species was an able climber in 
as much as it would rush at wire or other rough surfaces and clamber up, considered 
that it was not really a controlled cHmbcr m that it generally fell down. There seems 
little doubt that the slender mongoose climbs not only to escape pursuit but also to 
seek part of its food in the shape of nesting birds and arboreal mammals. Taylor also 
observed that this mongoose frequently rose to its hindfeet; and from this erect standing 
position could, if it wished to reach something above it, jump vertically as much as 
50 cm. 

Like other mongooses it consumes a large number of msects. Ansell (1965) records 
taking a specimen near a grass fire where it was busily eating fleemg grasshoppers; 
but when the stomach contents were examined they were found to contain, besides 
insects, the remains of a squirrel. Apart from this one example the question of food 
materials has not been very carefully investigated; mongooses of tliis genus are credited 



3l6 THE CARNIVORES OF WEST AFRICA 

with the usual range of prey — insects and their larvae, birds, eggs, rodents, lizards and 
snakes. As regards the last, Astley-Maberly (1959) wrote ot a southern African represen- 
tative of the genus that it will kill even large cobras, and Gerald Diirrcll (personal 
coinimmication) found that they would readily eat snakes so long as these were not 
too large. Slender mongooses are also said to be poultry thieves if they get the chance; 
and Cansdale (1946) relates, in relation to the forest form iitcldiiiiiii, an interesting Ghana 
version of the folk legend usually associated in other parts of Africa with Atihix. 
This is that GiilcrcUa lies on the gromid displaying its open anal scent gland pouch, 
thus arousing the curiosity of some nearby fowl, which, supposing this pinkish structure 
to be something edible, pecks at it, chokes and gets caught and killed. As with other 
carnivores, wild fruits are said to constitute part ot the dietary; and Gdhrclla has been 
accused of digging up groiuidnuts in fn'ms. 

There is little uiformation relating to breeding habits. The shelter, for raising the 
yoimg as much as for regular nightly use, is said to be made in hollow trees, in holes 
m logs or in the groimd amongst tree roots, or in crevices between boulders in rocky 
coiuitn'. The number born in a litter has been recorded as 2 or 3, occasionally 4. 
Regarding development of tlie young or parental care there is nothing on record. 
One specimen of Gii/crc//.; is known to have lived in a zoo to the age of nearly 6 years 
{Int. Zoo Yr. Bh., 2). 

GALERELLA SANGUINEA (Riippcll) Slender Mongoose 

Hcrpcslcs sangiiiiu'its Riippcll, 1836, Nciic il'irbclihiac :ii il-r Fmiim von Abyssiiiicii gt'licr{^, etc., pt 7, Sauge- 
thiere: 27, pi. 8 tig. i, pi. 10 fig. 3. Type locality Kordotaii, Sudan. The Latin adjective used as the 
specific name means bloody and was given in respect of tlie reddish tail tuft. 

Hcrpcstcs gracilis Riippell, 1836, Nckc IVirbclthicrc zii dcr Fitiaia von Ahyssinicti gcliorii;, clc, pt. 7, Sauge- 
thiere: 29, pi. 8 fig. 2, pi. 10 fig. 2. Type locality the valleys west ot Massawa, Eritrea. The specific 
name is the Latin word for slender, given with rcterencc to the build of the body. 

Cyniclis nielanmus Martin, 1S36, Proc. zoo]. Soc. Lond.: 56. Type locality Sierra Leone. Type in the iiritish 
Museum, No. 55.12.24.229, sex unknown; skin poor and with halt ot the tail missing; skull merely 
the anterior fragment of each jaw. This name is composed ot' the Greek words molas, mclanos black, 
.and oura tail, referring to the colour of the tip. Valid as a r.ice. 

Mungos mclanurm canus Wroughton, 1907, Aim. Mag. nat. Hist. (7) 20: 114-115. Type locality Cape 
Verde, Senegal. Type in the 13ritish Museum, No. 72. 12. 12. 5, ?; skin in poorish condition, skull 
with the right side and back of the hraincase missing. The Latin word conns means grey or hoary, and 
refers to the overall pelage colour. Valid as a race. 

Mnngos phocnicnins Thomas, 1912, Ann. Mag. nat. Hist. (8) 10: 2S0-281. Type locality Panyam, Bauchi, 
Northern Nigeria. Type in the Biitish Museum, No. 12.7.9.2, .^; skin and skull both in good con- 
dition except for the loss of both inner upper incisors. The specitic name was coined from the Creek 
words phoinix purplish-red, and onra tail, and refers to the tail tip. Valid as a race. 

