"•I"
BULLETIN OF
THE BRITISH MUSEUM
(NATURAL HISTORY)
ZOOLOGY
Vol. 26
1973-1974
BRITISH MUSEUM (NATURAL HISTORY)
LONDON: 1975
DATES OF PUBLICATION OF THE PARTS
No. i . . . .1 November 1973
No. 2 . . . . 15 November 1973
No. 3 . . . . . 7 February 1974
No. 4 22 May 1974
No. 5 24 May 1974
No. 6 7 June 1974
ISSN 0007-1498
Printed in Great Britain by John Wright and Sons Ltd. at The Stoncbridge Press, Bristol BS4 5NU
CONTENTS
ZOOLOGY VOLUME 26
PAGE
No. i. A revision of the large Barbus (Pisces, Cyprinidae) of east and
central Africa. Studies on African Cyprinidae Part II. By
K. E. BANISTER ......... i
No. 2. Some digenetic trematodes in fishes from the Bay of Biscay and
nearby waters. By R. A. BRAY ...... 149
No. 3. Observations on Trinema linear e Penard (Testacea : Protozoa).
By R. H. HEDLEY and C. G. OGDEN (7 Pis.) .... 185
No. 4. The indigenous earthworms (Megascolecidae : Oligochaeta) of
Tasmania. By B. G. M. JAMIESON (10 Pis.) .... 201
No. 5. The freshwater fishes of Rivers Mungo and Meme and Lakes
Kotto, Mboandong and Soden, West Cameroon. By E. TRE-
WAVAS (5 Pis.) 329
No. 6. Notes on some echinoderms from southern Africa. By A. M.
CLARK (3 Pis.) ......... 421
Index to Volume 26 ........ 489
INDEX TO VOLUME 26
The page numbers of the principal references and the new taxonomic names are printed in bold type.
abbreviatus, Procatopus . . 377, 381, 384
aberrans, Procatopus . 377, 379, 381, 384
abyssorum, Deretrema .... 167
abyssorum, Proctophantastes . 169-170, 172
abyssorum Steganoderma 151, 154, 167-ijo
Acanthamoeba . . . . .194
acanthifer phragmorus, Astropecten . 433
Acanthodrilinae . . 203, 212-216, 263
Acanthodrilini . . . 203, 212-216
acanthodriloides, Pseudocryptodrilus 204, 214,
296, 297, 298-299, 324, Pis. 61-63
Acanthophiothrix . . . 465-466, 467
acinus, Trinema . . . . .189
aculeatus, Prosorhynchus . . 159
acuminatus, Psilaster .... 434
acuticeps, Barbus 3, 6-9, 27, 73, 89, no, 128,
140, 144
acutirostris, Barbus brunellii . . 49, 60
acutisquama, Amphiura . . . 423, 458
adjecta, Amphiura . . 446, 447, 483
adustus, Marcusenius . . . 336, 338
aeglefinus, Gadus . . . . .169
affinis, Barbus ..... 48
afnnis brevibarbis, Barbus ... 48
affinis nedgia, Barbus . . . .48
africana, Asterias ..... 439
africana, Marthasterias .... 439
africana, Marthasterias glacialis . . 439
africanus, Asteracanthion . . . 439
Agrammobarbus ..... 94
ahli, Aphyosemion . . . . -375
ahli, Auchenoglanis . . 331, 356-358
ahlselli, Barbus .... 94-96
albella, Amphiura ..... 444
albertisi, Cryptodrilus 204, 208, 269, 270-271
albertisi, Megascolides . . . 207, 270
Alestes . . . . 331, 339-340, 410
alexandri, Ophidiaster .... 436
alluaudi, Astatoreochromis . . .127
alluaudi, Barbus . . . . 3, 9, 144
Alosa ...... 154, 178
alsophila, Perichaeta . . . .222
altera, Ophiopallas paradoxa . . 477, 478
altianalis, Barbus 3, 7-8, 9-2O, 22, 25-27, 65,
70, 73-74, 89, 94, 103, 106, no,
114, 127-128, 129-131, 134, 140, 144
altianalis, Barbus altianalis 9-10, 12, 15-16,
21, 23, 27, 144
altianalis altianalis, Barbus 9-10, 12, 15-16,
21, 23, 27, 144
altianalis eduardianus, Barbus 9, u, 15, 17-18,
20, 21-23, 27» XI4> I27. J44
altianalis labiosa, Barbus . . 9, 103
altianalis lobogenysoides, Barbus . .103
altianalis longifilis, Barbus . . 76, 103
altianalis paucisquamata, Barbus . 76, 103
altianalis radcliffii, Barbus 9-11, 13, 15, 17,
19-20, 22-27, I27. I29. X44
altianalis urundensis, Barbus . . in, 113
alticola, Barbus ..... 49
altidorsalis, Barbus . . . .119
Amphichilus . . . . . .461
Amphilepis ..... 448, 464
Amphilimna . . . 442—444, 479, 483
Amphiodia . . . . . 461, 481
Amphiophiura .... 423, 476
Amphioplus 450, 451, 452-461, 463, 464, 481,
483
Amphipholis . . . 450-452, 453, 461
Amphispina . . . . . 461, 481
Amphiura 423, 444-450, 452, 454, 458, 475, 483
Amphiuridae. .... 444-464
anacanthus, Astropecten . . . 432
Anamphiura .... 423, 478-479
Andreasenius 376, 377, 379, 380, 381, 387, 413
angolensis, Ophiacantha .... 442
angularis, Amphiura . . . 446, 447
angulosa, Stereomyxa . . . .194
annectens, Labeo .... 340, 341
Anseropoda ...... 423
Antedon ..... 427, 429
Anthosticte . . . . . .423
antinorii, Aplocheilichthys ... 72
Antipodrilus ...... 205
aper, Capros .... 154, 173
Aphyosemion 331, 332, 370-376, 411, 412, 413
apiocystis, Diporochaeta . . .223
apleurogramma, Barbus . . . .94
Aplocheilichthys . . . . .72
Aplocheilus ...... 369
appendiculatus, Hemiurus . . 154, 178
arabicus, Barbus .... 128, 131
aranea, Ophiopsila . . . .472
arenilega, Myxotheca . . . .195
Argilophilus . . . . . 218, 263
Argulus . . -35
aristulata, Ophiothrix . . 466-467, 468
aristulata, Placophiothrix . . 466, 467
arnoldi, Diporochaeta . . . .222
Astatoreochromis . . . . .127
Asteracanthion ..... 439
Asterias .... 439, 440, 467
Asteriidae ..... 439-440
Asterina .... 437~438. 483
4QO
INDEX
Asterinidae . . 437-438
Asterodiscides ..... 435
Asterodiscus ...... 435
Asteroidea ..... 431-440
Astrocladus .... 440-441, PL 3
Astroconus . . . . . .441
Astropecten . . 423, 431-434, 483, Pis. 1-2
Astropectinidae .... 431-434
athertonensis, Diporochaeta . . .221
athertonensis, Perionyx . . . .221
athi, Barbus ..... 94-96
Atopochilus 358, 366
Auchenoglanis . 331, 332, 356-358, 412, 4*3
australis, Barbus intermedius 4, 50, 70-71,
74, 75, 144
avesiculatus, Cryptodrilus 204, 208, 268, 269,
270, 271-273, 275, 323, Pis. 2-3
avicularia, Luidia ..... 431
Azygiidae . . . . . .167
babaulti, Agrammobarbus
babaulti, Barbus .
babaulti, Varicorhinus .
baccata, Ophiacantha
Bagridae
bakeri, Perichaeta .
ballayi, Auchenoglanis .
Banister, K. E.
Barbodes
94
4. 94-96
. 94
• 442
, 356-358
. 223
• 357
1-148
• 349
Barbus 1-148, 331, 332, 333, 341-343, 344-356,
410, 411, 412, 413
barbus, Barbus . . . . .134
Barilius ...... 25
barronensis, Perichaeta . . . .222
bassana, Perionychella . 203, 206, 253-255
bassanus, Megascolides . . 205, ,222, 253
bassanus, Plutellus .... 253
batesii, Barbus 128, 331, 342, 343-345, 346,
412, 413
batesii, Chiloglanis . 358, 359, 361, 364
batesii, Chromidotilapia 332, 392, 394, 395,
396, 397, 401
batesii, Labeo .... 340, 341
batesii, Marcusenius .... 338
batesii, Pelmatochromis . . . 393, 394
Bathybiaster ..... 434
Bathycreadium . . 151, 154, 160-162
baudoni, Barbus .... 332, 356
bayoni, Barbus .... 10-11
beauforti, Ophioteresis . . . 469, 483
beauforti, Ophiothela . . . 469, 483
Beirabarbus ...... 347
bemini, Tilapia .... 405, 411
Benthotrema . . . . .178
Beryx . ... 154, 170
bidiverticulatus, Oreoscolex 204, 210, 298,
303, 304, 305-307, 324, PL 67
bimaculatus, Hemichromis . . 411, 413
bingeri, Barbus . . . . 49, 56
bingeri, Capoeta . . . . .49
bingeri, Varicorhinus .... 49
biscayense, Bathycreadium 151, 154, 160-I&2
bispinosa, Ophiopsila . . 472-475, 483
bithecata, Perionychella 203, 206, 221, 225,
230, 232, 233-234, 256, PL 93
bithecata, Vesiculodrilus 203, 206, 221, 225,
230, 232, 233-234, 256, PL 93
bivittatum, Aphyosemion 331, 370-371, 411,
4J2, 413
bivittatus, Fundulus . . . .370
blackeri, Steringophorus 151, 154, 174-178
blennoides, Phycis . . 154, 162, 180-181
bottegi, Barbus ..... 49
boulengeri, Chromidotilapia . . .401
Brachyalestes . . . . -339
brachyistius, Brienomyrus 331, 332, 336-338,
410, 413, PL 2
brachyistius, Marcusenius . 335, 336, 338
Bray, R. A 149-1 83
bredoi, Varicorhinus . . 119-120, 123
brevibarbis, Barbus .... 48
brevibarbis, Barbus affinis ... 48
brevibarbis, Chiloglanis .... 364
brevispinis, Barbus . . . 342, 347
Brienomyrus 331, 332, 335-338, 41°, 4X3, PI- 2
Brissidae ...... 480
brucii, Barbus . . . . -37
brucii, Varicorhinus . . . 37, 123
brunellii, Barbus . . 49, 60, 65, 69-70
brunellii acutirostris, Barbus . . 49, 60
brunyensis, Cryptodrilus 204, 208, 268, 269,
270, 273-275, 323, Pis. 4-7
Brycinus 339
Bucephalidae .... 155-159
Bucephalopsis .... 154, 157
burniensis, Megascolex 204, 214, 319-321, 324
burtoni, Asterina .... 437-438
burtonii, Asterina . . . . -437
bynni, Barbus 3, 20, 25-26, 27-3&, 45-47, 50,
74, 78, 100, 114, 127-128, 140, 144
bynni, Cyprinus .... 27-28
bynni complex, Barbus . 128-129, 131-133.
139-140
bynni rudolfianus, Barbus . . 27-28
byrni, Barbus ..... 33
cabrae, Tilapia ..... 406
caelorhinchus, Coelorinchus . 154, 160, 167
caeruleus, Aphyosemion calliurus . -375
caeruleus, Megascolex . . . -319
caeruleus, Panchax calliurus . . -375
calabaricus, Calamoichthys . . .411
Calamoichthys . . . . .411
callichaeta, Woodwardia . . . 265
callipterus, Barbus 331, 333, 347-349, 352, 411,
413
callipterus,, Enteromius . . 347-349
calliurum, Aphyosemion . . 373, 375, 376
INDEX
491
calliurus caeruleus, Aphyosemion . . 375
calliurus caeruleus, Panchax . . . 375
Callochromis . . . . . 39 1
cameronense, Aphyosemion . . 375, 376
cameronensis, Aphyosemion . . 373, 375
cameron ensis, Panchax . . . -375
cameronensis, Tilapia .... 403
camerunensis, Labeo 331, 340-341, 413, PI. 3
camerunensis, Tilapia . 33 1, 332, 402-403,
405, 411, 413
camerunensis, Tilapia lata . . . 402
campestris, Cryptodrilus . . . 205, 307
campestris, Notoscolex .... 307
campestris, Oreoscolex . 204, 210, 303, 305,
306, 307-309, 324, Pis. 68-72
camptacanthus, Barbodes . . . 349
camptacanthus, Barbus . 331, 349-35O, 352,
354. 355. 4J3
camptacanthus, Puntius .... 349
canaliculata, Perichaeta . . . 217, 221
canaliculata, Perionychella . . .217
canaliculatus, Perionychella . . .216
Candida, Amphiura . . . 445, 483
canis, Barbus . . . . .134
canis, Tor ...... 134
capensis, Amphiura . . 446-447, 483
capensis, Astroconus . . . .441
capensis, Barbus . . . . .129
capensis, Dipsacaster sladeni . . . 434
capensis, Hacelia .... 435-436
capensis, Hacelia superba . . 435, 483
capensis, Ophiarachnella . . . 480
capensis, Ophiogymna .... 423
capensis, Ophiomyxa vivipara. . .482
capensis, Perionychella 203, 206, 220, 223,
224, 225-226, 228, 229, 256,
258, 260, Pis. 91-92
Capoeta . . . . . 49, 94
Capros ...... 154, 173
carnea, Ophiactis ..... 465
carpio, Cyprinus . . . . .128
castellanii, Acanthamoeba . . .194
caudifasciatus, Chromidotilapia . . 395
caudovittatus, Barbus 3, n, 36-41, 95, 106,
122-123, 126, 130, 139, 144
celiae, Aphyosemion 331, 371-373, 412, 413
Ceramaster .... 423, 435, 483
chadwicki, Decametra .... 426
Channa . . . . . .411
chapini, Varicorhinus . . . 123, 126
Chara ....... 194
Characidae . 331, 339-34°
chariensis, Labeo . . . . .341
chariensis nunensis, Labeo . . .341
Charitometridae .... 429-430
Chiloglanis . 331, 332, 358-361, 362-367, 413
chilotes, Barbus . . . . .119
chilotes sakaniae, Barbus ... 36
chlorotaenia, Barbus .... 354
chondriscus, Ceramaster . . . 435, 483
chondriscus, Ceramaster patagonicus . 435
Chromichthys .... 387, 388
Chromidotilapia . 331, 332, 888-393, 394-402,
404, 411, 413, Pis. 4-5
chuni, Democrinus. .... 482
Cichlidae .... 331, 387-410
cinereus, Muraenesox . . . .157
cinnamomeum, Aphyosemion . . -372
Clarias . . . 331, 356, 411, 412, 413
Clariidae . . . . 331, 356
Clark, A. M 421-487, 3 Pis.
clarki, Odinia ..... 482
codringtonii, Barbus . . . .119
coelorhynchi, Lepidapedon . . .166
Coelorinchus .... 154, 160, 167
Coitocaecum . . . . . .162
Colobometridae .... 424-426
Comanthus ..... 480, 481
comata, Ophiothrix .... 465
compinei, Barbus ..... 344
complanatum, Trinema . . . .189
complanatum globulosa, Trinema . .189
compressa, Amphiura . . 446, 447, 483
Conger . . 154-155, 157-158, 178, 181
conger, Conger . 154-155, 157-158, 178, 181
congicus, Chiloglanis . . . 365, 366
coniortodes, Amphioplus . . . 464
copei, Xenodermichthys . 154, 174-176
copelandi, Perichaeta . . . .220
Coptodon ...... 403
corynephora, Ophiomitrella . . 441, 482
Coryphaenoides . . . . .169
cribriformis, Amphilimna . 442-444, 483
Crinoidea ..... 424-430
Crotalometra . . 427-429, 430, 482
crucibulum, Distoma . . . .157
crucibulum, Gasterostomum . . . 157
crucibulum, Monostoma . . 157
crucibulum, Prosorhynchus 151, 154, 157-159
Cryptodrilus 204, 205, 208, 212, 214, 219, 221,
222, 241, 247, 257, 260, 265, 266-296,
3°3. 3°7, 309. 317. 323, Pis- 2-59
Ctenopoma . . . . .411, 413
Curimatus ...... 389
curtus, Barbus . . . • .116
cylindricus, Labeo . . . . .341
Cyprinidae . . . 1-148, 331, 340-356
Cyprinodontidae . . . 331, 367~387
cyprinoides, Mormyrus .... 335
Cyprinus 27-28, 128
daga, Barbus platystomus ... 48
dageti, Tilapia . . . 403
dainellii, Barbus . . . -49. 52, 55
dainellii macrocephalus, Barbus . . 49
Dalatias 154
danae, Ophiothela ..... 470
dartnalli, Cryptodrilus 204, 208, 269, 270,
273, 275-277, 323, Pis. 8-12
492
INDEX
davallia, Diporochaeta .... 220
davallia, Perionychella . . . .217
davidis, Antipodrilus . . . 205
decadactylus, Beryx . . . 154, 170
Decametra .... 423, 425, 426
decatheca, Perionychella . 203, 206, 255
decatheca, Plutellus . . 207, 222, 255
degeni, Barbus . . . 49~5°> 54~55
degeni leptorhinus, Barbus ... 49
dekkensis, Barbus platystomus . . 48
Democrinus ...... 482
dendyi, Perichaeta. . . . 217, 218
dendyi, Perionychella . . . 217 220
Deretrema ..... 167, 169
Dermechinus. .... 479-480
Derogenes .... 154, 180-181
deserti, Barbus ..... 348
diaphanus, Megascolides .... 296
diaphanus, Pseudocryptodrilus . 296, 299
Dichogastrini . . . 204, 300-302
dicksonia, Perichaeta .... 223
Difflugiella . . . . . .196
Digenea ..... 149-183
dilwynnia, Diporochaeta . . . 234
dilwynnia, Perichaeta . . 205, 221, 234
dilwynnia, Perionychella 203, 206, 225, 232,
234-236, 256, Pis. 94-96
dilwynnia, Vesiculodrilus 203, 206, 225, 232,
234-236, 256, Pis. 94-96
Diporochaeta 204, 217, 218, 219, 220, 221, 222,
223, 228, 234, 245, 257, 258, 259
Dipsacaster ...... 434
disneyi, Chiloglanis . 331, 361, 362, 363,
355-367, 413
Distoma ..... 157, 173
Distomum . . . . . .162
divergens, Distoma . . . .173
divergens, Fasciola . . . • . 173
divergens, Steringotrema 151, 154, 173-174
dividua, Amphiodia . . . .461
dividua, Ophiothela .... 470
Dolichoenterum . . 151, 154, 155-157
donyensis, Barbus .... 94~95
dubia, Perichaeta . . . . .220
dubia, Tilapia ..... 405
dubia, Tilapia mariae . . . 405, 406
dubius, Cryptodrilus . . . 269, 282
dubius, Microscolex . . . .213
duchesnii, Barbus ..... 48
duchesnii ibridus, Barbus ... 48
duchesnii maximus, Barbus ... 48
durbanensis, Decametra . . 423, 425, 426
dyscrita, Asterina .... 438, 483
dyscrita, Patiriella. .... 438
eburneensis, Barbus . . . 354, 355
Echinaster .... 438, 439, 483
Echinasteridae .... 438-439
echinata, Ophiothrix fragilis . . . 468
Echinidae
Echinodermata
Echinoidea .
Echinus
eduardianus, Barbus
479-480
. 421-487, 3 Pis.
479-480
• 479
. 9-n
eduardianus, Barbus altianalis 9, n, 15, 17-18,
20, 21-23, 27, IJ4. I27, 144
electricus, Malapterurus . . 331, 367, 413
electricus, Silurus ..... 367
elegans, Asterodiscides . . . 435
elegans, Asterodiscus .... 435
ellisi, Cryptodrilus . . . 205, 222, 257
ellisi, Perionychella . . 203, 206, 257
ellisi, Plutellus . . . . .257
elongata, Nicolla . . . . .162
elongatum, Bathycreadium . . 151, 162
elongatum, Lepidapedon 151, 154, 164-1 66
elongatum, Lepodora . . . .164
elongatum, Nicolla . . . .151
elongatum, Steganoderma . . .169
elongatus, Barbus ..... 48
elongatus, Chromichthys . . 387, 388
elongatus, Hemichromis . 331, 332, 387-388,
411. 413
enchelys, Trinema .... 188-189
Enteromius ..... 347-349
enteronephricus, Cryptodrilus 204, 208, 267,
268, 269, 273, 275, 276, 277-28o,
286, 290, 323, Pis. 13-19
Epiplatys . . 331, 367-370, 411, 413
eques, Lepidion . . 154, 160, 164, 167
erici, Diporochaeta . . . .221
erici, Perionyx . . . . .221
erlangeri, Barbus . . . 49, 55, 57, 72
esculenta, Tilapia . . . . .391
esculentus, Sarotherodon . . -391
ethiopicus, Barbus. . 3, 41-44, 9O-9X. 129,
139, 144
euchilus, Barbus ..... 36
Euglypha . . . 187-188, 192, 195-197
Euglyphacea. . . . . .187
Euglyphidae . . . . . .188
eumystus, Barbus ..... 49
Euryale ...... 440
euryale, Asterias ..... 440
euryale, Astrocladus . . 440-441, PL 3
euryplax, Ceramaster patagonicus . 435, 483
euzona, Diporochaeta . . . .222
evansi, Perionychella 203, 206, 221, 225, 230,
235, 236-237, 254, 256, 258
evansi, Vesiculodrilus 203, 206, 221, 225, 230,
235, 236-237, 254, 256, 258
excisum, Lecithocladium . . 154, 178
exigua, Asterina ..... 438
exigua, Patiriella ..... 438
fairbairnii, Barbus .
falcatus, Amphioplus
falcatus, Unioplus .
. 119
• 455-456, 458, 459, 460
455-456
INDEX
493
fallax, Alosa 154, 178
fasciata, Helicometra . . . 154, 159
fasciatus, Hemichromis 332, 387, 388, 411, 413
Fasciola 173
fastigatus, Cryptodrilus . . 269, 282, 288
faucium, Diporochaeta .... 223
Fellodistomidae .... 173-178
Fellodistomum . . . . .177
fergussonii, Barbus .... 9-11
Filosia ....... 187
finleyi, Chromidotilapia . 331, 392, 393-397,
400, 411, 413, PI. 5
finta, Alosa . . . . . .178
Fletcherodrilus . . . . 218, 221
flexicollis, Bathycreadium . . .162
flexuosa, Ophiactis . . . 464, 465
floridae, Hacelia ..... 436
formosum, Steganoderma . . . 169
forskalii, Labeo . . . . .341
foureaui, Barbus . . . 344, 346, 347
fragilis, Asterias ..... 467
fragilis, Ophiothrix . . 467-469, 483
fragilis echinata, Ophiothrix . . . 468
fragilis lusitanica, Ophiothrix . . . 468
fragilis pentaphylla, Ophiothrix . . 468
fragilis pentaphyllum, Ophiothrix . . 468
frenchi, Cryptodrilus . . . 221, 222
frenchi, Megascolex .... 326
frenchi, Vesiculodrilus . . . .222
frosti, Diporochaeta . . . .222
Fundulus . . . . . -370
furcatus, Amphioplus 452-453, 454, 481, 483
furcatus, Lymanella . 452-453, 481, 483
gadi, Lepidapedon . . . .166
Gadus ...... 166, 169
Gaidropsarus .... 154, 159
galeata, Trinema . . . . .189
galilaea, Tilapia ..... 408
galilaeus, Sarotherodon . . 331, 332, 404,
408-410, 412, 413
galilaeus, Sarotherodon galilaeus . 409, 411
galilaeus, Sparus ..... 408
galilaeus galilaeus, Sarotherodon . 409, 411
galilaeus multifasciatus, Sarotherodon . 408
gananensis, Barbus . 3, 32, 36, 44-47, ?8,
100, 128, 140, 144
Garra ..... 90, 131, 389
Gasterostomum . . . . 157
Geophagus . . . . . .391
georgei, Graliophilus .... 260
gibbosus, Amphioplus . . 452, 453, 454
gibbosus, Ophiophragmus . 452, 453, 454
gigas, Mithrodia . . . . .423
gippslandicus, Cryptodrilus . . . 222
glacialis, Marthasterias .... 439
glacialis africana, Marthasterias . . 439
glacialis rarispina, Marthasterias . . 439
glandifera, Perionychella 203, 206, 218, 221,
223, 230, 237-238, 254, 256
glandifera, Vesiculodrilus 203, 206, 218, 221,
223, 230, 237-238, 254, 256
glandulosa, Proctophantastes . . .170
glandulosum, Neosteganoderma . 151, 154,
170-173
glaucicaudis, Procatopus 331, 377, 379, 380,
381, 382, 383, 384, 386
globulosa, Trinema complanatum . .189
globulosa, Trinema lineare . . .189
Glyptometra . . . 428, 429-430, 482
Gnathonemus . . . . -338
goheeni, Mormyrus . . 334, 335, 413
Goniasteridae ..... 435
Gorgonocephalidae . . . 440-441
Gorgonocephalus ..... 440
gorgorensis, Barbus intermedius . . 48
gorguarii, Barbus . . 48, 55, 65, 68-69, 128
gorguarii macrophtalmus, Barbus . . 48
gracilescens, Bucephalopsis . . 154, 157
gracilis, Procatopus .... 380
gracilispina, Asterina .... 437
Graliophilus . 203, 204, 209, 212, 214, 218, 219,
259-26o, 261-262, 263, 265, 324, PI. i
grandis, Diporochaeta . . . .222
grandis, Notoscolex .... 303
grandis, Perionychella . . . .216
grandis, Vesiculodrilus . . . . 216
grandisquama natalensis, Amphiura 447-448
granulatus, Astropecten . . 431, 432, 433
granulatus natalensis, Astropecten 423, 431-433,
PI. i
gravidum, Lecithochirium . 154, 178-i8i
greenii, Labeo . . . . .341
gregorii, Barbus 3, 4, 10, 48, 50, 70, 72, 75, 143
Gromia ...... 195
Gromida ...... 187
gudaricus, Barbus ..... 49
guentheri, Atopochilus .... 358
guineensis, Tilapia .... 402, 403
guntheri, Chromidotilapia 332, 389, 390, 391,
392, 393, 394, 398-402, 411, 413
guntheri, Chromidotilapia guntheri . 331, 398,
399, 400, 401
guntheri, Hemichromis .... 398
guntheri guntheri, Chromidotilapia . 331, 398,
399, 400, 401
guntheri loennbergi, Chromidotilapia 331, 391,
392, 398, 399, 4°°. 401. 4", P1- 4
gymnophallum, Coitocaecum . . .162
Gymnophiura . . . 475, 476, 483
habereri, Barbus .
Hacelia
halli, Megascolex .
hamata, Ophiomitrella
Haplochilus .
Haplochromis
• 346, 347
435-436, 483
• 326
. 482
• 367
26, 389, 391
494
INDEX
harringtoni, Barbus
hastatus, Amphioplus
hastatus, Lymanella
haugi, Pelmatochromis
haugi, Tilapia
healesi, Woodwardia
Medley, R. H.
Helicometra .
hellyeri, Perionychella
452, 481, 483
. 481
402
405, 406
219
185-iQQ, 7 Pis.
154. 159
203, 2O6, 221, 225,
230, 235, 288-241, 247, 256
hellyeri, Vesiculodrilus . 203, 206, 221, 225,
230, 235, 238-241, 247, 256
Hemichromis 331, 332, 387-388, 398, 411, 413
Hemiuridae .
Hemiurus
Henricia
heteroporus, Plutellus
heudeloti, Tilapia .
hickmani, Perionychella
Hickmaniella
178-181
154, 178
438-439, 483
. 268
• 4°5
203, 206, 220, 223,
226-228, 229, 256
204, 2O5, 209, 212, 214, 270,
300-302, 324, Pis. 64-66
hindii, Barbus .... 94-97, 127
hobartensis, Cryptodrilus . 205, 221, 241
hobartensis, Perionychella 203, 206, 225, 230,
241-245, 250, 251, 253, 254, 255, 256, Pis. 97-98
hobartensis, Plutellus . . . .241
hobartensis, Vesiculodrilus 203, 206, 225, 230,
241-245, 250, 251, 253, 254, 256, Pis. 97-98
hoggi, Megascolex ..... 326
hollyi, Barbus . . . . .10-12
holubi, Barbus . . . . .129
horridus, Dermechinus . . . 479~48o
horridus, Echinus ..... 479
huguenini, Barbus .... 83
humilis, Barbus . . . . .114
hursensis, Barbus .... 49, 55-56
Hydrocynus . . . . . ' . 26
Hylopanchax ..... 382
Hypsopanchax .... 377, 382
ibridus, Barbus duchesnii ... 48
ilgi, Barbus ...... 49
illawarrae, Cryptodrilus .... 303
imparicystis, Oreoscolex .... 303
incana, Amphiura ..... 452
infrafasciatus, Haplochilus . . . 367
insularis, Cryptodrilus . . . 205, 241
insularis, Perionychella .... 255
insularis, Plutellus . . . . .241
insularis, Vesiculodrilus . . . .241
integer, Amphioplus 450, 451, 452, 453-455, 461
integer, Lymanella 450, 451, 452, 453-455, 461
integra, Amphipholis . . . -453
integra, Amphiura ..... 454
intermedia, Diporochaeta . . 219, 220
intermedius, Barbus 3, 14, 16, 27, 33, 44, 47-74,
91, no, 127-128, 130-132, 134, 140, 144
intermedius, Barbus intermedius 71, 74-75, 144
intermedius australis, Barbus
intermedius complex, Barbus .
114, 128-129,
intermedius gorgorensis, Barbus
intermedius group, Barbus
intermedius intermedius, Barbus
intermedius leptosoma, Barbus
intermedius microstoma, Barbus
inutilis, Ophiocirce
irregularis, Cryptodrilus .
irregularis, Diporochaeta
irregularis, Notoscolex .
irregularis, Oreoscolex
irregularis, Perichaeta
irregularis, Perionychella
204,
306
227
4. 50, 70-71,
74, 75, 144
73, 74- "0,
131-135, 140
. 48
• 9, 73- 129
. 71, 74-75,
144
. 48
. 48
423, 476-477
205, 309
. 228
• 309
210, 303, 305,
309-3IO, 324
205, 22O, 228
2O3, 2O6, 223,
, 228-229, 256
Jamieson, B. G. M. . 201-328, 102 Pis.
japonica, Polymixia . . . .172
jarsinus, Barbus ..... 48
jobaerti, Hypsopanchax .... 382
jubbi, Barbus ..... 82
jurupari, Geophagus . . . .391
kalki, Amphiura .... 445, 483
kassamensis, Barbus . . . 48, 55-56
kerguelensis, Leptoptychaster . . . 434
kershawi, Perionychella 203, 206, 220, 223,
224, 229-230, 256
kerstenii, Barbus . . . . .129
kimberleyensis, Barbus . . . .129
kingsleyae, Chromidotilapia . 388, 395, 402
kingsleyae, Ctenopoma . . .411, 413
kiogae, Barbus . . . . .10
kivuensis, Barbus . . . . .10
koreae, Amphiura ..... 450
kottae, Tilapia . 3 31, 332, 403, 404-405,
407, 408, 409, 411, 412
krapfi, Barbus .... 94-96, 98
Labeo . 135, 331, 332,
Labeobarbus .
labiatus, Barbus
labiosa, Barbus altianalis
Labyrinthula
lacustris, Diporochaeta .
lacustris, Perionychella .
lacustris, Perionyx.
lacustris, Procatopus
lacustris, Vesiculodrilus
lagensis, Barbus
225
382
340-341, 413, PL 3
. 9, 48
• 94-98, 127
9, 103
• 194
• 245
2O3, 2O6, 221, 223,
,231, 245-246, 256
207, 221, 245
331- 378, 379, 380,
, 383, 386-387, 412
203, 221, 223, 225,
231, 245-246, 256
83, 130, 132
INDEX
495
lamottei, Chiloglanis
Lanceabarbus
lancrenonensis, Barbus .
lapsus, Barbus
lata, Tilapia .
lata camerunensis, Tilapia
Lates ....
leai, Notoscolex
365, 366, 367
94
• 347
• 4- 94
402
402
26
207, 310
leai, Oreoscolex . 204, 210, 303, 305, 310-3 n
Lecithochirium . . . 154, 178-181
Lecithocladium .... 154, 178
Lepidapedon. . . 151, 154, 162-166
Lepidion . . . 154, 160, 164, 167
lepidus, Barbus ..... 355
Lepocreadiidae .... 162-166
Lepodora ..... 162, 164
Leptoptychaster ..... 434
leptorhinus, Barbus degeni ... 49
leptosoma, Barbus. . . .48, 50, no
leptosoma, Barbus intermedius . . 48
lestradei, Barbus .... 36-38, 95
Lethrinops . . . . . .391
liberiensis, Barbus . . 332, 354, 355
liberiensis, Marcusenius . . 336, 337, 338
licha, Dalatias . . . . .154
limicola, Ophionephthys . . 462, 463, 464
Limnodrilus ...... 205
lindti, Diporochaeta .... 223
lineare, Trinema . . 188-199, 7 Pis.
lineare globulosa, Trinema . . .189
lineare pellucida, Trinema . . .189
lineare terricola, Trinema . . .189
linearis, Amphiura .... 423, 475
linnellii, Barbus ..... 344
Lithodesium ...... 194
littoralis, Rhododrilus . 203, 213-2x6, 297
324, PI. 60
lobogenys, Barbus . . . 10, 14-15
lobogenysoides, Barbus altianalis . .103
lochensis, Perichaeta . . . .222
loennbergi, Chromidotilapia . . 332, 392
loennbergi, Chromidotilapia guntheri 331, 391,
392, 398, 399, 400, 401, 411, PI. 4
longianalis, Brienomyrus 332, 336, 337, 410
longianalis, Marcusenius . . . 336, 338
longinlis, Barbus 3, 4, n, 36, 76-79, 100, 106,
128, 140, 145
longinlis, Barbus altianalis . . 76, 103
longipinnis, Alestes 331, 339-340, 410, 413
longipinnis, Brachyalestes . . -339
longirostris, Barbus . . 10, 12, 16-17, J9
longispinis, Barbus .... 83
longissimum, Dolichoenterum 151, 154, 155—157
longus, Oreoscolex 204, 210, 303, 304, 305,
311-313, 324, Pis. 73-80
Lophaster . . . . . .438
Lophius ..... 154, 157
lowei, Polymixia . . . . .172
lowelli, Ophionephthys . . 462-464, 483
luetkeni, Ophiactis .... 465
Luidia . . . . . . .431
Luidiidae . . . . . .431
Lumbri'culus ...... 205
lusitanica, Ophiothrix fragilis . . . 468
Lymanella 450, 451, 452-455, 461, 481, 483
lymani, Ophiactis .... 464, 465
maclareni, Clarias ..... 356
macmillani, Barbus . . -49, 55, 57
macrobrachius, Oligometra serripinna . 426
macrocephala, Tilapia . . . 389, 408
macrocephalus, Barbus dainellii . . 49
macrolepidotus, Alestes . . 331, 339, 413
macrolepidotus, Brycinus . . -339
macrolepis, Barbus 3, 79-83, 130, 132, 139,
145
macronema, Barbus . . . 48, 68-69
macronema parenzani, Barbus . 48, 52
macrophtalmus, Barbus gorguarii . . 48
macrophthalma, Molva . . 154, 180-181
macropus, Barbus ..... 83
Macrurus . . . . . .169
maculata, Luidia . . . . .431
magnicirra, Antedon .... 427
magnicirra, Crotalometra 427-429, 430, 482
magnifica, Tropiometra . . 423, 426-427
magnus, Steringophorus . . . .178
makiensis, Garra ..... 90
Malapteruridae . . . . 331, 367
Malapterurus . . . 331, 367, 413
manni, Diporochaeta .... 222
manteri, Dolichoenterum . . .157
Marcusenius . 335, 336, 337, 338
marequensis, Barbus 4, 10, 82, 119, 123, 130
margaritae, Barbus .... 49
mariae, Barbus . 3, 83-87, 103, 129, 139, 145
mariae, Tilapia . 331, 332, 389, 404, 405-4o8,
411, 412, 413, PI. 4
mariae dubia, Tilapia . . . 405, 406
marlieri, Chiloglanis . . 365, 366, 367
Marthasterias ..... 439
mathoiae, Barbus .... 94-96, 98
matris, Barbus ..... 83
mawambi, Barbus ..... 94
mawambiensis, Barbus . . 95, no-in, 344
maximus, Barbus duchesnii ... 48
mediocincta, Diporochaeta . . . 222
mediocris, Cryptodrilus . . . 266, 269
meeki, Tilapia ..... 405
Megascolecidae . . 201-328, 102 Pis.
Megascolecinae . . . 203, 216-326
Megascolecini . . 204, 280, 302-326
Megascolex . . 204, 205, 212, 214, 217,
220, 318-326
Megascolides . 205, 207, 222, 253, 263, 270, 296
melanopleura, Tilapia .... 403
melanotheron, Sarotherodon . . 389, 390
Melitaea .... -47°
meneliki, Barbus .... .27-28
496
INDEX
mento, Barbus
Merluccius
merluccius, Merluccius .
micra, Woloszynskia
microbarbis, Barbus
microcephala, Tilapia
microcephalus, Brienomyrus
microcephalus, Mormyrus
microcotyleum, Lepidapedon
Micromesistius
microplax, Amphiodia
microplax, Amphispina .
micropogon, Chiloglanis .
154, 180-181
154, 180-181
. 194
3, 87-89, 139, 145
• 4°5
• 337
• 336, 338
. 166
154, 180-181
461, 481
461, 481
33i, 358, 359, 360,
361-365, 367, 413
• 338
72
. 213
microps, Stomatorhinus .
Micropterus ....
Microscolex ....
microstoma, Barbus intermedius . . 48
microterolepis, Barbus . 3, 44, 89-gi, 129-130,
139, M5
. 36
. 213
3, 5, 91-94, 114, 130-131,
133, 140, 145
. 469
. 480
• 423
331, 358-367
426
. 116
154, 180-181
miochilus, Barbus .
minutus, Rhododrilus
mirabilis, Barbus
mirabilis, Ophiothela
mirabilis, Spatagobrissus
Mithrodia
Mochokidae .
modica, Decametra
moeruensis, Barbus
Molva .
Monachocrinus
monkei, Auchenoglanis .
Monostoma .
montisarthuri, Megascolex
• 423
• 358
• 157
204, 214, 319,
320, 321-323, 324
montisarthuri, Perionychella 203, 206, 221,
225, 230, 241, 244, 246-247, 254, 257
montisarthuri, Vesiculodrilus 203, 206, 221,
225, 230, 241, 244, 246-247, 254, 257
montiskosciuskoi, Graliophilus . . 260
moorii, Barilius ..... 25
morhua, Gadus . . . . .166
Mormyridae . . . . 331, 334-338
Mormyrus . 331, 334-335, 336, 338, 413, PI. i
moroea, Diporochaeta . . . .257
moroea, Perichaeta . . 205, 223, 257
moroea, Perionychella . 203, 206, 257-258
mortoni, Cryptodrilus . 205, 219, 221, 247
mortoni, Perionychella . 203, 206, 221, 223,
225, 230, 244, 247-250, 254, 255, 257, 286,
Pis. 99-102
mortoni, Plutellus 247
mortoni, Vesiculodrilus . 203, 206, 221, 223,
225, 230, 244, 247-250, 254, 257, 286,
Pis. 99-102
mortoni, Woodwardiella .... 247
multicolor, Haplochromis . . . 389
multicolor, Pseudocrenilabrus . . . 389
multifasciatus, Sarotherodon galilaeus . 408
multiprostatus, Telmatodrilus .
mungoensis, Barbus . 331,
mungoensis, Labeo
Muraenesox . . .
myriaster, Conger .
Myxotheca ....
205
342, 343, 344-
345-347, 413
• 34i
• 157
• 157
• 195
nairobi, Barbus .... • 94-9 5
Nannaethiops . . . . .411
natalensis, Amphiura grandisquama 447-448
natalensis, Astropecten granulatus 423, 431-433,
PI. i
natalensis, Barbus . . . . .129
nedgia, Barbus . . .9, 48, 50, 54-55
nedgia, Barbus affinis . . . .48
nedgia, Labeobarbus . . . . 9, 48
nemoralis, Diporochaeta. . . . 223
Neosteganoderma . . 151, 154, 169-173
neumayeri, Barbus .... 95
neuvillii, Barbus ..... 49
Nicolla . .... 151, 161-162
nigra, Tilapia .... 72, 78
nigrofasciatus, Pelmatochromis . 393, 394
nigromarginatus, Procatopus . . 380, 384
niloticus, Lates . . . . .26
niloticus waterloti, Chiloglanis . 361, 363
Notoscolex 204, 207, 263, 280, 303, 307, 310, 317
nototaenia, Procatopus . 331, 376, 377, 379,
380, 381, 382, 383, 384, 385-386
novembris, Gymnophiura . 475, 476, 483
novemradiata, Anseropoda . . .423
nuda, Ophiothela ..... 469
nudum, Ophiopsammium . . 469, 483
nunensis, Labeo chariensis . . .341
nyongana, Tilapia ..... 403
obesus, Barbus . 10-11,
obliquae, Perionychella
obliquae, Vesiculodrilus
16, 19, 70, 127-128
203, 206, 221, 225,
249, 250-251, 257
2O3, 206, 221, 225,
249, 250-251, 257
obscura, Perichaeta . . . .223
obscurum, Aphyosemion . . . 375
obscurus, Channa . . . . .411
obscurus, Chromidotilapia . . .413
obtusispina, Perissasterias . . 440, 483
occidentalis, Barbus .... 347
occidentalis, Oligometra serripinna . 424-426
ochracea, Melitaea ..... 470
ochroleuca, Amphichilus. . . .461
ochroleuca, Amphioplus . . . .461
Odinia ....... 482
oeseri, Aphyosemion . 331, 332, 370, 371,
373-376, 411, 412, 413
oeseri, Panchax .... 373, 374
officeri, Cryptodrilus 204, 205, 208, 269, 280-282
INDEX
497
officer!, Notoscolex
officer!, Trinephrus
Ogden, C. G. .
ogowensis, Tilapia . .
Oligochaeta .
Oligometra .
olivacea, Amphilimna
Onchocerca .
Onos ....
Opecoelidae .
Ophiacantha .
Ophiacanthidae
Ophiactidae .
Ophiactis
Ophiarachnella
Ophidiaster .
Ophidiasteridae
Ophiocirce
Ophiocoma .
Ophiocomidae . .
Ophioderma .
Ophiogymna .
Ophioleucidae
Ophiomastix .
Ophiomisidium
Ophiomitrella
Ophiomusium
Ophiomyxa .
Ophionephthys
Ophionereis . . .
Ophiopallas .
Ophiophragmus
Ophiopsammium .
Ophiopsila
Ophiopteron .
Ophioteresis .
Ophiothela .
Ophiothrix .
Ophiotrichidae
Ophiura
Ophiuridae .
Ophiuroidea .
opisthogaster, Hickmaniella 204, 214, 270,
300, 301-302, 324,
Pis. 64-66
oreas, Barbus .... 48, 57
Oreasteridae ...... 435
Oreoscolex . 204, 205, 210, 212, 214, 280, 282,
298, 302-318, 324,
Pis. 67-90
ornata, Henricia ..... 439
Otodistomum .... 154, 167
oviformis, Gromia . . . . 195
oxycephalus, Barbus . . . .116
oxyrhynchus, Barbus 3, 36, 47, 50-51, 72, 78-80,
82, 87, 94-IOO, 103, 127-129, 131, 140, 145
pacei, Anthosticte .
pacei, Tethyaster .
280 Pachylometra ..... 429
. 280 pagenstecheri, Barbus 3, 101-IO3, 139-140, 145
185-199, Pis. 1-7 paludinosus, Barbus .... 129
404, 405 Panchax .... 373, 374, 375
201-328, 102 Pis. paradoxa, Ophiopallas . . 423, 477-478
423, 424-426 paradoxa altera, Ophiopallas . . 477, 478
442, 444 Paranepanthia ..... 437
332 parenzani, Barbus macronema . 48, 52
162 parva, Ophiactis .... 464, 465
159-162 patagonicus chondriscus, Ceramaster . 435
442 patagonicus euryplax, Ceramaster . 435, 483
441-444 Patiriella ..... 438, 483
464-465 paucispina, Ophiopsila .... 474
464-465 paucisquamata, Barbus altianalis . 76, 103
480 paucisquamatus, Barbus 3, 4, n, 40-41, 76,
. 436 103-io6, 140, 145
435-436 pectinatus, Amphioplus . . 456-459, 481
423, 476-477 pectinatus, Telmatodrilus . . . 205
472, 481 peculiaris, Oreoscolex . 204, 210, 303, 305,
470-475 306, 313-315, 324, Pis. 81-82
423 pedderensis, Perionychella 203, 206, 221, 225,
• 423 244, 251, 257
476-478 pedderensis, Vesiculodrilus 203, 206, 221, 225,
. 472 244, 251, 257
476 pellegrini, Geophagus . . . 391
441, 482 pellucida, Trinema lineare . . .189
476, 482 Pelmatochromis . 388, 391, 393, 394, 402
482 penardi, Trinema . . . . .189
462-464, 483 pentaphylla, Ophiothrix fragilis . . 468
. . . 480 pentaphyllum, Ophiothrix fragilis . . 468
423, 477-478 percivali, Barbus ..... 95
452, 453, 454 Perichaeta . 205, 217, 218, 220, 221, 222,
469, 483 223, 228, 234, 257, 258, 259, 324
470-475, 483 perince, Barbus 353
469 Perionychella 203, 204, 206, 209, 211, 212, 214,
469, 483 216-259, 260, 263, 265, 266, 286, Pis. 91-102
469-470, 483 Perionychini .... 203, 216-299
465-469, 483 Perionyx . . 207, 217, 218, 221, 245
465-470 Perissasterias . . . 439-440, 483
475-476, 483 perplexicans, Barbus . . -94, 96-97
475-476 perplexicans, Capoeta .... 94
440-479 perrieri, Monachocrinus .... 423
petersi, Gnathonemus .... 338
phalacrus, Diporochaeta . . . .221
phalacrus, Perionychella. . . .216
phalacrus, Perionyx . . . .221
Pheretima 217, 265
pheretima, Diporochaeta . . .222
phosphoreus, Microscolex . . .213
Phoxaster ...... 434
phragmorus, Astropecten . . 433, 483
phragmorus, Astropecten acanthifer. . 433
phragmorus, Astropecten polyacanthus 433-434,
Pis. 1-2
Phycis .... 154, 162, 180-181
pica, Ophiocoma . . . . .481
pietschmanni, Auchenoglanis . . 356, 357
pietschmanni, Barbus . . . .10-12
423 Pinguidrilus . 203, 204, 209, 212, 263-265,
423 294, 324
INDEX
154. J57
1-148, 329-420, 5 Pis.
.' . 466, 467
piscatorius, Lophius
Pisces .
Placophiothrix
plagiostomus, Barbus . . 48
plana, Ophiactis . . 464-465
platyacanthus, Amphioplus . . .461
platyrhinus, Barbus 3, 44, 106-iog, 126-127,
129, 139, M5
platystomus, Barbus . . . . 48
platystomus, Barbus platystomus . . 48
platystomus daga, Barbus . . 48
platystomus dekkensis, Barbus . . 48
platystomus platystomus, Barbus . . 48
platystomus prognathus, Barbus . . 48
platystomus vatovae, Barbus . . 48, 65
Plecostomus .... . 389
Pleionogaster ... . 300
plenum, Benthotrema . . . .178
pleuromelas, Sarotherodon . . . 409
plumosus, Procatopus . . 379, 380
Plutellus . 204, 207, 218, 222, 241, 247, 253,
255, 257, 260, 263, 265, 266, 268
pojeri, Barbus ..... 36
polyacantha, Perissasterias . 439-44O, 483
polyacanthus, Astropecten . . 433, 483
polyacanthus, Astropecten polyacanthus 433-
434, PI. 2
polyacanthus phragmorus, Astropecten 433-434,
Pis. 1-2
polyacanthus polyacanthus, Astropecten 433-
434- PL 2
polyarthra, Porphyrocrinus . . . 482
polylepis, Barbus . . . . .129
Polymixia . . . . . .172
polymixiae, Neosteganoderma . . 151, 172
polymixiae, Proctophantastes . . 151, 172
polymorphus, Amphioplus . . 463, 464
polynephricus, Cryptodrilus 204, 205, 267, 268,
269, 270, 271, 273, 275,
280, 282-283, 284, 286
polynephricus, Cryptodrilus polynephricus 204,
208, 281, 283, 284-288, 289,
291, 292, 323, Pis. 20-31
polynephricus, Trinephrus . . .284
polynephricus ad urethrae, Cryptodrilus
polynephricus 204, 208, 281, 286, 291-293,
323, Pis. 32-43
polynephricus polynephricus, Cryptodrilus 204,
208, 281, 283, 284-288, 289,
291, 292, 323, Pis. 20-31
polynephricus polynephricus ad urethrae,
Cryptodrilus . 204, 208, 281, 286, 291-293,
323, Pis. 32-43
polynephricus urethrae, Cryptodrilus 204, 208,
280, 281, 286, 287, 288-291, 292, 323,
Pis. 44-59
polyodon, Chiloglanis . . 365, 366, 367
Polypterus . . . . . .26
Porphyrocrinus ..... 482
porrecta, Ophionereis .... 480
poutassou, Micromesistius . 154, 180-181
pritchardi, Megascolex . . . 217, 220
Procatopus . 331, 332, 333, 376-384, 385-387,
411, 412, 413
procatopus, Barbus .... 49
Proctophantastes . . 151, 167, 169-170, 172
prognathus, Barbus platystomus . . 48
Prosorhynchus . . 151, 154, 157-159
proteus, Acanthophiothrix . 46S-4&6, 467
proteus, Ophiothrix . . 466-466, 467
proteus, Placophiothrix . . • . . 466
Protozoa . . . . 185-199
Pseudocrenilabrus . . . . .389
Pseudocryptodi ilus . 204, 205, 212, 214,
296-299, 324, Pis. 61-63
pseudoexigua, Patiriella .... 438
Psilaster ...... 434
pulchellum, Ophiomisidium . . . 476
pulchellum, Ophiomusium . . .476
pumilus, Phoxaster .... 434
Puntius ...... 349
purpureus, Vesiculodrilus . . . 222
quadrispinus, Lophaster .
queenslandica, Rhododrilus
quinquemaculata, Ophiothrix
rachiaea, Lepodora
rachion, Distomum
rachion, Lepidapedon
radcliffii, Barbus .
radcliffii, Barbus altianalis
. 438
216
•~ . • 468
162
. 162
151, 154, 162-I&4
9-1 1, 23, 144
9-1 1, 13, 15, 17,
19-20, 22-27, 127, 129
ramosa, Stereomyxa . . . .194
rarispina, Marthasterias glacialis . . 439
rathkei, Epiplatys sexfasciatus . 367, 369
regani, Pelmatochromis .... 402
reticulata, Henricia . . . 438-439
reticulatus, Echinaster . . 438, 439, 483
rhinoceros, Barbus. .... 83
rhipidiophorus, Argulus .... 35
Rhizopodea . . . . - .187
Rhododrilus . 203, 204, 209, 212-2 16, 263, 297,
324, PI. 60
richardi, Diporochaeta .... 222
richardsoni, Benthotrema . . .178
richea, Diporochaeta .... 258
richea, Perichaeta . . . 2 05, 223, 258
richea, Perionychella . 203, 206, 258-259
roseipinnis, Procatopus . . . .380
rotunda, Euglypha . 187, 192, i95-J97
ruandae, Varicorhinus . . • '; 87, 89
ruasae, Barbus 3, 7-9, 27, 73, 109-1 n, 128,
rubra, Megascolex . .
rudolfianus, Barbus bynni
rueppelli, Barbus ..
rufoviride, Lecithochirium
rupestris, Coryphaenoides
.326
.27-28
48, 55
178-179
.169
INDEX
499
rupestris, Macrurus .
ruspolii, Barbus . .
rusticus, Cryptodrilus
ruwenzorii, Varicorhinus
169
27-28, 35
267, 269
9, "4
saccarius, Cryptodrilus .
sakaniae, Barbus chilotes
salmoides, Micropterus .
sanagaensis, Tilapia
santaisabellae, Aphyosemion
santa-isabellae, Aphyosemion
Sarotherodon
• 303
• 36
72
• 4°9
373- 374. 376
• 373
331. 332, 388, 389, 390, 391, 404,
405, 408-410, 411, 412, 413
scheeli, Aphyosemion . . . 373, 376
schoutdeni, Chromidotilapia . . . 397
sclateri, Antedon ..... 429
sclateri, Glyptometra . 428, 429-430, 482
sclateri, Pachylometra .... 429
scolecoidea, Diporochaeta . . .259
scolecoidea, Perichaeta . 205, 217, 223, 259
scolecoidea, Perionychella 203, 204, 206, 259
scolecoides, Perionychella . . . 222
Scomber ..... 154, 178
scombrus, Scomber . . . 154, 178
scutata, Amphilepis .... 464
sedecimalis, Diporochaeta . . .223
seguensis, Barbus ..... 347
seminuda, Ophiopsila . . 476-472, 474
semperi, Ophiopsammium . . . 469
sentus, Echinaster ..... 439
sepositus, Echinaster .... 439
serripinna, Oligometra . . . 423, 426
serripinna macrobrachius, Oligometra . 426
serripinna occidentalis, Oligometra . 424-426
sexfasciatus, Epiplatys .331, 867-370, 411, 413
sexfasciatus rathkei, Epiplatys . 367, 369
sexfasciatus togolensis, Epiplatys . . 369
sexthecatus, Oreoscolex 204, 210, 282, 303,
305, 315-317, 324, Pis. 83-86
sigillatus, Diporochaeta
sigillatus, Perionyx
Silurus .
silvestris, Hylopanchax
similis, Amphipholis
similis, Procatopus
221
221
• 367
. 382
450, 451, 461
33i. 332, 377. 378, 379,
Simsia .
simsoni, Amphiura
simsoni, Cryptodrilus
simsoni, Megascolides
simulans, Cryptodrilus .
Simulium
sladeni, Dipsacaster
sladeni capensis, Dipsacaster
Solasteridae .
somereni, Barbus
Sparus .
380, 381, 382, 383, 384-385, 386
218
448-449
204, 208, 269, 270,
273, 277, 293-296, 323
205
• 3°3
• 332
• 434
. 434
. 438
Spatagobrissus
spenceri, Diporochaeta .
squamifera, Ophiopsila .
stappersii, Barbus . 3,
stappersii, Varicorhinus .
steeli, Megascolex .
Steganoderma
Stelleroidea .
Stereomyxa .
Steringophoru s
Steringotrema
Stomatorhinus
strata, Amphipholis
striolata, Ophiacantha .
sublineatus, Barbus
superba, Hacelia .
superba capensis, Hacelia
surkis, Barbus 16, 47, 49,
susanae, Barbus
svenssoni, Barbus .
sylvaticus, Megascolex .
Synodontis .
. 480
• 223
• 472
116-1 19, 130, 139, 145
• 36-37, 123
326
151, 154, 167-170, 172
431-479
• 194
151, 154, i74-J78
151, 154, 173-174
• 338
450-452
. 442
352-355
• 436
• 435, 483
51-52, 55,69, 127-128
4. M3
• 356
326
26, 358
Tamaria
tanensis, Barbus
tanensis, Lanceabarbus .
tanganicae, Varicorhinus
tanjilensis, Cryptodrilus .
tanjilensis, Perichaeta
tapirus, Mormyrus .
Tasmaniaedrilus
tasmaniaensis, Tasmaniaedrilus
tasmanianus, Megascolides
tasmanianus, Notoscolex
tasmanianus, Pinguidrilus
tasmanianus, Plutellus
tasmanica, Perichaeta
tasmanicus, Megascolex
• 436
50, 72, 94-96, 100
94
126
222
22O
331, 334-335, 413, PI. i
• 205
• 205
263
204, 207, 263
203, 268-265,
294, 324
. 263
205, 324
204, 205, 214, 319, 320,
3. 9, 92-94. HI-US,
130-131, 133, 140, 145
. 408
324-326
Telmatodrilus ..... 205
telopea, Diporochaeta .... 223
tenuibrachia, Comanthus wahlbergi . .481
tenuis, Cryptodrilus .... 269
terrareginae, Perionyx . . . .221
terricola, Trinema lineare . . .189
tesselatus, Cryptodrilus . . 205, 219, 266
tesselatus, Woodwardiella . . .266
tessellatus, Plutellus .... 266
tessellatus, Woodwardiella . 203, 214, 266
Testacea ..... 185-199
Tethyaster ...... 423
tetrastigma, Barbus .... 355
Thalassometridae .... 427-429
tholloni, Tilapia . . . 389, 404, 405
thysi, Barbus . 331, 350-356, 410, 411, 413
Tilapia. 72, 78, 134, 331, 332, 388, 389, 391,
402-408, 409, 411, 412, 413, PI. 4
timida, Ophiopsila ..... 475
500
INDEX
tisdalli, Megascolides .... 222
togolensis, Epiplatys sexfasciatus . . 369
Tor . -134
trachypterus, Barbus . 3, 40-41, 106, 116,
119-123,140, 145
Trachyrhynchus . 154, 160-161, 164-166,
167-168, 180-181
trachyrincus, Trachyrhynchus. 154, 160-161,
164-166, 167-168, 180-181
Trematoda 149-183
Trewavas, E. 329-420, 5 Pis.
tricirratus, Onos . . . . .162
trifolium, Amphiophiura . . 423, 476
triglochis, Ophiothrix . . 468, 469, 483
trimeni, Ophiura . . . 475-476, 483
Trinema . . . 185-199, Pis. 1-7
Trinephrus .... 269, 280, 284
tripapillatus, Groliophilus . 203, 214, 260,
261-262, 324, PI. i
trispilos, Barbus . . 332, 350, 352-355
trispinosus, Ceramaster . . . .423
Tritogenia . . . . . .301
tropidolepis, Barbus 3, 44, 108-109, 123-126,
127, 129, 139, 145
Tropiometra .... 423, 426-427
Tropiometridae .... 426-427
tunnackensis, Perionychella . 203, 206, 221,
225, 230, 251, 252, 253, 257
tunnackensis, Vesiculodrilus . 203, 206, 221,
225, 230, 251, 252, 253, 257
Tylochromis . . . . . . 391
udekemianus, Limnodrilus
uncinatus, Plutellus
undulatum, Lithodesium
Unioplus
unitaeniatus, Nannaethiops
• 205
222
. 194
455-456
411
urethrae, Cryptodrilus polynephricus 204, 208,
280, 281, 286, 287,
288-291, 292, 323, Pis. 44-59
urethrae, Cryptodrilus polynephricus poly-
nephricus ad . 204, 208, 281, 286, 291-293,
323, Pis. 32-43
urundensis, Barbus . . . in
urundensis, Barbus alt ianalis . . in, 113
variegatus, Lumbriculus .... 205
vatovae, Barbus platystomus . . 48, 65
veliporum, Otodistomum . . 154, 167
verrucosum, Euryale .... 440
verrucosus, Astrocladus .... 440
verrucosus, Gorgonocephalus . . . 440
versluysii, Barbus ..... 344
Vesiculodrilus 203, 204, 206, 212, 216, 217, 218,
219, 221-223, 225, 230, 231, 232, 233-253,
254, 256, 257, 258, 263, 286, Pis. 93-102
vexillifer, Aphyosemion . . . 374, 375
vexillifer, Bathybiaster .... 434
vexillifer, Panchax . . 373, 374, 375
victoriae, Barbus . . . . .119
victoriae, Cryptodrilus . . . .222
vivipara capensis, Ophiomyxa. . . 482
volpinii, Barbus . . . . 49, 52
voltae, Barbus ..... 356
volvens, Megascolides . . . .222
vulgaris, Gaidropsarus . . . 154, 159
wahlbergi, Comanthus .... 480
wahlbergi, Ophioderma .... 423
wahlbergi tenuibrachia, Comanthus . .481
walhallae, Perichaeta . . . .223
walkeri, Clarias . . 331, 356, 411, 413
waterloti, Chiloglanis niloticus . 361, 363
weldboroughi, Perionychella 203, 206, 220,
223, 230-232, 256, 260
wellingtonensis, Cryptodrilus . . 205, 317
wellingtonensis, Notoscolex . . -317
wellingtonensis, Oreoscolex 204, 210, 303, 305,
316, 317-318, 324, Pis. 87-90
willsiensis, Cryptodrilus . . . .222
Woloszynskia . . . . .194
woodi, Graliophilus .... 260
Woodwardia . . . . 217, 219, 265
Woodwardiella . 203, 204, 209, 212, 214,
217, 219, 247, 263, 265-266
Xenodermichthys .
yunnanensis, Barbus
154. 174-176
valida, Anamphiura . . 423, 478-479 zaphiri, Barbus
Varicorhinus . 4, 9, 36-37, 49, 87, 89, 94, 114, zillii, Tilapia .
119-120, 123, 124, 126, 135 Zoogonidae .
varicus, Derogenes. . . 154, 180, 181 zuaicus, Barbus
49
403, 404, 405
167-173
49
A REVISION OF THE LARGE
BARBUS (PISCES, CYPRINIDAE)
EAST AND CENTRAL AFRICA
K. E. BANISTER
BULLETIN OF
THE BRITISH MUSEUM (NATURAL HISTORY)
ZOOLOGY Vol. 26 No. i
LONDON: 1973
A REVISION OF THE LARGE BARBUsL*
(PISCES, CYPRINIDAE) OF EAST AND CENTRAL
AFRICA '
STUDIES ON AFRICAN CYPRINIDAE
PART II
JC
KEITH EDWARD BANISTER/^
Pp. 1-148; 105 Text-figures
BULLETIN OF
THE BRITISH MUSEUM (NATURAL HISTORY)
ZOOLOGY Vol. 26 No. i
LONDON: 1973
THE BULLETIN OF THE BRITISH MUSEUM
(NATURAL HISTORY), instituted in 1949, is
issued in five series corresponding to the Departments
of the Museum, and an Historical series.
Parts will appear at irregular intervals as they
become ready. Volumes will contain about three or
four hundred pages, and will not necessarily be
completed within one calendar year.
In 1965 a separate supplementary series of longer
papers was instituted, numbered serially for each
Department.
This paper is Vol. 26 No. i of the Zoological series.
The abbreviated titles of periodicals cited follow those
of the World List of Scientific Periodicals.
World List abbreviation :
Bull. Br. Mus. nat. Hist. (Zool.)
Trustees of the British Museum (Natural History), 1973
TRUSTEES OF
THE BRITISH MUSEUM (NATURAL HISTORY)
Issued i November, 1973 Price £6-95
A REVISION OF THE LARGE BARBUS
(PISCES, CYPRINIDAE) OF EAST AND CENTRAL
AFRICA
STUDIES ON AFRICAN CYPRINIDAE
PART II
By KEITH EDWARD BANISTER
CONTENTS
SYNOPSIS ..........
INTRODUCTION .........
Notes on counts and measurements .....
Barbus acuticeps .........
Barbus alluaudi .........
Barbus altianalis .........
Barbus bynni ..........
Barbus caudovittatus .........
Barbus ethiopicus .........
Barbus gananensis .........
Barbus intermedius .........
Barbus longifilis .........
Barbus macrolepis .........
Barbus mariae .........
Barbus microbarbis .........
Barbus microterolepis ........
Barbus mirabilis .........
Barbus oxyrhynchus .........
Barbus pagenstecheri .........
Barbus paucisquamatus ........
Barbus platyrhinus .........
Barbus ruasae ..........
Barbus somereni .........
Barbus stappersii .........
Barbus trachypterus .........
Barbus tropidolepis .........
DISCUSSION ..........
ACKNOWLEDGEMENTS ........
REFERENCES ..........
APPENDIX i An artificial key to the species described in this paper
APPENDIX 2 Gazetteer ........
APPENDIX 3 A replacement name for Barbus gregorii Norman
APPENDIX 4 Study material .......
INDEX
4
4
5
6
9
9
27
36
41
44
47
76
79
83
87
89
9i
94
101
103
1 06
109
in
116
119
123
127
135
135
139
141
143
143
145
4 K. E. BANISTER
SYNOPSIS
, il the previously described species and subspecies of the large Barbus (the species usually
described as possessing parallel striae on their scales) from east and central Africa are examined.
The area under consideration is bounded in the north by the mouth of the Nile and in the
south by the Zambezi system. The western boundary is the upper reaches of the Congo system
adjacent to the rift valley and the eastern boundary is the coast.
It is shown that many nominal species can be synonymized when a sufficiently large series
of specimens is available to show that the characters formerly used to distinguish them form
a continuous series within a polytypic species. Consideration is given to the taxonomic value
of these characters and their variability is demonstrated. Particularly variable are the lips,
body depth and head length. This is a reflection both of eco-phenotypic factors and of
allometry.
The quaternary palaeogeography of this part of Africa is considered in an attempt to
elucidate the present distribution of certain species.
Two former subspecies are raised to specific rank (Barbus longifilis and Barbus paucisquamatus)
and one new subspecies is described (Barbus intermedius australis). Two replacement names
are included ; Barbus lapsus for Barbus babaulti Pellegrin 1935 and Barbus susanae for Barbus
gregorii Norman 1923.
INTRODUCTION
THIS revision was initiated by an unsuccessful attempt to identify satisfactorily the
large Barbus species collected by the Sandhurst Army College Ethiopian Expeditions
of 1964 and 1966. The use of the keys in Boulenger's Catalogue of African Fresh-
water Fishes (1911-1916) showed that some species were so imprecisely denned that
an individual specimen could be assigned to any one of several different species.
Many of the species described by early workers on African Barbus species were
based on one or a few poorly preserved individuals, and it is only now, when
sufficient material is available, that an attempt can be made to redefine some
species allowing for growth and natural variation.
Worthington (i932a) demonstrated that lip shape, formerly considered a signifi-
cant specific character in the large Barbus species, was extremely variable. This
aspect was greatly extended by Groenewald (1958) who was able to demonstrate
the great range in Barbus species lip form engendered by the environment in certain
species. In the course of this work he cast doubts upon the validity of the inclusion
of certain species in the closely related genus Varicorhinus.
Jubb (1961, 1963, 1965, I967a, 1968), Crass (1960) and Jackson (1961) have
continued this study and have partially confirmed Worthington's (i932a) views on
the variability of the large Barbus species. Barbus marequensis A. Smith has been
shown by the South African workers' efforts to be a widespread and variable species.
An analogous situation exists in eastern and central Africa. Many large Barbus
species have been described, but as more specimens were collected the boundaries
between the species became obscured as new data rendered the old specific defi-
nitions untenable. New descriptions of species must, therefore, incorporate the
range of intraspecific variation if the definition of the species is to have any validity
and reflect the status of the populations in the field.
Several authors, e.g. Worthington (i932a), Pellegrin (1935) and Bini (1940), have
described Barbus subspecies and varieties. The subspecies considered in this paper
A REVISION OF THE LARGE AFRICAN BARBUS 5
conform to the definition of Mayr (1949) although it has not always been possible
to conform to the '75 per cent distinguishable' convention suggested by Mayr.
Geographical isolation, on its own, without any supporting morphological or colour
differences is not considered sufficient justification for the establishment of sub-
species. I would like to have used Mayr's definition of superspecies. I am unable
to do so as I cannot prove that the species I consider to form such taxa would
constitute a monophyletic assemblage and hence I use the terms supra-specific
complexes or groups. This does not assume monophyly although hopefully one
day it may be proved.
There are still many problems that I have been unable to consider in this work ;
ecological information is only rarely available (and the effect of the environment
could well be an important factor influencing the phenotypic variation) and colour
patterns are based almost entirely on preserved specimens.
This revision remains, therefore, essentially that of a museum worker, but it is
very possible that further information from live specimens may lead to the estab-
lishment of subspecific taxa especially in the widespread species.
There are some geographical areas that have presented many difficulties, par-
ticularly where watersheds are close together. By unfortunate quirks of geography
or international politics these areas are inadequately sampled and some of my
subsequent conclusions about the status of the specimens must be regarded as
tentative. These problems are mentioned in the text. A lot of time was spent in
trying to pin-point the localities where specimens had been collected. The localities
were often given the name of the local village, but these villages were frequently
only temporary settlements of nomadic tribes. Many town and village names were
changed when countries were occupied or invaded and collections made at the same
site under different regimes bear different locality names. Where such localities
have been traced they are incorporated in the gazetteer (Appendix 2) and as much
information as I can give, in physiographical details or in grid references, is included.
The limits of distribution of the species described in this paper are from the Nile
in the north down to, but not including, the Zambezi system. The eastern limit
is the eastern coast of Africa and the western limit is the western rift valley. Some
species whose range extends to the west of the rift valley, into the Upper Congo are
included, e.g. the Upper Congo species Barbus mirabilis is included because of its
relationship with the species that live in the area covered.
The arrangement of species in the descriptive part of the text is alphabetical.
Notes on counts and measurements
The standard length (S.L.) was taken in the usual manner (see Banister 1972).
The lateral line count (LL) was taken from the first pore-bearing scale behind the
head to the scale lying lateral to the end of the hypurals. The body depth (D) is
the maximum body depth, usually to be found just in front of the dorsal fin. The
anterior limit for the head length (H) and snout length (Snt) was the premaxillary
symphysis, with the premaxillae retracted. In 'rubber-lipped' fishes this measure-
ment was only taken with difficulty. The posterior limit for the snout is the
6 K. E. BANISTER
anterior margin of the orbit. The term mouth width (MW) is a shorthand notation
for the width of the lower jaw at its widest point. The pectoral fin length (Pet.)
is the total length of the fin, measured in a straight line from the base of the first
ray to the distal extremity of the fin. The measurement is taken in this manner
because of the ease of so doing with dial calipers, which were used on all fish except
the smallest when dividers were used. The caudal peduncle length (CP1) is the
horizontal distance from the posterior angle of the base of the last anal fin ray to the
end of the hypurals, and the caudal peduncle depth (CPd) is the least depth of that
part. The interorbital width (10) was measured as the least distance apart of the
bony edges of the interorbital space. The eye diameter (I) had to be taken as the
horizontal diameter of the orbit ignoring the skin around the eye. This was because
in many of the more ancient specimens the skin had markedly shrivelled away from
the eye and the measurements would not otherwise have been comparable with those
of recently preserved fish. The dorsal spine (DSp), strictly the last unbranched ray
in the dorsal fin, was measured from its base to the proximal articulation (if any
were present). The reason for this was to overcome the inaccuracy caused by
damage to the flexible tips and, although an arbitrary point, experience showed it
to be moderately consistent. Severely damaged spines were not measured. With
any measurements which were repeatable on both sides of the fish (e.g. anterior
barbel, Ab, or posterior barbel, Pb), the larger was taken except in a few cases when
gross deformity was obvious.
The majority of measurements were taken to the nearest millimetre. The
exceptions were some measurements made on small fishes ; these were taken to the
nearest half millimetre. The limits of accuracy do not justify calculations taken to
more than one place of decimals. The mean is symbolised by x, the standard
deviation by s.d., the standard error by s.e. and the mode by m. The height of a
pharyngeal tooth is taken as the parameter at 90 degrees to the plane of the pharyn-
geal bone, the length of the pharyngeal tooth is the parameter of a transverse
section of the tooth in the plane of the anterior edentulous process and the width of
the tooth is the parameter at 90 degrees to the length. The code letters for the
museums from whose collections the specimens came are listed at the start of
Appendix 4.
All the drawings of the pharyngeal bones show i) the bone directly from above
with the bone lying flat on its edentulous surface and 2) an occlusal view of the
inner row of teeth with the bone at 90 degrees to the position in i), i.e. drawn from
above with the bone mounted vertically on its posterior edentulous process. For
convenience these are referred to in the text and figure captions as, respectively, the
dorsal and lateral views.
Barbus acuticeps Matthes 1959
Barbus acuticeps Matthes, 1959, Folia scient. Afr. cent. 5 (3) : 62 ; Matthes, 1962, Annls Mus. r.
Afr. cent. Ser. 8vo 111 (2) : 81, pi. i, fig. b.
HOLOTYPE. A fish of 210 mm S.L. from the Nyawarongo river, Lake Rugwero,
Rwanda. Specimen no. M.A.C.T. 130313.
A REVISION OF THE LARGE AFRICAN BARBUS
FIG. i. Barbus acuticeps from Matthes 1962.
DESCRIPTION. The description is based on nine specimens, the holotype, four
paratypes (130310-12, 130314) of S.L. 43 mm, 121 mm, 92 mm and 202 mm
respectively, and four other specimens M.A.C.T. 12910 (S.L. 131 mm), 13045-6
(119 and 116 mm S.L.) and 172421 (242 mm S.L.).
My standard length measurements are consistently shorter than those of Matthes
(1962). I attribute this to different concepts of standard length.
L
D
9
H
9
I
9
10
9
MW
8
Pet
8
CP1
9
CPd
9
Snt
8
Ab
8
Pb
8
27-8
28-7
6-3
7-1
5-8
21*7
17-6
n-4
8-3
4-8
5-6
s.d.
4'i
3'5
i'3
1-6
i'3
2-O
1-7
i'5
1-2
2-1
s.e.
0-7
0-6
o-5
0-4
0-8
0-8
range
43 -242 mm
23-6-33-7
24-8-35-6
4-3- 9-3
4-7-10-4
4-5- 7'9
19-5-25-8
15-2-20-6
9-3-I3-7
6-4-10-4
2-2- 6-9
2-9- 8-6
All measurements are expressed as percentages of the standard length.
The dorsal profile of the head is concave and a pronounced nuchal hump is present.
The snout is pointed and the mouth slopes upwards. Otherwise the body and head
shape resembles that of Barbus altianalis (sensu lato) and Barbus ruasae.
Squamation. It was not possible to obtain scale counts for the two smallest
specimens. The scales have parallel striae and there are 27 (1.3), 28 (f.i), 29 (f.2)
or 30 (f.i) scales in the lateral line. Between the dorsal mid-line and the lateral
line there are 4-5 (f.5) or 5-5 (f.2) scale rows. In only two specimens was it possible
to count the number of scale rows between the lateral line and the ventral mid-line,
8 K. E. BANISTER
in both there are 5-5 rows. Between the lateral line and the base of the pelvic fin
there are 2 (f.2), 2-5 (f.2) or 3 (f.i) scale rows. There are 12 scales encircling the
caudal peduncle.
Dorsal fin. There are 8 (f .5) or 9 (£.3) branched rays. The fourth unbranched
ray is ossified into a smooth, straight spine (x = 22-3 ; s.d. = 2-9 ; s.e. = i-o ;
range 17-4-267). There is no sheath of scales at the base of the dorsal fin. The
dorsal fin origin is slightly behind that of the pelvic fins. The anal fin has three
simple rays and five branched rays.
Gill rakers. In five specimens examined, three had eight gill rakers on the lower
limb of the first gill arch, the other two had ten and eleven.
Pharyngeal bones and teeth. I have not been able to study the pharyngeal bones
but they were described by Matthes (1962) as ' Falciformes, non soudes, attaches
1'un a 1'autre par une symphyse ligamenteuse. Us sont peu elargis et les branches
montantes sont comprimees et assez greles. Dents pharyngiennes obliquement
tronquees et legerement excavees, avec une pointe anterieur recourbee vers 1'arriere,
en 3 rangees au nombre de 2, 3, 5-5, 3, 2, les 2me et 3me dents de la rangee interne
un peu plus fortes. '
Coloration. Described by Matthes (1962) as completely silvery with pearly
reflections when alive. The operculum is yellowish with metallic glints, the eye is
yellowish, the pupil ringed with gold. The fins are clear, yellowish ; the dorsal
and caudal are greyish, yellowish and salmon-pink.
DISTRIBUTION. Four of the specimens came from the Nyawarongo river (to the
north of Lake Rugwero), the fifth specimen in the type series came from the falls of
Rusumu on the Kagera (Rwanda). The other specimens came from Lake Ihema,
Kagera ; the Nyabugogo river (an affluent of the Lusine river, Burundi) and from
Matale (? Burundi).
DIAGNOSIS AND AFFINITIES. >Barbus acuticeps is dubiously distinct from Barbus
altianalis. The principal differences between Barbus acuticeps and Barbus altianalis
are i) the concave dorsal profile of the head and the nuchal hump of the former
species, 2) the antero-dorsal gape of the mouth of the former species which contrasts
with the, usually, subterminal mouth in Barbus altianalis, 3) fewer gill rakers in
Barbus acuticeps (8-n) compared with 10-14 (most frequently 11-13) m Barbus
altianalis, 4) slightly fewer scales in the lateral line series in Barbus acuticeps (27-30)
compared with 28-36 (most frequently 30-34) in Barbus altianalis, 5) a longer
dorsal fin spine in Barbus acuticeps (x = 22-3, range 17-4-26-7) compared with
x = 16-9, range 9-0-30-4, in Barbus altianalis.
It must be remembered that Barbus acuticeps is known from very few specimens
and that more are necessary to confirm whether or not its continued separation from
Barbus altianalis is justified.
Barbus ruasae (p. 109) also from Rwanda to some extent resembles Barbus
acuticeps in general appearance but is readily separable on the lower number of
scales in the lateral line series [25 (f-9), 26 (f.i)]. The pharyngeal teeth of Barbus
acuticeps are less molariform than those of Barbus ruasae, but this could be the
A REVISION OF THE LARGE AFRICAN BARBUS g
result of diet. Barbus ruasae does not have a mouth opening antero-dorsally.
Barbus acuticeps is considered here to belong to the Barbus intermedium group (see
p. 128).
Barbus alluaudi Pellegrin, 1909
Barbus alluaudi Pellegrin, 1909, Bull. Soc. zool. Fr. 34 : 155 ; Pellegrin, 1910, Mem. Soc. zool.
Fr. 22 : 287, pi. 14, fig. 2 ; Banister, 1972, Bull. BY. Mus. nat. Hist. (Zool.) 24 (5) : 261-290.
1cm
FIG. 2. Barbus alluaudi Holotype.
This specific name has, I believe, been given to hybrids between Barbus somereni
and Varicorhinus ruwenzorii [Banister (1972)].
Barbus altianalis Boulenger, 1900
Barbus altianalis Blgr., 1900, Ann. Mag. nat. Hist. (7) 6 : 479 ; Blgr., 1901, Trans, zool. Soc.
London, 16 : 144, pi. 13, fig. i ; Blgr., 1911, Cat. Afr. Fish 2 : 36 ; Norman, 1925, Occ. Pap.
Boston Soc. nat. Hist. 5 : 189.
Barbus altianalis altianalis : Worthington, 1932, /. Linn. Soc. (Zool.) 38 (258) : 124-127 (Lake
Kivu and Ruzizi river specimens) ; Poll, 1953, Result, scient. Explor. hybrobiol. Lac Tanganika
(3) 5A : 88.
Barbus altianalis radcliffii : Worthington, 1932, ibid. : 124-127 (Lake Victoria specimens).
Barbus altianalis eduardianus : Worthington, 1932, ibid. : 124-127 (Lakes Edward and George
specimens) .