Hcrpostcs phocnicnrns saharae Thomas, 1925, Aim. Mag. nat. Hist. (9) 16: 1S9. Type locality Aoudcras, 
Asben, Niger, 835 metres. Type in the British Museum, No. 25. 5. 12. 12, o; skin in good condition 
but a little of the end of the tail missing, skull good except for one imier upper incisor missing. 

Distribution, hi view of the continued argument regarding the validity or other- 
wise of riputed species it is impossible to give any clear indication of die tull distribution 
of Sdiioiiiiiiit. As the species is understood in this present work, and explained in the 



GALERELLA 3I7 

taxonomic section on page 309, it is known in West Africa from Cape Verde to Nigeria, 
and is said to occur also in at least the Bamcnda area of upper Camcroun; but, as regards 
this last, in the absence of specimens it is impossible to say which form is concerned. 
Racial identification is impossible also in respect of sightings in the Borgu Game Reserve 
of western Nigeria reported by G. S. Child (private communication). Extralimitally, 
siiuguinca must, by definition, range to the Red Sea; and Schwarz's mustchi from lower 
Camerouii is certainly part of the same complex; but how far south in the continent 
the species reaches is, at the moment, entirely a matter of opinion. 

Description. The species conforms in all essentials to the broad general description 
already given for the genus; but the races into which it is subdivided are, within this 
framework, so diverse that any more detailed description must be reserved for each 
of these. 

Habits. The habits so far as known have been dealt with above, and there is nothing 
that can be specifically added for sniii^uinca. 

Skull (fig. 40). There arc no differences from the general generic characters given 
above. 

Taxonomy. In the broad review of the taxonomy of the genus the question of 
species was necessarily gone into and there is not much on this account that can be 
usefully added here. Whether there are different species in South and West Africa or 
whether they are all merely local forms of one species it is at present quite impossible 
to determine with any approach to finality; and it would seem that not only must 
more material be forthcoming but that, also, cytological and serological studies must 
be made before any fully satisfactory solution to the problem can be achieved. For 
South Africa, for example, Roberts in 195 1 recognised 7 independent species, none of 
which was 5cH;(j»/;/e<r, wliile in 1953 Ellerman, Morrison-Scott & Hayman cast the 
same 7 forms into 2 species, classing 4 of them as races o( Siingiiinca. 

Really, much the same confusion exists subspecifically. Something like 45 names 
are recognised as possibly having some validity in various parts of the continent — or, 
at least, in the absence of a general revision have not so far been seriously questioned. 
The fict is that with a speckled pelage and the wide diversity' of ecological and climato- 
logical conditions existing in Africa an almost infuiite variety of colour forms can 
exist. How many of these it is possible and useful to distinguish with a definite name is 
another matter. Li West Africa, as elsewhere, two distinct colour forms stand out, 
those with a black tail tip and those in which it is rufous; but whether tliis has specific 
significance or is merely local or phasal nobody knows. Ruiming alongside these two 
main types are pelage colour variations. Several of these have been accorded subspecific 
names and are separately described below; but their precise taxonomic value is an 
open question. 

The West African races listed earher can be differentiated by the following key. 

KEY TO THE WEST AFRICAN RACES OF G. SAXGIIXEA 
(previous key page 245) 

I. Tail tip red . . . ■ . . . . . . . . .2 



3l8 TllK CAKNIVORIiS OF WIST AIKKA 

Tail tip black ........... 3 

2. Underside pale yellowish-brown ...... phoenictira 

Underside white .......... saharue 

3. Dorsal pelage basically deep rich red, speckled with black . . melaniira 
Dorsal pelage basically butiish ........ cana 

Galcrella sanguinca phoenicura (Thomas) Western Red-tailed Mongoose 

The two West African red-tail-tippcd Gdlcrcllii arc, as exemplified by the types, both 
distinctly darker and redder than the niajorir}' of Sudanese scvi^iiiiiicd, in respect of the 
tail tip as well as of the dorsal pelage. The nominate, Sudanese, race is basically buff 
above; the tail tip pale orange-ochre rather than red; in phovuiaira (Plate 9) the dorsal 
pelage has a considerable amount of red in it, particularly in the mid-lower back; 
and the tail tip is rich rusty-red. The belly is buffy'. hi this form the bristlc-hairs average 
about 12 mm in length; their basal zone is buff, about 8 mm; there is next a deep 
brown zone of about 5 mm; the subterminal band is yellow-gold, also about 5 mm; 
and the deep-brown tip is about 4 mm. 