Barbus altianalis var. labiosa Pellegrin, 1935, Revue Zool. Bot. afr. 28 (3) : 376-385 (only the
holotype, from Lake Kivu).
Barbus (Labeobarbus) nedgia : Hilgendorf, 1888, Sber. Ges. naturf. Freunde Berl. : 88 (from
Lake Victoria, misidentification, not Barbus nedgia Riippell ; fide Blgr., 1911).
Barbus eduardianus Blgr., 1901, Ann. Mag. nat. Hist. (7) 8 : 12 ; Blgr., 1907, Fish Nile : 230,
pi. 43, fig. 2.
Barbus fergussonii Blgr., 1901, Ann. Mag. nat. Hist. (7) 8:12; Blgr., 1907, Fish Nile : 230,
pi. 43, fig. i.
Barbus radcliffii Blgr., 1903, Ann. Mag. nat. Hist. (7) 12 : 218 ; Blgr., 1907, Fish Nile : 212,
pi. 37, fig. i.
io K. E. BANISTER
Barbus marequensis : Pellegrin, 1904, Mem. Soc. zool. Fr. 17 : 17 (misidentification, not Barbus
marequensis Smith, specimen from Kavirondo, Lake Victoria, fide Blgr., 1911).
Barbus lobogenys Blgr., 1906, Ann. Mag. nat. Hist. (7) 17 : 435 ; Blgr., 1907, Fish Nile : 210,
pi. 36, fig. i.
Barbus bayoni Blgr., 1911, Annali Mus. civ. Stor. nat. Giacomo Doria (3) 5 : 77.
Barbus kivuensis Pappenheim, 1914, Wiss. Ergebn. dt. ZentAfr. Exped. Zool. (3) 5 : 237.
Barbus gregorii : Blgr., 1916, Cat. Afr. Fish 4 : 230 (only the specimens from the Malawa river).
Barbus pietschmanni Lohberger, 1929, Anz. Akad. Wiss. Wien 66 : 92-94.
Barbus hollyi Lohberger, 1929, Anz. Akad. Wiss. Wien 66 : 92-94.
Barbus obesus Worthington, 1929, Proc. zool. Soc. Lond. (3) : 433.
Barbus kiogae Worthington, 1929, Proc. zool. Soc. Lond. (3) : 434.
Barbus longirostris Worthington, 1929, Proc. zool. Soc. Lond. (3) : 435.
NOTES ON THE SYNONYMY. Barbus altianalis is a widespread species occupying
three isolated areas. The fish of each of these areas, although in morphometric
characters scarcely different modally from one another, were given subspecific
status by Worthington (iQ32a).
Barbus altianalis was first described from two fish from Lake Kivu and one from
the Ruzizi river. All are large specimens ; the smallest (from the Ruzizi) has a
S.L. of 306 mm while the other two are 331 and 406 mm.
Barbus eduardianus was described from one large fish from Lake Edward and
Barbus fergussonii from two small fishes from the same lake. The latter two
specimens have unusually shallow bodies when compared with equal-sized specimens
even from the same waters.
Barbus radcliffii was based on a skin from Lake Victoria. Barbus lobogenys was
based on two 'rubber-lipped' specimens from Lake Victoria. Barbus bayoni was
described from a very large fish (S.L. 640 mm) from near Jinja on Lake Victoria.
Barbus pietschmanni and Barbus hollyi, both from Lake Victoria, were based
respectively on one and two specimens.
The holotype and sole example of Barbus obesus from Lake Kioga (fig. 3) is a fat,
deep-bodied fish with a low number of lateral line scales (28). The scales are
slightly lobed and the specimen gives the impression of a rather deformed individual.
The two syntypes of Barbus longirostris (fig. 4) from the same locality are large
specimens (388 and 410 mm S.L.) with shallow bodies and heads longer than the
mean for the species. They are linked with more typical Barbus altianalis by
members of the type series of Barbus kiogae. No morphometric or meristic way
can be found for separating the populations from Lake Kioga from those of Lake
Victoria. Large fish from Lake Victoria not infrequently approximate to the facies
of Barbus longirostris and Barbus kiogae (P. H. Greenwood : pers. comm.). The
description of the holotype of Barbus kivuensis differs hardly at all from that of
Barbus altianalis from the same lake.
Norman (1925) commented that the specimens of Barbus altianalis from Lake
Edward, collected by Dr John Phillips, were identical with Barbus radcliffii from
Lake Victoria.
Worthington (iQ32a) was able to gather together sufficient material to realize that
a single species is found in the three areas, Barbus altianalis altianalis from Lake
Kivu and the Ruzizi river ; Barbus altianalis radcliffii from Lake Victoria and the
A REVISION OF THE LARGE AFRICAN BARBUS
ii
FIG. 3. Barbus altianalis (Holotype of Barbus obesus re-drawn from Worthington 1929).
Victoria Nile ; and Barbus altianalis eduardianus from Lakes Edward and George
and the Kazinga channel.
Boulenger's Barbus raddiffii, Barbus lobogenys and Barbus bayoni were synony-
mized by Worthington (i932a) in Barbus altianalis raddiffii, and Barbus eduardianus
and Barbus fergussonii in Barbus altianalis eduardianus. An important conclusion
of Worthington's work was the realization that Barbus species can exhibit remark-
able intraspecific variation, particularly with regard to the development of the lips
and he suggested that further extensive synonymizing might be necessary when
larger samples became available. However, he did not synonymize the species he
described from Lake Kioga because with the material at his disposal they remained
apparently distinct. Scatter diagrams comparing body depth and head length of the
Lake Victoria and Lake Kioga populations are shown below (figs. 6 and 7). Pel-
legrin (1933, 1935) described several varieties of Barbus altianalis from Kivu region :
only one of these is considered here as belonging to this species (for the others see
under Barbus caudovittatus, Barbus paucisquamatus and Barbus longifilis] . A slight
element of confusion was introduced by Pellegrin's use of 'Region du Kivu' ; this
is not always the Lake Kivu basin but in some cases refers to sites in the Congo
system.
Greenwood (1966) synonymized Barbus pietschmanni and Barbus holly i with
Barbus altianalis raddiffii.
I am retaining Worthington's three subspecies, certainly on geographical grounds
but also because of the consistent colour differences between the Barbus subspecies
of Lake Victoria and those of Lake Edward-George. Dr P. H. Greenwood (pers.
comm.) has noted that specimens of Barbus altianalis eduardianus of over 100 mm
S.L. from Lake Edward-George have pink caudal fins whereas those of Lake Victoria
never do. I have no information about the specimens from Lake Kivu.
12
K. E. BANISTER
FIG. 4. Barbus altianalis (figured specimen of Barbus longirostris,
re-drawn from Worthington 1929).
FIG. 5. Barbus altianalis altianalis (Holotype from Boulenger 191 la).
LECTOTYPE. A fish of 331 mm S.L. from Lake Kivu (one of the three in
Boulenger's type series). B.M. (N.H.) Reg. No. 1906.9.6 : 13.
DESCRIPTION. The description is based on a total of 213 specimens ; 15 of these
are from Lake Kivu and the Ruzizi river (size range 46-410 mm S.L.), 130 are from
Lakes Edward and George, the Kazinga channel and Lake George feeder streams
(size range 68-436 mm) and 68 are from Lakes Victoria and Kioga (size range
99-640 mm). The combined morphometric data for all three populations are as
follows, and include all the holotypes except for Barbus pietschmanni and Barbus
hollyi.
A REVISION OF THE LARGE AFRICAN BARBUS
CD
CD
) *
1^
O
_,
o
"•o
f-H
O
1/5
H-H
»w
•
'•»
_^ •
.0
£
8
"bo
g
.0
to
i— i
CN
CD
-2
I "
-§
i
.p
00
0
0
'oo
d
•
i\
cS
• •
§
en •
CD
o
'•0
en
A
-C
*
p
+* •** •
(3
i • *
"*
-(-> sSb
'•*
H-3 S^>
p
CN
en *«?*
11
o
"CN
en ^
'3 •§
_8
^ ^
-8
bo g
Co
rC ^
iO
rC '2
p
00
ll
.0
00
rH S
8.5
*
•<>
T3 ^ *
p
'"H *^
*
p
11 • •
CO Q
.0
O "§
•£
-4J iVN
.0
i i ^3
en '•M
CN
en -C
_8
r^ °
-1
CD "*H
rd O
\ .
en
<4-l S3 * •*
p en
§
JO
oo
0 .0
p
'oo
o|
o "«J
CD -P
«.
8
43 "3 . *
o
"o
bo ™
CO 3
^J
£ JT^
p-l JI_
£>
« R,
y A
.0
8 a
IH
!_,
'
.p
CN
CD CO
J.2
0
"CN
CD CO
OH bO
CO -2
*
_8
en ^^
0
en "
»
CO
d T3 *
ft
*. . *
fi
J2 C3 •
-t-> CO •
be • •
p
co
£ co
C .5
o ^5
. t
p
o
H O •
O
o
'd 'C
>^-§
• 1*
o
^ IT1 * . * *
p
T3 .^
..
•^
CD J>.
O ^>
" ". "•"
p
N
en •
CD CD • . • %
p
CN
^ "
• v
.
"*~* rt
*
O
O
1
* *
~CN
"o J *
-a
*s
*••
0
S ' * "
p
g
00
oo
CO
u
*
•
•o
.2
P
O
bo
.2
p
M
p
IH
.
"*
CD
4-> •
"*
CD
. .
o
-t->
CO *
o
"co
a«
CN
O .
CN
o
in
Z
-1
— 1 • •
-8
i
o ">• •
^
_
o
CN i CO ' -O ' 4 " CN°°
CJ ob |ts^>u!>4r!)<N'i ickoorsyiiA^fO00
o
<"> CO CN CN ^ *"*
£ co n m o o co
fe
14 K. E. BANISTER
x s.d. s.e. range
L 46 -640 mm
D 29-4 2-4 0-4 22-4-37-4
H 26-1 1-6 o-i 22-5-31-4
I 7-6 i-o o-i 4-1-11-5
IO 8-4 i-o 0-7 6-4-11-2
MW 6-2 0-9 o-i 4-7- 9-5
Pet 21-8 1-4 o-i 17-5-25-0
CP1 16-9 1-4 o-i 13-1-19-8
CPd 12-0 0-9 o-i 9-5-16-2
Snt 8-3 0-8 o-i 5-8-10-9
Ab 4-3 i-i o-i 2-1-11-6
Pb 5-4 1-3 o-i 2-3-12-4
The body shape is extremely variable, especially in large fish. This is more
noticeable in Lake Kioga than elsewhere and a graph showing the variation in body
depth to length can be seen in fig. 7. It is possible that the deeper-bodied specimens
are females but most of the specimens that I have seen had been gutted (see below).
The mouth is sub-terminal and usually in the shape of a narrow horse-shoe.
'Rubber-lipped' forms have been found (e.g. the holotype of Barbus lobogenys - see
fig. 8). The snout tends to overhang the mouth in fishes of medium size.
Dorsal fin. The dorsal fin has four unbranched rays. The last unbranched ray
is enlarged into a smooth, straight or slightly curved, well-ossified spine with articu-
lations only at the tip. The dorsal fin spine is negatively allometric ; x = 16-9,
s.d. = 3-2, s.e. = 0-2, range = 9-0-30-4 (percentage of the S.L.) for the whole
sample (fig. 9). This range is too large to be useful as a diagnostic character but it
can be seen from fig. 9 that there are three convenient size ranges of fishes which can
be treated separately : 0-170 mm, x = 19-2, s.d. = 3-0, s.e. = 0-3, range = 13-7-
30-4 ; 171-250 mm S.L., x = 16-8, s.d. = 1-7, s.e. = 0-2, range = 12-8-20-6 ;
above 251 mm S.L., x = 13-4, s.d. = 2-3, s.e. = 0-3, range = 9-9-20-2. It is
interesting to note that these three S.L. ranges are the same for Barbus intermedius
(see p. 68).
The branched rays number 8 (1.30), 9 (f.i77) or 10 (f.6). The relative propor-
tions of fishes with 8, 9 or 10 rays does not differ detectably in the three subspecies.
The dorsal fin origin is situated in front (rare) or more commonly directly above or
just behind the vertical from the origin of the pelvic fins.
The anal fin has three simple rays and five branched rays ; the unbranched rays,
although thickened, are much more flexible than those in the dorsal fin.
Pharyngeal bones and teeth. The pharyngeal teeth almost always number
2.3.5. ~ 5-3-2- The one exception to this is a specimen from Lake George, S.L.
208 mm B.M. (N.H.) Reg. No. 1971.8.11 : 33-39 which is abnormal and the teeth
number 1.1.2.3.5. ~ 5-3-2.1.
The pharyngeal bones may vary in thickness in fishes of the same size. The
variability in the length and thickness of the bones and the teeth is a striking feature
of this species and Barbus intermedius. Both species have remarkably variable
body forms and there appears to be a loose correlation between body depth and the
A REVISION OF THE LARGE AFRICAN BARBUS
FIG. 8. Barbus altianalis radcliffii (Type of Barbus lobogenys from Boulenger 191 la).
..•• v
~ JV
• ••»
40 8° 100 2° 40 60 80 200 20 40 60 80 300 20 40 60 80 400 20 4° 60 80 500 20 40 60 80 600 20 *° 6°
FIG. 9. Scatter diagram of the dorsal spine length as a percentage of the standard length
against the standard length for Barbus altianalis. (= Barbus altianalis altianalis,
Barbus altianalis radcliffii, Barbus altianalis eduardianus) .
16
K. E. BANISTER
FIG. 10. Dorsal and lateral views of the left pharyngeal bone of the holotype of
Barbus altianalis altianalis.
tip-to-tip length of the pharyngeal bone. The very deep-bodied specimen of Barbus
altianalis (i.e. Barbus obesus) and the deep-bodied Barbus intermedius (i.e. Barbus
surkis] both have pharyngeal bones that are very much longer and stouter than
those of less deep-bodied fishes of the same standard length.
A series of ten fishes from Lake Victoria and ten from Lakes Edward and George
were examined to see what relationship was present between the length or depth of
the body and the size of the pharyngeal bone. The results from this small trial
sample must be used with considerable restraint but using the coefficient of variation
technique there are indications that the length of the pharyngeal bone is closely
related to the depth of the body. I do not want to pursue this matter any further
or in any detail but simply mention the possibility that this correlation is worth
further study.
The pharyngeal bone of a long-headed specimen (Barbus longirostris] is shown in
fig. 12.
Diet presumably must have an effect upon the stoutness of the pharyngeal bones
and upon the shape of the teeth. The stomach of the deep-bodied Barbus inter-
medius mentioned above (and p. 69) is full of gastropod mollusc shells, which, if
this were its normal diet, could explain the molariform teeth and the stout bone as
having been developed as a result of breaking open strong shells. Regrettably, the
stomach contents of the deep-bodied Barbus altianalis were unidentifiable and there
were no traces at all of shells. The extent to which diet affects the pharyngeal teeth
A REVISION OF THE LARGE AFRICAN BARBUS
FIG. 1 1 . Dorsal and lateral views of the left pharygeal bone of a specimen of
Barbus altianalis eduardianus.
FIG. 12. Dorsal and lateral views of the left pharyngeal bone of Barbus altianalis
radcliffii (one of the types of Barbus longirostris S.L. 410 mm).
18
K. E. BANISTER
A-F.J.
10mm
FIG. 13. A series of pharyngeal bones from specimens of Barbus altianalis eduardianus,
to show the variation in shape with the growth of the fish. The standard lengths (mm)
of the specimens are : A = 90 ; B = 100 ; C = 101 ; D = 125 ; E = 141 ; F = 171 ;
G = 205 ; H = 227 ; I = 258 ; ] = 330.
A REVISION OF THE LARGE AFRICAN BARBUS
FIG. 14. A series of pharyngeal bones from specimens of Barbus altianalis radcliffii.
Compare with the series opposite. Standard lengths (mm) of the specimens are : A = 84 :
B = 102 ; C = 118 ; D = 125 ; E = 145 ; F = 160 ; G = 197 ; H = 228 ; I = 255 ;
J = 343; K = 410; L = 388. K is from the holotype of Barbus longirostris ; L is from
the holotype of Barbus obesus.
20 K. E. BANISTER
in these species cannot be determined at the moment, but further considerations
are mentioned on p. 127.
Consistent differences were noted in the pharyngeal bones and teeth of specimens
of Barbus altianalis radcliffii and Barbus altianalis eduardianus of the same size
(figs. 13 and 14). The pharyngeal bones of Barbus altianalis radcliffii are consis-
tently stouter than those of Barbus altianalis eduardianus and the angle of the
anterior process is different.
About 15 per cent of the pharyngeal bones examined had a molariform second
tooth in the inner row. This phenomenon is as common in Lake Victoria as it is in
Lakes Edward and George and is apparently unrelated to size, the molariform tooth
being found in a specimen of only 121 mm S.L.
The change in the shape of the pharyngeal bones and teeth with the increase in
the size of the fish can be seen in figs. 13 and 14. These should be compared with
the series of pharyngeal bones of Barbus bynni (fig. 19) - a species with a less variable
body form.
Cyprinid pharyngeal bones and teeth have frequently been considered good
characters for distinguishing species. The value of this character must now be
reconsidered in the light of the observed variability of these characters in Barbus
altianalis. There are two possible interpretations of the evidence : that either the
degree of variation in the pharyngeal bones and teeth has not been fully realized
and that it is a dubious character at the species level ; or that certain species only
have variable pharyngeal bones and teeth and that for these species the very
variation is a good character. Insufficient evidence is available at the moment
to expand the arguments for either of the two interpretations.
Squamation. The lateral line follows a gentle curve. There are from 28 to 36
scales in the lateral line ; 28 (£.3), 29 (f.6), 30 (f.2o), 31 (f.4i), 32 (£.57), 33 (£.41),
34 (f.26), 35 (f.g), 36 (f.2) (fig. 59). The number of scales in the lateral line series
could not be counted on all the specimens. There are 6-5 (less frequently 5-5,
rarely 4-5) scales between the dorsal mid-line and the lateral line and 5-5 (less
frequently 6-5, rarely 4-5) scales between the lateral line and the ventral mid-line.
The number of scale rows between the lateral line and the origin of the pelvic fin
varies, between 2 and 3-5. The number of scales around the caudal peduncle
varies from 12 to 14.
Gill rakers. There are from 10 to 14 (most frequently 11-13) gill rakers on the
lower limb of the first gill arch in all three subspecies.
Coloration. Generally the body colour in live fish is from a silvery-grey to a
bronzy-green, the scales frequently having dark bases, the operculum slightly
bronzy. In fresh specimens from Lake George the pectoral fins are faintly pink
with a darker posterior half. The pelvics are grey suffused with pink ; the anal fin
membrane is dark at the base. The caudal fin is salmon-pink with a dark posterior
edge in specimens of over 100 mm S.L. from Lakes Edward and George. This is a
diagnostic character separating the Lake Victoria and the Lakes Edward and
George subspecies.
A REVISION OF THE LARGE AFRICAN BARBUS 21
Barbus altianalis altianalis Boulenger
A general description of the species is given on p. 12. For comparative purposes
morphometric data of the 15 specimens (S.L. 46-410 mm) from Lake Kivu and the
Ruzizi river are given below.
x s.d. s.e. range
L 49 -410 mm
D 26-9 2-3 0-6 22-6-30-3
H 25-6 1-4 0-4 23-4-28-0
I 7-0 2-1 0-5 4-5-10-2
IO 7-8 1-2 0-3 6-6-10-3
MW 6-0 0-8 0-2 4-6- 7-7
Pet 22-0 i-i 0-3 20-2-23-8
CP1 15-6 1-3 0-3 13-1-17-6
CPd n-o 0-7 0-2 10-1-11-4
Snt 8-5 0-9 0-2 7-3-10-1
Ab 4-5 i-i 0-3 2-4- 6-6
Pb 5-9 1-2 0-3 4-6- 8-3
DSp 16-9 2-3 0-6 13-3-23-1
Lateral line count 30 (1.5) ; 31 (f.2) ; 32 (f.3) ; 33 (f.3) ; 34 (f.2).
Barbus altianalis eduardianus Boulenger
HOLOTYPE. A fish of 371 mm S.L. B.M. (N.H.) Reg. No. 1906.9.7 : 41 from
Lake Edward, Uganda.
DESCRIPTION. The description is based on 130 specimens. A general descrip-
tion is given above. The morphometric data in detail are as follows.
x s.d. s.e. range
L 68 -436 mm
D 29-6 2-i 0-2 22-4-37-0
H 25-8 1-6 o-i 22-5-31-4
I 8-0 1-5 o-i 4-1-11-5
IO 8-1 0-9 o-i 6-4-10-8
MW 6-0 0-7 o-i 4-4- 8-3
Pet 22-1 1-3 o-i 18-8-25-0
CP1 17-1 1-2 o-i 14-4-19-7
CPd 12-4 0-9 o-i 10-1-16-2
Snt 8-1 0-7 o-i 5-8-10-1
Ab 4-2 0-7 o-i 2-2- 6-4
Pb 5-4 0-9 o-i 3-7- 7-3
DSp 17-5 2-4 0-2 11-7-24-0
Lateral line count 28 (f.i), 29 (f.5), 30 (f.g), 31 (f.2i), 32 (f.39), 33 (1.22),
34 (f.i7) 35 (f.6), 36 (f.2). The lateral line scales could not be counted on all the
specimens examined.
K. E. BANISTER
It is perhaps unwise to compare this sample with the much smaller sample from
Lake Kivu but the general agreement in mean values indicates that the two popu-
lations are extremely similar.
Barbus altianalis radcliffii Boulenger
HOLOTYPE. A skin of 530 mm S.L. B.M. (N.H.) Reg. No. 1904.5.19 : 13 from
Lake Victoria.
DESCRIPTION. The description is based on 68 specimens (S.L. 90-640 mm). A
general description is given above. The morphometric data in detail are given
below.
x s.d. s.e. range
L 90 -640 mm
D 29-6 2-7 0-3 24-0-37-4
H 26-7 1-7 0-2 23-8-29-8
I 7-0 i-i o-i 4-1- 9-3
IO 9-0 i-o o-i 7-5-11-2
MW 6-6 i-o o-i 4-7- 9-5
Pet 21-3 1-6 0-2 17-5-24-6
CP1 16-8 1-7 0-2 13-2-19-8
CPd 11-7 0-9 o-i 9'5-13'Q
Snt 8-7 0-7 o-i 7'5-io-g
Ab 4-5 1-6 0-2 2-1-11-6
Pb 5-3 1-9 0-2 2-3-12-4
DSp 15-9 4-2 0-5 9-0-30-4
Lateral line count 28 (f.2), 29 (f.i), 30 (f.6), 31 (f.i8), 32 (f.i5), 33 (f.i6),
34 (f-7)> 35 (f-3).
The mean values for the measurements taken differ little from those of the other
two populations. The absence of the red/pink caudal fin in Barbus altianalis
radcliffii has been mentioned above.
DISTRIBUTION. Barbus altianalis is found in Lake Kivu and the adjacent part
of the Ruzizi river, Lake Victoria including the Lake Kioga basin and Lakes Edward
and George. Although Lake Victoria and Lakes Edward and George are still con-
nected, migration along the connecting waterways is not easy at the moment. The
presence of Barbus altianalis in Lake Kivu, which is isolated from Lake Victoria
and Lakes Edward and George, and its absence from Lake Albert are facts which
must be explained. A certain amount is known about the geological history of the
lakes and I have attempted to assemble and correlate recorded events from each
lake and to try to explain the distribution of this species.
During the early Pliocene, the region now occupied by Lake Victoria was drained
to the west by the upper tributaries of the Congo system (Cooke 1958). This
drainage system persisted well into the Pleistocene (Kendall 1969). Stages in the
formation of the western rift valley interrupted this flow and the upwarp of the
A REVISION OF THE LARGE AFRICAN BARBUS
FIG. 15. Barbus altianalis radcliffii (the type of Barbus radcliffii,
from Boulenger 191 za).
FIG. 16. Map of the distribution of the three subspecies of Barbus altianalis : B. altianalis
altianalis = %, B. altianalis eduardianus = A, B. altianalis radcliffii = |.
24 K. E. BANISTER
eastern shoulder of the rift caused the rivers to reverse their flow and pond-up in
the shallow depression between the eastern and western rifts. This led to the for-
mation of the present-day Lake Victoria which Temple (1969) dates from the mid-
Pleistocene. Tectonic movements have altered the lake since its original formation.
Lacustrine deposits at Nsongezi in the Kagera valley (to the west of Lake Vic-
toria) are about 35 m above the present lake surface, and some no km inland of its
western shore. The dates of the Nsongezi deposits are in doubt. Kendall (1969)
cites authors who have given dates ranging from 60 ooo years B.P. to 10 ooo years
B.P. Doornkamp and Temple (1966) suggest that continued tectonic warping
has caused the eastward drift of the western shore of the lake. Climatic events
have probably had an effect as well. Kendall (1969) states that from before
14 500 years B.P. to about 12 ooo years B.P. Lake Victoria was a closed basin and
the water had a high concentration of salts. From about 12 ooo years B.P. to
about 10 ooo years B.P. the lake had an exit, which could have been to the south,
but its locality is by no means certain. Around 10 ooo B.P. to about 9500 B.P. the
basin became closed again and then drained to the north cutting the Nile gorge.
The water level in the lake rose after 12 500 B.P. There followed a period of aridity
around 10 ooo B.P., the water level fell and the outlet was lost. A wetter climate
ensued, the lake level rose, overflowed to the north and established the Nile outflow.
It is difficult to correlate these events because tectonic movements, climatic changes
and the rate of discharge from the lake have all had an effect upon the lake size,
level and position. The date of cessation of the tectonic movements is not known
with certainty but there are three horizontal raised beaches 3, 12 and 18 m above
the lake level which were presumably formed after earth movements had ceased.
The most recent of these has been dated at 3720 years B.P. (Stuvier et alii 1960).
Kendall (1969) argues that they all must be post 12 500 B.P. which may well indi-
cate that tectonicity had stopped by that time. The different beach levels are
possibly related to stages in the erosion of the Nile gorge.
Much less is known about Lake Kivu and very little absolute dating has been
done. Lake Kivu resulted from the effects of eruptions of the Mufumbiro (or
Bufumbiro) volcanoes (to the north of the present lake) when lava flows dammed
the rift valley. Prior to this unknown date the drainage in this part of the rift is
usually thought to have been northwards, but there may have been a proto-Lake
Kivu. Cahen (1954) describes the origin of Lake Kivu as the result of the landscape
changes which affected a network of rivers flowing S.W. to N.E. which curved
northwards and ponded-up to form the lake. This direction of flow is possibly not
compatible with the ideas of Cooke (1958) who describes the drainage as roughly
E. to W. but this may have been due to local topographical features. The lake later
rose in height and spilled over southwards to drain into Lake Tanganyika cutting
the Ruzizi gorge. Kendall (1969) noted that circa 12 450 B.P. Lake Kivu was
100 m higher than its present level This date was established by radio-carbon
dating on mollusc shells on a raised beach. Possibly this immediately pre-dates
the formation of the Ruzizi river.
The critical date of the eruption of the Mufumbiro volcanoes is unknown. Cahen
(1954) states that it was before the end of the alleged Kamasien ; Bishop (pers.
A REVISION OF THE LARGE AFRICAN BARBUS 25
comm.) dates the eruptions as upper Pleistocene and Fryer (1968) suggests a date
of about 100 ooo years B.P. Lake Kivu has been connected to Lake Tanganyika
for a minimum period of 12 ooo years and although Barbus altianalis is found in the
Ruzizi river at the Lake Kivu end it has not been recorded from Lake Tanganyika
(Poll 1953 ; Marlier 1953). Barilius moorii has made the journey in the other
direction (Poll 1950). Why only Barilius moorii has been able to ascend the Ruzizi
river is a question that I am unable to answer. It may be that the ascent of the
river is easier for some species than the descent is for others. There are also rapids
on the Ruzizi river which must limit the amount of faunal interchange. All that
one can say at the moment is that for at least the last 12 ooo years only one species
of fish is known to have successfully navigated the Ruzizi river.
A lake occupying the site on the floor of the western rift valley now occupied by
Lakes Edward and George began to form at the time of the uplift of the rift wall.
This corresponds with fig. 2B in Doornkamp and Temple (1966), and Bishop (1971)
dates it as late Pliocene. Rivers drained onto the floor of the rift and a lake
resulted. There was probably more than one lake, for the extensive Kaiso beds
(later Pliocene) are lacustrine as, largely, are the overlying Semliki deposits (mid- to
later Pleistocene) . The full extent of the ' Kaiso ' and ' Semliki ' lakes both in time
and area is not completely known. The Kaiso beds are up to 8000 feet thick
(Brown 1956) and represent possibly three million years of comparative stability.
Bishop (1969) noted that the Kaiso and Semliki deposits are separated in time by
renewed movement of the rift valley boundary faults. The Semliki beds are found
north of Lake Albert indicating a very large area for the ' Semliki' lake. The upper-
most Semliki beds are later Pleistocene with an age in excess of 10 ooo years B.P.
At the end of the Semliki period soil formation occurred and about 8000 years
to 10 ooo years B.P. the Katwe volcanoes ceased erupting and the present forms of
Lakes Edward, George and the Kazinga channel became defined. The Kazinga
channel runs through the Katwe lavas, but probably dates from the time that soil
formation occurred, i.e. the post-Semliki period. The form of the channel was
little altered by the volcanic activity (Bishop 1969 : 45).
The Ishango beds, which were deposited during the active phase of the Katwe
volcanoes (8000 years to 10 ooo years B.P., Bishop, op. cit.}, have yielded fossil
Barbus species remains at Ishango on the Semliki river exit from Lake Edward
(Greenwood 1959) . From Site IX in Makalian strata, fossils comparable with Barbus
bynni and Barbus altianalis, then dated Mesolithic or Holocene, were described by
Greenwood (op. cit.}. Site VI at Ishango, in the upper Semliki beds (Katanda for-
mation), dated mid-Pleistocene [but older than the Nsongezi deposits (Bishop 1969)],
yielded unidentifiable Barbus species fragments, the oldest known from this area.
Today the fish fauna of Lake Edward is different from that of Lake Albert, the
former lacking species present in the latter. The Semliki river connects the two
lakes ; probably the rapids and possibly the cooler streams flowing down from the
Ruwenzori mountains (Fryer 1968) prevent any faunal interchange. The fauna
of Lake Edward is poor, lacking representatives of some widespread African fish
families, e.g. Polypteridae, Characidae, Mochokidae, Centropomidae and Masta-
cembelidae, as well as some specifically Nilotic species present in Lake Albert, e.g.
26 K. E. BANISTER
Lates niloticus and Barbus bynni. This situation has been described in detail by
Greenwood (1959).
Some of the species now absent from Lake Edward have been found fossilized in
the Ishango beds. Kendall (1969) states that the major fish extinctions probably
occurred within the last 18 ooo years. Polypterus, Lates, Synodontis and Barbus cf.
bynni were present some 8000-10 ooo years ago. Characins disappeared much
earlier in the fossil record, the latest known fossil Hydrocynus occurring in the Kaiso
beds (probably lower Pleistocene). Certainly, at the time the Ishango beds were
deposited, the fish fauna of Lake Edward was closer to that now in Lake Albert
than it is today.
The Ishango beds lie on, and are also covered by, volcanic rocks (Bishop 1969).
So, it would seem very possible that some species were killed by the outpourings of
the Katwe eruptions. It is less likely that a drying up of the lake caused the
extinctions because there is no evidence for such an event and Kendall (1969)
suggests that in the Lake Victoria basin the period from 9500 B.P. to 6500 B.P. was
marked by high water levels and the levels fell only slightly from 6500 B.P. onwards.
It is uncertain whether Lake Edward became completely denuded of fish life at
the end of the Ishango period. If it did then the lake has been recolonized ; if it
did not, then some species must have escaped, or must have remained unaffected by
the eruptions. Dr P. H. Greenwood (pers. comm.) reports having seen fish behaving
normally within a hundred metres of a lava stream entering Lake Kivu, this suggests
that perhaps an increase of toxic products in the water is more lethal than the heat-
ing effect of the lava. Possible means by which various species could have survived
are discussed by Greenwood (1959). They include migration into feeder streams and
different degrees of resistance to toxicity and low oxygen concentrations.
If there was an invasion from Lake Victoria it must have occurred since 8000 B.P.
(the date of cessations of the Katwe volcanoes). However, since that time there is
no evidence that Lake Victoria has been subjected to any great earth movements
(see above) and it had already 'gained its outlet via the Nile gorge. This would
limit any rises in water level to about 18 m above the present level. The watershed
between Lakes Edward and George and Lake Victoria is extremely low. Now the
Katonga river is blocked by Papyrus swamp but a few metres rise in water level
might possibly enable fishes to travel from Lake Victoria into Lakes George and
Edward. The fact that no, e.g. Haplochromis, fossils have been found in the
Ishango beds does not necessarily mean that they have invaded the lake in post-
Ishango times. The fossils could yet await discovery. Present studies (Dr P. H.
Greenwood, pers. comm.) have suggested that Lake Edward Haplochromis species
are not as closely related to particular Lake Victoria Haplochromis species as had
been thought, which possibly militates against the idea of an invasion from Lake
Victoria.
It seems much more likely that the fauna of Lakes Edward and George and Lake
Kivu was originally widespread in the upper reaches of the old Congo system and
became isolated in the present lake basins as a result of tectonicity. It would be
very difficult to explain satisfactorily how Barbus altianalis could have migrated
into Lake Kivu from either Lake Edward or Lake Victoria. At the time of the
A REVISION OF THE LARGE AFRICAN BARBUS 27
highest known level reached by Lake Kivu, 12 450 B.P., Lake Victoria was not at
its highest. However, it must be noted that Lake Victoria possibly had an outlet
(see above) whose direction is unknown. The general topography of the Lake Kivu
basin makes invasion from the east or north extremely unlikely.
If, as judged by their breeding habits (Whitehead 1964), Barbus species are
fundamentally fluviatile fishes, then it is more likely that Barbus altianalis originally
lived in the upper reaches of the Congo system in the area now occupied by Lakes
Kivu, Edward, George and Victoria. Since the suggested modifications in its range
the Barbus altianalis populations have undergone no significant morphological
changes but the colour difference in the caudal fin is noted above.
DIAGNOSIS AND AFFINITIES. Barbus altianalis has three recognized subspecies
which are meristically and morphometrically very similar. The pink caudal fin
in Lakes Edward and George subspecies Barbus altianalis eduardinanus separates
this subspecies from Barbus altianalis radcliffii of Lake Victoria. No information
is available on the live colour of Barbus altianalis altianalis from Lake Kivu.
Barbus altianalis is, in gross morphology, similar to Barbus intermedius but the
two species are separable on the modal lateral line count (see fig. 59), 26-34 (most
frequently 28-32 ; mode 29) in Barbus intermedius ; 28-35 (most frequently 30-34,
mode 32) in Barbus altianalis. The last unbranched ray in the dorsal fin is slightly
shorter in Barbus altianalis (9-0-30-4 per cent S.L., x = 16-9) than in Barbus inter-
medius (6-9-33-9, x = 20-1). The anterior and posterior barbels are shorter in
Barbus altianalis (Ab 2-1-11-6 per cent S.L., x = 4-3 ; Pb 2-3-12-4, x = 5-4) than
in Barbus intermedius (Ab i-8-n-i, x = 5-8 ; Pb 1-9-12-1, x = 6-9).
For a comparison of Barbus acuticeps and Barbus altianalis see under the former
species.
Barbus ruasae, from Rwanda, is somewhat similar in appearance to Barbus
altianalis but can be distinguished easily from it by the fewer lateral line scales
(25-26 against 28-35).
Barbus altianalis is considered here to belong to the Barbus intermedius group.
Barbus bynni (Forsskal) 1775
Cyprinus bynni Forsskal, 1775, De scrip. Anim. : 71.
Barbus bynni : Blgr., 1911, Cat. Afr. Fish 2 : 26, fig. 8 (includes a full bibliography up to this
date).
Barbus ruspolii Vinciguerra, 1896, Annali Mus. civ. Stor. natn. Giacomo Doria (2) 17 : 29.
Barbus meneliki Pellegrin, 1905, Bull. Mus. Hist. nat. Paris 16 : 293
Barbus bynni vudolfianus Worthington, 1932, /. Linn. Soc. (Zool.) 38 : 132.
NOTES ON THE SYNONYMY. The type specimen of Barbus bynni is no longer
extant (Klausewitz & Nielsen 1965) and a neotype has been selected (see below).
The holotype of Barbus ruspolii is in a very poor state and has not been seen by me,
but other specimens referred by Vinciguerra to Barbus ruspolii have been examined.
The holotype and unique specimen of Barbus meneliki in the Paris Museum has been
studied.
28
K. E. BANISTER
FIG. 17. Barbus bynni (from Boulenger 191 la).
The similarity of Barbus bynni to Barbus ruspolii was noted by Boulenger (1907)
and by Worthington (i932a). The latter author was also of the opinion that Barbus
meneliki from Lake Rudolf was close to his Barbus bynni rudolfianus. Barbus
meneliki is very similar in many respects to Barbus bynni rudolfianus from the same
lake. The greatest difference is the presence of 29/30 lateral line scales in the former
against 34 or 35 in the latter. Barbus bynni from Lake Albert have a lateral line
range of 28 to 35 scales. Barbus meneliki is based on the holotype and unique
specimen so either it is an extremely rare fish in the lake or it is a member of Lake
Rudolf's Barbus bynni population with fewer scales than usual. As I have seen
only five Barbus bynni specimens from Lake Rudolf the latter suggestion seems the
more likely especially when the information is compared with the lateral line scale
count range in Lake Albert fish. The mouth of Barbus meneliki has a horny covering
on the lower jaw, although this has not been seen in other Lake Rudolf specimens of
Barbus bynni, it is not an unusual phenomenon in Barbus species (Groenewald
1958). In all other respects (e.g. dorsal spine length, caudal peduncle shape and
scale striations) Barbus meneliki and Barbus ruspolii both closely resemble Barbus
bynni hence I consider them conspecific. Boulenger (1907) remarked that little
work had been done on the morphometric variation in Barbus bynni, both within
one population and between isolated populations. What little I have been able to
do is detailed below.
NEOTYPE. Klausewitz and Nielsen (1965) listed Cyprinus bynni as being 'no
longer in existence from Forsskal's original collection'. I have also tried but
without success to trace this specimen and must agree with Klausewitz and Nielsen
that the specimen is, at least, lost. Therefore I have selected as a neotype
B.M. (N.H.) specimen No. 1907.12.2 : 1230, a fish of 152 mm S.L. collected from
A REVISION OF THE LARGE AFRICAN BARBUS
29
the Nile at Asswan. Forsskal did not state the size of his specimen nor its exact
locality, so I have selected this fish as it appears to be a modal representative of the
fishes from the lower Nile.
DESCRIPTION. The description is based on a total of 63 fish ; 36 from the lower
and White Niles ; ten from Lake Albert ; five from the Blue Nile ; seven from
Lake Abaya and five from Lake Rudolf. A further specimen doubtfully from the
Webi Shebeli will be treated separately.
The body is shaped like an elongate rhomboid, the dorsal profile rises sharply and
evenly from the snout to the occiput and then more steeply to the origin of the dorsal
fin. The profile of the belly slopes down to a point in front of the ventral fins.
The body is compressed, more so than in most other species under consideration.
The compressed caudal peduncle is short and deep and is quite characteristic for
Barbus bynni.
The scales bear sinuous longitudinal striae.
The most distinctive features are the massive dorsal spine and the caudal
peduncle. At least one specimen (a fish from the lower Nile) shows the development
of 'rubber lips'.
The following tables show the range, mean, standard deviation and standard
error of the measurements taken for the five populations.
Range of standard lengths of the populations examined (in mm)
No. of specimens
Nile
59-351
36
Lake Albert
136-349
10
Blue Nile
95-246
5
Lake Abaya
236-329
7
Lake Rudolf
196-233
5
Total
59-351
63
Body depth (this and all subsequent data are expressed as percentages of the S.L.)
Locality
Nile
Lake Albert
Blue Nile
Lake Abaya
Lake Rudolf
Total
n
36
10
5
7
5
63
32-9
34'6
34'3
35-o
33-6
33-6
s.d.
2-5
2-7
1-9
2'5
2-3
2-5
s.e.
0-4
0-8
0-8
i-o
I'D
range
26-0-37-7
31-0-39-6
31-6-36-2
31-4-37-7
30-0-36-0
26-0-39-6
Locality
Nile
Lake Albert
Blue Nile
Lake Abaya
Lake Rudolf
Total
n
36
10
5
7
5
63
Head length
x s.d.
24-0
24-7
25*3
24-9
24-3
24-4
1-7
i-o
0-7
i'3
0-6
s.e.
0-3
o-3
o-3
o-5
o-3
0-8
range
21-4-30-6
23-6-26-9
24-4-26-3
23-1-26-2
24-0-25-3
21-4-30-6
K. E. BANISTER
Locality
Nile
Lake Albert
Blue Nile
Lake Abaya
Lake Rudolf
Total
n
36
10
5
7
4
62
Eye diameter
x
7-0
6-7
7-2
5-7
7-0
7-0
s.d.
i-i
0-7
s.e.
0-2
0-4
O-2
0-2
0-2
range
4-3-10-2
5-7- 8-3
6-9- 8-4
5-3- 6-3
5-5- 7'3
4-3-10-2
Locality
n
X
Nile
36
8-1
Lake Albert
10
8-6
Blue Nile
5
8-2
Lake Abaya
7
8-4
Lake Rudolf
4
8-0
Total
62
8-2
Interorbital width
s.d.
0-8
i-o
0-8
0-4
o-3
0-8
s.e.
o-i
o-3
0-4
o-i
O-I
o-i
range
7-0-11-7
7-0-10-5
7-2- 8-7
8-1- 8-8
7'7- 8'3
7-0-11-7
Mouth width
Locality
n
X
s.d.
s.e.
Nile
36
5'5
0-7
o-i
Lake Albert
10
6-0
0-6
0-2
Blue Nile
5
5'9
0-4
O-2
Lake Abaya
7
6-2
0-7
o-3
Lake Rudolf
4
5-o
0-7
o-3
Total
62
5'7
0-7
O-I
range
4-2-7-3
4-9-6-9
5-4-6-5
5-4-6-9
4-3-5-8
4-2-7*3
Pectoral fin length
Locality
n
X
s.d.
s.e.
range
Nile
36
21-3
i-3
0-2
18-4-24-3
Lake Albert
10
21-5
i-o
o-3
19-0-22-6
Blue Nile
5
24-5
i-3
0-6
23-1-25-8
Lake Abaya
7
24-2
1-9
0-7
21-6-26-4
Lake Rudolf
5
21-0
1-2
o-5
19-6-22-5
Total
63
21*9
1-8
O-2
18-4-26-4
Locality
Nile
Lake Albert
Blue Nile
Lake Abaya
Lake Rudolf
Total
36
10
5
7
4
62
Caudal peduncle length
x s.d.
17-0
16-4
16-2
17-1
17-1
16-9
0-7
1-8
0-6
s.e.
0-2
o-3
0-7
range
13-5-21-1
14-7-19-7
15-4-17-0
15-5-20-6
16-6-17-5
13-5-21-1
A REVISION OF THE LARGE AFRICAN BARBUS
Caudal peduncle depth
Locality
Nile
Lake Albert
Blue Nile
Lake Abaya
Lake Rudolf
Total
Locality
Nile
Lake Albert
Blue Nile
Lake Abaya
Lake Rudolf
Total
Locality
Nile
Lake Albert
Blue Nile
Lake Abaya
Lake Rudolf
Total
Locality
Nile
Lake Albert
Blue Nile
Lake Abaya
Lake Rudolf
Total
n
36
10
5
7
5
63
n
36
10
5
7
5
63
34
10
5
7
5
61
34
10
5
7
5
61
M'3
13-5
13-6
13-4
13-8
s.d.
1-6
0-9
o-5
0-7
o-i
Snout length
x s.d.
7'7
7-8
8-1
8-2
8-0
0-7
0-9
o-3
0-4
o-3
0-7
Anterior barbel
4-6
2'4
6-3
5'7
2-5
4'4
s.d.
1-4
0-8
i-i
1-4
Posterior barbel
x s.d.
6-5
6-1
4-0
5'5
i-o
i-i
O-2
1-2
s.e.
0-3
o-3
0-2
o-3
o-i
0-2
s.e.
o-i
0-3
o-i
o-i
o-i
O-I
s.e.
0-2
o-i
0-4
0-2
0-2
O-I
s.e.
o-i
o-3
0-2
o-i
o-i
o-i
range
11-7-19-4
13-1-15-4
12-6-18-0
12-6-14-6
13-0-13-9
11-7-19-4
range
6-7-10-5
5-7- 9-2
7-7- 8-4
7-6- 8-8
7'5- 8-3
5-7-IQ-5
range
2-3-6-0
1-2-4-6
5-7-6-9
4-6-6-1
2-1-2-9
1-2-6-9
range
4-2-8-4
3-2-6-9
6-1-7-4
5-7-6-8
3-9-4-3
3-2-8-4
Of particular note is the ratio of the caudal peduncle length to its depth, which
has a mean value of 1-23. Only one specimen that I examined had a ratio greater
than 1-40.
The figures presented here show that the barbels are shorter in fishes from Lakes
Rudolf and Albert than they are in the other localities but the significance, consid-
ering the inequality of the samples, is low.
On the basis of the evidence from morphometrics any attempt to maintain or
establish sub-specific categories is rather unsound. Any such attempts must be
supported by evidence not available to the museum worker.
Squamation. The scales of Barbus bynni have slightly irregular parallel striae
(fig. 93). There are from 28 to 37 scales in the lateral line series. The range and
K. E. BANISTER
the mean (to the nearest whole number) of scales on the lateral lines from the
different localities are given below.
Locality
Nile
Lake Albert
Blue Nile
Lake Abaya
Lake Rudolf
Total
33
32
3i
32
33
33
range
31-37
28-35
28-33
31-34
28-35
28-37
Six and a half (less frequently 5-5) scales are present between the dorsal mid-line
and the lateral line. Five and a half (rarely 6-5 or 4-5) scales are present between
the lateral line and the ventral mid-line. The number of scales between the lateral
line and the base of the ventral fin varies from 2 to 4-5 ; then, 2 (f.i), 2-5 (f.2o), 3 (f.2o),
3-5 (f.i4), 4 (f-7), 4-5 (f.i). There is no apparent correlation between locality and
this count. There are from 12 to 16 scales around the caudal peduncle ; 12 (f.8),
13 (f.22), 14 (f.i7), 15 (1.7), 16 (f.7). The fish from the Blue Nile had 12 scales
around the caudal peduncle whilst those from Lake Rudolf had 16. Two of the
Lake Albert fish also had 16. The samples are too small though to use such a
variable character to establish a population taxonomically.
Dorsal fin. The dorsal fin has 8 (f.i), 9 (1.59), 10 (f.3) branched rays, the fourth
unbranched ray is ossified into a straight, massive spine.
Dorsal spine.
White Nile
Lake Albert
Blue Nile
Lake Abaya
Lake Rudolf
Total
36
10
5
7
5
63
29-3
30-9
36-4
35'3
s.d.
3'4
2-5
5-o
3'9
5'2
4'3
s.e.
0-6
0-8
2-2
i-5
2-3
0-6
range
23-7-38-4
24-9-35-2
29-5-40-0
28-8-39-3
30-0-43-0
23-7-43-0
The anal fin has three unbranched rays and five branched rays.
Pharyngeal bones and teeth. The pharyngeal teeth (fig. 18) number five in the
inner row. The first tooth is small and angled towards the second tooth. This
tooth becomes mammilliform with increasing size but is almost always higher than
it is long (cf. Barbus gananensis). Small concavities develop on the posterior face
of the crown on teeth 3, 4 and 5. These three teeth become successively more
slender and the fifth tooth becomes markedly hooked in large fishes. The three
teeth of the second row and the two teeth of the third row resemble the last tooth
in the inner row in shape. A series of pharyngeal bones (fig. 19) shows the variation
in shape with the growth of the fish.
Gill rakers. There are 12 to 1 6 gill rakers on the lower limb of the first gill arch.
Coloration. Described by Boulenger (191 la) as yellowish or silvery, grey or olive
on the back, fins yellow or orange, anal and caudal pink or red, ends of dorsal and
A REVISION OF THE LARGE AFRICAN BARBUS
33
FIG. 1 8. Dorsal view of the left pharyngeal of -B. bynni of 280 mm S.L.
caudal rays sometimes blackish. Amirthalingam (1965), describing Barbus bynni
from the Sudan, states that the colour is very variable, but in live fish of about
250 mm is frequently as follows : body silvery, dark above and light below ; all
fins with deep-orange streaks ; dorsal and caudal fins edged with black ; anal,
pelvic and pectoral fins edged with white.
DISTRIBUTION. Specimens from Lake Abaya, Lake Rudolf and Lake Albert
have been examined. Sandon (1950) states that Barbus bynni is common in the
Blue Nile and I have seen specimens from the Metti and Gudar rivers and from
Rosaires. Sandon (op. cit.) also refers to the widespread distribution of Barbus bynni
throughout the White Nile, although the B.M. (N.H.) collections have no specimens
from between Khartoum and Lake Albert. This species is common throughout
the lower Nile.
Specimens nos. 92-101 listed by Boulenger (igua : 28) from Lake Baringo are
not referable to this species, but belong to Barbus intermedius.
Peel (1900 : 336) records Barbus byrni (sic) from the Webi Shebeli in Somaliland.
Donaldson-Smith (1897) collected three specimens of a Barbus species identified as
Barbus bynni by Giinther (1896) ; the two small specimens from Stony Brook, the
Errer river (Webi Shebeli system) (Academy of Natural Sciences Philadelphia, Nos.
14541, 14542), are not Barbus bynni but the large specimen is. The two small fishes
are mentioned with details of the locality in Donaldson-Smith's account of his
expedition but there is no mention of the large fish. Giinther (1896) gives its
locality as the Shebeli river. This specimen is now in the collection of the Philadel-
phia Academy of Natural Sciences No. 16710, where the locality Thebeti, East
Africa, has been given to it. Thebeti could well be a mis-reading of Shebeli, for the
34
K. E. BANISTER
FIG. 19. A series of pharyngeal bones from specimens of B. bynni to show the variation
with the size of the fish. Standard lengths (mm) of the specimens are A = 58 ; B = 98 ;
C = 140 ; D = 260 ; E = 235 ; F = 171 ; G = 280 ; H = 350.
A REVISION OF THE LARGE AFRICAN BARBUS
35
^rr^A^Jo-/
FIG. 20. Map of the distribution of B. bynni.
writing on the label inside the gill cover is very faint. This fish is certainly Barbus
bynni but its locality is in some doubt and it is not included with the data presented
above. Its measurements, however, will appear at the end of this section.
Vinciguerra (1898) gives Bissan Guarrica as a locality for Barbus ruspolii. As
far as I can find out this is a tributary of the Sagan river which drains into Lake
Stephanie.
Lakes Rudolf and Abaya are now isolated from the Nile system. Fryer (1968)
suggests that the Southern Ethiopian rift valley lakes drained into Lake Rudolf and
Lake Rudolf was connected to the Nile via the Sobat river (or via the present
Lotagipi and Kenamuki swamps as suggested by Cooke 1958). Therefore Barbus
bynni could have been isolated in the distal parts of what was once a more extensive
Nile system. Argulus rhipidiophorus , a parasitic crustacean discussed by Fryer
(op. cit.) is also found in the Nile, Lake Rudolf and the Ethiopian rift valley lakes.
Greenwood (1959) has recorded remains of fishes closely resembling Barbus bynni
from an epi-Pleistocene site at Ishango near the source of the Semliki river on Lake
Edward. Barbus bynni is now extinct in Lake Edward (see p. 26).
36 K. E. BANISTER
Barbus bynni is confined to the Nile system and is also found in lakes that were
once connected to the Nile.
DIAGNOSIS AND AFFINITIES. Barbus bynni has some gross morphological charac-
ters in common with Barbus gananensis, Barbus oxyrhynchus and Barbus longifilis.
The characters in common include a rhomboidal compressed body, a strong, straight,
smooth dorsal spine and parallel, sinuous striae on the scales. The species listed
above are considered to be closely related and comprise the Barbus bynni complex
which is discussed further on page 128.
Barbus bynni can be distinguished from Barbus gananensis by the length and
strength of the dorsal spine (longer and stronger in the former species) and by the
shape of, particularly, the second tooth in the inner pharyngeal row. In Barbus
gananensis this tooth is higher and longer than in Barbus bynni (see fig. 28). The
anterior edentulous process of the pharyngeal bone is relatively longer in Barbus
bynni than in Barbus gananensis.
The other two species are well separated geographically from Barbus bynni.
Barbus longifilis can easily be distinguished by its barbels (Ab, x = 8-9, range
8-3-9-6 ; Pb, x = ii'3, range 9-5-13-4) which are much longer than those of Barbus
bynni (Ab, x = 4-4, range 1-2-6-9 > Pb, x = 5-5, range 3-2-8-4).
Barbus oxyrhynchus has fewer scales in the lateral line series than has Barbus bynni
(21-29, most frequently 22-25 against 28-37, most frequently 31-33). The dorsal
spine is usually thinner and shorter in Barbus oxyrhynchus than in Barbus bynni
(x = 25-3 in Barbus oxyrhynchus, x == 31-1 in Barbus bynni).
Barbus bynni P.A.S. No. 16710 (see above)
L = 349 mm ; D = 31-0 ; H = 21-8 ; 1 = 5-4; IO = 8-6 ; MW = 5-1 ; Pet
= 23-2 ; CP1 = 14-6 ; CPd = 13-4 ; Snt = 6-9 ; Ab = 4-0 ; Pb = 4-9 ; DSp
= 29-3-
Dorsal fin IV-g. Twelve scales around caudal peduncle ; 31 in lateral line,
5-5 between dorsal mid-line and lateral line, 5-5 from lateral line to ventral mid-
line ; 2-5 between lateral line and base of pelvic fin.
Barbus caudovittatus Boulenger 1902
Barbus caudovittatus Blgr., 1902, Annls. Mus. r. Congo Beige Zool. (i) 2 : 32 ; Poll, 1953, Result.
scient. Explor. hydrobiol. Lac Tanganika 3 (5 A) : 85.
Varicorhinus stappersii Blgr., 1917, Ann. Mag. nat. Hist. (8) 20 : 364.
Barbus euchilus Blgr., 1919, Proc. zool. Soc. Lond. : 400.
Barbus miochilus Blgr., 1919, Proc. zool. Soc. Lond. : 401.
Barbus lestradei David, 1936, Revue Zool. Bot. afr. 27 (2) : 150, fig. i.
Barbus chilotes sakaniae Poll, 1938, Revue Zool. Bot. afr. 30 (3) : 413, fig. 13.
Barbus pojeri Poll, 1944, Bull. Mus. r. Hist. nat. Belg. 20 (3) : 2, fig. 3.
NOTES ON THE SYNONYMY. Poll (1946) synonymized Barbus euchilus and Barbus
miochilus and then in 1953 included Barbus pojeri within the expanded Barbus
euchilus. After examination of the type specimens I agree with Poll, although it is
worth noting that the dorsal fin spines of Barbus euchilus and Barbus miochilus are
more strongly ossified than is the norm in Barbus caudovittatus.
A REVISION OF THE LARGE AFRICAN BARBUS 37
Barbus lestradei had been recognized as a separate species by authors since its
first description by David (1936) . The type specimen in the Central African Museum
at Tervuren has IV-g rays in the dorsal fin (not II-Q) and 27 scales in the lateral
line. It is a large fish (S.L. 373 mm) as are almost all fish referred to Barbus lestradei.
The body is dark coloured, but the measurements, the weakly ossified dorsal spine,
overall body shape and the indications of darkening in body colour in some Barbus
caudovittatus leave little room for doubt that Barbus lestradei is a synonym of
Barbus caudovittatus. Varicorhinus stappersii was described from two specimens of
255 mm and 191 mm S.L. respectively from Nganza on the Lufuko river, Lake
Tanganyika basin B.M. (N.H.) Reg. No. 1920.5.25 : 36-37, and two specimens from
the same locality, M.A.C.T. No. 14222 of S.L. 175 and 184 mm. The morphometric
data for the two B.M. (N.H.) specimens (not included below) are D = 29-0, 29-3 ;
H = 20-8, 22-5 ; I = 47, 5-8 ; IO = 8-8, 8-6 ; MW = 9-4, 8-6 ; Pet = 21-2,
20-9 ; CP1 = 19-2, 19-4 ; CPd = n-8, 11-5 ; Snt = 7-5, 8-1 ; Ab = 3-9, 4-7 ;
Pb = 5-5, 6-5 ; DSp = 9-4, io-o ; D. fin IV-io, IV-g. The morphometric data
of these two fish are given separately to facilitate a comparison with the morpho-
metric data of the other specimens of Barbus caudovittatus. There are 12 scales
around the caudal peduncle, 28 in the lateral line, 4-5, 5-5 in the transverse series
and 2-5 between the lateral line and base of the pelvic fin. The M.A.C.T. specimens
agree closely with these measurements.
The mouth is broad with a straight, horny edge to the lower jaw. The width of
the mouth is a natural consequence of achieving the 'Varicorhinus' facies, as per-
haps is the decrease in head length. The barbels are conspicuous and the pharyngeal
bones are identical with those of Barbus caudovittatus. These features and the
similarity between the morphometric data of Varicorhinus stappersii and Barbus
caudovittatus suggest that the two specimens of Varicorhinus stappersii are highly
modified 'sector' forms of Barbus caudovittatus (Groenewald 1958). The situation
here parallels the case of Varicorhinus brucii and Barbus brucii, two species, display-
ing different facies, that were eventually synonymized by Jubb (1968). It is not
within the scope of this paper to discuss the validity of the genus Varicorhinus, but
there is sufficient awareness of the variability of Barbus species mouth parts to
suggest that it is not unlikely that some African Varicorhinus species are highly
modified individuals of various polymorphic Barbus species. The specimens from
Koki, recorded by Poll (1953), are only doubtfully Barbus caudovittatus. The
three fishes examined, B.M. (N.H.) 1955.12.20 : 870-872, ex Poll, are small,
43-47 mm S.L. with 10 or u branched rays in the dorsal fin, a well-ossified dorsal
spine, 24-25 scales in the lateral line and no dark marks on the caudal fin. Although
these Koki specimens come from within the distributional limits of Barbus caudo-
vittatus I have too little information to decide whether they are a separate species
or a distinct population of Barbus caudovittatus. They have not been considered to
be Barbus caudovittatus in this account.
HOLOTYPE. A fish of 77 mm S.L., M.A.C.T. No. 1168, from Banzyville. Another
specimen listed as 'one of the types' is in the collections of the British Museum
(Natural History) No. B.M. (N.H.) 1901.12.26 : 26.
K. E. BANISTER
FIG. 21. Barbus caudovittatus (Holotype, from Boulenger
DESCRIPTION. The description is based on 47 specimens,
data are expressed in tabular form as follows :
The morphometric
L
D
H
I
IO
MW
Pet
CP1
CPd
Snt
Ab
Pb
28-8
26-6
7-5
9-4
7-3
21-0
16-8
12-2
8-9
4'4
5-8
s.d.
3'4
2-0
1-6
1-6
0-8
1-4
i-i
i-i
1-6
s.e.
O-2
0-2
0-1
0-2
O-2
0-2
0-2
O-2
O-2
range
35 -580 mm
24-1-37-5
21-4-31-4
4-7-11-4
6-5-12-4
5-7- 8-9
18-1-24-7
12-8-19-7
8-6-15-8
6-5-11-4
2'3- 7'5
3-1-11-3
Small specimens have a shallow compressed body, but this becomes deeper and
thicker in larger fish. The mouth is small, inferior and frequently horse-shoe
shaped. Thicker lips have been observed in some larger specimens from Lake
Tanganyika, whilst in a specimen from the Luilu river (Sankuru system - Kasai ;
M.A.C.T. No. 78925), the 'rubber-lipped' condition is approached.
The ventral profile of the body, from the jaw articulation to the anus, is straight
or slightly convex. The illustrated specimen (the holotype) is not typical in this
respect. The contrast between the ventral and dorsal profiles is usually greater in
larger fishes.
Dorsal fin. The dorsal fin has four unbranched rays. In one fish from the
Luberizi river (Ruzizi area) there are only three unbranched rays. David (1936)
recorded only two unbranched rays in the type specimen of Barbus lestradei, but on
examination of the type I found that there are four. The last unbranched ray is
scarcely ossified and in the majority of specimens articulations persist almost to its
base. A specimen from Nyundeulu (an affluent of the Lualaba), M.A.C.T. No.
A REVISION OF THE LARGE AFRICAN BARBUS
39
FIG. 22.
1cm
Dorsal and lateral views of the right pharyngeal bone from a large
specimen of B. caudovittatus of 420 mm S.L.
124937, has an ossified fourth dorsal ray with a length of 17 per cent of the standard
length.
The branched rays range from 8 to u in number ; 8 (f.3), 9 (f-40), 10 (f.3),
ii (f.i). The fishes with 10 and n branched rays in the dorsal fin all come from
the river at Elizabeth ville (= Lubumbashi). The last branched ray is commonly
longer than the few rays immediately preceding it.
The anterior edge of the dorsal fin is in advance of the insertion of the pelvic fin.
Because of the softness of the last unbranched ray it was impractical to measure its
length using the same standards as for the other species.
The anal fin has three unbranched and five branched rays.
Squamation. The lateral line has 24-30 scales, 24 (f.3), 25 (f.6), 26 (f.12), 27 (f.i3),
28 (f.8), 29 (f.4), 30 (f.i). There is no evidence for any geographical variation in
these figures. There are 12 scales round the caudal peduncle and 4-5 (r. 5-5) scale
rows between the dorsal mid-line and the lateral line and 4-5 (r. 5-5) between the
lateral line and the ventral mid-line. Two and a half (r. 3) scale rows are present
between the lateral line and the base of the pelvic fin.
Pharyngeal bones and teeth. The pharyngeal bones of a large specimen are shown
in fig. 22 and of a small specimen in fig. 23.
The crowns of the teeth are hooked in small fishes but become molariform in large
fish. No exception to the pharyngeal formula 2. 3. 5. -5. 3. 2 was found.
Gill rakers. The number of gill rakers on the lower limb of the first gill arch is
between 11 and 15 in the specimens examined.
40 K. E. BANISTER
FIG. 23. Left pharyngeal bone from a specimen of B. caudovittatus of no mm S.L.
Coloration. In small specimens (less than about 50 mm S.L.) a small black spot
is present on the caudal peduncle at the base of the caudal fin. This spot is not
present on larger specimens but instead the dorsal and ventral edges of the caudal
fin become darker. In fishes of from about 70 to 170 mm S.L. the dark bands on
the caudal fin are conspicuous. Above this size the bands are less obvious as a
result of the general darkening of the caudal fin.
In most specimens dark pigment is present on the distal parts of the anal, pelvic
and pectoral fins. This is most noticeable in medium-sized fishes.
The body is brown in alcohol-preserved specimens. Large fishes are darker
than small fishes.
DISTRIBUTION. Specimens have been examined fom Elizabethville (= Lubum-
bashi) (Lualaba Congo) ; Banzyville ( = Mobayi) (Ubangi Congo) ; Avakubi (Ituri) ;
Tshikapa, Luembe, and Tchitatu rivers (upper Kasai) ; Bushiame and Luilu rivers
(Sankuru system) ; Dundo, Angola ; Stanleyville ( = Kisangani) (Wamba river) ;
Yangambe (Isalowe river - Uele system) ; Luberizi river (Ruzizi river) and
from Lake Tanganyika.
Poll (1953) thinks that the presence of Barbus caudovitattus (as defined here) in
Lake Tanganyika is fortuitous and that it essentially is a fluviatile species.
DIAGNOSIS. Barbus caudovittatus is a slender-bodied fish characterized by a very
soft, poorly ossified last simple ray in the dorsal fin and a dark band on each lobe
of the caudal fin. It is most likely to be confused with Barbus trachypterus and
Barbus paucisquamatus which it superficially resembles in body shape, scale counts,
and mouth position and mouth shape. It is not yet possible to determine whether
these three species are related as I have seen too few specimens of the latter two
species. The dorsal spines of Barbus trachypterus (x = 18-6, range 16-1-21-3) and
Barbus paucisquamatus (x = 19-7, range 14-1-22-5) are much longer and better
ossified than is normal in Barbus caudovittatus, but a few specimens have developed
A REVISION OF THE LARGE AFRICAN BARBUS
FIG. 24. Map of the distribution of B. caudovittatus.
spines of up to 17 per cent S.L. Dark pigment is present on the distal parts of the
pectoral, pelvic and anal fins in Barbus caudovittatus and Barbus paucisquamatus
but on the proximal parts of these fins in Barbus trachypterus. The barbels are much
shorter in Barbus caudovittatus (Ab, x = 4-4, range 2-3-7-5 ; Pb, x = 5-8, range
3-1-11-3) than in Barbus paucisquamatus (Ab, x = 7-5, range 5-8-10-6 ; Pb,
x = 8-6, range 6-2-11-6).
Barbus ethiopicus Zolezzi 1939
Barbus ethiopicus Zolezzi, 1939, Boll. Pesca Piscicolt. Idrobiol. Roma 15 : 369, fig. i.
HOLOTYPE. A fish of circa 230 mm S.L. from Lake Zwai. This specimen has
not been seen by me but is the only one described by Zolezzi. The specimen is
housed in the Laboratorio centrale di Idrobiologia in Rome.
DESCRIPTION. The description is based on three specimens of 240, 250 and
258 mm collected in Lake Zwai by Mr Eric Hamblyn. B.M. (N.H.) Nos. 1971.7.12 :
1-3-
K. E. BANISTER
'I-
A REVISION OF THE LARGE AFRICAN BARBUS 43
The range of ratios (expressed as percentage of S.L.) for the three specimens is
as follows.
D 23-2-24-6 CP1 15-8-17-0
H 25-8-27-1 CPd 8-4-10-0
I 5-8- 6-2 Snt 8-8- 9-4
IO 8-0- 8-8 Ab 3-6- 4-2
MW 7-0- 7-9 Pb 3-6- 4-7
Pet 16-3-18-7
Calculation of the standard deviation and the standard error with only three
specimens was not considered a profitable action.
The elongate body is round in cross-section and the caudal peduncle is long in
relation to its depth. The mouth is wide and the gape is terminal. The body was
covered with a thick layer of mucus which had to be removed before scale counts
could be taken. There are 39 (= 18 + 21) vertebrae.
Dorsal fin. The dorsal fin has four unbranched rays. The last unbranched ray
is weakly ossified as a smooth spine from 8-4 to 9-6 per cent of the standard length.
The three specimens have 6 (f.i), 7 (f.i) or 8 (f.i) branched dorsal fin rays. The
holotype has eight. This degree of variation is unusual and the number of branched
rays is low for the large Barbus species of East Africa although not infrequent in
the small Barbus species with radiately striated scales. The dorsal fin is inserted
slightly in front of the pelvic fins.
The anal Jin has three unbranched and five branched rays.
Squamation. The lateral line has 46 (f.i), 48 (f.i) or 52 scales (48 and 50 on
either side of the holotype), a number much higher than in any other large Barbus
species of the region. There are 18 scales round the caudal peduncle (18-20 in the
holotype according to Zolezzi). Between the dorsal mid-line and the lateral line
there are 9-5 (f.i) or 10-5 (f.2) scale rows and 10-5 (f.2) or 11-5 (f.i) between the
lateral line and the ventral mid-line. Five and a half scale rows are present between
the lateral line and the base of the pelvic fin.
Pharyngeal bones and teeth. The pharyngeal teeth number 2. 3. 5. -5. 3. 2 (fig. 26),
the first two teeth in the inner row are conical with small depressions on the posterior
face of the crown. In succeeding teeth the concavity becomes larger and as the
stem of the tooth becomes thinner the whole tooth assumes a spatulate appearance
with a recurved tip.
There are 10 (f.i) or n (f.2) gill rakers on the lower limb of the first gill arch in the
three specimens examined.
Coloration. The preserved fishes are olive grey-brown on the back and flanks to
just below the lateral line. The ventral part of the flanks and the belly are orangeish
to pinkish-brown. The dorsal fin and the dorsal lobe of the caudal fin are olive-
brown. The other fins are pale brown.
DISTRIBUTION. Known only from Lake Zwai, Ethiopia.
DIAGNOSIS AND AFFINITIES. Barbus ethiopicus is easily distinguished from the
other species described in this paper. It has a shallow, almost cylindrical body
44 K. E. BANISTER
FIG. 26. Left pharyngeal bone of Barbus ethiopicus. Dorsal and lateral views.
with more than 46 scales in the lateral line. Barbus microterolepis , known by one
specimen from a stream flowing into Lake Zwai, has only 40 scales and a much more
compressed body.
Small-scaled Barbus species from the region under discussion in this paper are
confined to lake basins, e.g. Barbus tropidolepis and Barbus platyrhinus of Lake
Tanganyika. However, it is not possible to decide, in these cases, whether the
small scales are an adaptation to a primarily lacustrine environment or not. The
geological history of the Lake Zwai basin is not known in enough detail to determine
how long the lake was isolated from the major rivers of the area and hence whether
Barbus ethiopicus represents a relict population from a once widespread small-
scaled species (and here it must be noted that there are fluviatile small-scaled
species in South Africa) or whether it is a modification of, for example, the Barbus
intermedius stock with which it is now sympatric. Barbus intermedius as well as
Barbus microterolepis is found in Lake Zwai. The possible relationships of the latter
species to Barbus ethiopicus are discussed on p. 91.
The evidence for the affinities of Barbus ethiopicus is inconclusive. There are no
characters which preclude its descent from a Barbus intermedius stock but equally
there are no characters which positively suggest a relationship with such a stock.
In the absence of positive information, the affinities of Barbus ethiopicus are left in
abeyance and it is not, in this paper, considered to be related to Barbus intermedius.
Barbus gananensis Vinciguerra 1895
Barbus gananensis Vinciguerra, 1895, Annali Mus. civ. Stor. nat. Giacomo Doria (2) 15 : 53,
pi. 5, fig. 2 ; Vinciguerra, 1897, Annali Mus. civ. Stor. nat. Giacomo Doria (2) 17 : 362.
HOLOTYPE. The holotype M.S.N.G. (Museo Storia Naturale Genova) No. 17525
from the River Ganana (= Juba) between Lugh and Bardera in Somaliland was
A REVISION OF THE LARGE AFRICAN BARBUS
45
^^TO^^^g^^.
FIG. 27. Barbus gananensis (the holotype after Vinciguerra - in Boulenger 191 la).
figured by Vinciguerra. It is now in extremely poor condition and I have not been
able to examine it although it was examined for me by Dr E. Tortonese. I have
though, through the kindness of Dr E. Tortonese, been able to examine two paralecto-
types. These are small fish of 50 and 39 mm S.L. and they came from the Auata
river (= Awata). A further specimen from the type locality, a fish of 176 mm S.L.,
M.S.N.G. No. 17339, was studied and this specimen is illustrated (fig. 28) to compare
with the original rather poor figure. A fourth fish, identified as Barbus gananensis
from the Citerni collection from the Upper Ganana, M.S.N.G. No. 17343, is not
referable to this species.
The description is therefore based upon three specimens.
DESCRIPTION. As a result of the description being based on two small fish and
one larger specimen, certain morphometric data show a large range. Only the
range is shown as other calculations could be misleading.
L
D
H
I
10
MW
39 -176 mm
28-2-32-0
25-0-28-2
7-4-10-2
6-0- 8-5
4'5- 7'7
Pet
CP1
CPd
Snt
Ab
Pb
23-1-24-0
16-5-18-0
13-1-15-0
7-7- 8-0
2-6- 6-3
6-0- 7-9
The calculations are expressed as percentages of the standard length.
The body is compressed and similar in outline to the body of Barbus bynni,
although rather less deep. The caudal peduncle is not so square as in Barbus bynni,
the ratio of the CPl/CPd is from 1-2 to 1-3 in Barbus gananensis against i-o to 1-2
in Barbus bynni of the same size.
Squamation. The scales bear parallel striae, slightly less sinuous than is typical
for Barbus bynni. There are 29 (f.2) or 31 (f.i) scales in the lateral line and 12 scales
K. E. BANISTER
FIG. 28. Barbus gananensis, a specimen of 176 mm S.L.
around the caudal peduncle. The other scale counts were only possible for the
largest specimen and here there are 5-5 scales between the dorsal mid-line and the
lateral line, 5-5 scales between the lateral line and the ventral mid-line and 2-5
scales between the lateral line and the insertion of the pelvic fin.
Dorsal fin. The dorsal fin has four unbranched rays and nine branched rays.
The base of the fin is enveloped by a sheath of scales. The last unbranched ray
forms a smooth ossified spine from 18-0 to 22-7 per cent of the standard length.
This is rather less than in similar-sized specimens of Barbus bynni. The spine in
the small specimens is conspicuously less well ossified than equal-sized Barbus bynni
specimens. The origin of the dorsal fin is slightly in advance of the origin of the
pelvic fin.
FIG. 29. A comparison of the left pharyngeal bones from a specimen of B. gananensis
of 176 mm S.L. (left bone) with the left pharyngeal bone from a Barbus bynni of
171 mm S.L.
A REVISION OF THE LARGE AFRICAN BARBUS 47
The anal fin has three unbranched rays and five branched rays. The last
branched ray is markedly bifurcated.
Pharyngeal bones and teeth. The contrast between the pharyngeal bones of the
176 mm S.L. Barbus gananensis and a 171 mm S.L. Barbus bynni can be seen in
fig. 29. The bone is thicker and deeper in the former and the teeth of the inner row
are higher. The second tooth of the inner row in particular is much longer in Barbus
gananensis than in Barbus bynni. These differences are equally clear in the pharyn-
geal bones of the smaller Barbus gananensis when compared with equal-sized
Barbus bynni. The teeth of the second and third rows are only marginally more
robust than in comparable Barbus bynni material.
Gill rakers. There are 12 gill rakers on the lower limb of the first gill arch in the
largest specimen ; the gill rakers could not be counted in the smaller specimens.
Coloration. In alcohol, the young fish are silvery flanked, with a brown back.
The largest fish is sandy brown on the flanks and belly, darker on the back.
DISTRIBUTION. Collected from the Juba river (between Lugh and Bardera) and
from the Awata river (upper Juba). Also it is reported from the confluence of the
Ganana Dulei and Bas Nabor. It has not been recorded from the Webi Shebeli
where Barbus bynni is found.
AFFINITIES. A species that is very close to Barbus bynni, differing in certain
features mentioned above. A larger collection of Barbus gananensis may show
modal differences further distinguishing it from Barbus bynni. The general im-
pression culled from these three fish is that they resemble a Barbus bynni stock
which has begun to approach the facies of the Barbus oxyrhynchus from the Athi and
Tana systems immediately to the south, particularly in possessing slightly longer
pectoral fins than is typical for Barbus bynni at these sizes, a reduced dorsal fin spine
and no less than nine branched rays in the dorsal fin. However, this is based solely
on an extremely small sample of fish and this trend (if it exists) must remain hypo-
thetical until more information is available about Barbus gananensis.
Barbus intermedius Riippell 1837
Barbus surkis Riippell, 1837, Mus. senckenb. 2 : 5, pi. i, fig. if ; Blgr., 1902, Ann, Mag. nat.
Hist. (7) 10 : 424 ; Blgr., 1907, Fish Nile : 226, pi. 42 ; Blgr., 1911, Cat. Afr. Fish 2 : 65.
Barbus intermedius Rupp., 1837, Mus. senckenb. 2 : 7, pi. i, fig. 2 ; Giinther, 1864, Proc. zool.
Soc. Lond. : 91 (non B. intermedius Riipp. -fide Blgr., 1911 : 45) ; Blgr., 1902, Ann. Mag.
nat. Hist. (7) 10:424; Blgr., 1907, Fish Nile : 228, pi. 29, fig. 2; Blgr., 1911, Cat. Afr.
Fish 2 : 59, fig. 39.
f In Tracts on Natural History, vol. 4, in the Zoological Department library of the British Museum
(Natural History) (Cat. No. 5 q T r) there is a separate of this Riippell paper bearing the date 1835.
The title page, pagination and figures are identical with those in the 1837 Mus. senckenb. except that the
title page bears the inscription 'Ausgezogen aus dem II Band des Museum Senckenbergianum. Frankfurt
am Main, 1835'.
As the cover of the second volume of the Mus. senckenb. bears the date 1836 and the title page the
date 1837, I am assuming that the 1835 date is either a typographical error or an over-optimistic estimate
of the date of publication. Until firm evidence to the contrary is available I shall continue to refer to
Riippell 1837.
48 K. E. BANISTER
Barbus intermedius leptosoma : Bini, 1940, Missione Stud. Lago Tana, 3 (2) : 167, pi. 6, fig. i
(vide B. leptosoma Blgr., 1902).
Barbus intermedius microstoma Bini, 1940, Missione Stud. Lago Tana, 3 (2) : 168, pi. 6, fig. 2.
Barbus intermedius gorgorensis Bini, 1940, Missione Stud. Lago Tana, 3 (2) : 168, pi. 6,
fig- 3-
Barbus affinis Riipp., 1837, Mus. senckenb. 2 : 8, pi. i, fig. 3 ; Blgr., 1902, Ann Mag. nat. Hist.
(7) 10 : 425 ; Blgr., 1907, Fish Nile : 216, pi. 29, fig. i ; Blgr., 1911, Cat. Afr. Fish 2 : 47,
fig. 28.
Barbus affinis brevibarbis : Bini, 1940, Missione Stud. Lago Tana, 3 (2) : 162, pi. i, fig. 2 (vide
B. brevibarbis Blgr., 1902).
Barbus affinis nedgia : Bini, 1940, Missione Stud. Lago Tana, 3 (2) : 163, pi. 2, figs, i and 2
(vide Labeobarbus nedgia Riipp., 1837).
Barbus gorguari i Riipp., 1837, Mus. senckenb. 2 : 9, pi. i, fig. 4 ; Blgr., 1902, Ann. Mag. nat. Hist.
(7) 10 : 424 ; Blgr., 1907, Fish Nile : 237, pi. 44 ; Blgr., 1911, Cat. Afr. Fish 2 : 75, fig. 53.
Barbus gorguarii macrophtalmus Bini, 1940, Missione Stud. Lago Tana, 3 (2) : 177, pi. 12, fig. 2.
Barbus elongatus Riipp., 1837, Mus. senckenb. 2 : n, pi. 2, fig. i.
Labeobarbus nedgia Riipp., 1837, Mus. senckenb. 2 : 14, pi. 2, fig. 3.
Barbus nedgia (Riippell) : Gunther, 1868, Cat. Fish 7 : 104 ; Blgr., 1902, Ann. Mag. nat. Hist.
(7) 10 : 426 ; Blgr., 1907, Fish Nile : 223, pi. 41, fig. 2 ; Blgr., 1911, Cat. Afr. Fish 2 : 51,
fig- 32-
Barbus gregorii Blgr., 1902, Ann. Mag. nat. Hist. (7) 10 : 422 ; Blgr., 1911, Cat. Afr. Fish 2 : 45.