Further collecting may well show that there is iiuergrading. The type of pliociiimra 
was collected at Panyam in the Doka woodland at about 1250 metres on the Bauchi 
Plateau, north Nigeria; a second, younger, specimen taken at Kabwir about 50 km 
further east, on the edge of the plateau at about 760 metres and vegetation more of a 
Sudan woodland type, and assigned also, by Thomas, to phoaiiciira, is not so red, though 
it is considerably darker than typical lowland Sudanese examples oi sanQ^u'mca. But it is 
closely matched by specimens from the lebel Marra mountains of Sudan, taken at 
2000-2500 metres, though the tail of the West African animal is of a slightly deeper 
red and the belly rather more yellowish. 

Galcrella sanguinea saliarac (Thomas) Air Red-tailed Mongoose 

This differs very little from the above. The dorsal pelage is a more orangey-red; 
the tail tip deep chestnut through the admixture of a number of black hairs; and the 
underside is pure white. This last is the only clear difference between it and ]i//oc);/c»rii, 
though the few available measurements indicate it as a somewhat smaller animal. The 
annulation of the bristle-hairs in this form appears to be very irregular; but their mean 
length is about 18 m . 

Only two specimens are known, both from Air on the southern edge of the Sahara: 
one from Aouderas, the other from Tchsiderak, both in the Sub-desert zone, at an 
altitude of 750-S50 metres. 

Galcrella sanguinea mclaniira (Martin) Forest Slender Mongoose 

This IS without doubt the commonest race of West Africa. Specimens exist in die 
British Museum from Daru, Njala and lumamed localities in Sierra Leone; Mount 
Barclay in Liberia; Beoumi in the Ivory Coast; Bibianaha, Kumasi, Mampong, 
[uaso, Tarkwa and a number of unspecified places in C'.hana; a total of 20 specimens 




Forest Slender Mongoose {Cakrclla saiii;iiiiica mclaniira): Western Ucd-tailcd Mongoose (GakrcUa saiiiiiiii 
plhH-iiiaini) ; Guine.i Slender Mongoose {Galcrclla saiiguinca cann) 



GALERELLA 319 

as compared with a maxiinuin ot z of any other form. Kuhii (1965) also records this 
race from a large number of places in Liberia. All known collecting localities are in the 
high forest. 

This race (Plate 9) differs from all others in its dark, rich orange-red dorsal colouring. 
Most specimens, but not all, exhibit a taint pattern ot narrow black transverse banding. 
The bristle-hairs average about 19 mm in length. Their amuilation pattern is more 
complex than in the other races by reason of the fact that the normally pale basal and 
subterminal zones are each divided by a dark intrusion so that each instead of being a 
single unit becomes three zones. Thus, the basal zone, totalling on the average 8 mm 
is divided up, ncUow, brown, yellow, 4.^ mm, i-\ mm, 2 mm. Distally of this lies 
the normal dark ring (deep brown), in this case exceptionally narrow, naeasuring about 
2 mm. The succeednig pale zone becomes threefold, orange, deep brown, orange, 
i^ mm, 2^ mm, 2 mm; succeeded finally by the tip which in this case is pale brown, 
about 3 mm. It will therefore be seen that instead of the normal four zones there are in 
inclauura eight; and the deep, warm coloration is brought about by the amount of 
brown, orange and yellow in place of the white, butt or gold of the paler forms. 

The underside is a little variable, sometimes orangey but mostly rather like the 
back but of a paler tone; the hinder part of the belly is more sparsely haired. The 
upper side of the distichous tail is like the back, but the lower side has a medial longitud- 
inal unspeckled stripe of reddish-orange; the tip, for some 50 to 80 mm, is jet-black. 
The legs and feet have much the same coloration as the back. The size is relatively large 
as in phoeiiiaira. The skull is shown in fig. 40. 

Galerella saiiguinea cana (Wroughton) Guinea Slender Mongoose 

The type of this, which came from Cape Verde a hundred years ago, has an excep- 
tionally fmely speckled dorsal pelage with much less red in it than any of the other 
West African forms, and a jet-black tail tip (Plate 9). In broad terms it can be charac- 
terised as buff ticked with black rather than reddish ticked with black. The bristle-hairs 
have an average length of about 17 mm. The pale basal region is in this race divided 
into three by an intrusive pale brown band, there thus being a total of six instead of the 
usual four zones. The effect of fme speckling is due to the fact that all of these are 
narrow; from the base outwards they run as follows, with their mean lengths: white 
3^ mm, pale brown 3! mm, white 2 mm, black 4 mm, white il mm, tip, black, 3 mm. 

So far as can be told from the poor skin the underside is buffish, lightly speckled 
except over the hinder part of the belly; but the chest, throat and chin are greyish due 
to the pure white subterminal and basal bands and grey underfur. The rype locality 
lies within the Guinea zone. 

There are other specimens which, while not matcloing exactly, nevertheless fall 
into this category; that is to say with a black tail tip and predominantly buffish rather 
than predominantly reddish dorsal pelage. It would be possible, picking on slight 
colour differences both above and below, to give these separate distinguishing names; 
but it will almost certainly h: found diat they all intergrade. In this connexion it may 
be repeated here that there arc two skins, both collected at Erkowit near Suakin on 



320 THE CARNIVOKES OF WEST AFRICA 

the Rod Sea coast of Sudan, the one cold grey, the other warm buff; and these sccni to 
demonstrate that there may be marked colour variation within a single locality let 
alone differences that may be expected within a wide-ranging cline. While all of the 
specimens mentioned in this section may be clearly recognised as pertaining to the 
'\'M(i7i.'.-" category, not one of them corresponds more than approximately to the 
colour plate accompanying RiippeH's type description of that form. It can be argued 
that if all these animals do constitute part of a transcontinental cline then cana is synony- 
mous with i;ri!n7/5; but it is retained here for the time being as a convenient distinguish- 
ing label in West Africa. 

The other two West African specimens in the British Museum that can be assigned 
to this race are No. 50.1571 from Lawra (Ghana, Doka woodland); and No. 7.7.8.72 
from the River Fafa (Central African Repubhc, Doka woodland, c. 6°3oN i8°2oE). 
hi the former the speckling is coarser than in the type and the dorsal colour is not 

Tabic 19; Numerical data for Galcrclla saiigiiiiica 

Vegetation 

Nuinher m mean 
Coiidylobasal length 
Basilar length 
Palanlar length 
Zygomatic breadth 
Upper cheekteeth breadth 
Nasals, length 
Interorbital breadth 
Postorbital constriction 
Braincase breadth 
Tootiirow (f — m-) 
p* length 
m' breadth 
HI- breadth 
mi length 
ni2 length 

Head & body 
Tail 

Hindfoot 
Ear 

RATIOS (per cent) 

Tail/head & body 

Zygom. br./condylob. 1. 

Braincasc/condylob. I. 

Braincase/zygom. br. 

Palatilar l./condylob. I. 

Interorb./postorb. 

p^lc—iifi 27-0 27-5 28-1 26-8 



ihocniciirii 


saharac 


inclainira 


cana 


Doka— 


Subdesert 


Forest 


Guinea- 


Sudan 






Doka 


I 


2 


8 


I 


(56 -8 


60-4 


64-9 


— 


61-3 


55-6 


60-2 


— 


33-5 


28-8 


33-3 


29-4 


32-7 


30-6 


33-3 


28-5 


20-7 


19-2 


21-0 


18-7 


— 


— 


14-0 


13-6 


12-2 


10-8 


ti-8 


lO-I 


9-7 


10-2 


II-4 


II-9 


25-0 


24-2 


2.V2 


— 


24-4 


22*2 


23-1 


21-3 


6-6 


6-1 


6-5 


5-7 


6-0 


5-5 


5'9 


5-1 


3-0 


2-9 


3-4 


2-9 


5-6 


5-4 


5-8 


5-3 


3-1 


2-7 


3-2 


2-4 


330 


325 


324 


— 


289 


-77 


264 


— 


61 


56 


59 


?c.48 


28 


23 


27 


~ 


gs 


85 


81 





49 


51 


SI 


— 


37 


40 


39 


— 


76 


79 


76 


— 


50 


48 


51 


— 


126 


106 


103 


85 



GALERISCUS » 321 

far from the Kabwir specimen of phcaiicura; but it differs from this latter by its black 
tail tip and dull orange belly. The latter is closer to the type dorsally; but the under- 
side is dull orange as far forward as the throat. 