Barbus rueppelli Blgr., 1902, Ann. Mag. nat. Hist. (7) 10 : 427 ; Blgr., 1911, Cat. Afr. Fish 2 : 26,
fig- 7-
Barbus macronema Blgr., 1902, Ann. Mag. Nat. Hist. (7) 10:427; Blgr., 1911, Cat. Afr.
Fish 2 : 25.
Barbus macronema var. parenzani Zolezzi, 1940, Boll. Pesca. Piscic. Idrobiol. 15 : 372.
Barbus harringtoni Blgr., 1902, Ann. Mag. nat. Hist. (7) 10 : 429 ; Blgr., 1911, Cat. Afr. Fish
2 : 61, fig. 40.
Barbus plagiostomus Blgr., 1902, Ann. Mag. nat. Hist. (7) 10 : 429 ; Blgr., 1907, Fish Nile : 232,
pi. 38, fig. 2 ; Blgr., 1911, Cat. Afr. Fish 2 : 68, fig. 47.
Barbus jar sinus Blgr., 1902, Ann. Mag. nat. Hist. (7) 10:429; Blgr., 1911, Cat. Afr. Fish
2 : 62, fig. 41.
Barbus platystomus Blgr., 1902, Ann. Mag. nat. Hist. (7) 10 : 430 ; Blgr., 1907, Fish Nile : 234,
pi. 40, fig. 2 ; Blgr., 1911, Cat. Afr. Fish 2 : 73, fig. 51.
Barbus platystomus var. vatovae Zolezzi, 1939, Boll. Pesca. Piscic. Idrobiol. 15 : 370.
Barbus platystomus platystomus : Bini, 1940, Missione Stud. Lago Tana, 3 (2) : 170, pi. 8, fig. i.
Barbus platystomus daga Bini, 1940, Missione Stud. Lago Tana, 3 (2) : 171, pi. 8, fig. 2.
Barbus platystomus dekkensis Bini, 1940, Missione Stud. Lago Tana, 3 (2) : 172, pi. 9, fig. i.
Barbus platystomus prognathus Bini, 1940, Missione Stud. Lago Tana, 3 (2) : 173, pi. 9, fig. 2.
Barbus brevibarbis Blgr., 1902, Ann. Mag. nat. Hist. (7) 10 : 431 ; Blgr., 1907, Fish Nile : 219,
pi. 35, fig. 2 ; Blgr., 1911, Cat. Afr. Fish 2 : 49, fig. 29.
Barbus kassamensis Blgr., 1902, Ann. Mag. nat. Hist. (7) 10 : 431 ; Blgr., 1911, Cat. Afr. Fish
2: 30, fig. ii.
Barbus oreas Blgr., 1902, Ann. Mag. nat. Hist. (7) 10 : 432 ; Blgr., 1911, Cat. Afr. Fish 2 : 38,
fig. 19 ; Pietschmann, 1913, Jb. Ver. Naturk. Wiesbaden 66 : 189.
Barbus leptosoma Blgr., 1902, Ann. Mag. nat. Hist. (7) 10 : 432 ; Blgr., 1907, Fish Nile : 220,
pi. 40, fig. i ; Blgr., 1911, Cat. Afr. Fish 2 : 50, fig. 30.
Barbus duchesnii Blgr., 1902, Ann. Mag. nat. Hist. (7) 10 : 433 ; Pellegrin, 1905, Bull. Mus.
Hist. nat. Paris 11 : 291 ; Blgr., 1907, Fish Nile : 208, pi. 35, fig. i ; Blgr., 1911, Cat.
Afr. Fish 2 : 31, fig. 12.
Barbus duchesnii maximus Bini, 1940, Missione Stud. Lago Tana, 3 (2) : 165, pi. 4, fig. 2.
Barbus duchesnii ibridus Bini, 1940, Missione Stud. Lago Tana, 3 (2) : 166, pi. 5.
Barbus mento Blgr., 1902, Ann. Mag. nat. Hist. (7) 10 : 434 ; Blgr., 1911, Cat. Afr. Fish 2 : 41,
fig. 22.
A REVISION OF THE LARGE AFRICAN BARBUS 49
Barbus hursensis Blgr., 1902, Ann. Mag. nat. Hist. (7) 10 : 434 ; Blgr., 1911, Cat. Afr. Fish
2 : 46, fig. 27.
Barbus degeni Blgr., 1902, Ann. Mag. nat. Hist. (7) 10 : 435 ; Blgr., 1907, Fish Nile : 222,
pi. 41, fig. i ; Blgr., 1911, Cat. Afr. Fish 2 : 50, fig. 31.
Barbus degeni leptorhinus Bini, 1940, Missione Stud. Lago Tana, 3 (2) : 164, pi. 3, fig. 2.
Barbus erlangeri Blgr., 1903, Proc. zool. Soc. Lond. (2) : 329, pi. 29 ; Blgr., 1911, Cat. Afr.
Fish 2 : 29, fig. 10.
Capoeta bingeri Pellegrin, 1905, Bull. Mus. Hist. nat. Paris 11 : 292.
Barbus ilgi Pellegrin, 1905, Bull. Mus. Hist. nat. Paris 11 : 293.
Barbus neuvillii Pellegrin, 1905, Bull. Mus. Hist. nat. Paris 11 : 294.
Barbus zuaicus Blgr., 1906, Ann. Mag. nat. Hist. (7) 17 : 562 ; Blgr., 1911, Cat. Afr. Fish 2 : 64,
fig. 43-
Barbus macmillani Blgr., 1906, Ann. Mag. nat. Hist. (7) 17 : 562 ; Blgr., 1911, Cat. Afr. Fish
2 : 67, fig. 46.
Barbus zaphiri Blgr., 1906, Ann. Mag. nat. Hist. (7) 17 : 563 ; Blgr., 1907, Fish Nile : 235,
pi. 45, fig. i ; Blgr., 1911, Cat. Afr. Fish 2 : 74, fig. 52.
Barbus bottegi Blgr., 1906, Ann. Mag. nat. Hist. (7) 17 : 563 ; Blgr., 1911, Cat. Afr. Fish 2 : 70,
fig. 48.
Barbus gudaricus Blgr., 1906, Ann. Mag. nat. Hist. (7) 17 : 564 ; Blgr., 1907, Fish Nile : 241,
pi. 34, fig. 2 ; Blgr., 1911, Cat. Afr. Fish 2 : 40, fig. 21.
Barbus eumystus Blgr., 1906, Ann. Mag. nat. Hist. (7) 17 : 565 ; Blgr., 1907, Fish Nile : 215,
pi. 38, fig. i ; Blgr., 1911, Cat. Afr. Fish 2 : 44, fig. 25.
Barbus margaritae Blgr., 1906, Ann. Mag. nat. Hist. (7) 17 : 565 ; Blgr., 1911, Cat. Afr. Fish
2 : 39, fig. 20.
Barbus alticola Blgr., 1906, Ann. Mag. nat. Hist. (7) 17 : 566 ; Blgr., 1911, Cat. Afr. Fish 2 : 55,
fig- 35-
Barbus bingeri : Blgr., 1911, Cat. Afr. Fish 2 : 69.
Barbus procatopus Blgr., 1916, Ann. Mag. nat. Hist. (8) 17 : 244.
Barbus volpinii Parenzan, 1940, Boll. Idrobiol. Cacc. Pesca. Afr. orient, ital. 1 : 9, fig. i.
Barbus brunellii Bini, 1940, Missione Stud. Lago Tana, 3 (2) : 173
Barbus brunellii acutirostris Bini, 1940, Missione Stud. Lago Tana, 3 (2) : 174.
Barbus dainellii Bini, 1940, Missione Stud. Lago Tana, 3 (2) : 175.
Barbus dainellii macrocephalus Bini, 1940, Missione Stud. Lago Tana, 3 (2) : 176.
Varicorhinus bingeri : Berlin and Esteve, 1947, Catalogue des Types des Poissons, Paris 4 : 44.
NOTES ON THE SYNONYMY. Although the specific name surkis appears before
intermedius in Riippell (1837), I have selected the latter name as the senior synonym
according to article 24 of the International Code. The reasons for this are :
a) Ruppell's text states that Barbus surkis has 36 scales in the lateral line but his
illustration (pi. i, fig. i) shows only 32 scales ; the text and figure of Barbus inter-
medius agree with the holotype ; b) the holotype of Barbus surkis is untraceable
but the holotype of Barbus intermedius exists as specimen 6778 in the Senckenberg
Museum, Frankfurt-am-Main ; c) the specific name intermedius is a great deal more
apposite for this variable species than is surkis.
A synonymy as extensive as this, involving 50 nominal species and subspecies
and resulting in one species with two subspecies requires a great deal of justification.
The evidence will be presented below.
Just over a hundred years after Riippell described the deep-bodied Barbus surkis
Bini (1940) described the elongated, predatory-looking Barbus brunellii. These
two species, ironically the first and (to date) the last described from Lake Tsana, are
very different in shape (see figs. 31, 32 and 46) and a simple comparison of an
5o K. E. BANISTER
example of each form would undoubtedly result in their being placed in separate
species. It is only when the extremes are compared with other Ethiopian specimens
that it becomes evident that one variable species is present instead of many, scarcely
distinct species. The degree of variability is great and is reflected throughout many
parts of the bodies of species ; head length may be singled out as one of the most
variable factors. Because of this plasticity of form it has proved impossible to
detect more than two statistically recognizable groups within the whole sample.
It had been suspected by Worthington (ig32a) that this would be the case from the
practical difficulties he encountered in trying to distinguish one nominal species
from another using the published descriptions.
The majority of species were described between 1902 and 1906, usually from a
very few, often poorly preserved specimens. The species were denned on the basis
of small differences in barbel lengths, lip formation, scale counts and body propor-
tions ; differences that may be tenable with few specimens but lose their significance
with a large series. Subsequent workers have tried with increasing difficulty to fit
newly acquired specimens into the framework of species established by Boulenger
(igna, 1916). Bini (1940) reduced several of Boulenger's species to subspecific
rank but at the same time described a further two species. Worthington (i932a :
127) from his work on the Kenyan and Ugandan lakes prophetically commented,
' thus, when a large series of specimens is collected from Lake Tsana and the upper
tributaries of the Blue Nile, it is probable that the series of species endemic to the
region - Barbus leptosoma, Barbus degeni, Barbus nedgia etc., will prove to be one
and the same species since they differ from each other almost solely in the degree of
lip development and in the size of the eye and dorsal spine'.
The graphs and calculations presented below will, I hope, show the essentially
normal distributions of the characters heretofore considered trenchant, and that it is
biologically more reasonable to suggest the presence of one variable species rather
than a multitude of almost indistinguishable, ill-defined or indefinable species.
The illustrations reproduced from Boulenger (igna) and Bini (1940) will help to
supplement the calculations and show the various intergrades between the most
extreme forms.
Worthington (i932a : 124) noted that the fishes listed by Boulenger (1911) as
Barbus bynni from Lake Baringo should be placed with Barbus gregorii ( = Barbus
intermedius).
Mann (1971) suggests that Barbus gregorii from Lake Baringo is rightfully placed
with Barbus tanensis (= Barbus oxyrhynchus) . His comparative material was
limited to small fishes (less than 60 mm S.L.) which can give rise to misleading
conclusions if compared with larger fish. In the scale counts, nature of the scale
striae and in most morphometric details the Lake Baringo fishes are identical with
Barbus intermedius from Ethiopia. However, the dorsal spine is modally longer
than in the other Barbus intermedius populations (although the spine is not as thick
as in Barbus tanensis} and the pectoral fin is longer. Because of these characters
and because Lake Baringo is isolated and the most southern locality for this species,
I think it would be valid to accord sub-specific status to the Barbus intermedius
population in Lake Baringo and I propose the name Barbus intermedius australis.
A REVISION OF THE LARGE AFRICAN BARBUS
FIG. 30. Barbus intermedius (Holotype from Riippell 1837).
FIG. 31. Barbus intermedius (Holotype of B. surkis from Riippell 1837).
It is perhaps worth noting that the longer dorsal spine and longer pectoral fin which
distinguish the Lake Baringo population are two characters possessed in greater
degree by Barbus oxyrhynchus.
HOLOTYPE. A dried specimen of 296 mm S.L. No. 6778, Senckenberg Museum,
Frankf urt-am-Main .
DESCRIPTION. The description is based on a total of 454 specimens ; 179 from
Lake Tsana, 58 from Lake Baringo, 44 from Lake Abaya, 32 from Lake Zwai, 32
from the Hawash river, 30 from Lake Stephanie, 29 from the Blue Nile, 24 from
K. E. BANISTER
FIG. 32. Barbus intermedius (figure of B. surkis from Boulenger 191 la).
FIG. 33. Barbus intermedius (Holotype of B. dainellii from Bini 1940).
Lake Rudolf and the Omo river, 21 from the Webi Shebeli river, four from Lake
Langano and one from Lake Orsodi.
All traceable holotypes have been examined ; those not seen were Barbus surkis
(Frankfurt, missing since the war), Barbus volpinii (not traceable), Barbus macronema
var. parenzani (not traceable) and Barbus dainellii (lost through poor preservation) .
n
X
s.d.
s.e.
range
L
454
66 -489 mm
D
454
28-4
2-3
o-i
20-7-37-7
H
454
26-4
2-2
O-I
19-0-31-7
I
454
5'9
1-2
0-05
3-5-10-2
10
454
8-0
0-8
0-03
4-7-12-3
MW
450
5'9
I-O
0-04
3-6- 9-1
Pet
452
21-8
1-8
o-i
15-6-28-0
CP1
454
16-9
1-6
o-i
11-7-22-9
CPd
453
n-6
1-2
o-i
8-0-15-0
Snt
454
8-5
I-I
0-05
5-5-12-7
Ab
449
5-8
1-9
o-i
i-8-ii-i
Pb
45i
6-9
2-1
o-i
1-9-12-1
A REVISION OF THE LARGE AFRICAN BARBUS 53
B
n . i mi-*!:'! ; I
5-
imMmwn
BN
WS
20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35
FIG. 34. Histograms of the distribution of the body depth (as a percentage of the standard
length) in B. intermedius populations from various localities. B = Lake Baringo ;
O = Omo river ; H = Hawash river ; A = Lake Abaya ; BN = Blue Nile ;
T = Lake Tsana ; WS = Webi Shebeli River ; Z = Lake Zwai ; S = Lake Stephanie.
54
K. E. BANISTER
70-
60-
50-
40-
30-
20-
10-
TrmTmTTTnTm
20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35
FIG. 35. Barbus intermedius : a composite histogram of the body depth for specimens
from all the localities treated separately in Fig. 34. The distribution of body depth
throughout the whole B. intermedius sample is normal.
FIG. 36. Barbus intermedius (Holotype of B. degeni from Boulenger 191 la).
FIG. 37. Barbus intermedius (figured specimen of B. nedgia in Boulenger 1911 a).
A REVISION OF THE LARGE AFRICAN BARBUS
55
wwmmmtim
FIG. 38. Barbus intermedius (Holotype of B. rueppelli in Boulenger 191 la).
FIG. 39. Barbus intermedius (figured specimen of B. gorguarii from Boulenger 191 la).
The morphometric data do not adequately reflect the variation in body form
shown by this species. A series of examples of different facies is shown in figs. 36
to 44. Names synonymized above are used below for convenience.
The body varies in shape from the slender Barbus dainellii form (fig. 33) to the
deep Barbus surkis form (fig. 32). It can be seen, though, from the histograms that
in the whole sample the body depth has a normal distribution (figs. 34, 35). There
is a little difference in body depth distribution in different localities, e.g. the Lake
Tsana population is modally less than the Lake Baringo and Omo river populations.
The mouth and lips exhibit a considerable degree of variation. 'Rubber-lipped'
forms are common, e.g. the forms described as Barbus degeni (fig. 36) and Barbus
nedgia (fig. 37) . Barbus rueppelli (fig. 38) and Barbus gorguarii (fig. 39) have large
mouths with the gape at 45 degrees to the horizontal ; however, Barbus kassamensis
(fig. 40) and Barbus hursensis (fig. 41) link this mouth type with the ventral or sub-
terminal mouths of the forms referred to Barbus oreas (fig. 42) and Barbus erlangeri
(fig. 43). The ventral mouth with a horny lower lip is found in Barbus macmillani
56 K. E. BANISTER
(fig. 44) and Barbus bingeri. Although the differences seem quite distinct when
isolated examples are considered, examination of a large number of specimens shows
that the mouth types grade into each other without any sharp divisions. The form
of the mouth in this species is valueless as a taxonomic character.
Four barbels are invariably present ; the Lake Stephanie population has modally
slightly longer barbels than any other population, but this is very possibly a reflec-
tion of the size range of the sample. The barbels are relatively smaller in larger
fishes, partly because of negative allometry and partly because of the increased
chance of physical damage. Histograms of the barbel length in different popu-
lations are shown in fig. 45. The head length varies considerably. The forms with
FIG. 40. Barbus intermedius (Holotype of B. kassamensis Boulenger 191 la).
FIG. 41. Barbus intermedius (Holotype of B. hursensis from Boulenger 191 za).
A REVISION OF THE LARGE AFRICAN BARBUS
57
FIG. 42. Barbus intermedius (Holotype of .B. oreas from Boulenger 191 la).
FIG. 43. Barbus intermedius (Holotype of B. erlangeri from Boulenger 191 la).
FIG. 44. Barbus intermedius (Holotype of B. macmillani from Boulenger 191 la).
K. E. BANISTER
1
55-
r
50-
45-
40-
35-
30-
25-
15-
1O
5-
T
HTITTTT1 Lrrrn ,
ws
15-
10-
5-
10-
5-
1234567
9 10 11 12
A REVISION OF THE LARGE AFRICAN BARBUS
is-
le-
s'
10-
5- [JTTT
TT
H
.[„
T
'1 1 ! 1 ! I 1 I
15-
10-
5-
ID-
S'
~~^~
BN
- - - - - II 1 1 1 1 . i llffirilHI
mn
: i , ^^ . T . . : : . : 1
1
8
10 11 12
FIG. 45. Histograms of the distribution of the anterior and posterior barbel lengths in
Barbus intermedius populations from various localities. The lengths are expressed as
percentages of the standard length. The shaded column represents the anterior barbel.
The locality codes are as in Fig. 34.
6o
K. E. BANISTER
FIG. 46. Barbus intermedius (figured examples of B. brunelli (above) and
B. brunelli acutirostris from Bini 1940).
90-
80-
70-
60-
50-
40-
30-
20-
10
19 20 21 22 23 24 25 26 27 28 29 30 31
FIG. 47. Histograms of the distribution of the head length expressed as a percentage
of the standard length for the entire Barbus intermedius sample.
A REVISION OF THE LARGE AFRICAN BARBUS
61
15-
10-
5-
10-
5-
PTTTTTTfliiililihTTTTin1
mrmTTTTTTJTTTTTmr^TO.
10-
5-
BN
30-
25-
20-
15-
10-
5-
5-
WS
10-
5-
19 20 21 22 23 24 25 26 27 28 29 30 31
FIG. 48. Histograms of the distribution of the head length expressed as a percentage of the
standard length for B. intermedius populations from various localities. Locality coding
as in Fig. 34.
62
K. E. BANISTER
15-
10-
5-
10-
5-
n,;- :i ,,.,
BN
50-
45-
40-
35-
30-
25-
20-
15-
10-
5-
WS
10-
5-
10-
5-
15 16 \7 18 19 20 21 22 23 24 25 26 27 28 29
FIG. 49. Histograms of the distribution of the pectoral fin length expressed as a percentage
of the standard length in populations from various localities. Locality coding as in
Fig- 34-
A REVISION OF THE LARGE AFRICAN BARBUS
100-
90-
80-
70-
60-
50-
40-
30-
20-
10-
15 16 17 18 19 20 21 22 23 24 25 26 27 28 29
FIG. 50. Histograms of the distribution of the pectoral fin lengths for the entire sample
of B. intermedius. Lengths are expressed as a percentage of the standard length.
100-
90-
80-
70-
60-
50-
40-
•
30-
20-
10-
wm
26 27 28 29 30 31 32 33 34 35
FIG. 51. Histogram of the distribution of the lateral line count for the entire
sample of Barbus intermedius.
64
K. E. BANISTER
15-
10-
5-
10-
5-
ID-
S'
'Tllmnm—
BN
40-
35-
30-
25-
15-
10-
MIL-,
ws
10-
5-
llllllil'iillll!!!
26 27 28 29 30 31 32 33 34 35
FIG. 52. Histograms of the distribution of the lateral line scale counts of Barbus intermedius
populations from various localities. Locality coding as in Fig. 34.
A REVISION OF THE LARGE AFRICAN BARBUS
40-
30-
20
10-
10 II 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33
FIG. 53. Histogram of the distribution of the dorsal spine lengths, expressed as a
percentage of the standard length of the entire Barbus intermedius sample.
the longest heads (Barbus brunellii, fig. 46, and Barbus gorguarii, fig. 39) are usually
large mouthed but the correlation is not perfect. The head length distribution is
normal (fig. 47) when the whole sample is considered, but the small samples from
some localities, e.g. the Blue Nile, give a chance bimodal or discontinuous distri-
bution (fig. 48). No significance can be attached to this.
The pectoral fin is modally longest in the Lake Baringo sample (fig. 48). The
histogram for the entire sample (fig. 50) is bimodal, suggesting a significant difference
in the Baringo fishes and these (see below) have been accorded sub-specific status.
Squamation. The scales bear numerous longitudinal striae. The lateral line
scale count range for the whole sample is from 26 to 35 (fig. 51). The range and
frequencies for each population show no significant variation (fig. 52). The lateral
line count is adequate for distinguishing a Barbus intermedius sample from a Barbus
altianalis sample, but is of no value for characterizing the subspecies of Barbus
intermedius. There are almost always twelve scales around the caudal peduncle.
The only exceptions are three specimens from Lake Tsana that have 13. Between
the dorsal mid-line and the lateral line there are 4-5 (f.24), 5 (f.i), 5-5 (f.38i),
6 (f.2), or 6-5 (f.33) scale rows. Between the lateral line and the ventral mid-line
there are 4 (f.i), 4-5 (f.i6o), 5 (f.2), 5-5 (1.231), 6 (f.i), 6-5 (f.3o) or 7-5 (f.i) scale
rows. Between the lateral line and the base of the ventral fin there are 1-5 (f.2),
2 (1.9), 2-5 (1246), 3 (f.4i), 3-5 (f.i37), 4 (f.i) or 4-5 (f.3) scale rows. The number of
scale rows is, not infrequently, different on either side of the specimen and in many
cases scales had been lost and it was impossible to count the original number.
Zolezzi (1940) gives 39-40 scales in the lateral line series of Barbus platystomus
var. vatovae. The only specimen in the collections of the Stazione Idrobiologia,
Roma (where Zolezzi's specimens were lodged), that is attributed to this variety and
is of the same length as Zolezzi's holotype has only 30 lateral line scales on one side
and 31 on the other. This specimen is not labelled as type material but agrees with
the only specimen described by Zolezzi in all respects except for the lateral line scale
count.
Dorsal fin. The origin of the dorsal fin varies from just in front to just behind
the vertical to the origin of the pelvic fins.
There are four unbranched rays. The last of these is ossified into a smooth spine,
x = 20-1, s.d. = 4-9, s.e. = 0-2, range 6-9-33-9 (percentage of the S.L.). The
range is unusually large and reflects the size range of the specimens examined and
66
K. E. BANISTER
tuo
9
,-
S m
cx .
O SuO
ae
i s
.
o, O
8 >>
O <U
* j
O CD
43 o
r-l !-
2 1>
£ (O,
°C
4-> rt
1/1
1T3 V)
A
O> ><j
g £
c3
O
4->
CO
3
A REVISION OF THE LARGE AFRICAN BARBUS
67
'•' '.*'.-•' :>
£
(50
a
c
i
I
M
<U
^3
00
d
"
68
K. E. BANISTER
FIG. 56. A comparison of the left pharyngeal bones from a shallow-bodied specimen of, A,
B. intermedius (previously identified as B. gorguarii) of 329 mm S.L. with a deep-bodied
specimen, B, (previously identified as B. macronema) of 332 mm S.L.
the negative allometry of the dorsal fin spine (see below). A histogram of the dis-
tribution of the length of the dorsal fin spine (fig. 53) shows that the whole sample is
bimodal. The fishes with the relatively longer spines all came from Lake Baringo
(fig. 54). The length of the spine serves as one of the distinguishing characters of
this population (see above). "Without the Lake Baringo fishes the dorsal spine has
a mean length of 19-1 and the same range as in the whole sample (6-9-33-9). These
values cover too wide a range to be useful as a diagnostic character. A graph of
the dorsal spine length (expressed as a percentage of the S.L.) against the standard
length (fig. 55) shows that the spines are relatively shorter in longer fishes. For fish
of less than 170 mm S.L. the dorsal spine values are x = 22-6; s.d. = 3-4; s.e. = 0-3;
range 16-3-33-7 '> f°r fishes of 171-250 mm S.L., x = 18-9; s.d. = 2-3; s.e. = 0-2;
range 13-8-24-5; for fishes of more than 251 mm S.L., x = 14-2; s.d. = 2-4;
s.e. = 0-2; range = 6-9-19-7.
Almost the same number of specimens have eight branched dorsal fin rays as
have nine. No other number was observed except in cases which were obviously
the result of physical damage.
A REVISION OF THE LARGE AFRICAN BARBUS
69
FIG. 57. The pharyngeal bones from a specimen of Barbus intermedius
(previously identified as B. surkis) of 348 mm S.L.
The anal fin invariably has three unbranched rays and five branched rays.
The gill raker count was only taken on 47 specimens. The range is from 10 to 16
on the lower arm of the first gill arch. The only factor which might have some
significance is that the long-headed fishes tend to have 10-12 gill rakers whilst the
shorter-headed fishes have 12-16 gill rakers. Not enough specimens were examined
to attribute any real significance to this difference.
Pharyngeal bones and teeth. The pharyngeal bones vary in shape. Generally at
an equal S.L. fishes with a proportionately longer head have thinner pharyngeal
bones than shorter-headed specimens ; fig. 56 contrasts the pharyngeals of a speci-
men of the ' gorguarii' form, a fish of 329 mm S.L., with the pharyngeals of a specimen
of the 'macronema' form of 323 mm S.L.
An even greater difference is immediately noticeable in the deep-bodied speci-
mens ; fig. 57 shows the pharyngeals from a specimen of the ' surkis ' form of
348 mm S.L. The bone in this case is extremely thick and the teeth are molariform.
A radiograph of this specimen [B.M. (N.H.) No. 1902.12.13 : 229] shows its stomach
to be full of gastropod mollusc shells. Possibly the difference in pharyngeal bone
strength reflects differences in diet and age. The left pharyngeal bone of a slender-
bodied Barbus brunellii is shown in fig. 58.
The pharyngeal teeth number, without exception, 2. 3.5. -5.3. 2. In fishes
< 180-200 mm S.L., the teeth are long with recurved, mammilliform crowns (as
shown in fig. 58) whilst larger and bulkier fishes have teeth like those shown in fig.
57. Needless to say these examples are linked by an almost continuous series of
intermediate shapes. I am unable to offer anything more than the most tentative
elucidation of this variation (see p. 127).
K. E. BANISTER
FIG. 58. Pharyngeal bones from a specimen of Barbus intermedius
(one of Bini's B. brunellii).
Little is known about the effect of age and diet upon cyprinid pharyngeal teeth.
The range in pharyngeal tooth form shown by Barbus intermedius is paralleled to
some extent by that in Barbus altianalis where the deep-bodied form hitherto
referred to Barbus obesus possesses singularly massive pharyngeal bones (see p. 14).
Coloration. This is very variable. In life the body colour ranges from silvery-
grey to blue, through green, olive and brown, to bronze. Some specimens have
been reported with pinkish or yellowish bellies. The fins have been described as
whitish tinged with crimson, deep-green, slaty-grey, yellowish and yellow-green. It
is quite possible that the colour may be a constant difference distinguishing popu-
lations, but there is not enough information available at the moment. In alcohol-
preserved specimens the body is grey, brown or bronzy, the scales are usually darker
at the base and the fins more or less the colour of the body.
DISTRIBUTION. Barbus intermedius is widely distributed throughout Southern
Ethiopia and into Northern Kenya, certainly as far as Lake Baringo (fig. 60). It has
been recorded from the following rivers : Omo system - Zendo, Gibe, Wondinak
and Ergino rivers ; Webi Shebeli system - Errer, Jerrer, Iraro and Modjo rivers ;
Hawash system - Kassam, Hurso, Akaki and Gota rivers ; Blue Nile system -
Wama, Urgessa, Gudar, Metti, Juju, Didessa and Mogre rivers ; Euasso Nyiro
system - Uaso Narok ; Juba system - Awata river ; Lake Zwai basin - Suksuki
and Maki rivers ; Lakes Abaya and Ganjule basins - Zeissi, Sire, Ganda, Elgo,
Alaba and Burka rivers ; Lake Stephanie basin - Zuja, Sagan, Gato and Bar] a
rivers ; Lake Rudolf basin - Kerio and Ngeng rivers. It has also been recorded
from Lakes Zwai, Tsana, Baringo, Langano, Abaya, Ganjule, Orsodi and Stephanie.
There has been a certain amount of confusion over the presence of Barbus gregorii
(= Barbus intermedius australis) in the Athi-Tana system as Mann (1971) noted.
A REVISION OF THE LARGE AFRICAN BARBUS 71
The localities for the syntypes are given by Boulenger (191 la) as : i Kiroruma
(Upper Tana), Leikipia ; 2-4, Guasso el Narua (Baringo), Leikipia ; 5-6, Guasso
Nyuki (Naiwasha), Njemps Ndogo and a skeleton, 7, Kibwesi river (Athi). All of
these specimens were collected by Professor J. W. Gregory's expedition. Mann
(ibid.) was unable to trace these localities with any certainty, but he points out that
the Leikipia plateau drains into the northern Euasso Nyiro and Lake Baringo, not
into the Tana system. Gregory (1896) in his account of his travels gives grid
references for Guasso Nyuki and Guasso el Narua. The former is a small stream
at o°28' N, 36°o8' E, slightly east of south of Lake Baringo, the latter is at the foot
of the Leikipia escarpment, to the east of Lake Hannington at o°i6' N, 38°i8' E.
Guasso Nyuki is nowhere near Lake Naivasha. The map of Gregory's route crosses
these rivers at the grid references given. The problem of Kiroruma is not so easily
FIG. 59. The distribution of B. intermedius : A = B. intermedius intermedius,
• = B. intermedius australis.
72 K. E. BANISTER
settled. Mann (ibid.) says that it is not on any modern maps. Gregory's (loc. cit.}
gazetteer gives the following information ; Kiroruma = Kiloluma o°4o' S, 37°3o' E.
The river is marked on his map and is in the upper Tana system, parallel to the
Thika-Thika river, Gregory's route did not pass through that grid reference (accord-
ing to his route map) and at the nearest was 24 km away from his location of the
Kiroruma river. From Gregory's text (1896 : 199) it appears that he was in a great
hurry at the time that he was in this area and there is no mention of specimens
having been collected. The Kiroruma is separated from the Leikipia escarpment
and plateau by Mount Kenya and no specimens of Barbus intermedius (sensu lato)
have been collected from neighbouring rivers in the Tana system (but see below).
At the moment this matter cannot be resolved.
Specimen No. 7 of Barbus gregorii in Boulenger (191 la : 46) from Kibwesi is
another problem. It bears the B.M. (N.H.) Reg. No. 1893.12.2 : 36. This number,
according to the register, is one of 16 specimens of Barbus tanensis ( = Barbus
oxyrhynchus) brought back by the Gregory expedition. Barbus intermedius and
Barbus tanensis were the only Barbus spp. brought back by the expedition and they
are easy to distinguish, especially so as the specimens are large. The other five
specimens bearing the locality Kibwesi are undoubtedly Barbus tanensis. This is
inexplicable. Three of Gregory's specimens of Barbus tanensis were recorded as
having come from Guasso el Narua (see above). This species has not been found
there since and I am at a loss to explain its alleged presence there. Perhaps the
locality was incorrectly recorded, there are enough inconsistencies in this collection
to cast a shadow of doubt over some of the localities. The Lake Baringo basin is
the most southerly definitive record of Barbus intermedius, but further south in the
rift valley before the Aberdare mountains and the Maui (or Mau) escarpment lie
Lakes Hannington, Elementaita, Nakuru and Naivasha. The first three of these
are extremely alkaline. Lake Naivasha according to Worthington (i932b, 1933)
and Copley (1948) has a small, cyprinodont, Aplocheilichthys antinorii, as its only
indigenous fish, but Tilapia nigra was introduced in 1925 and Micropterus salmoides
was introduced later. Copley (1948) mentions that no Barbus species are present
in Lake Naivasha but Parenzan (1939) lists Barbus gregorii ( = Barbus intermedius)
as present in the lake. Whether this locality of Parenzan's was based on first-hand
evidence or taken from Boulenger's (igna) list I cannot find out.
The significance of Aplocheilichthys antinorii is that it also lives in Southern
Ethiopia, i.e. the distribution is similar to that of Barbus intermedius. Cooke (1958),
in his reconstructions of the lower Pliocene drainage of east Africa, shows volcanic
highlands present in the Aberdare region (to the immediate south of Lake Naivasha)
and the 'Nile' and 'Athi-Tana' drainage areas much as they are today. This
ancient separation ol the two watersheds largely precludes the presence of many
species in common.
The northern Euasso Nyiro river presents certain problems, not the least of which
is the paucity of specimens. From this river in the region of Chanlers Falls come
the three Barbus erlangeri ( = Barbus intermedius) specimens whilst from the eastern
extremity of the system in the Nero-Narok and Ngau-Narok systems associated
with the Lorian swamp come 17 specimens of Barbus oxyrhynchus. The presence of
A REVISION OF THE LARGE AFRICAN BARBUS
90-
73
80-
70-
1 , " '
60-
50-
i
40-
30- -TTTTTTTT
20-
i
10-
i
1 ' 1
I'NIr! i 1
I i i ! ! ' : |
26 27 28 29 30 31 32 33 34 35 36
r
10-
20-
30-
40-
50-
60-
FIG. 60. Histograms comparing the distribution of the lateral line counts of
Barbus intermedius (above) with B. altianalis (below).
' Nilotic ' and an ' Athi-Tanan ' species in the same river system is surprising, but the
watersheds across the divide are low. This area is now semi-desert, but about the
time that Lake Baringo was connected to Lake Rudolf the water table must have
been much higher and water connections must have linked the two river systems in
this region.
The subsequent drop of water level has left the Euasso Nyiro system as a relict
area containing examples of the formerly contiguous faunae.
DIAGNOSIS AND AFFINITIES. Barbus intermedius is a variable species and bears a
close gross morphological resemblance to Barbus altianalis, to which I suggest it is
very closely related. These two species (and Barbus acuticeps and Barbus ruasae]
form a group of closely related species referred to here as the intermedius group or
complex.
Typically Barbus intermedius has a shallow compressed body, a caudal peduncle
longer than it is deep, numerous more or less parallel striae on the scales and a well-
ossified dorsal spine of moderate length and stoutness.
74 K. E. BANISTER
It is difficult to distinguish a specimen of Barbus altianalis from Barbus intermedium
if the locality is unknown. There are modal differences in the populations which
are listed under Barbus altianalis.
Both species are polytypic and can show considerable variation, particularly in
body shape and mouth form. However, it has always proved possible to distinguish
an unusually deep-bodied member of the Barbus intermedius group from a member
of the typically deep-bodied Barbus bynni group because members of the latter
group have more compressed bodies and longer, stronger dorsal spines.
Barbus intermedium is not, as far as I know, found in the same localities as Barbus
altianalis. The nearest approach of these two species appears to be in the poorly
collected Karasuk area to the north-east of Lakes Victoria and Kioga where the
headwaters of the Lake Rudolf basin streams and the Lake Victoria and Kioga
basin streams are very close.
The intermedium complex is discussed further on p. 128. The two subspecies of
Barbus intermedium can be distinguished, not only by locality, but also by the modally
much longer dorsal spine of Barbum intermedium australis (x = 24-9 against 19-1 in
Barbus intermedium intermedium}. Barbus intermedium australis also has a longer
pectoral fin (x = 23-4 against 21-5) and longer barbels (Ab, x = 7-1 against 5-6 ;
Pb, x = 8-4 against 6-7). Regrettably it is not always possible to place an individual
specimen, lacking locality data, in the correct group.
Barbus intermedius intermedius Riippell
A general description is given on p. 51. Them orphometric data for this sub-
species are :
n t x s.d. s.e. range
D 388 28-1 2-8 o-i 20-7-37-7
H 388 26-5 2-3 o-i 19-0-31-7
I 388 6-0 1-4 o-i 3-5-10-2
IO 388 8-0 0-9 0-04 4-7-11-6
MW 386 5-9 i-o 0-05 3-6- 9-1
Pet 388 21-5 2-0 o-i 15-6-29-0
CP1 388 17-0 1-6 o-i 11-7-22-9
CPd 388 11-5 1-2 o-i 8-6-15-0
Snt 388 8-6 1-2 0-05 5-4-12-7
Ab 385 5-6 1-8 o-i i-8-ii-i
Pb 387 6-7 2-1 o-i 1-9-12-1
DSp 385 19-1 4-2 0-2 6-9-33-9
Not all specimens examined are included in the morphometric data above.
The size range of the specimens is 94 to 489 mm S.L. The lateral line count
ranges from 26 to 34; 26 (f.7), 27 (f.27), 28 (f.49), 29 (f.85), 30 (1.85), 31 (£63),
32 (f.36), 33 (f.25), 34 (f.5).
Between the dorsal mid-line and the lateral line there are 4-5 (f.24), 5 (f.i),
5'5 (f-325)> 6 (f.2) or 6-5 (f.32) scale rows. Between the lateral line and the
A REVISION OF THE LARGE AFRICAN BARBUS 75
ventral mid-line there are 4 (f.i), 4-5 (f.iog), 5 (f.2), 5-5 (f.226), 6 (f.i), 6-5 (£.29) or
7-5 (f.i) scale rows. Between the lateral line and the base of the ventral fin there
are 1-5 (f.2), 2 (£.9), 2-5 (£.242), 3 (£.40), 3-5 (£.85), 4 (f.i) or 4-5 (£.3) scale rows.
DISTRIBUTION. The distribution is as on p. 70 except for Lake Baringo.
Barbus intermedius australis ssp. nov.
HOLOTYPE. A fish of 128 mm S.L., No. 18 in 1932.6.13 : 191-200, from a jar
labelled Barbus gregorii, Lake Baringo, in the collection of the B.M. (N.H.). This
specimen was selected as being close to the mean for most morphometric characters,
and therefore is typical of the population.
DESCRIPTION. The description is largely as for the nominate subspecies (see
p. 51). The morphometric data in detail are as follows and based on 58 specimens,
S.L. 66-388 mm.
n x s.d. s.e. range
L 66 -388 mm
D 58 30.0 2-2 0-3 26-3-35-1
H 58 26-2 1-7 0-2 23-0-31-0
I 58 6-0 i-o o-i 4-4- 9-1
IO 58 8-1 i-o o-i 6-1-12-3
MW 58 5-5 0-7 o-i 4-1- 7-3
Pet 58 23-4 1-5 O-2 2I-O-27-2
CP1 58 16-1 1-4 0-2 11-7-18-7
CPd 58 12-7 0-9 o-i 9-5-14-2
Snt 58 8-4 0-9 o-i 6-7-10-7
Ab 58 7-1 1-2 0-2 4-8-10-0
Pb 58 8-4 1-3 0-2 6-0-10-9
DSp 58 24-9 3-9 0-5 15-6-32-3
The number of scales in the lateral line ranges from 27 to 33 : 27 (£.4), 28 (f.n),
29 (f.i8), 30 (£.9), 31 (f.8), 32 (f.6), 33 (f.2). Only one specimen has 6-5 scale rows
between the dorsal mid-line and the lateral line, the rest have 5-5. Between the
lateral line and the ventral mid-line there are 4-5 (£.51), 5-5 (£.4), or 6-5 (f.i) scale
rows. Between the lateral line and the base of the pelvic fin there are 2-5 (f-4),
3 (f.i), or 3-5 (£.52) scale rows.