There is no record of Schwarz's cxtralimital imistcU (Efulen, lower Cameroim) 
having been captured within the boundaries taken for this present work. Dorsally this 
form is close to cana\ but below it is appreciably darker, a drab brown; and the tail 
lacks any of the bright orange which marks its lower aspect in cana. 



Genus GALERISCUS Thomas, 1894 
Four-toed Mongooses 

Bdeo^ak Peters, 1850, Spenersche Z., 25th June 1850(111 which were published the Sber. Ges. naturf. Freunde 
Bcrl. for the meeting of i8th June 1850; subsequently made more readily available in the volume of 
reprinted reports, 1839-1859, issued by the Gesellchaft Naturforschender Freunde zu Berlin in 1912); 
and 1852, Mber. K. pretiss. Altad. iViss., 81. Type species Bdi'ogale crasskaiida Peters, as designated by 
Thomas, 1882: 81, Mozambique. In part, of recent authors. This generic name was derived from the 
Greek words bdco to stink, and gale weasel. 

Galerisais Thomas, 1894, Ann. Mag. nat. Hist. (6) 13: 522. Type species Galeriscus jacksoiii Thomas, from 
Kenya. The name is a diminutive of Galcra, Thomas having likened the external appearance of the 
species on which the genus was founded to the grison of Brazil, then known as Galictis (Galera) alla- 
mandi. 

This genus covers two known species of mongoose, jacksoni from East Africa and 
uigripcs from West Africa, the one from montane vegetation at 2500 metres, the 
other from the closed forest block. 

Taxonomy. Galtriscus was erected by Thomas in 1894 to accommodate a new 
small carnivore from Kenya, to which he at the same time gave the specific name 
jacksoni. Later, Pocock (1916a) published a note pointing out that this genus "falls as a 
synonym oi Bdi'cgalc'; but while confirming that in his opinion this was correct he 
did not, in fact, claim this idea to have originated with lumsclf but attributed it to a 
change of view (unpublished) on the part of Thomas himself Thomas, on the other 
hand, would not seem to have concurred; for he subsequently (1928) asserted that the 
notion of synonymy, which extended beyond the mere genus, actually came from 
Matschie and from Pocock who considered Galeriscus jacksoni to be the same as the 
mongoose then currently known as Bcico(ialc »/^'n;!C5. The paper in which Thomas 
(1928) dealt with this subject was, indeed, one of the last he ever wrote, and in it he, 
without reservation, expressed his mature taxonomic opinion that Galeriscus, type 
spccKS jacksoni, was clearly genetically distinct from Bdeogalc, type species crassicanda. 
This view seems to have gone unquestioned up to at least 1941, for G. M. Allen (1939) 
listed the two genera separately in his Checklist, and Hill & Carter (1941) also used 
Galeriscus as a fully independent genus. Simpson (1945), however, included Galeriscus 
in B(/((n;.ilc; and Ellerman, Morrison-Scott & Hayman (1953) listed it, incidentally, 
as nothing more than a vahd subgenus. Since then it has become customary to sink, 
or partially sink, Galeriscus — e.g. Grasse (1955), Dekeyser (1955), Walker (1964). 



322 THE CARNIVORES OV WEST AFDICA 

In point ot flict no one appears ever to have publislied any reasoned argument for 
equatnig Giilcrisais with Bdcot^iilf. Thomas, originally led astray, to use his own words, 
bv a poor and incomplete specimen, believed his newly named GtilcrisctisjdcL'Soni to be 
related to the grison, belonging to an entirely different section of the carnivores, the 
Miistelidae; and it was not until he acquired better material that he came to see that it 
was really a viverrid. Matschie's correction was a bald assertion in a list of errata 
(1895: 147). When Pocock (1916a) published his brief statement of synonymy he cited 
as confirming Thomas's alleged view the single character of the structure of the ear, 
which was "like that of Miimios rather than of Crisoii". In effect, all diis scrap ot 
evidence amounted to was not the synonymy of Galcrisciis with BJco\^<)lc but that the 
former was, in fact, a viverrid not a mustelid, a very different matter. 

Thomas based his view ot the validity ot Galcriiciis on both external and cranial 
characters. Regarding the tirst, he consitiered that between "the two groups, iii(;riih's 
and jdihsoiii on the one hand and craisiciuidii and its allies on the other, there is a difference 
of general appearance and coloration so conspicuous that if such a comparison had 
ever before been made a suspicion of the generic distinction of the two must have 
arisen". He went on to cite build, tail-shape, and colour; but the differences of form 
are not, in fact, so striking as one might be led to suppose, except possibly in the tails. 
Colour, ot course, is of little moment in generic differentiation. The pelage ot some 
old West African iii{;rijH's is, in its extreme shortness, sharply distinct from Bdcofialc; 
but m yoiuiger examples as well as in East Atncan iiiiiripcs and in jacksiVii this distinction 
disappears; and the apparent difference may be nothing more than a matter of age, 
as discussed below. 

The main case for the separate recognition of Gdkrifius must rest on cranial and 
dental characters. Thomas drew attention to a few important differences; but there arc 
more than he cited. The cranial distinctions are those of proportion. Li iii/cinjii/c the 
posterior part of the skull is very much longer; in fict, if the glenoid fossa of a Bdcogiilc 
skull is held level with that o[ Giihrisais it will be seen that the occipital condyles are 
almost level too, in spite of the very much shorter overall skull length. On the other 
hand the rostrum is relatively shorter. Expressing this in terms of percentages of the 
condylobasal length: die distance from the glenoid tossa to the condyle is luider 
30 per cent in Gakrisciis and about 33 per cent in /ii/ccn'ii/e. The length of skull lying 
anterior to the front margin of the orbit is 3 1 to 34 per cent in Giilniscini and 28 or 
29 per cent in Bdcooalc; and die nasals are 24 per cent and 18 per cent. Other cranial 
differences are to be seen in the postorbital constriction, which 111 Giikrisciis is situated 
appreciably more posterior to the processes than in Bdco^^alc; and in the shape of the 
postdental palate, which is roughly square in the former but somewhat longer than 
broad in the latter. 

What is possibly the most signitlcant distinction between the two genera lies in 
the posterior cheekteeth, which m Galcrisais are ot a much heavier build and squarer 
shape, having none of the transverse narrowing ot ft/ccijii/c; /»- is a much more import- 
ant, larger, tooth. 

Taking all diese various differences into account and especially the clearly distinct 
characters of the teeth, the present author feels diat, despite the apparently inter- 



GALERISCUS 323 

mediate role sometimes played by j<icksoui (though certainly not in the teeth and cranial 
proportions detailed above), Thomas was justified in regarding Gakrisais as validly 
separable from Bdco^^alc at generic level. 

Distribution. As thus defined, Galcrisais has two species and a known distribution 
from the Cross River basin (eastern Nigeria) to the Congo and, possibly (Hill &: 
Carter, 1941) northern Angola; and in the east as far as the highlands of Kenya. The 
one species, nigripvs, is essentially a high forest, the other, jackscni, a motuitain species. 

Description. These are moderate-sized to very big mongooses; nigripcs, as it occurs 
in West Africa, is, in fact the largest of all. The pelage is mostly short to very short, 
dependent to some extent upon age, and is characterized by very dense underfiir 
amongst wliich stand fairly long, aimulated bristle-hairs which, when the coat is in 
repose, more or less conceal the underfur. The general overall dorsal colour is variable, 
usually a mediinn to palish grey but it may be deep brown or even red. The legs and 
feet are always deep blackish-brown, contrasting, except in the unusual dark forms, 
very markedly with the back. This striking feature is shared, in West Africa, with 
Ichncumia; but Gakrisais has only 4 digits on each foot. 

Skull. A detailed description of the skull as it occurs in West Africa is given below 
in connexion with the species nifiripcs; here a few of the main generic characters are 
briefly given; some have already been discussed in the taxonomic section above. 
There is always a well-developed sagittal crest in both sexes; the interorbital breadth 
is always much greater than the postorbital constriction; and the circumorbital ring 
seems never to be complete. 

GALERISCUS NIGRIPES (Pucheran) Black-footed Mongoose 

Bdeogah nigripcs Pucheran, 1855, Rcinic Mag. Zoo]. (2) 7: in. Gaboon. The specific name is compounded 
from the Latin words mgro to be black, and pes foot. 

Distribution. This, the largest of West African m