The distinguishing characters of the two subspecies are the longer dorsal spine,
longer pectoral fins, slightly deeper body and longer barbels in Barbus intermedius
australis.
Lake Baringo is an alkaline lake, and it is impossible to say whether the Baringo
population displays its particular phenotype as a result of the environment or as a
result of the genotype. There are very slight indications that the fishes from the
Omo river and Lake Rudolf incline slightly towards the Baringo facies but it must
be remembered that the Lake Rudolf and Omo river sample is extremely small.
DISTRIBUTION. This species is known from Lake Baringo, Kenya.
76 K. E. BANISTER
Barbus longifilis Pellegrin 1935
Barbus altianalis var. longifilis Pellegrin, 1935, Revue Zool. Bot. afr. 27 (3) : 376-385 (part, not
the specimen from Nyabarongo).
NOTES ON THE DETERMINATION. This species, from the upper reaches of the
Luhoho Congo, is not a variety of Barbus altianalis, but a well-defined species. One
of Pellegrin's type series (M.H.N.P. No. 35-75) is better referred to Barbus pauci-
squamatus. Poll (1939 : 69) synonymized Barbus altianalis var. longifilis with
Barbus altianalis var. paucisquamata, a move which does not take into account the
much longer dorsal spine, the much longer barbels and the much deeper body of the
former variety.
LECTOTYPE. A fish of 173 mm S.L. (M.H.N.P. No. 35-150). This is the least-
damaged specimen of Pellegrin's type series from the Loama river.
DESCRIPTION. The description is based on nine specimens, 132-247 mm S.L.
x s.d. s.e. range
D 31-2 1-6 0-5 29-2-34-0
H 24-6 0-7 0-2 23-5-25-3
I 6-5 0-5 o-i 6-0- 7-2
IO 8-0 0-5 0-2 7-0- 8-6
MW 5-5 0-6 0-2 4-5- 6-3
Pet 23-1 i-o 0-3 21-7-24-2
CP1 17-0 i-o 0-3 15-8-18-3
CPd 11-4 0-7 0-2 9-9-12-1
Snt 8-3 0-6 0-2 7-5- 9-3
Ab 8-9 0-7 0-2 8-3- 9-6
Pb 11-3 i-i 0-4 9-5-13-4
Barbus longifilis has a pointed snout. Except for a slight nuchal hump the dorsal
profile is straight from the snout to the origin of the dorsal fin. The mouth is ventral
with thin fleshy lips ; the barbels are characteristically long.
Dorsal Jin. Has IV-g (f.8) or IV- 10 (f.i) rays, the last unbranched ray is solidly
ossified into a thick straight spine (x = 25-6, s.d. = 2-7, s.e. = i-o, range 22-0-
30-3). The dorsal fin origin is usually in front of the insertion of the pelvic fins. A
low sheath of large scales surrounds the base of the dorsal fin.
The anal fin has three unbranched and five branched rays.
Squamation. The striae on the scales are slightly irregular, either parallel or
slightly converging, the lateral line has 25 (f.i), 26 (f.3), 27 (£3) or 28 (f.2) scales.
There are 4-5 scales between the dorsal mid-line and the lateral line and 5-5 between
the lateral line and the ventral mid-line. There are 2-5 (f-7) or 3 (f.2) scale rows be-
tween the lateral line and the base of the pelvic fins and without exception 12 scales
are present around the caudal peduncle.
There are between 10 and 12 gill rakers on the lower arm of the first gill arch in
the specimens examined.
Pharyngeal bones and teeth. The pharyngeal tooth formula is 2. 3. 5. -5. 3. 2. The
first tooth in the inner row (fig. 62) is conical and directed posteriorly, the second
A REVISION OF THE LARGE AFRICAN BARBUS
77
s
K. E. BANISTER
FIG. 62. The left pharyngeal bone from the lectotype of Barbus longifilis.
tooth is taller and more flattened laterally than the first. A small concavity on the
posterior face of the crown creates a slight hook, the hook and the concavity pro-
gressively enlarge on teeth 3, 4 and 5. The teeth of the second and third rows are
small, less extreme versions of the posterior tooth on the inner row.
Coloration. The body of preserved specimens is brassy, slightly darker on the
back than on the belly. A band of dark-brown pigment is present on the middle
third of the fin membrane of the dorsal and pelvic fins and extends from the middle
to the end of the pectoral, anal and caudal fins, the density of the pigment varies
considerably from specimen to specimen, especially the caudal fin.
DISTRIBUTION. The specimens examined came from the Loama and Kansehete
rivers, Congo.
DIAGNOSIS AND AFFINITIES. Barbus longifilis is a member of the Barbus bynni
group (see p. 36). The much longer barbels distinguish this species from Barbus
bynni and Barbus gananensis. In other respects it most closely resembles Barbus
oxyrhynchus and were the two species found in the same area it may be difficult to
assign some individual specimens (especially preserved material) to either of the two
species. The longer barbels and more heavily pigmented fins in Barbus longifilis
will usually enable it to be distinguished from Barbus oxyrhynchus where locality
data are unavailable.
Cooke (1958 : 26) presents some evidence to suggest that some Athi-Tana faunal
elements had been able to move across the Lake Victoria area. He cites the presence
of the Athi river species Tilapia nigra which occurs in the mid-Pleistocene beds at
Rawi in the Kavirondo Gulf. The relationship of Lake Victoria to the headwaters
of the Congo in Pleistocene times is discussed on page 22. There is just a possibility
A REVISION OF THE LARGE AFRICAN BARBUS
79
that there could have been a movement of Barbus oxyrhynchus (or its ancestor) from
the Athi into the Congo before the rift valley broke the connection (see fig. 4 in
Cooke, op. cit.) and that Barbus longifilis represents a surviving population descended
from this migration.
Barbus macrolepis Pfeffer 1889
Barbus macrolepis Pfeffer, 1889, Jb. hamb. wiss. Anst. 6 (2) : 17 ; Pfeffer, 1893, Jb. hamb. wiss.
Anst. 10 (2) : 34, pi. i, fig. i ; Pfeffer, 1896, Thierw. O-Afr. Fische : 63.
LECTOTYPE. The lectotype was selected by Ladiges et alii (1958) as a fish of
108 mm S.L. (this specimen was kindly measured for me by Dr W. Ladiges), Reg.
No. H.330 from Mbusini on the Wami River, Tanzania.
DESCRIPTION. The description is based on 20 fishes, from 48 to 243 mm S.L.
L
D
H
I
10
MW
Pet
CP1
CPd
Snt
Ab
Pb
x
27-6
7-6
8-9
6-8
21-7
14-6
14-8
7-9
3'5
5-8
s.d.
3'5
2-3
i-4
0-9
0-8
1-6
1-9
i'4
1-7
i'3
2-5
s.e.
0-8
o-5
o-3
0-2
0-2
0-4
0-4
o-3
0-4
o-3
0-6
range
48 -243 mm
29-9-38-2
24-4-31-3
4-6-10-1
7.3-10-4
5-3- 8-3
18-5-24-6
11-4-20-3
12-9-17-2
5-8- 9-0
1-6- 6-3
1-4- 9-0
FIG. 63. Barbus macrolepis (from Boulenger igua).
8o
K. E. BANISTER
5 mm
FIG. 64. A scale from the shoulder of B. macrolepis to show the position of the striae.
The three paralectotypes examined have the overall body shape of the figured
example. The larger fish have a straighter ventral profile and a more humped
back. The mouth is sub-terminal or just ventral, horse-shoe shaped ; in all the
specimens the lips are slightly thickened and fleshy.
Of four specimens radiographed, three had 36 vertebrae and one had 37.
Dorsal fin. Has four unbranched rays and 9 (f.y) or 10 (f.i3) branched rays.
The last unbranched ray is weakly ossified with persistent articulations distally.
The length of the non-articulated part ranges from 8-7 to 20-8 per cent S.L. with a
mean value of 15-3. The whole ray though, when unbroken, is 35 per cent of the
S.L. which gives a high dorsal fin, with a markedly concave dorsal margin. This
ray is more elongated in larger fish. The syn types have a very small sheath of scales
at the base of the dorsal fin. -This sheath is not present in the larger fish ; as there
is no sign of physical damage it presumably has been lost as a result of growth. The
leading edge of the dorsal fin is slightly in advance of the pelvic fin.
The anal fin has three unbranched rays and five branched rays. The last ray of
the latter is in some cases markedly bifurcated.
Squamation. There are 22 (f.7), 23 (f.5), 24 (f.4) or 25 (f.i) scales in the lateral
line ; 3-5 (f.6) or 4-5 (f.i2) scales between the dorsal mid-line and the lateral line
and 3 (f.i), 3-5 (f-7) or 4-5 (f.i) scales between the lateral and the ventral mid-line.
On several specimens the scales could not be counted reliably. One and a half (f.4)
or 2-5 (f.i6) scale rows are present between the lateral line and the insertion of the
pelvic fin. There are 12 scales around the caudal peduncle.
The striae on the scales are characteristic (fig. 64). They are comparatively
sparse on the exposed portion of the scale and converge towards the posterior edge
of the scale. In this respect they contrast significantly with those of Barbus
oxyrhynchus (fig. 81).
Pharyngeal bones and teeth (figs. 65 and 66). The first tooth of the inner row is
small, conical with a small spoon-shaped depression at the crown. The second
A REVISION OF THE LARGE AFRICAN BARBUS
81
tooth is much thicker and higher with a hooked crown. The teeth become pro-
gressively thinner posteriorly and the hook and the depression become more exag-
gerated. The pharyngeal tooth formula is 2.3. 5. -5.3. 2.
Gill rakers. There are 12-14 broad, hooked gill rakers on the ventral limb of the
first gill arch.
Coloration. Preserved specimens are light brown on the back and a paler silvery-
brown on the lower part of the flanks. The scales are dark edged. The caudal
and dorsal fins are brownish, the other fins are hyaline.
FIG. 65. The left pharyngeal bone from a specimen of B. macrolepis of 94 mm S.L.
FIG. 66. The left pharyngeal bone from a specimen of B. macrolepis of 243 mm S.L.
82
K. E. BANISTER
Juba
FIG. 67. A map of the distribution of Barbus macrolepis.
DISTRIBUTION. This species is known from Katare (or Kotare), Malagarasi
swamp ; from the Wami, Ruaha and Rufiji rivers, Tanzania.
The distribution of this species is rather unusual in that it is confined to a narrow
belt between 5 and 8 degrees South but covers a wide longitudinal range from
Malagarasi to near the Tanzanian coast. The Malagarasi river drains westwards into
Lake Tanganyika whilst the other two rivers flow eastward into the Indian Ocean.
The Malagarasi has certain Congo faunal affinities and its geological history
suggests that it was once part of the Congo system. The eastward flowing rivers
do not have any Congo links.
DIAGNOSIS AND AFFINITIES. Barbus macrolepis is a distinctive species. The
body is deep, the last simple dorsal fin ray is tall and weak, the caudal peduncle is
nearly as deep as it is long and the striae on the scales converge markedly. The
striae on the scales of Barbus marequensis are parallel, but the tall crescentic dorsal
fin of the larger specimens is also present in specimens of Barbus marequensis from
the Zambezi river to the south of the Rufiji river.
The more heavily ossified last simple ray of the dorsal fin of Barbus oxyrhynchus
in the rivers to the north of the Wami and the parallel sinuous striae on the scales
are easily distinguished points of difference from Barbus macrolepis.
Poll (1967 : 181) is of the opinion that Barbus jubbi has some marked similarities
to Barbus macrolepis, principally in the high number of dorsal fin rays and the low
number of scales in the lateral line series. However, the striae on the scales are
quite different, as is the dorsal fin spine and at the moment I am inclined to think
that the resemblances are spurious.
A REVISION OF THE LARGE AFRICAN BARBUS 83
A very much greater degree of morphological similarity exists between Barbus
macrolepis and Barbus lagensis from Nigeria. Both species have ten unbranched
rays in the dorsal fin ; a high but weak fourth unbranched ray in the dorsal fin ; a
comparable low number of scales in the lateral line series and very similar striations
on the scales.
I have not seen sufficient Barbus lagensis material to comment further on the
relationships of these two species but both seem very different from the other
African Barbus species and if the characters in common are not the result of con-
vergence, then the real possibility exists that these two species are related and if so
then their distribution may indicate that they are relicts of an early invasion of
Barbus species (see p. 132). I have not seen any other African Barbus species that
have converging striae on their scales and the combination of a deep body, large
scales and a high dorsal fin without a well-ossified dorsal spine is also unique. Some
of these characters are found in some Asian Barbus species, e.g. converging striae
are found in Barbus longispinis Gunther, this Celanese species has striae which
converge more with age but it has a strong dorsal spine. Barbus macropus Blkr.
and Barbus huguenini Blkr. respectively from Borneo and Sumatra are deep-bodied
fishes with large scales and few converging striae ; however, they have a serrated
dorsal spine. I have not been able to find any mainland Asiatic species which have
all the characters under discussion, but it seems that the presence of converging
striae is more frequent in Asiatic Barbus species than it is in African Barbus species.
I have not been able to examine as many Indian species as I would have liked but
the occurrence of the converging striae in some of the Asiatic island species suggests
that it may be a primitive feature (or at least an ancient one) which is present in
two African species. At the moment I cannot with any certainty align Barbus
macrolepis with any Asiatic species because it is impossible to show that the simi-
larities in the pattern of scale striae are not due to convergence. If it is not due to
convergence, then it is possible that there may be some fairly close relationship
between a group of Asiatic Barbus species and a pair of African species. This idea,
though, assumes that the differences in the dorsal spines are of less significance than
the similarities in the scale striations and sadly this is a matter on which I have no
information at all.
Barbus mariae Holly 1929
Barbus mariae Holly, 1929, Anz. Akad. Wiss. Wien 66 (4) : 34 ; Copley, H., 1958, Common
Freshwater Fishes of E. Africa : 78-80.
Barbus rhinoceros Copley, 1938, //. E. Africa Uganda nat. Hist. Soc., 13 : 191.
NOTES ON THE SYNONYMY. It may well eventually prove necessary to synonymize
Barbus mariae with Barbus matris, Holly, 1928 [Zoo/. Anz. Leipzig 85 (1-2)] from
the Athi river at Nairobi. Holly's description of Barbus matris is very similar to
that of Barbus mariae but I am unwilling to proceed in this matter without examin-
ing the holotype of Barbus matris and at the time of writing this has not been
located.
LECTOTYPE. Holly described this species on the basis of two specimens of
280 mm and 295 mm total length from the Kitui river (Athi system) in Kenya. I
84 K. E. BANISTER
have not seen either of these specimens which are supposed to be in the Natural
History Museum, Vienna, but they could not be located by Dr P. Kahsbauer who
kindly searched for them. The larger specimen is designated the lectotype on the
assumption that both specimens are extant.
DESCRIPTION. The description is based on five specimens in the B.M. (N.H.)
collections of standard lengths, 86, 105, 112, 117 and 342 mm from the Athi river.
x range
D 26-9 24-4-29-5
H 30-7 30-1-32-5
I 8-5 5-5-10-4
10 6-8 5-8- 8-5
MW 5-2 4-5- 7-0
Pet 22-5 21-9-22-5
CP1 17-3 16-1-19-6
CPd 10-6 9-3-12-9
Snt 9-9 8-9-10-4
Ab 2-7 1-2- 3-8
Pb 5-3 4-5- 5-8
All measurements are expressed as percentages of the standard length. With this
particular sample it was not considered useful to calculate the standard deviation
and the standard error.
Barbus mariae is a distinctive species. The upper jaw has a remarkable median
protrusion (the 'rhinoceros horn' of Copley 1958) when the mouth is open. This
is caused by the fish having a large kinethmoid (sensu Alexander 1966). The
antero-ventral edge of the labial part of the premaxilla is gently curved and over-
hangs the lower jaw giving a 'clupeoid' appearance to the jaws.
The body is long and thin, the dorsal fin has its origin in the posterior half of the
body more or less vertically above the insertion of the pelvic fin.
Dorsal fin. It has TV-g rays. The fourth unbranched ray is heavily ossified,
smooth and from 17-5 to 32-6, x = 29-0, of standard length (negatively allometric).
The dorsal margin of the fin is markedly concave.
Anal Jin. With III-5 rays.
Squamation. Holly gives 30 or 31 for the lateral line scale count of the specimens
he described. The specimens I examined have 27 (f.i), 28 (f.i), or 29 (f.3) scales.
The difference is probably attributable to the different techniques used in counting
these scales. The figures given by Holly agree with the complete number of scales
in the lateral line series if 27 or 28 are present to the end of the hypurals. There
are 12 scales around the caudal peduncle, 4-5-5-5 between the dorsal mid-line and
the lateral line and 4-5 between the lateral line and the ventral mid-line. One and
a half or 2 scales were present between the lateral line and the base of the pelvic fin.
Pharyngeal bones and teeth. The pharyngeal tooth formula is 2.3. 5. -5. 3. 2. The
first tooth in the inner row is about two-thirds of the length of the second. The
second tooth is the widest. All in this row are unicuspid and recurved and the
A REVISION OF THE LARGE AFRICAN BARBUS
in
£
cc
vO
K. E. BANISTER
FIG. 69. The left pharyngeal bone from a specimen of Barbus mariae of 342 mm S.L.
teeth become more hooked, taller and thinner posteriorly. Only the pharyngeal
teeth of the largest specimen were examined.
Gill rakers. On the largest specimen 9 + 1. The gill rakers were widely spaced
and hooked forward. This number is lower than is usual among the large Barbus
species under consideration in this revision.
Coloration. Holly's (op. cit.) remarks on the coloration match the colour of the
preserved specimens I examined. The back is dark ochre-brown shading to a
silvery-yellow in the belly. The fins are very light brown and the scales have dark
edges. Copley (1958) states that the live fish are olive-brown on the back and
silvery on the belly.
DISTRIBUTION. The five specimens 1936.12.22 : 35-39 in the B.M. (N.H.) col-
lections are from the Athi river. Specimen No. B.M. (N.H.) 1937.6.4 : 16 has only
the locality Kenya. Copley (1958) records it from the Athi and Tana systems, he
also mentions that they can weigh up to 48 Ib.
A REVISION OF THE LARGE AFRICAN BARBUS 87
DIAGNOSIS AND AFFINITIES. Barbus mariae is most easily distinguished by the
'rhinoceros horn '-like development of the kinethmoid and the 'clupeoid' appear-
ance of the upper jaw. Barbus mariae is sympatric with Barbus oxyrhynchus but
lacks many of its characters and I am unable, at the moment, to comment further
on its relationships.
Barbus microbarbis David and Poll 1937
Barbus microbarbis David & Poll, 1937, Annls. Mus. r. Congo Beige Zool. (i) 3 (5) : 261 (only
the holotype).
HOLOTYPE. A fish of 216 mm S.L. from Lake Luhondo, Rwanda ; M.A.C.T.
No. 41847. The two paratypes (M.A.C.T. Nos. 41848, 41849) do not belong to
this species, nor do any of the other specimens in the M.A.C.T. collection (Nos.
56449-56455, 56479, 92966-92968, 94219-94220). This species is split because
the holotype has a sub-terminal mouth, five teeth in the inner pharyngeal row, a
smoothly curved pharyngeal bone outline, nine branched rays in the dorsal fin and
15 gill rakers on the lower limb of the first gill arch (in contrast to the ventral mouth,
horny covered lower jaw, four teeth on the inner pharyngeal row, an oddly shaped
pharyngeal bone, 10 branched dorsal fin rays and 19-20 gill rakers on the lower limb
of the first gill arch of the paratypes and the other M.A.C.T. specimens). All the
specimens previously included in this species are now considered to belong to
Varicorhinus ruandae Pappenheim, & Boulenger, 1914.
DESCRIPTION. The description is based on the holotype. D = 31-0 ; H = 23-6 ;
I = 5-6 ; IO = 9-2 ; MW = 7-4 ; Pet = 19-9 ; CP1 = 157 ; CPd = n-i ; Snt = 3-3.
The mouth is sub-terminal, the lower jaw is curved without a horny margin and both
anterior and posterior barbels are present although very short. Scales on the
posterior part of the body have parallel striae but are radiately striated on the
shoulders ; 32 scales in the lateral line, 12 around the caudal peduncle, 5-5 between
the dorsal mid-line and the lateral line ; 5-5 between the lateral line and the ventral
mid-line and 2-5 between the lateral line and the base of the pelvic fin.
The dorsal fin has four unbranched rays, the last is moderately thickened and
ossified into a smooth spine, the non-articulated part of which is n-o per cent of
the standard length. There are nine branched rays and the dorsal margin of the
fin is slightly concave.
The anal fin has three unbranched and five branched rays. David & Poll
(1937) give six branched rays but they were apparently misled by the complete
bifurcation of the last ray.
Pharyngeal bones and teeth. The right pharyngeal bone is shown in fig. 71 ; the
fifth tooth in the inner row is notable for the presence of a ridge across the concave
crown. All the teeth are hooked.
There are 16 short, broad gill rakers on the lower limb of the first gill arch.
Coloration. Described by David & Poll (op. cit.} as dark grey on the back, lighter
grey below. Scales with a black border. The colour now is dark brown on the
back and lighter brown below.
88
K. E. BANISTER
I
I
I
A REVISION OF THE LARGE AFRICAN BARBUS
FIG. 71. The right pharyngeal bone from the holotype of Barbus microbarbis.
DISTRIBUTION. This species is known only from Lake Luhondo, Rwanda.
DIAGNOSIS AND AFFINITIES. Barbus microbarbis is characterized by having very
short barbels and a wide sub-terminal mouth with a gently curving lower jaw. In
these two characters this species approaches members of the ill-defined genus
Varicorhinus Riippell but is clearly separate from Varicorhinus ruandae (see above)
with which it is sympatric.
It is unwise to speculate too much on the basis of one specimen. Barbus micro-
barbis could be an abnormal specimen of Barbus altianalis or Barbus acuticeps.
Groenewald (1958) has shown how the mouth shape can change markedly. This
individual could be a hybrid between a Barbus species and a Varicorhinus species.
Without more specimens and a greater knowledge of the fauna of the area the
relationships and nature of this species must be left in abeyance.
Barbus microterolepis Boulenger 1902
Barbus microterolepis Blgr., 1902, Ann. Mag. nat. Hist. (7) 10 : 426 ; Blgr., 1911, Cat. Afr.
Fish 2 : 23, fig. 5.
HOLOTYPE. A fish of 118 mm S.L. from the Maki river, Lake Zwai, Ethiopia,
B.M. (N.H.) Reg. No. 1902.12.13 : 220.
DESCRIPTION. The description is based on the only known specimen of this
species. All measurements are expressed as percentages of the standard length.
K. E. BANISTER
FIG. 72. Barbus microterolepis holotype (from Boulenger 191 za).
D = 27-1 ; H = 25-4 ; I = 7-6 ; IO = 7-6 ; MW = 5-9 ; Snt = 7-6 ; Pet = 21-2 ;
CP1 = 16-9 ; CPd = 10-2 ; Ab = 5-1 ; Pb = 6-8.
The shape of the body can be seen in fig. 72.
Squamation. The lateral line has 40 scales. There are 7-5 scale rows between
the dorsal mid-line arid the lateral line and 6-5 scale rows between the lateral line
and the ventral mid-line. There are four scales between the lateral line and the
base of the pelvic fin. Around the caudal peduncle there are 15 or 16 scales. The
scales bear numerous fine parallel striae.
Dorsal fin. The dorsal fin origin is slightly in front of the pelvic fin insertion. It
has four unbranched rays, the last of which is ossified into a smooth spine of length
22 '9 per cent. There are eight branched rays the last of which is bifid.
The anal fin has five branched rays and three unbranched rays.
Pharyngeal bones and teeth. The first tooth in the inner row (fig. 73) is small and
conical, the second tooth is longer and thicker with a recurved, unicuspid crown.
The other three teeth in this row become progressively thinner and more hooked.
There are no molariform teeth. The pharyngeal tooth formula is 2. 3.5. -5. 3. 2.
Gill rakers. Ten moderately stout, slightly curved gill rakers are present on
the lower limb of the first gill arch. Three gill rakers are present on the upper limb.
Coloration. Boulenger (op. cit.) describes the colour of the preserved specimen
as olive above, silvery beneath.
DISTRIBUTION. This species is known only from the Maki river, which is a fast-
flowing stream rising in the hills of Gouaza and emptying into Lake Zwai.
AFFINITIES. Any conclusions about this species, based on one specimen, must
be regarded as tentative. The Zwai basin contains two other endemic cyprinids,
Barbus ethiopicus and Garra makiensis (Blgr. 191 la; Menon 1964). The geology
of the area is not known in enough detail to be able to decide whether the endemism
A REVISION OF THE LARGE AFRICAN BARBUS
FIG. 73. The left pharyngeal bone from the holotype of Barbus microterolepis .
is the result of a long isolation or not. There are several possibilities concerning
Barbus microterolepis. It could be an abnormal specimen of Barbus intermedius in
which the number of scales has been increased by a chance mutation. Barbus
intermedius lives in the Zwai basin and differs from Barbus microterolepis in the size
of the scales. Barbus microterolepis could be a valid species descended from a
common ancestry with Barbus intermedius having an increase in scale numbers. If
this is so one must conclude that the species must be rare or inhabiting inaccessible
regions as only one fish has ever been collected. A third possibility is that Barbus
microterolepis is a hybrid between the small-scaled Barbus ethiopicus of Lake Zwai
and Barbus intermedius. There is no evidence at the moment that Barbus micro-
terolepis is not a valid species but more specimens are needed before any firm con-
clusions regarding its affinities can be reached.
Barbus mirabilis Pappenheim and Boulenger 1914
Barbus mirabilis Pappenheim & Blgr., 1914, Wiss. Ergebn. dt. ZentAfr. Exped. Zool. (3) 5 : 239.
HOLOTYPE. A fish of 353 mm S.L., Z.M.B. Reg. No. 19059.
DESCRIPTION. The description is based on the holotype as I have not been able
to find any other specimens which have been referred to this species. D = 30-8 ;
H = 24-6 ; 1 = 4-2; IO = io-o ; MW = 7-1 ; Pet = 19-8 ; CP1 = 13-8 ;
CPd = 1 1 -9 ; Snt = 8-5 ; Ab = 4-7 ; Pb = 5-4. All these measurements are
expressed as percentages of the standard length.
K. E. BANISTER
FIG. 74. Barbus mirabilis holotype.
The shape of the body can be seen in fig. 74 and it presents a rather bulky
appearance.
Squamation. The striae on the scales are slightly diverging, less so on the shoulder
scales than on the flanks or belly. There are 28 scales in the lateral line (Pappenheim
and Boulenger's count of 31 is the total number). Five and a half scale rows are
between the dorsal mid-line and the lateral line and the same number between the
lateral line and the ventral mid-line. Three scale rows lie between the lateral line
and the base of the pelvic fin. The scales on the chest are reduced in size and are
less well formed. There are 12 scales around the caudal peduncle.
The mouth is sub-terminal and crescentic with a continuous lower lip.
Dorsal fin. Has four unbranched rays. The last ray is ossified into a smooth
spine, the non-articulated part of which is 9-6 per cent of the standard length.
There are ten branched rays. The origin of the dorsal fin is in front of the pelvic
fins. There is no sheath of scales at the base.
Coloration. The preserved fish is pale brown in colour with a darker lower lobe
of the caudal and a dark edge to the pectoral fins. The back and sides above the
lateral line are slightly darker than the belly.
Gill rakers. There are 10 gill rakers on the lower arm of the first gill arch.
Pharyngeal bones and teeth. The pharyngeal teeth (figs. 75 and 76) have slightly
hooked crowns. There is little enlargement of the second tooth in the inner row.
The pharyngeal tooth formula is 2.3. 5. -5.3.2.
DISTRIBUTION. Known only from Mawambi on the Ituri river (Congo system).
AFFINITIES. In its general shape, coloration, striations of the scales, scale and
gill raker counts and pharyngeal teeth, Barbus mirabilis closely resembles Barbus
somereni. The barbels are shorter than in Barbus somereni but the unique specimen
of Barbus mirabilis is much bigger than the largest Barbus somereni examined.
Barbus somereni also usually lacks the dark edge to the pectoral fin.
A REVISION OF THE LARGE AFRICAN BARBUS
93
Without more specimens I cannot synonymize these two species nor state their
relationship with any degree of certainty. It seems possible that, although Barbus
mirabilis comes from the Congo system on the other side of the rift valley, it was
found only some 150 km from the Ruwenzori mountains where Barbus somereni is
common and could represent a population of Barbus somereni that lived in the
westward-flowing rivers of that area and became isolated when the rift valley formed.
There is a certain amount of circumstantial evidence to support this idea. Barbus
FIG. 75. The dorsal aspect of the left pharyngeal bone from the holotype of
Barbus mirabilis.
5mm
FIG. 76.
The lateral aspect of the left pharyngeal bone from the holotype of
Barbus mirabilis.
94
K. E. BANISTER
somereni has a wide, scattered distribution, usually at high altitudes suggesting
that it was once a widespread species and changes in the environment (whether
tectonic or climatic) or competition have driven it into higher altitude streams.
What little geological evidence is available (see under Barbus altianalis] suggests
that the Congo watershed extended further eastwards than it does now and that its
eastern extremities were cut off and their direction changed by the rift valley
formation. It would therefore seem possible that an old-established large Barbus
species (which this must be if the arguments have any veracity at all) could be
found on both sides of the rift valley. More specimens from Mawambi are badly
needed before this matter can be pursued further. I have not been able to examine
Pappenheim & Boulenger's (op. cit.) unique specimen of Barbus mawambi also from
Mawambi but from their description and figure I am tempted to think that this
species could be the same as Barbus mirabilis.
Barbus oxyrhynchus Pfeffer 1889
Barbus oxyrhynchus Pfeffer, 1889, Jb. Hamb. Wiss. Anst. 6 (2) pi. 8 ; Pfeffer, 1896, Thierw.
O-Afr. Fische : 64.
Barbus tanensis Gunther, 1894, Proc. zool. Soc. Lond. : go, pi. n ; Blgr., 1911, Cat. Afr. Fish
2 : 58 ; Fowler, 1936 (Part), Proc. Acad. nat. Sci. Philad., 88 : 287, fig. 50 (as Barbus
(Lanceabarbus) tanensis) .
Barbus hindii Blgr., 1902, Proc. zool. Soc. Lond. (2) : 222, pi. 16, fig. i ; Pappenheim & Blgr.,
1914, Wiss. Ergebn. dt. ZentAfr. Exped. Zool. (3) : 239 ; Pellegrin, 1909, Mem. Soc. zool.
Fr. 22 : 281-298.
Barbus (Capoeta) perplexicans Blgr., 1902, Proc. zool. Soc. Lond. (2) : 223, pi. 16, fig. 2 ; Pellegrin,
1909, Mem. Soc. zool. Fr. 22 : 281-298.
Barbus labiatus Blgr., 1902, Proc. zool. Soc. Lond. (2) : 223, pi. 17, fig. i.
Barbus krapfi Blgr., 1911, Cat. Afr. Fish 2 : 54.
Barbus mathoiae Blgr., 1911, Cat. Afr. Fish 2 : 66.
Barbus ahlselli Lonnberg, 1911, K. svenska Vetensk. Akad. Handl. 47 (6) : 39.
Barbus athi Hubbs, 1918, Fieldiana 12 (2) : 9-16.
^Barbus babaulti Pellegrin, 1926, Bull. Soc. zool. Fr. 51 : 384.
Barbus nairobi Holly, 1928, Zool. Anz. 75 (3-4) : 1-4.
Barbus donyensis Holly, 1929, Sber. Akad. Wiss. Wien (4) : 32-35.
Varicorhinus babaulti : Berlin & Esteve, 1947, Catalogue des Types des Poissons Paris : 45.
NOTES ON THE SYNONYMY. Pfeffer (1889) described Barbus oxyrhynchus from
two small specimens from the Rufu river (Pangani system). Barbus tanensis was
described from much larger fish (from 188 to 285 mm S.L.) from the Thika-Thika
river (Tana system). Barbus hindii, Barbus perplexicans, Barbus labiatus, Barbus
krapfi and Barbus mathoiae, all came from the Athi, Tana and Pangani rivers or the
Lorian swamp, and they were separated principally on slight differences in propor-
f Pellegrin (1935, Revue Zool. Bot. afr. 27: 382) subsequently described a small Barbus from Lake Kivu
under the name Barbus (Agrammobarbus) babaulti. Greenwood (1962) synonymized this species with
Barbus apleurogramma, Blgr. 191 la. The International rules of Zoological Nomenclature (Arts. 53 and
60) require one to give a replacement name for Barbus (Agrammobarbus) babaulti, Pellegrin 1935, which
is here rejected as a primary junior homonym of Barbus babaulti Pellegrin 1926; I propose that Pellegrin's
Lake Kivu species be called Barbus lapsus.
A REVISION OF THE LARGE AFRICAN BARBUS 95
tion, especially of the lips, dorsal spine and body depth. The same characteristics
were used to differentiate the other species from the same localities, viz. Barbus
ahlselli, Barbus athi, Barbus babaulti, Barbus nairobi and Barbus donyensis.
Fowler (1936) was of the opinion that the continued separation of Barbus hindii,
Barbus tanensis and Barbus mathoiae was unjustified. Allowing for the known
vagaries of lip development (Groenewald 1958) and for the allometry of various
parts of the body with growth it seems reasonable to conclude that all these nominal
species from a few adjacent rivers are preferable to the same species. Certainly
the adults have the same general appearance and, as the histograms indicate, normal
variation (i.e. Poisson distribution) of various characters is present when large series
of specimens are examined. The only species over which I have any doubts is
Barbus labiatus (known only from two specimens) which has a shallower body and a
slightly longer head than the others. But with only two specimens available I
cannot attribute too much significance to slight differences.
It has not proved possible, using the characters considered useful in this paper,
to divide Barbus oxyrhynchus into smaller groups. This is only to be expected as
the overall range of morphometric characters is not very great and the head waters
of the Athi and Tana systems are extremely close.
One fish (Coll. M.A.C.T. No. 47341) from the Malagarasi river was attributed to
Barbus krapfi by David (1937) and to Barbus lestradei (= Barbus caudovittatus] by
Poll (1953). Barbus oxyrhynchus has not been found in the Malagarasi, yet I
cannot easily place this specimen with any of the Malagarasi species. The pectoral
fin is slightly shorter than in Barbus oxyrhynchus, but in other features it corresponds
well. I can only assume either that Barbus oxyrhynchus is found in the Malagarasi
but is extremely rare or that this specimen is a deformed member of another species
if it is not a new species.
A specimen in the Stockholm Museum, No. 8061, identified as Barbus percivali
Blgr. ( = Barbus neumayeri Fischer, fide Greenwood 1962 : 178) from Kibonoto on
the Sanya river, belongs to Barbus oxyrhynchus.
There are two fish from the Ruaha river (Rufiji system) in the collection of the
Central Africa Museum, Tervuren, which belong possibly to this species. The
Rufiji is to the south of the Pangani and its fauna is very poorly known. These two
fish are extremely deep bodied, their standard lengths are 156 and 164 mm with
body depths of 38-4 and 40 per cent respectively. Their other features are within
the Barbus oxyrhynchus range except for the dorsal spines which are short and
articulated distally (respectively 11-5 and 15-2 per cent). Although at the moment,
these specimens are considered as belonging to this species they are not included in
the data given in the description below. Further specimens may show that the
Rufiji fishes are of a different species or are a discrete population of Barbus oxyrhyn-
chus. Pappenheim & Boulenger (1914) record Barbus hindii from the Ituri river
but the very low lateral line count would seem to refer these fish to Barbus
mawambiensis.
LECTOTYPE. The lectotype, a fish of 48 mm S.L. from the Pangani river, Ham-
burg Museum, No. 11339, was selected by Ladiges et alii (1958). I have not been
able to examine this specimen but it was compared for me by Dr W. Ladiges who
96
K. E. BANISTER
FIG. 77. Barbus oxyrhynchus paralectotype.
kindly sent me the largest paralectotype (40 mm S.L.) which is figured above
(fig- 77)-
DESCRIPTION. The description is based on 108 specimens, including, in addition
to the types of Barbus oxyrhynchus, the types of Barbus tanensis, Barbus hindii,
Barbus perplexicans, Barbus labiatus, Barbus krapfi, Barbus mathoiae, Barbus
ahlselli, Barbus babaulti and Barbus athi. The size range of the specimens is from
28 to 369 mm S.L. Because the lectotype is a juvenile fish figs. 78, 79 and 80 show
larger specimens indicating the variations produced by growth in this species.
The morphometric data are expressed in tabular form as follows and unless
otherwise stated are expressed as a percentage of the standard length.
D
H
I
IO
MW
Pet
CP1
CPd
Snt
Ab
Pb
26-9
7-6
8-9
6-4
24-4
16-9
12-8
8-5
5-9
7-1
s.d.
2-8
2-5
1-9
i-i
0-9
1-9
1-4
i-i
0-9
1-4
1-4
0-2
O-2
O-I
O-I
O-2
O'l
O'l
O'l
O-I
O'l
range
26-1-39-4
23-0-32-6
4-4-n-i
6-9-12-5
4-3- 8-7
21-3-28-6
14-1-20-0
10-8-15-5
6-3-11-0
3-1-10-7
3-6-12-4
The standard deviation is higher than for many species described in this paper
because of the size range and marked allometry of the specimens examined. This
is especially noticeable in the body depth ; in fishes of less than 100 mm S.L.
(measurements expressed as a percentage of the standard length unless otherwise
A REVISION OF THE LARGE AFRICAN BARBUS
97
FIG. 78. Barbus oxyrhynchus (type of B. labiatus from Boulenger 191 la).
FIG. 79. Barbus oxyrhynchus (type of B. hindii from Boulenger 191 xa).
stated), x = 29-3 ; s.d. = 2-2 ; s.e. = 0-4 ; range = 26-1-35-0 ; the upper end of
this range is extended by two small fishes from the Athi river which have a pro-
nounced keel leading up to the dorsal fin. Fishes of S.L. 101-170 mm, x = 32-2 ;
s.d. = 1-6 ; s.e. = 0-2 ; range = 29-5-35-8. Fishes of S.L. above 171 mm,
x = 33'6 ; s.d. = 2-7 ; s.e. = 0-5 ; range = 28-4-39-4. In this case the lower end
of the range is extended by the shallow-bodied specimens previously referred to
Barbus labiatus (see below).
The typical adult fish has a fairly flat ventral profile. The dorsal profile rises
sharply from the snout to the origin of the dorsal fin after which it follows a sloping
concave course down to the caudal fin. The same type of body shape is shown by
the two specimens attributed to Barbus labiatus except that the body is less deep.
The snout is obtusely pointed. The mouth ranges from ventral and curved in
most specimens through the wide, cutting mouth of the types of Barbus perplexicans
to the rubber lips of the type of Barbus labiatus. The mean length for the pectoral
fin is sufficiently great for it to serve as a diagnostic character for the species.
98
K. E. BANISTER
FIG. 80. Barbus oxyrhynchus (type of B. krapfi from Boulenger 191 la).
FIG. 81. A scale from a specimen of Barbus oxyrhynchus (the type of B. mathoiae)
to show the distribution of the striae.
The caudal peduncle is compressed, short and deep with a mean length/depth
ratio of 1-32.
Squamation. The scales bear many parallel or slightly converging striae (fig. 81).
The lateral line count is low, from 21 to 28 scales; 21 (f.i), 22 (f.io), 23 (f.J-5),
24 (f-36), 25 (f.3o), 26 (f.8), 27 (f.6), 28 (f.i). The type of Barbus labiatus has
24 on one side and 27 on the other. The lateral line follows a rather dipping course.
There are 4-5 (f.ioo) rarely 3-5 (f.6) or 5-5 (f.2) scales between the dorsal mid-line
and the lateral line and 4-5 (f.8i) rarely 3-5 (f.i6) or 5-5 (f.3) scales between the
lateral line and the ventral mid-line. This count was unobtainable from some
A REVISION OF THE LARGE AFRICAN BARBUS 99
specimens. There are 2-5 (f.65), 2 (f./j.) or 1-5 (f-32) scales between the lateral line
and the base of the pelvic fin. Again this count was not possible in some specimens.
Around the caudal peduncle there are 12 (1.99) or n (f.5) scales ; this count too
was unobtainable for some specimens.
Dorsal fin. There are four unbranched rays. The last of these is ossified into a
long, moderately broad, straight or gently curving spine. This fourth ray has a
mean length of 25-3 with a range from 16-6 to 35-7 per cent, s.d. = 4-6, s.e. = 0-5.
There are from eight to ten branched rays : 8 (1.5), 9 (f.82), 10 (f.2i) ; the types
of Barbus oxyrhynchus are somewhat atypical in having only eight branched rays.
The dorsal margin is strongly concave and the base of the dorsal fin frequently has
an enveloping sheath of scales.
The anal fin has three unbranched rays and five (very rarely six) branched rays.
Pharyngeal bones and teeth. The pharyngeal tooth formula is constant at
2. 3. 5. -5. 3. 2. The first tooth in the inner row is tall, thin and has a slightly hooked
crown. It is slightly angled towards the second tooth. This tooth is longer than
wide (fig. 83) and again hooked distally. Teeth 3, 4 and 5 become progressively
thinner and develop articular surfaces on the posterior face of the crown. The
fifth tooth is distinctly recurved but the terminal hook remains. The teeth of the
second and third rows resemble the fifth tooth of the ventral row in general shape
but are much smaller.
FIG. 82. Scatter diagram of the dorsal spine length as a percentage of the standard length
against the standard length for the entire sample of Barbus oxyrhynchus.
K. E. BANISTER
FIG. 83. The left pharyngeal bone from Barbus oxyrhynchus (a syntype of B. tanensis).
The gill rakers range from n to 16 on the lower limb of the first gill arch.
Coloration. Copley (1958) states that the colour of Barbus tanensis varies in live
fish from olive-green to deep olive-brown on the back shading to silver on the belly.
Fowler (1936) figures Barbus tanensis with a dark spot on the caudal peduncle ; I
have not seen this in any of the specimens I examined. The colour in alcohol is
sandy-brown on the back shading to a pale brown to silver on the belly. The scale
margins are frequently a deeper shade of brown than the centre of the scale.
DISTRIBUTION. Specimens are known from the Rufu river, Pangani system ;
the Kibwesi, Makindu, Tsavo, Regati, Mathoiya and Thika-Thika rivers, Athi-Tana
systems ; the Rufiji river ; Nero-Narok system, Lorian swamp (in the Northern
Euasso Nyiro system). This species is widespread throughout the Athi and Tana
systems (Copley 1958).
The locality G. el Narua given in Boulenger (191 la : 57) is the subject of
some dispute and is discussed in detail on p. 71.
I have not had the opportunity to examine any Barbus specimens from Mozambique
north of the Zambezi so it is possible that the range of Barbus oxyrhynchus may
extend further south.
DIAGNOSIS AND AFFINITIES. Barbus oxyrhynchus is closely related to Barbus
bynni, Barbus gananensis and Barbus longifilis and the points of differences between
Barbus oxyrhynchus and the three other species will be found on pp. 36 and 129.
Comments on clinal phenomena within this group of species and exemplified by
Barbus oxyrhynchus are found on p. 129.
A REVISION OF THE LARGE AFRICAN BARBUS 101
Barbus pagenstecheri Fischer 1884
Barbus pagenstecheri Fischer, 1884, Jb. hamb. wiss. Anst. 1 : 30 ; Pfeffer, 1896, Thierw. O.-Afr.
Fische 5 : 65 ; Boulenger, 1911, Cat. Afr. Fish 2 : 72.
LECTOTYPE. The original description was based on two specimens (Nos. H34I
and H342 in the Hamburg Museum of 315 and 217 mm S.L. respectively). Boulen-
ger (191 la) redescribed the species on the basis of the smaller specimen and implied
that this specimen was the lectotype (he used the word ' type ' and mentioned that
a larger specimen had also been referred to this species). Ladiges et alii (1958),
in a non-revisional work, designated the larger specimen (H34i) as the lectotype.
Here I follow Boulenger in accepting the smaller specimen as the lectotype and
regarding the larger specimen as the paralectotype.
DESCRIPTION. The description is based on the only two specimens referred to
this species. The morphometric data are shown below and in each case the lectotype
comes first. L = 217, 315 mm ; D = 26-3, 29-8 ; H = 24-0, 30-8 ; I = 6-4, 4-8 ;
IO = 7-4, 10-2 ; MW = 6-9, 8-9 ; Pet = 20-3, 24-8 ; CP1 = 17-5, 17-5 ; CPd
= io-i, 12-3 ; Snt = 6-9, 9-8 ; Ab = 4-1, 5-1 ; Pb = 5-2, 6-4.
Unless otherwise stated all measurements are expressed as percentages of the
standard length.
5cm
FIG. 84. Barbus pagenenstecheri lectotype.
The body (fig. 84) is elongated, the dorsal profile rises gradually and smoothly
from the snout to the origin of the dorsal fin. The caudal peduncle is about half as
long again as it is deep. The mouth is sub-terminal with a sharp edge to the lower
jaw in the lectotype but rubber lips are developed in the paralectotype. The snout
is blunter in the lectotype than in the paralectotype. Both specimens have a clearly
defined rostral groove. Numerous small tubercules are present on the snout and
the cheeks of both specimens.
Squamation. There are 26 or 27 scales in the lateral line series, 4-5 (f.2) scales
between the dorsal mid-line and the lateral line, 4-5 (f.2) scales between the lateral
line and the ventral mid-line and 2-5 (f.2) scales between the lateral line and the
IO2
K. E. BANISTER
5 mm
FIG. 85. Barbus pagenenstecheri - dorsal and lateral views of the left pharyngeal bone.
base of the pelvic fin. Twelve scales encircle the caudal peduncle. The exposed
parts of the scales bear numerous, sinuous, more or less parallel striae.
Dorsal fin. There are four unbranched rays and eight branched rays in the dorsal
fin. The last unbranched ray is weakly ossified into a smooth spine, 14-3, 13-7 per
cent S.L. The dorsal margin of the fin is slightly concave. The dorsal fin origin
is slightly in advance of the pelvic fin origin.
The anal fin has three unbranched rays and five branched rays.
Pharyngeal bones and teeth. Pharyngeal bones are present only in the paralecto-
type. The pharyngeal teeth number 2.3.5. -5.3. 2. The teeth are widely spaced
(fig. 85). The lateral flange is confined to the angle ot the bone level with the
fourth and fifth teeth of the inner row.
Gill rakers. There are 12 or 15 gill rakers on the lower limb of the first gill arch.
Coloration. The colour of alcohol-preserved specimens is dark brown on the back
paling to yellow-brown on the ventral surface. The edges of the scales on the back
and dorsal part of the flanks are lighter than the centres. The fins are dark grey-
brown.
DISTRIBUTION. Fischer (1884) collected the fish from a stream flowing down
Mount Kilimanjaro. Bailey (1969) gives the locality as in the Pangani system.
There is no evidence for this because only the streams on the south face of Mount
Kilimanjaro flow into the Pangani. The streams on the eastern face flow into the
Tsavo river whilst the streams on the northern and western faces have no outlet
to the sea. All of the streams are covered by the locality description ' German East
Africa'. Detailed information on the expeditions' collecting sites is lacking.
DIAGNOSIS AND AFFINITIES. Barbus pagenstecheri is a species of uncertain affini-
ties, its distribution is localized and it is surrounded by waters containing Barbus
A REVISION OF THE LARGE AFRICAN BARBUS 103
oxyrhynchus. I have not seen any specimens, nor have I been able to find any
reliable records, of Barbus oxyrhynchus from the Kilimanjaro region and it is there-
fore possible that the two specimens referred to Barbus pagenstecheri may represent
a local population of Barbus oxyrhynchus. The differences between these two fishes
and equal-sized specimens of Barbus oxyrhynchus are marked. The former fishes
have shallow bodies and weak dorsal spines whilst the latter have deeper bodies
and strong dorsal spines. The lateral line scale count in Barbus pagenstecheri is
at the upper end of the Barbus oxyrhynchus range and the scale striations are similar
in both species. Barbus mariae from the Athi river can easily be distinguished from
Barbus pagenstecheri by its 'rhinoceros horn' (see p. 84). With only two specimens
available it is, I feel, preferable to maintain the species Barbus pagenstecheri rather
than to regard these specimens as local variants of Barbus oxyrhynchus. Further
collections from the Kilimanjaro region are needed to reach a satisfactory con-
clusion.
Barbus paucisquamatus Pellegrin 1935
Barbus altianalis var. labiosa (part) Pellegrin, 1933, Bull. Soc. zool. Fr. 58 : 169 (only the
paratype, from Loama).
Barbus altianalis var. paucisquamata Pellegrin, 1935, Revue Zool. Bot. afr. 27 (3) : 378.
Barbus altianalis var. longifilis (part) Pellegrin, 1935, Revue Zool. Bot. afr. 27 (3) : 378 (only
the specimen from Nyabarongo).
Barbus altianalis var. lobogenysoides Pellegrin, 1935, Revue Zool. Bot. afr. 27 (3) : 380.
NOTES ON THE SYNONYMY. The specimens of the varieties of Barbus altianalis
described by Pellegrin (op. cit.) which came from rivers in the Congo system consti-
tute a species distinct from Barbus altianalis.
The name labiosa, although the oldest infra-specific name, is not available for this
newly defined species as the holotype of that variety (a fish from Lake Kivu) is a
specimen of Barbus altianalis altianalis. The type series of Barbus altianalis var.
paucisquamata is homogeneous and the infra-specific name paucisquamata (when
the gender is changed to form paucisquamatus - Article 30) is available and is used
accordingly.
There are certain inaccuracies in the original description of the specimens of
Barbus altianalis var. paucisquamata. Pellegrin (1935 : 379) lists three specimens
of lengths 130 + 35 = 165, 140 + 40 = 180, 130 + 35 = 165. The measurements
presumably refer to the standard length, ' tail fin ' length and the total length. The
three syntypes (Paris Museum Nos. 35-76, 35-77, 35-78 are of 124, 164 and 116 mm
S.L. respectively. M.A.C.T. specimen 42932 is registered as a 'co-type', there
is a label with this fish saying 'co-type, don. de Mus. Paris Loc. Kivu, rec. Guy
Babault'. This fish is of 177 mm S.L.
LECTOTYPE. Specimen 35-76 in the Paris Museum is designated the lectotype.
This specimen of 124 mm S.L. is closest in size to any of the Pellegrin measurements.
It is also the only specimen with a precise locality which is Kitembo, Nyabarongo
river.
io4
K. E. BANISTER
d
fi
A REVISION OF THE LARGE AFRICAN BARBUS 105
DESCRIPTION. The description is based on 12 specimens from 80 to 248 mm S.L.
x s.d. s.e. range
L 80 -248 mm
D 26-8 1-6 0-5 24-0-30-0
H 25-5 1-3 0-4 23-8-27-6
I 7-2 1-2 0-3 5-4- 9-4
IO 7-6 0-9 0-2 6-3- 8-5
MW 5-8 0-4 o-i 5-2- 6-2
Pet 21-2 0-9 0-3 2O-2-22-8
CP1 18-0 1-2 0-3 i6'4-i9'7
CPd ii-o i-o 0-3 9-4-13-0
Snt 8-0 0-9 0-3 6-8- 9-3
Ab 7-5 1-4 0-4 5-8-10-6
Pb 8-6 1-7 0-4 6-2-11-2
All measurements are expressed as percentages of the standard length. The body
is compressed. The dorsal profile of the snout is curved, the ventral profile is flat.
The mouth is ventral and horse-shoe shaped. The lower lip is well defined and
varies between continuous with a slight mental lobe and medially discontinuous.
The upper lip forms the ventral edge of the snout. The barbels are conspicuously
long.
Squamation. The lateral line has 27 (f.3), 28 (f.4) or 29 (f-5) scales. Between
the dorsal mid-line and the lateral line there are 4-5 (1.12) scales. Only seven
specimens were in sufficiently good condition for the scales between the lateral line
and the ventral mid-line to be counted ; five specimens have 4-5 scales and two
specimens have 5-5 scales. Two and a half scales are present between the lateral
line and the pelvic fin insertion. There are 12 scales around the caudal peduncle.
The striae on the scales are numerous and more or less parallel. The striae of
scales on the upper part of the body tend to diverge slightly whilst scales on the
lower part of the body tend to have striae that are parallel or slightly converging.
Dorsal fin. The dorsal fin has four unbranched rays, the last one is ossified into a
smooth, straight or slightly curved spine (x = 19-7 ; s.d. = 2-7 ; s.e. = 0-8 ;
range 14-1-22-5). There are nine (f.n) or eight (f.i) branched rays. The dorsal
margin of the fin is only very slightly concave. The origin of the dorsal fin is in
front of the origin of the pelvic fins.
Pharyngeal bones and teeth. No exceptions to the pharyngeal tooth formula
2.3.5. -5.3. 2. were found. The teeth are shown in fig. 87. The anterior edentulous
process has its distal half at an angle to the proximal half.
Gill rakers. There are n (f.3) or 10 (f.3) on the lower limb of the first gill arch
in the six specimens examined.
Coloration. All the preserved specimens are dark fish. Dark-brown pigment is
present on the distal parts of the dorsal, anal, pelvic and pectoral fins. On the
caudal fin the pigment is concentrated at the margins.
io6
K. E. BANISTER
FIG. 87. Barbus paucisquamatus - left pharyngeal bone from the lectotype.
The body is dark brown but the scales have a noticeably high reflectivity. The
skin of the cheeks and snout has a silvery layer below the brown pigment. The
dorso-lateral surface of both barbels, particularly the anterior barbel, is pigmented.
DISTRIBUTION. Known from the Loama river, the Nyabarongo river and the
Luhoho Congo.
DIAGNOSIS AND AFFINITIES. Barbus paucisquamatus is a fish with 27-29 scales
in the lateral line series, dark pigment widespread over the body but not obscuring
a very silvery layer beneath and scales on which the striae vary with the position of
the scale on the body.
Its affinities are uncertain as there is an absence of clear-cut similarities or dis-
tinctions with other species. It is sympatric with Barbus longifilis from which it
can be distinguished by its shorter barbels, shallower body and weaker dorsal spine.
Although Barbus paucisquamatus was described as a variety of Barbus altianalis,
the caudal peduncle is conspicuously longer and shallower than in that species,
and the overall impression gained from the body shape and colour is that there is no
close relationship between Barbus altianalis and Barbus paucisquamatus.
The heavy pigmentation and a similar mouth are also found in Barbus caudo-
vittatus (p. 40) but at the moment I do not have enough information to comment
further on the possibilities of a relationship between these two species, nor between
Barbus paucisquamatus and the morphologically somewhat similar Barbus trachy-
pterus.
Barbus platyrhinus Boulenger 1900
Barbus platyrhinus Blgr., 1900, Ann. Mag. nat. Hist. (7) 6 : 479.
HOLOTYPE. A fish of 347 mm S.L. from Usamburu, Lake Tanganyika, B.M.
(N.H.) Reg. No. 1906.9.6 : 12.
A REVISION OF THE LARGE AFRICAN BARBUS
107
DESCRIPTION. Based upon seven specimens, the holotype and six fishes M.A.C.T.
Nos. 89789-92, 130658-60.
L
D
H
I
IO
MW
Pet
CP1
CPd
Snt
Ab
Pb
29-6
257
7-2
9-8
6-7
21-4
16-5
12-2
8-8
3'4
4'5
s.d.
2-4
I-O
0-9
0-9
0-9
i'3
i-i
0-8
0-6
0-7
s.e.
0-9
0-4
0-4
0-4
0-2
range
57 -347mm
26-1-32-3
24-2-27-2
6-3- 9-1
8-2-10-9
5-7- 8-1
19-6-23-7
15-2-18-7
10-9-13-7
7-6- 9-5
2-1- 4-3
Barbus platyrhinus is a heavy-bodied fish, the bulky appearance becoming more
exaggerated in larger specimens. The mouth is ventral and the snout has a slightly
bulbous profile above the upper jaw.
FIG. 88. Barbus platyrhinus holotype (from Boulenger 191 la).
Squamation. The lateral line follows a fairly straight course. The dip in the
middle of the line is shallow. The lateral line scale counts are 38 (f.2), 39 (1.3),
40 (f.i), 41 (f.i). There are 6-5 (f.i) or 7-5 (f.6) scales between the dorsal mid-line
and the lateral line. There are 6-5 (f.i), 7-5 (f.2), 8-5 (f.4) scales between the
lateral line and the ventral mid-line. There are 3-5 (f.2) or 4-5 (f.4) scales between
the lateral line and the base of the pelvic fin. Sixteen (f.3) or 18 (f.4) scales encircle
the caudal peduncle, the scales are longitudinally striated.
Dorsal fin. The dorsal fin has four unbranched rays. The last unbranched ray
ranges from 10-1 to 24-6 per cent S.L. with a mean value of 18-1 per cent. There
are 8 (f.i), 9 (f.5) or 10 (f.i) branched rays. The origin of the dorsal fin is slightly
in advance of the origin of the pelvic fins.
The anal fin has three unbranched rays and five branched rays.
io8
K. E. BANISTER
FIG. 89. The pharyngeal bones from the holotype of Barbus platyrhinus.
Pharyngeal bones and teeth. The only pharyngeal bones that I have been able to
examine are from the holotype and are apparently aberrant (fig. 89) ; the pharyngeal
formula is 4.3.2. The inner row lacks the first tooth present in other species, the
first tooth present in Barbus platyrhinus is the homologue of the usual second tooth
and is referred to as the second tooth. The second tooth is large and molariform.
The third tooth has a rounded crown wider than long with a slight concavity in
the posterior face of the left-hand tooth and a deep concavity in the anterior face
of the right-hand tooth. Jhe fourth tooth is smaller than the preceding, the
anterior edge of the crown is its highest point and the posterior face is concave.
This tooth is wider than it is long. The fifth tooth is slight and its spatulate crown
is very close to the preceding crown. On the left-hand bone the first tooth of the
middle row is missing but the scar is present. Its counterpart is mamilliform and
wider than long. It is physically closer to the anterior tooth of the dorsal row than
to the second tooth of the middle row. The second and third teeth become pro-
gressively smaller. They are complete on the left-hand bone but either aberrant
or in the process of being replaced on the right-hand bone. The two small teeth in
the outer row are wider than long with a concave posterior face.
The pharyngeal bones themselves are short and thick.
There are between n and 13 broad, bilobed gill rakers on the lower part of the
first gill arch.
DISTRIBUTION. This species has been recorded at Usamburu and Uvira, Lake
Tanganyika and from the Koki river, an affluent of Lake Tanganyika. Poll (1953)
believes that it is rare in the lake and may be commoner in the affluent rivers.
DIAGNOSIS AND AFFINITIES. This species is sympatric with Barbus tropidolepis
to which, at the very least, it is closely related. Both species have a high number
of lateral line scales and similarly shaped pharyngeal teeth. The major difference
A REVISION OF THE LARGE AFRICAN BARBUS 109
between them is the development of the fatty ridges on the scales in Barbus tropido-
lepis and the universal presence of barbels in Barbus platyrhinus. A larger series of
specimens of this species and some information on their ecology may show that they
are not a valid species, but for the moment it must be retained as there is, in some
characters, no overlap. I have not yet found any fish which cannot with certainty
be placed in either Barbus platyrhinus or Barbus tropidolepis , but the specimens
examined have been few in number. The possibility that Barbus platyrhinus is a
hybrid must also be borne in mind.
Barbus ruasae Pappenheim and Boulenger 1914
Barbus ruasae Pappenheim and Blgr., 1914, Wiss. Ergebn. dt. ZentAfr. Exped. Zool. (3) 5 : 238,
pi. 2, fig. 2.
LECTOTYPE. The lectotype is the smaller of the two type specimens of Pappen-
heim & Boulenger, a fish of 155 mm total length and 126 mm S.L. This is the
specimen figured by the authors. The lectotype and paralectotype (a fish of
128 mm S.L.) are in the Zoological Museum, Berlin, both registered as ZMB 19053.
The locality is Mkunga in Rwanda.
DESCRIPTION. The description is based on five specimens, the two from Mkunga
(see above) and three specimens, M.A.C.T. Nos. 91755, 91756, 92214, from the
Nyabugogo, an affluent of the Lusine about 64 km SSE of the type locality.
x s.d. s.e. range
L 103 -135 mm
D 25-6 1-6 0-7 23-7-28-1
H 27-0 0-9 0-4 25-8-28-1
I 8-6 0-6 0-3 8-0- 9-7
IO 7-3 0-3 o-i 7-0- 7-8
MW 6-5 i-o 0-4 5-6- 8-0
Pet 21-8 2-2 I-O 18-3-23-4
CP1 16-2 1-3 0-6 14-7-18-4
CPd 10-6 0-5 0-2 10-1-11-5
Snt 8-4 0-5 0-2 7-8- 8-9
Ab 5-3 1-7 0-7 2-4- 7-1
Pb 6-1 1-8 0-8 3-2- 8-0
All measurements are expressed as a percentage of the standard length. The
mouth is ventral, the anterior margin of the lower jaw is curved and the upper and
lower lips developed but not lobed. The barbels in the lectotype show signs of
damage. The head is pointed, the dorsal margin of the body rises in a smooth
curve from the snout to the origin of the dorsal fin. The lateral line count is 25 (f .9)
or 26 (f.i) (both sides of the fish considered). The smallest of the M.A.C.T. speci-
mens has the count of 26 on one side. There are 4-5 scale rows between the dorsal
mid-line and the lateral line and 4-5 from there to the ventral mid-line. Two and a
half scales are between the lateral line and the base of the pelvic fin and 12 scales
K. E. BANISTER
FIG. 90. Barbus ruasae lectotype.
encircle the caudal peduncle. The striae on the scales are parallel and much wider
apart on the scales of the pectoral region than on the scales of the caudal peduncle.
The dorsal fin has four unbranched rays, the last is ossified into a strong smooth
spine (x = 21-0; s.d. = 2-9; s.e. = 1-3; range 18-6-26-0). There are 8 (f.2),
9 (f.2) or 6 (f.i) branched rays. The dorsal fin with six branched rays showed
obvious signs of damage. The origin of the dorsal fin is above or slightly anterior
to the origin of the pelvic fin. A low sheath of scales surrounds the base of the
dorsal fin.
The anal Jin has three unfrranched and five branched rays.
Pharyngeal bones and teeth. The pharyngeal bones are moderately slender, the
teeth are in three rows of 5.3.2. on each bone. The second tooth in the inner row
of the largest M.A.C.T. specimen (No. 992214) is more molariform than in the
lectotype.
There are from 10 to 13 gill rakers on the lower limb of the first gill arch.
Coloration. In alcohol the fish present a silvery appearance. The back is a mid-
brown which shades to silver on the belly. The centres of the scales are bright
silvery and the edges are darker. Traces of dark pigment remain in the middle of
the caudal fin, the distal edge of the dorsal fin, the anal fin and the posterior margins
of the pectoral and pelvic fins.
DIAGNOSIS AND AFFINITIES. Barbus ruasae is a member of the Barbus intermedium
complex and as such it is compared with and distinguished from Barbus acuticeps
on p. 8, Barbus altianalis on p. 27 and Barbus intermedium on p. 140.
Pappenheim & Boulenger (op. cit.} thought, prophetically, that Barbus ruasae
showed affinities to Barbus leptosoma (= Barbus intermedium).
Very low lateral line counts are found in Barbus mawambiensis. This species is
found in the not-too-distant Ituri and Ja rivers which although in the Congo system
do not flow into the Lake Victoria basin. I do not have enough information to
A REVISION OF THE LARGE AFRICAN BARBUS
in
determine whether or not there is any relationship between Barbus ruasae and
Barbus mawambiensis.
My knowledge of the Barbus species of the Congo is insufficient to even hazard
any opinions on the relationships of the species concerned.
Barbus somereni Boulenger 1911
Barbus somereni Blgr., 1911, Ann. Mag. nat. Hist. (8) 8 : 369 ; Greenwood, 1966, The Fishes of
Uganda 2nd ed. : 69 ; Banister, 1972, Bull. Br. Mus. nat. Hist. (Zool.) 24 (5) : 261-290.
Barbus altianalis urundensis David, 1937, Revue Zool. Bot. afr. 9 (4) : 414.
Barbus urundensis Poll, 1946, Annls. Mus. r. Congo Beige (i) 4 (3) : 185-188.
HOLOTYPE. A fish of 172 mm S.L., B.M. (N.H.) Reg. No. 1911.7.26 : i from the
Sibwe river, Ruwenzori mountains, Uganda.
DESCRIPTION. The description is based on 51 specimens, 66-279 mm S.L. All
measurements are expressed as percentages of the standard length.
D
H
T
10
MW
Pet
CP1
CPd
Snt
Ab
Pb
x
29-5
29-9
7-6
8-5
7'5
22-2
II-6
8-6
7-8
8-6
s.d.
2-9
'4
•4
•i
0-8
•7
•3
0-7
0-6
i-o
1-2
s.e.
0-4
O-2
O-2
0-1
O-I
0-2
O-2
O-I
O-I
O-I
0-2
range
25-4-38-4
21-6-28-8
5-2-10-6
6-7-11-7
6-3- 9-2
18-4-28-0
13-9-19-0
10-4-13-2
7-6-10-2
5-6- 9-6
6-3-11-6
The body is slightly compressed and becomes relatively deeper in larger fishes.
This and the relative decrease in the diameter of the eye with an increase in the
FIG. 91. Barbus somereni holotype (from Boulenger 1916).
K. E. BANISTER
1
I
£ 8.
r1— < h^
JJ f>->
+->
be O
S "3
•£*
II
I ?
ll
I
A REVISION OF THE LARGE AFRICAN BARBUS 113
standard length are the only obvious signs of allometric growth. In the other
measurements, the scatter of the points precludes the determination of the direction
of the allometry.
The mouth is sub-inferior and usually has continuous, moderately developed lips,
but a 'rubber-lipped' specimen was collected in the Sibwe river.
Three fish from the Malagarasi river were described by David (1937) as Barbus
altianalis urundensis. I find that her three syntypes (M.A.C.T. Reg. Nos. 46963-
46965) have IV-8 (f.i), IV-g (f.2) rays in the dorsal fin (not III-9 as described)
and have 31 (f.i), 32 (f.i), 33 (f.i) scales in the lateral line. Her counts were of
the total number of scales.
Squamation. The number of scales in the lateral line ranges from 26 to 34 :
26 (f.i), 27 (f.2), 28 (13), 29 (f.5), 30 (f.n), 3i (f-i2), 32 (f.7)> 33 (f-7)> 34 (f.i).
The pattern of striae on the scales varies between that which is regarded as typi-
cally radiate and that which is typically parallel. There is a general tendency for
the shoulder scales in this species to be of the latter type and the belly scales to be
of the former. The striae are, however, more numerous than is usual for the
classical radiately striated Barbus (see fig. 93).
There are 5-5 (sometimes 4-5 rarely 6-5) scales between the dorsal mid-line and
the lateral line and 4-5-6-5 (rarely 7-5) scales between the lateral line and the ventral
mid-line. The specimens from the Sibwe and Tokwe rivers (Lake George affluents)
have 3-5 scales between the lateral line and the pelvic fin, whilst those from Mahembe
and the Mutamphu river (Kagera system) have 2-5 scales (rarely 2 or 3).
There are 12 scales round the caudal peduncle.
Dorsal fin. There are four unbranched rays, the last is ossified with a smooth
posterior margin. This is also tnie for the type specimen although Boulenger (ignb)
describes only three unbranched rays. The length of the last unbranched ray varies
from 8-7 to 21-2 per cent of the standard length (x = 13-8). There are nine or ten
branched rays except in one fish where there are eight.
The anal fin has three unbranched and five branched rays.
Coloration. The ripe-running males in the Sibwe river have a deep olive-brown
back which changes sharply into an ochrous yellow colour on the flanks and belly.
The dark olive on the back is continued into the lower lobe of the caudal fin. The
upper caudal lobe and the dorsal fin are pale brown. This pattern of a dark back
and dark lower caudal lobe persists in fishes which are sexually inactive and is also
visible in the great majority of preserved specimens.
GUI rakers. There are 8-n gill rakers on the lower limb of the first arch. The
rakers are bilobed with the sharply triangular outer lobe the larger.
The pharyngeal bones and teeth. The teeth are slightly hooked with no significant
enlargement or molarization of the second tooth on the inner row (fig. 94). The
pharyngeal tooth formula is 2. 3. 5. -5.3. 2.
DISTRIBUTION. Specimens are known in the Ruwenzori area from the following
rivers : Sibwe river, Mubuku river, Tokwe river and Kirimia river. In the Sibwe
and Mubuku rivers which flow into the northern end of Lake George, Barbus somereni
K. E. BANISTER
1 mm
B,C.
FIG. 93. Scales from the shoulders of A) Barbus bynni, B) B. somereni, C) B. humilis to
compare the striae on the B. somereni scales with the typically 'parallel' and 'radiate'
striae patterns shown by the other two species.
lives in the faster-flowing reaches before the rivers meander into the lake. Green-
wood (1966) gives 5500 feet as the upper limit of Barbus somereni distribution.
Barbus somereni is not common in the Mubuku river (see Banister 1972).
Barbus somereni has also been found in the Nyawarongo (at Mahembe) and
Akianaru (Rwanda) rivers, in the Mutamphu river (12 km from Astrida on the
road to Shangugu), in the Chirangobwe river (Lake Kivu basin), Mwogo river
(Kagera system), upper Malagarasi river (Burundi), Kitenge river (Ruzizi) and from
the Nyamagana and Nyakagunda rivers (Burundi).
DIAGNOSIS AND AFFINITIES. The relationship between Barbus somereni (and its
relative Barbus mirabilis) and the other African Barbus species is uncertain. The
body shape, the last simple dorsal fin ray, the number of dorsal fin rays, the nature
of the striae on the scales and the colour pattern are all points of difference between
Barbus somereni and the members of the Barbus bynni and Barbus intermedius com-
plexes (especially Barbus altianalis), which are the closest groups geographically.
Barbus somereni is a species which lives at higher altitudes than, for example,
Barbus altianalis eduardianus in the Lake George- Ruwenzori Mountain region.
This could suggest that Barbus somereni has been displaced by Barbus altianalis
and, if so, then Barbus somereni is a longer established resident of the area. This
view is strengthened by the presence of Barbus mirabilis on the other side of the
rift valley.
Barbus somereni is possibly related to Varicorhinus ruwenzorii. These two
species live in the same area, have the same colour pattern and it is suggested
(Banister 1972) that they hybridize.
A REVISION OF THE LARGE AFRICAN BARBUS
FIG. 94. The dorsal and lateral aspects of the left pharyngeal bone from Barbus somereni.
8*
FIG. 95. A map of the distribution of Barbus somereni.
u6 K. E. BANISTER
Barbus stappersii Boulenger 1915
Barbus stappersii Blgr., 1915, Revue zool. afr. 4 (2) : 165 ; Blgr., 1920, Revue zool. afr. 8 (i) : 13 ;
Jackson, 1961, Fishes of Northern Rhodesia : 57.
Barbus curtus Blgr., 1915, Revue zool. afr. 4 (2) : 165 ; Blgr., 1920, Revue zool. afr. 8 (i) : 14 ;
Jackson, 1959, Occ. Pap. natn. Mus. Sth. Rhod. No. 236 : 298 ; Soulsby, 1960, Nth. Rhod. J.
4 (4) : 329, fig. 10.
Barbus oxycephalus Blgr., 1915, Revue zool. afr. 4 (2) : 165 ; Blgr., 1920, Revue zool. afr. 8 (i) : 15 ;
Soulsby, 1960, Nth. Rhod. J. (4) : 329, fig. n.
Barbus moeruensis Pellegrin, 1922, Revue zool. afr. 10 (3) : 273 ; Pellegrin, 1928, Annls. Mus. r.
Congo Beige (i) 3 : 42.
NOTES ON THE SYNONYMY. Jackson (1961) synonymized Barbus curtus, Barbus
oxycephalus and Barbus moeruensis with Barbus trachypterus Blgr. 1915- These
four species and Barbus stappersii were each described from single specimens, all
from Lake Mweru, and all except Barbus trachypterus are large fish, respectively
their standard lengths are 233 mm, 264 mm, 594 mm, 128 mm and 283 mm.
These nominal species fall into two obvious groups, one with the caudal peduncle
as deep as it is long and the other with the caudal peduncle much longer than deep.
The former group consists of the types of Barbus stappersii, Barbus oxycephalus,
Barbus curtus and Barbus moeruensis whilst the latter group contains Barbus
trachypterus. The caudal peduncle depth is the most trenchant difference between
the two groups although there are others. The size difference has had no effect
upon the dimensions of the caudal peduncle, a specimen of 112 mm S.L. referable
to Barbus stappersii has a caudal peduncle deeper than it is long.
Barbus trachypterus is considered here to be a valid species and is described on
p. 119.
HOLOTYPE. A fish of 283 mm S.L., M.A.C.T. No. 14250, from Lake Mweru.
DESCRIPTION. The description is based on nine specimens ranging from 103 to
594 mm S.L.
x s.d. s.e. range
D 35-6 3-8 i-i 30-0-40-7
H 25-5 -2 0-4 25-0-28-2
I 6-6 -3 0-4 4-1- 8-7
IO 9-8 -4 0-4 7'7-I2>3
MW 7-6 -4 0-4 6-2-10-8
Pet 24-0 -2 0-7 18-6-26-0
CP1 15-1 -4 0-5 13-0-17-9
CPd 15-4 -7 0-5 12-0-17-4
Snt 9-1 0-8 0-2 8-0-10-8
Ab 2-7 0-9 0-3 1-8- 3-9
Pb 3-6 0-7 0-2 2-8- 4-5
They are stocky, deep fish with moderately compressed bodies. The ventral
profile is gently convex from the mouth to the anal fin whilst the dorsal profile
ascends steeply towards the dorsal fin. The mouth is terminal, the lips moderately
developed and with a median lobe on the lower lip. All the examined specimens
A REVISION OF THE LARGE AFRICAN BARBUS
117
o
I
I
I
09
I
ct
o
o
n8
K. E. BANISTER
FIG. 97. The dorsal and lateral aspects of the left pharyngeal bone of the holotype of
Barbus stappersii.
have lips conforming to this pattern, except for a specimen from Kilwa in which the
lips are thinner.
The caudal peduncle is as deep as it is long.
Dorsal fin. The leading edge of the dorsal fin is slightly in advance of the origin
of the pelvic fin. A low basal sheath of scales is present. The anterior edge of the
dorsal fin is high and curved so that the concave dorsal margin is positioned almost
vertically. The dorsal spine is moderately well ossified, the measurements of the
non-flexible basal part in the nine specimens are x = 16-0, s.d. = 3-88, s.e. = 1-37
and the range is 9-1 to 21-8 per cent. There are 9 (f.2) or 10 (i.j] branched rays.
The anal fin has three unbranched rays and five branched rays.
Squamation. The scales have numerous parallel striae. The lateral line has
between 23 and 28 scales : 23 (f.i), 24 (f-3), 25 (f.2), 26 (f.i), 27 (f.i), 28 (f.i).
There are 4-5 scale rows between the dorsal mid-line and the lateral line and
4-5 (f.8) or 5-5 (f.i) scales between the lateral line and the ventral mid-line. Two
and a half scales are present between the lateral line and the pelvic fin base. There
are 12 scales around the caudal peduncle.
Pharyngeal teeth and bones. The pharyngeal tooth formula is 2. 3. 5. -5.3. 2., the
pharyngeal bone (fig. 97) is moderately slender. The crowns of the inner row of
teeth are curved . There is little enlargement of the second tooth on the inner row,
the first tooth on that row is small with a slightly spoon-shaped crown and it is
angled towards the second tooth. The succeeding teeth in that row become higher,
more slender and more recurved. The tip of the fifth tooth is hooked forwards.
A REVISION OF THE LARGE AFRICAN BARBUS
^ 0
119
FIG. 98. A map of the distribution of Barbus stappersii.
Gill rakers. There are 13 curved gill rakers on the lower limb of the first gill arch.
Coloration. The colour in alcohol-preserved specimens is uniformly brown, with
the centre of each scale a richer, deeper brown than the posterior margin.
DISTRIBUTION. This species is known from Kilwa, on Lake Mweru, and from the
Lubumbashi region.
DIAGNOSIS AND AFFINITIES. Barbus stappersii is a species easily recognized by
its short, deep caudal peduncle, low number of scales in the lateral line series (23-28)
and by the high anterior edge of the dorsal fin.
Its affinities are uncertain. In some respects (short, deep caudal peduncle and
high dorsal fin) it resembles specimens of Barbus marequensis A. Smith from the
Zambezi river (the forms described as Barbus victoriae Blgr., Barbus altidorsalis
Blgr., Barbus chilotes Blgr., Barbus codringtonii Blgr. and Barbus fairbairnii Blgr.).
A low watershed is the only barrier between the Upper Zambezi and the Lake
Mweru basin. Until I have examined more specimens of Barbus marequensis and
Barbus stappersii I am reluctant to comment further on any relationship between
these two species.
Barbus trachypterus Boulenger 1915
Barbus trachypterus Blgr., 1915, Revue zool. afr. 4 (2) : 164 ; Jackson (partim) 1961, Fishes of
Northern Rhodesia : 58.
Varicorhinus bredoi Poll, 1948, Bull. Mus. r. Hist. nat. Belg. 24 (21) : 9.
HOLOTYPE. A fish of 128 mm S.L., M.A.C.T. No. 11380, from the hydrographical
station, Lake Mweru.
120 K. E. BANISTER
DESCRIPTION. The description is based on 17 specimens from 54 to 239 mm S.L.
n x s.d. s.e. range
L 54 -239 mm
D 17 27-9 1-6 0-4 25-6-31-4
H 17 25-4 2-5 0-6 21-8-28-8
I 17 5-4 i-o 0-2 3-8- 7-4
IO 17 7-7 1-3 0-3 6-0-11-4
MW 17 6-4 0-7 0-2 5-4- 7-6
Pet 17 2I-O 1-4 0-3 18-9-22-6
CP1 17 17-4 1-3 0-3 14-5-19-3
CPd 17 10-7 0-9 0-2 9-1-11-8
Snt 17 8-2 i-i 0-2 6-5-11-4
Ab 16 3-5 i-o 0-2 2-4- 5-5
Pb 16 4-9 1-4 0-3 2-6- 8-0
All measurements are expressed as percentages of the standard length.
The body is slightly compressed with a level or gently convex ventral profile and
a dorsal profile which rises evenly from the snout to the dorsal fin origin (fig. 99).
The mouth is ventral and semi-circular. The upper lip is visible in the ventral view
as a thin, fleshy surround to the mouth. The thin rostral flap reaches down to the
level of the mouth. The anterior edge of the lower jaw is strongly convex in outline.
In small fishes a fleshy lower lip, often with a small mental lobe, is present, but in
larger fishes the tendency is for the lower jaw to have a flat, cutting anterior margin.
One specimen (M.A.C.T. No. 129097) has well-developed 'rubber lips'.
Tubercles are present on the snout and cheeks of several specimens. The
tubercles are comparatively small, but widespread and tend to coalesce. In speci-
mens M.A.C.T. Nos. 165254-165256 they are present on the snout, cheeks, oper-
culum and anal fin rays. They are also present on the anal fin rays of the holotype.
In three comparatively fresh specimens (M.A.C.T. Nos. 165254-165256), the peri-
toneum is black.
Squamation. The scales possess numerous parallel or, at the most, slightly con-
verging striae. There are from 27 to 31 scales in the lateral series : 27 (1.4),
28 (f-4), 29 (f-5), 30 (f.2), 31 (f.2). Twelve scales are invariably present around the
caudal peduncle. There are 4-5 (f.i5) or 5-5 (f.i) scales between the dorsal mid-line
and the lateral line and 5-5 (f.io) or 4-5 (f.3) scale rows between the lateral line and
the ventral mid-line except in one specimen (M.A.C.T. No. 129100 of standard length
168 mm) where the scales on the chest are conspicuously reduced. There are 2-5 or
3 scales between the lateral line and the base of the pelvic fins.
Dorsal Jin. Except for the holotype of Varicorhinus bredoi the dorsal fin has four
unbranched rays. This specimen now has only two unbranched rays although Poll
(1948) records three. There is little doubt that the reduction in the number of
simple rays in this specimen is the result of physical damage. The last unbranched
ray is ossified into a smooth, often slightly curved spine : x = 20-0; s.d. = 2-48;
s.e. = 0-6; range = i6-i-25-o. The dorsal fin is slightly in advance of the pelvic
fin origin. There are 8 (f.3), 9 (f.i2) or 10 (f.2) branched rays.
The anal fin has three unbranched and five branched rays.
A REVISION OF THE LARGE AFRICAN BARBUS
c/i
E
00
o
E
'o
a
K. E. BANISTER
FIG. 100. The dorsal and lateral aspects of the left pharyngeal bone from the figured
specimen of Barbus trachypterus.
Pharyngeal bones and teeth. The pharyngeal tooth formula is 2.3. 5. -5.3. 2. The
pharyngeal teeth are small, slender and crowded (fig. 100). The second tooth of the
inner row has a tendency to become molariform ; in six of the nine specimens
examined the crown is rounded and flattened at the top. The pharyngeal bone is
thick and stout, but less so than in Barbus caudovittatus specimens of the same size
(cf. figs. 22 and 23).
Gill rakers. The gill rakers number between n and 13 on the lower arm of the
first gill arch.
Coloration. The colour pattern is invariably different from that of Barbus
caudovittatus with which some of these specimens have been confused. The black
upper half of the sides are mid-brown with darker brown bases to the scales. The
lower part of the flanks, chest and belly are pinkish-brown. Dark pigment is
present on the proximal half of the membrane of the dorsal, anal, pelvic and pectoral
fins. The caudal fin has a uniform mid-brown colouring. The colour notes are
based on alcohol-preserved specimens (three of which are comparatively recent),
but no difficulty was experienced in separating this species, on colour alone, from
Barbus caudovittatus. In the latter species the dark pigment is found on the distal
half of the pectoral, pelvic and anal fins. The colour pattern is sometimes bleached
out in specimens that have been badly preserved or have been preserved for a long
time.
DISTRIBUTION. The specimens examined came from Lake Mweru, from between
Kolwezi and Jadotville on the Lualaba, Upper Katanga, from Mwena, a tributary
of the Lufira, Upper Katanga, from Gombela, Upper Katanga, from Kabiyashi on
A REVISION OF THE LARGE AFRICAN BARBUS
123
FIG. 101. A map of the distribution of Barbus trachypterus.
the Luanza, Upper Katanga, from Ngundeulu, Elizabethville (= Lubumbashi),
and Kando, near Tenke, Upper Katanga.
AFFINITIES. The tubercles, the ventral mouth and the crowded pharyngeal
teeth are reminiscent of the condition found in many species of the genus Varico-
rhinus. It is interesting to note that Poll (1948) thought that the relationships of
Varicorhinus bredoi lay with Varicorhinus stappersii (here considered to be a synonym
of Barbus caudovittatus} and Varicorhinus brucii (which was considered to be a
variant of Barbus marequensis by Groenewald 1958). Barbus trachypterus specimens
have often been confused with Barbus caudovittatus specimens but can be dis-
tinguished by the presence of a stronger dorsal spine and by a different colour
pattern. Both Barbus trachypterus and Barbus caudovittatus have a ventral, curved
mouth and possess a colour pattern unlike many of the east African species. The
affinities of Barbus trachypterus could well lie with Barbus caudovittatus but much
more needs to be known about the Barbus species of the southern and eastern parts
of the Congo before a more informed conclusion can be drawn.
Barbus tropidolepis Boulenger 1900
Barbus tropidolepis Blgr., 1900, Annls. Mus. r. Congo Beige Zool. 1 : 133, pi. xlix ; Poll, 1953,
Result, scient. Explor. hydrobiol. Lac Tanganyika 3 (5 A) : 74.
Varicorhinus chapini Nichols & LaMonte, 1950, Proc. biol. Soc. Wash. 63 : 175 (fide Poll, 1952,
Revue Zool. Bot. afr. (46) 3-4 : 222).
I24
K. E. BANISTER
FIG. 102. Barbus tropidolepis 'Type' specimen (from Boulenger 191 la).
LECTOTYPE. Boulenger described this species from three fish from Usambura,
Lake Tanganyika [B.M. (N.H.) Reg. Nos. 1906.9.6 : 19-21]. The smallest speci-
men, 239 mm S.L., is recognized as the lectotype.
DESCRIPTION. The description is based on 47 fish of 99 to 365 mm S.L. The
morphometric data are expressed in tabular form below.
D
H
I
10
MW
Pet
CP1
CPd
Snt
28-8
27-0
9-4
9-0
6-4
20-8
14-9
n-7
8-4
s.d.
2-7
1-4
0-8
i-o
1-2
0-9
I-O
s.e.
0-6
0-2
O-2
0-2
O-2
range
25-0-33-7
24-8-30-1
7-0-11-0
7-4-11-4
5-2- 8-8
19-1-23-9
12-9-17-7
10-1-13-7
5-2- 9-3
The eye is large and frequently protuberant. The characteristic body shape is
shown in fig. 102. Although the anterior barbel is invariably absent and the
posterior represented by no more than a small protrusion (as in many African
Varicorhinus species) the mouth is no wider than in most Barbus. Worthington &
Ricardo (1937) noted that the degree of lip development varies from continuous to
discontinuous. A few specimens have thickened lips but in none of the fish examined
were ' rubber lips ' or ' Varicorhinus-like ' lips developed. The mouth is ventral
under an obtuse snout.
Squamation. The lateral line has from 39 to 44 scales : 39 (f.2), 40 (f-9), 41 (f.n),
42 (f.i-3), 43 (f.g), 44 (f-3). There are 8-5 (rarely 7-5) scales between the dorsal mid-
line and the lateral line and 8-5 (rarely 7-5, very rarely 9-5) between the lateral line
A REVISION OF THE LARGE AFRICAN BARBUS
125
FIG. 103. The dorsal and lateral aspects of the right pharyngeal bone from
Barbus tropidolepis.
and the ventral mid-line. Five and a half scale rows (rarely 4-5) lie between the
lateral line and the pelvic fin base. Around the caudal peduncle are 16 (f-34),
17 (f.6) or 18 (1.7) scales. The most remarkable feature of the scales of Barbus
tropidolepis is the presence of longitudinal folds of fat which form well-marked
ridges on the body. These are most frequently found below the lateral line and
from the middle to the posterior of the body. Not all specimens have them, e.g.
B.M. (N.H.) Nos. 1936.6.15 : 596-598, fishes of 156 to 200 mm S.L. lack the ridges
whilst they are present in other fish of less than 100 mm S.L.
Dorsal fin. It has four unbranched rays [not three as reported by Boulenger
(191 la) and Worthington & Ricardo (1937)]. The last unbranched ray is ossified
to form a smooth stout spine which varies in length from 20-8 to 30 per cent
(x = 25-8, s.d. = 2-7, s.e. = 0-6). There are nine (rarely 10) branched rays. The
dorsal margin of the fin is frequently markedly concave with its posterior corner
extended a little (see fig. 102). The dorsal fin origin is in advance of or above the
origin of the pelvic fin.
The anal fin has three unbranched rays and five branched rays.
Pharyngeal bones and teeth. The pharyngeal teeth are molariform (fig. 103). The
first tooth in the inner row is veiy small and in a few cases is absent although a
small pinnacle of bone marks its site. The second tooth is large with a slightly
126
K. E. BANISTER
5 mm
FIG. 104. The pharyngeal bone of Barbus tropidolepis positioned to show the alignment
of the tooth rows.
concave crown. The third and fourth teeth are wider than they are long with
concave posterior faces. The fifth tooth is conical and may be recurved. The
alignment of the second and third rows is rather unusual in that the first tooth of
the second row is slightly displaced dorsally so that both rows seem to radiate from
that tooth. In the other species under consideration the second and third rows
are distinct and parallel (cf. figs. 103 and 104). The pharyngeal teeth resemble
those of Barbus platyrhinus (see- p. 108) .
Coloration. Brown or olive-brown above, lighter brown or silvery below in
preserved fishes.
DISTRIBUTION. This species is endemic to the Lake Tanganyika basin. Poll
(I953) gives a full list of the localities in this region where Barbus tropidolepis has
been captured.
DIAGNOSIS AND AFFINITIES. Barbus tropidolepis appears to occupy a rather
isolated position among East African Barbus ; only Barbus platyrhinus has some
features in common. The large number of scales, the development of ridges of fat
on the scales and the shape of the pharyngeal teeth are indicative of the dissimi-
larities between this species and those of the neighbouring areas and it is clearly
distinct from Barbus caudovittatus, the other large Barbus species in the lake.
It has certain features, in common with Varicorhinus tanganicae, viz. a high
number of lateral line scales, a thick dorsal spine, large eyes and a similar snout.
The phyletic significance of these similarities is doubtful. The number of pharyngeal
teeth is the same but the alignment is different.
On the other hand, the unique specimen of Varicorhinus chapini is definitely a
small Barbus tropidolepis with a slightly unusual mouth. The measurements of this
small fish do not differ from equal-sized Barbus tropidolepis specimens.
A REVISION OF THE LARGE AFRICAN BARBUS 127
DISCUSSION
Particularly noteworthy is the remarkable degree of variation in lip form, body
shape and dorsal spine strength within some of the species described above (e.g.
Barbus altianalis and Barbus intermedius}.
These two species were represented by very large series of specimens (213 and 454
fishes respectively), so whether the same degree of variability would be shown by
other species when equally large samples were studied or whether these two species
are inherently more variable than the other described species cannot at the moment
be determined. Barbus bynni (59 specimens) shows much less variation. The same
is true of Barbus oxyrhynchus, which although more variable in body form than
Barbus bynni (cf. the type of Barbus labiatus, fig. 78, with the type of Barbus hindii,
fig. 79) does not approach the degree of variability shown by Barbus intermedius.
Barbus oxyrhynchus is represented by 108 specimens, i.e. more than Barbus bynni
and less than Barbus altianalis of Barbus intermedius. Without large series of
specimens this matter cannot be taken further.
The intraspecific variation of the pharyngeal teeth is as large as the interspecific
variation (excluding Barbus tropidolepis and Barbus platyrhinus, both of which have
a very large degree of molarization of the pharyngeal teeth). It seems that the
molarization of, particularly, the second tooth of the inner row is of no significance
in establishing the identity of a specimen. The range in pharyngeal teeth shapes
found in Barbus intermedius (figs. 56, 57 and 58) and Barbus altianalis (figs. 13 and
14) show this well. The seemingly random occurrence of a molariform second tooth
in the inner row is shown by the series of pharyngeal bones of two of the subspecies
of Barbus altianalis (figs. 13 and 14, also p. 20) .
It was suggested above (p. 16) that the shape and strength of the pharyngeal
bones and teeth might be influenced by the diet. The pharyngeal bones of Barbus
altianalis radcliffii are stouter than those of Barbus altianalis eduardianus. Whether
or not this is an effect of dietary differences could be checked comparatively easily
by field studies.
It is known that water-snails are much less abundant in Lakes Edward and
George than in Lake Victoria, and the differences between the pharyngeal teeth and
bones of the populations of the cichlid Astatoreochromis alluaudi in these lakes has
been associated with this fact. Greenwood (1964) has shown the differences in the
stoutness of the pharyngeal teeth and bones to be phenotypic. Comparative data
on the diets of the populations of Barbus altianalis in these lakes are lacking, but
the striking similarity between the two phenomena is suggestive.
The presence of massive pharyngeal bones in large, deep-bodied specimens
(p. 16) is possibly directly related to body depth. The ' surkis' form of Barbus
intermedius has been shown to have been feeding on gastropod molluscs and the
likelihood of the ' obesus ' form of Barbus altianalis radcliffii also having gastropods
as an important part of its diet has been mentioned above, although no identifiable
remains were found in its alimentary tract. As only the deep-bodied examples of
populations which consume gastropods display this phenomenon, it is more likely
that the length of the pharyngeal bone is associated with the depth of the body as
was noted on p. 16.
128 K. E. BANISTER
Wunder (1939) experimented on Common Carp (Cyprinus carpio L.) and observed
that controlled feeding could produce a 'hunger' form and a 'fattened' form. The
former resemble the 'gorguarii' form of Barbus intermedium, with a shallow body and
a relatively long head. The fattened form resembled the ' surkis' form of Barbus
intermedium (fig. 32) or the ' obesus ' form of Barbus altianalis (fig. 3) in having a deep
body, a relatively short head and stubby fins.
It is not possible to say whether greater body depth in a fish induces longer (but
presumably not necessarily stouter) pharyngeal bones or whether extra food pro-
duces a deep-bodied fish in which the pharyngeal bone size relates to the kind and
amount of food rather than to body depth. It has not yet been possible to determine
the effect that a deeper body has on the depth of the head. It would be interesting
to know if the 'gorguarii' and 'surkis' forms of Barbus intermedium represent poorly
fed and well-fed populations. This could possibly be answered by field studies, but
until that time one can do little but accept the various forms as different phenotypes
or eco-phenotypes of variable species. It is important to note that the deep-bodied
forms have so far only been found in lakes whilst most of the other body and lip
forms occur throughout the range of the species. There is, regrettably, insufficient
information available to comment further on the variation of these Barbus species.
As in the small Barbus with radiately striated scales described by Greenwood
(1962), supra-specific complexes are discernible in the large Barbus described above.
These complexes (see below) are difficult to define precisely since they are based on
the rather amorphous (although useful) concept of general appearance. It must
also be borne in mind that the species described here represent only a fraction of the
total number of species of the large African Barbus. The extent and composition
of the supra-specific assemblages will doubtless have to be re-evaluated when the
phyletic interrelationships of the African Barbus are better understood.
Two complexes can be recognized amongst the species studied :
1) The Barbus intermedium complex. This comprises Barbus intermedium, Barbus
altianalis, Barbus acuticeps and Barbus ruasae which inhabit most of the rivers of
southern and eastern Ethiopia and northern Kenya, the Blue Nile system (including
Lakes Rudolf and Baringo), the Lake Victoria basin and the lakes in the adjacent
parts of the western rift valley. The lateral line scale counts range from 25 to 36
but are most frequently 28 to 32. Typically, the body is shallow, the caudal
peduncle substantially longer than deep and the dorsal spine is smooth, strongly
ossified and shorter than in the Barbus bynni complex (see below). The scales
possess many parallel or slightly converging striae. At least two of the component
species (Barbus intermedium and Barbus altianalis) are noteworthy for their high
phenotypic variability. The species in this group show strong gross morphological
resemblances to the Barbus of western and southern Saudi Arabia (e.g. Barbus
arabicus Trewavas, 1939 and some as yet undescribed species from Aden) as well
as to species like Barbus batesii Blgr. 1903 from South Cameroons. The significance
of their distribution will be discussed below.
2) The Barbus bynni complex. This group contains Barbiis bynni, Barbus ganan-
ensis, Barbus oxyrhynchus and Barbus longifilis which inhabit the White and
Albertine Niles, lakes at the fringe of the Blue Nile system (Abaya and Rudolf),
A REVISION OF THE LARGE AFRICAN BARBUS 129
the eastern part of the Juba river, the Athi and Tana rivers and the Loama and
Kansehete rivers (upper Congo to the west of Lake Kivu). The lateral line scale
count range is from 21 to 37, the most southerly species (Barbus oxyrhynchus)
having a lower range than the rest, modally 24-25 against 31-33. The trend
towards larger scales in southerly forms has been noted before by Greenwood (1962)
for Barbus paludinosus and Barbus kerstenii. The scales have fewer striae than in
the Barbus intermedius complex and the striae are more sinuous The caudal peduncle
is deeper in relation to its length than in the Barbus intermedius group and the dorsal
spine is noticeably long and straight (p. 28). Barbus oxyrhynchus has one or two
more branched rays in the dorsal fin than the other species in this group. Barbus
mariae of the Athi river which is sympatric with Barbus oxyrhynchus does not
seem to have much in common with the bynni complex, except a long dorsal spine
and a similar lateral line scale count range.
One feature, the presence of small scales, is common to several lacustrine species :
Barbus tropidolepis, Barbus ethiopicus, Barbus platyrhinus and Barbus microterolepis
(see below). It is not suggested that these species are related, merely that the
character has been acquired independently by the species that live in, and probably
evolved in, lakes. Small-scaled, large Barbus species are not necessarily lacustrine
though, as some of the fluviatile South African species (e.g. Barbus polylepis] show.
It is possible that the modally higher lateral line scale count in Barbus altianalis
amongst the Barbus intermedius complex is connected with its confinement in lake
basins. However, its confinement in lake basins is not the same as the confinement
in lakes of Barbus tropidolepis etc. There are populations of Barbus altianalis
radcliffii which appear to be permanent inhabitants of the Kagera river (Dr P. H.
Greenwood, pers. comm.). Superficially these fishes do not appear to differ from
those in Lake Victoria, but no detailed investigations have been carried out. The
lacustrine Barbus altianalis radcliffii do not spend all their time in the lake, they
display their fluviatile ancestry by ascending rivers to breed (Whitehead 1959).
Barbus tropidolepis, it should be noted, also ascends rivers to breed (Poll 1953)
but does not apparently live in rivers outside of the breeding season. Nothing is
known about the habits of Barbus ethiopicus, Barbus platyrhinus, nor Barbus
microterolepis.
In this context there are some ideas of Farquharson (1962) that must be con-
sidered. He discussed the distribution in South Africa of the small-scaled Barbus
species (e.g. Barbus natalensis, Barbus holubi, Barbus capensis and Barbus kimberley-
ensis] with parallel scale striations. From their widespread distribution and endem-
ism he suggested that an ancient dispersion is indicated. If this is so, then it is
possible that the small-scaled lacustrine species mentioned above are relicts of this
dispersion southwards (see above). I have not been able to compare specimens of
the small-scaled South African species with the small-scaled species from Lakes
Tanganyika and Zwai.
Farquharson (ibid. : 247) discusses the influence of water temperature on scale
number and lists a series of Barbus species arranged in order of localities from ' hot '
to 'cold' showing the increase in the number of lateral line scales. Regrettably,
the situation is not as simple as this as, for example, the review by Taning (1952)
130 K. E. BANISTER
shows. There may be some basic truth in Farquharson's idea but at the moment
there is insufficient data to put forward a general case.
It is suggested here that Barbus altianalis came from the same stock as Barbus
intermedium and that the smaller scales developed after it was confined to the lake
basins, subsequent to the rift valley formation (see p. 22). Barbus microterolepis
could also have been derived from Barbus intermedium in this fashion (assuming that
it is a good species and not a hybrid or genetical abnormality - see p. 91).
No other supra-specific groups are recognizable among the species described. The
affinities of Barbus stappersii would seem to be with the Zambezi species Barbus
marequensis. Barbus caudovittatus is probably associated with Congo species and
hence outside the scope of this paper. Too few specimens of some of the other
species are available to be able to evaluate their characters, but a revision of the
Congo Barbus may give some indications of their relationships.
The problematical relationships of Barbus macrolepis may not be solved until the
Barbus species of West Africa are revised. The remarkable similarity of this species
to Barbus lagensis from Nigeria is noted above (p. 83).
Barbus somereni and Barbus mirabilis do not seem to be related to any of the other
species described in this paper. Again, a study of the Congo fauna may reveal
species related to these two.
It is, perhaps, interesting to compare the distribution of the small Barbus species
with radiately striated scales with that of the large Barbus species described above.
The small Barbus species are widely distributed and several species are found in
widely separated river systems. By contrast the most widespread of the large
Barbus species described in this paper are found in palaeogeographically closely
related and/or adjacent river systems. Presumably the small Barbus species are
able to travel from one river system to another via the swamps which frequently
form the watersheds. Bell-Cross (1965) noted that the only Barbus species living
in the watershed on the Muhinga plain which separates the Congo from the Zambezi
system are small, 'radiately striated' species. The large Barbus species would
seem to be less successful at negotiating the small streams, seasonal ponds and
swamps at the headwaters of the river systems. Their larger size must be a con-
tributing factor, certainly none have been found at the extremities of river systems.
Bell-Cross (1960) suggested that the reason why some fishes did not cross watersheds
was not necessarily because of their physical inability to do so but because of some
other factor like behavioural inhibitions. Regrettably there have been very few
observations on this problem. Darlington (1957 : 78-80), however, argues that
given enough time primary fresh-water fishes (e.g. Ostariophysi) can overcome
almost any physical barriers limiting their range. Perhaps the time scale involved
is not large enough to be applicable to the large Barbus species in this context.
If the habitats of the large Barbus species are likely to be isolated, the possibility
exists that the relationships of various species might prove useful as indicators of
the relationship of river systems and vice versa.
If the rate of phenotypic change is less than the time taken for rivers to be dis-
sociated then it is feasible to think that the now-separated populations of fishes
would be detectable as members of the same taxon. The general principle of this
A REVISION OF THE LARGE AFRICAN BARBUS 131
idea is borne out in certain instances, e.g. between Barbus altianalis and Barbus
intermedius, between Barbus somereni and Barbus mirabilis and between the mem-
bers of the Barbus bynni complex. The geographical evidence for this idea in the
Barbus bynni complex is lacking in detail, but there is nothing geographically incon-
sistent in the former association of their river systems (see p. 22).
Giinther (1869), writing on the Nile fishes collected by Petherick, noted that the
upper Nile fauna is related more closely to that of the Palestinian and West African
rivers than to the fauna of Lake Nyasa (Malawi) and the Zambezi river. By 1880
Giinther had expanded the horizons of the affinities of the African fresh-water fish
fauna. He thought that as some families had more representatives in India than
in Africa they probably originated in India or derived from an Indian stock. He
also considered the fauna of Jordan and Syria to have so many African representa-
tives that he included them in the African region as well as in his Euro-Asiatic region.
Gregory (1896) reviewed the evidence for the fauna! similarities between the Nile
and Jordan rivers and concluded that the Jordan river has species in common with
the upper, but not with the lower Nile. Because of this, he conjectured that in the
past a river flowed from Lakes Albert and Victoria, which then had no outlet to the
Nile, into Lake Rudolf and then via the Omo and Hawash rivers across the Afar
depression into the Red Sea. The Red Sea trough would, he thought, at that time
have been filled with fresh water and with an extension of the Jordan river flowing
into it.
Nichols & Griscom (1917), with a great many more species at their disposal than
were available to Gregory, divided the fish fauna of Africa into six regions and com-
mented on the relationships of one faunal area to another. It is probably significant
that the boundary line between their ' Nile and North East Africa ' region and their
'East African' region follows closely the boundary between the ranges of Barbus
intermedius and Barbus oxyrhynchus. The fish of North West Africa, they noted,
have affinities with the European fishes. The fauna of the 'Nile and North East
Africa' region had affinities both with the West tropical Africa region (Nichols
and Griscom postulate extensive contact in the past between the Nile and the
Congo) and northern or Asiatic forms. The fauna (particularly the Cyprinidae)
of the Nile and North East Africa region ' . . . seem invariably to have entered Africa
from the north-east. They predominate in the Nile basin and East Africa, reaching
South Africa but are much more poorly represented in West Africa. '
Menon (1964) in his revision of the cyprinid genus Garra decided that the African
Garra species came into Africa in a series of waves, each succeeding wave dispersing
the forms that came in the preceding wave. His postulated routes for this were
either via the Sinai peninsula or via Saudi Arabia and Somaliland before these
two countries were separated by the Red Sea. The evidence for this is somewhat
sparse, the idea being based on the general pattern of the distribution of cyprinids
in Asia and the pattern of distribution of fishes in Africa with the most archaic forms
in the west (Darlington 1957 : 60).
There exist in south-western Saudi Arabia and Aden some Barbus species (Barbus
arabicus Trewavas 1939 - and some undescribed species) which would fit in well
with the Barbus intermedius supra-specific complex. The question that cannot yet
132 K. E. BANISTER
be answered is whether the Arabian species represent relict populations of the
migration of the Barbus intermedius ancestral stock into Africa or whether they
represent a radiation from Ethiopia into Arabia. In either case the increase in the
salinity of the Red Sea (possibly during the Pliocene - see below) severed contact
between the two groups.
The geological history of the Red Sea is not known in great detail, but enough is
known to enable a brief history to be constructed. Dubetret (1970) considers the
Red Sea cut across the Precambrian rocks forming the African and Arabian shield.
Up to the Miocene it seems that marine incursions along the incipient Red Sea
trench were confined to the northern end (north of Quasir 27 °5' N) where upper
Cretaceous marine deposits are found. The Neogene history is better known.
Dubetret (op. cit.} dates the Red Sea from the Miocene. Foundering from the
Lower Miocene resulted in the Mediterranean Sea flowing into the Gulf of Suez to
approximately the southern end of the Red Sea. A land barrier existed at the
south separating the Mediterranean water from the Indian Ocean water. During
the late Miocene and Pliocene an upward earth movement in the Gulf of Suez area
cut off the Red Sea from the Mediterranean.
In the Pliocene the southern land barrier sank and allowed Indian Ocean water
to flow into the inland sea. The northern land barrier prevented any contact with
the Mediterranean.
Botros (1971) thinks that a series of fresh-water lakes was present in the southern
part of the Red Sea trench during the Oligocene and late Eocene, whilst the northern
end was an arm of the Mediterranean Sea. He is also of the opinion that it was
possible for the land bridge across the southern end of the Red Sea to have become
apparent again during the Pleistocene as a result of the lowering of the sea level
during the Ice Ages.
It seems then that what is 'now Arabia (Roberts 1970 : fig. 2) did not separate
from the horn of Africa until the end of the Miocene or early Pliocene arid that the
two land masses may have been in contact again during the Pleistocene. It can
therefore be surmised that there was enough contact over a long period of time for
Asiatic cyprinids to have migrated into Africa or vice versa.
The pattern of distribution of the supra-specific complexes defined above (fig.
105) refines the problem but does not solve it. The geographically most compact
supra-specific complex (the Barbus intermedius complex) is in northeast Africa (and it
is suggested, in Saudi Arabia). This is ringed to the west and south by the rather
more fragmented and scattered members of the Barbus bynni supra-specific complex.
Still more scattered are the localities for Barbus somereni and Barbus mirabilis.
This distribution could be interpreted as the result of a series of invasions from the
northeast. Barbus lagensis and Barbus macrolepis, if they are as closely related as
the available evidence suggests, could represent the remnants of an early scattering
of species caused by these invasions (see p. 83).
The reasons for this apparent effect are unknown, but it could be related to the
formation of topographical changes in the Ethiopian highlands. The tectonic
movements must have had a profound effect in altering the courses of rivers, linking
some and separating others. This could have happened more than once and over a
A REVISION OF THE LARGE AFRICAN BARBUS
133
FIG. 105. A map of the distribution of A) the Barbus intermedius supraspecific complex
(hatched area), B) the B. bynni supraspecific complex (dark area), C) B. somereni and
B. mirabilis (solid circles).
9
134 K- E- BANISTER
long period of time, which could ultimately cause speciation within a previously
uniform population. Unfortunately the geological history of this region is poorly
known. Migration west and south would scatter the daughter species. Any sub-
sequent waves of immigration in this region might hence be subject to the same
kinds of isolation as the earlier waves and further supra-specific complexes would
result.
There is evidence to suggest that fishes have been able to move from northeast
Africa into the middle east. The presence of Tilapia species in Jordan is an example
of this. The only conclusions that can be drawn about the large Barbus species is
that they were at some stage able to move between the horn of Africa and Saudi
Arabia. It would need fossil evidence to show in which direction the movement
took place. Brown (1970) reports the presence of Barbus species and Tilapia
species in the late Oligocene or Miocene of Jizan (Saudi Arabia, just north of the
Yemen border).
The large Barbus species of Africa are, as a group, quite distinct from the majority
of species living in the Tigris and Euphrates rivers (the nearest region with a good
cyprinid fauna) . The Tigris and Euphrates species mostly resemble Barbus barbus L.
of Europe in possessing relatively small scales and a serrated dorsal spine. Of the
smooth dorsal spined species from Syria, Barbus canis C. & V. (Tor canis of Karaman
1971) most closely resembles the African species, especially the Barbus intermedium
complex. An isolated population of Barbus canis from Khamis Mushyat (i8°i7' N,
42°34' E - about 400 km southeast of Jeddah, Saudi Arabia) is the population closest
to the African continent. The members of this population are less like the African
Barbus species than are the Syrian populations as they lack the anterior pair of
barbels. The fish fauna of Saudi Arabia is sparse and too poorly collected for
opinions to be formed on whether the absence of the anterior barbels in the Khamish
Mushyat population is a local' phenomenon or part of a clinal phenomenon. The
Barbus species of Aden are, as has been mentioned above, extremely similar to
Barbus intermedius.
The presence of Barbus species in Saudi Arabia which closely resemble some of
the African species presents two possibilities about the origin of the African
representatives.
Firstly, did the Barbus species now typical of Africa evolve in Africa from dis-
similar ancestors of which now no trace remains? If so, then the Arabian and
Syrian smooth-spined Barbus species could represent the relicts of a migration from
Africa. The centre of this particular radiation could well have been in Ethiopia as
the African and Arabian land masses must have been in contact. Therefore the
status of the Arabian and Syrian species would be the same as Barbus altianalis and
Barbus ruandae, i.e. the members of the Barbus intermedius supra-specific complex.
The other possibility is that the facies characteristic of the African Barbus species
were to some extent present in the ancestral stock. If this were so, and if the
ancestral stock was not originally African, could the Arabian and Syrian species of
the Barbus canis type represent remnants of the ancestral stock?
Not enough is known at the moment to be able to decide which of these two
possibilities (if either) represents the truth and it is quite possible that the notion of
A REVISION OF THE LARGE AFRICAN BARBUS 135
a series of migrations all in the same direction is far too simple. The ideas in the
previous two paragraphs refer only to the putative last migration that produced the
species forming the Barbus intermedium supra-specific complex. Of the even earlier
migrations and scatterings conjectured from the distribution of the extant species
there is no trace, but this does not affect the equally likely (on available evidence)
alternative ideas on the relationship of the Barbus intermedium complex to the
relevant Barbus species of Saudi Arabia and Syria.
However, no firm conclusions can be drawn until the fossil record and the phylo-
geny of this group of Barbus species are better known. For a full understanding
of the phylogeny and distribution of the African Barbus species, the relationship of
the small 'radiately striated' Barbus species to the large Barbus species must be
determined as well as the relationships of the African Barbus species sensu lato to
the Barbus species of Europe and the Near East. The distributional problems
relate principally to the role of the closely related cyprinid genera, i.e. why should
there be a relative paucity of large Barbus species in west Africa but a comparative
abundance of Varicorhinus species and Labeo species? The reverse situation exists
in the east. In all probability the solutions to these problems lie as much in the
sphere of the field worker as in the sphere of the museum worker.
ACKNOWLEDGEMENTS
Many people have helped me in the production of this paper and it is with con-
siderable pleasure that I take this opportunity of thanking them all.
My senior colleagues Dr P. H. Greenwood and Dr E. Trewavas have spent many
hours discussing speciation and zoogeography and have generously given me the
benefit of their knowledge and experience of African fish. Dr M. Poll of the Central
African Museum, Tervuren, Belgium, gave me facilities to study the collections in
his charge as well as readily sending specimens. Dr W. Klausewitz of the Sencken-
berg Museum, Frankfurt, Dr E. Tortonese of Genoa, Dr C. Karrer of Berlin and
Dr J. C. Tyler of Philadelphia respectively allowed me to examine the collections
of Riippell, Vinciguerra, Pappenheim and Donaldson-Smith. Through the good
offices of Dr Letitzia Ferrara of the Stazione Idrobiologia, Rome, I was able to
examine Dr G. Bini's extensive collections from Ethiopia.
I want to express my gratitude to Miss M. A. McLellan for the help she gave me
in preparing the graphs and with the calculations. The drawings of the pharyngeal
bones are the work of Mr G. J. Howes and the uncredited drawings of whole fish are
the work of Mrs S. Chambers. Mr J. Chambers prepared some alizarin specimens.
Dr W. W. Bishop has helped considerably in advising me and giving me infor-
mation on stratigraphical and palaeogeographical matters. The staff of the Royal
Geographical Society deserve my heartiest thanks in tracing localities that I had
thought were untraceable.
REFERENCES
ALEXANDER, R. McN. 1966. The function and mechanisms of the protrusible upper jaws
of two species of cyprinid fish. /. ZooL, Lond. 149 : 288-296.
I36 K. E. BANISTER
AMIRTHALINGHAM, C. & YASSAA KHALIFA, M. 1965. A Guide to the Common Commercial
Fresh Water Fishes in the Sudan. Game and Fisheries Dept., Republic of Sudan.
BAILEY, R. G. 1969. The non-cichlid fishes of the eastward flowing rivers of Tanzania, East
Africa. Revue Zool. Bot. afr. 30 (1-2) : 171-199.
BANISTER, K. E. 1972. On the cyprinid fish Barbus alluaudi Pellegrin ; a possible intergeneric
hybrid from Africa. Bull. Br. Mus. not. Hist. (Zool.) 24 (5) : 261-290.
BELL-CROSS, G. 1960. Observations on the movements of fish in a fish ladder in Northern
Rhodesia. Third Symposium on Hydrobiology and Inland Fisheries Problems of Major
Lakes. C.S.A./C.C.T.A. Publication No. 63 : 113-125.
1965. Movement of fish across the Congo-Zambesi watershed in the Mwinilunga district
of Northern Rhodesia. Proc. Cent. Afr. Sci. Med. Congr. : 415-424.
BINI, G. 1940. I pesci del Lago Tana. Missione Stud. Lago Tana 3 (2) : 138-206.
BISHOP, W. W. 1963. The later Tertiary and Pleistocene in eastern equatorial Africa. In :
African Ecology and Human Evolution : 246-275, eds Howell, F. C. and Bouliere, C.
Werner Gren Foundation, New York.
— 1969. Pleistocene stratigraphy in Uganda. Mem. geol. Surv. Uganda 10.
— 1971. The later cenozoic history of East Africa in relation to Hominid evolution.
In : The Late Cenozoic Glacial Ages : 493-527, ed Turkenan, K. L. Yale University.
BOTROS, G. A. 1971. Fishes of the Red Sea. In : Oceanogr. Mar. Biol. Ann. Rev., ed. Barnes,
H. 9 : 221-348. George Allen & Unwin Ltd, London.
BOULENGER, G. A. 1907. Zoology of Egypt : the Fishes of the Nile. Pub. for the Egyptian
Government by Hugh Rees, London.
191 la. Catalogue of the Fresh-water Fishes of Africa in the British Museum (Natural
History], 2. London.
191 ib. Descriptions of two new African barbels. Ann. mag. nat. Hist. (8) 8 : 369-
370-
1916. Catalogue of the Fresh-water Fishes of Africa in the British Museum (Natural History],
4. London.
BROWN, G. F. 1970. Eastern margins of the Red Sea and the coastal structures in Saudi
Arabia. In : A discussion on the structure and evolution of the Red Sea, Gulf of Aden
and Ethiopian rift junction, organised by FALCON, N. L. et alii. Phil. Trans. Roy. Soc.
Lond. A 267 : 75-89.
BROWN, J. M. 1956. Geophysics. In : Harris, N., et alii. Oil in Uganda. Mem. geol.
Surv. Uganda, 9.
CAHEN, J. 1954. Giologie du Congo Beige, Liege.
COOKE, H. B. S. 1958. Observations relating to the quaternary environment in east and
southern Africa. 73 pp. DuToit Memorial Lecture No. 5. Geol. Soc. S. Africa Annex. 60.
COPLEY, H. 1948. Lakes and Rivers of Kenya. Longman Green & Co. Nairobi.
1958. Common Freshwater Fishes of East Africa. Witherby, London.
CRASS, R. S. 1960. Notes on the freshwater fishes of Natal with descriptions of four new
species. Ann. Natal Mus. 14 (3) : 405-458.
DARLINGTON, P. J., JR. 1957. Zoogeography : the Geographical Distribution of Animals.
John Wiley & Sons Inc., New York.
DAVID, L. 1936. Contribution a 1'etude de la faune ichthyologique du Lac Tanganika.
Revue Zool. Bot. afr. 23 (2) : 149-160.
1937- Poissons de 1'Urundie. Revue Zool. Bot. afr. 29 (4) : 413-420.
& POLL, M. 1937. Contribution a la faune ichthyologique du Congo Beige. Annls
Mus. r. Congo Beige (i) 3 (5) : 193-292.
DE HEINZELIN, J. 1955. Le fosse tectonique sous le parallele d'lshango. Explor. Pare natn.
Albert Miss. J. de Heinzelin de Braucourt 1.
DONALDSON-SMITH, A. C. 1897. Through Unknown African Countries. Edward Arnold,
London and New York.
DOORNKAMP, J. C. & TEMPLE, P. H. 1966. Surface, drainage and tectonic instability in part
of southern Uganda. Geogrl. J . 132 (2) : 238-252.
A REVISION OF THE LARGE AFRICAN BARBUS 137
DUBERTRET, L. I9jo. Review of the structural geology of the Red Sea and surrounding areas'
In : A discussion on the structure and evolution of the Red Sea, Gulf of Aden and
Ethiopian rift junction, organised by FALCON, N. L. et alii. Phil. Trans. Roy. Soc. Lond.
A 267 : 9-20.
FARQUHARSON, F. L. 1962. The distribution of cyprinids in South Africa. Ann. Cape prov.
Mus. 2 : 233-251.
FISCHER, J. G. 1884. Uber einige afrikanische Reptilien, Amphibien und Fische des
Naturhistorischen Museums. Jb. hamb. wiss. Anst. 1 : 3-39.
FOWLER, H. W. 1930. Fishes obtained by the Gray African Expedition 1929 with notes
on other species in the Academy Collection. Proc. Acad. nat. Sci. Philad. 72 : 27-83.
— 1936. Zoological results of the George Vanderbilt African Expedition of 1934. Part III,
The freshwater fishes. Proc. Acad. nat. Sci. Philad. 78 : 243-335.
FRYER, G. 1968. The parasitic Crustacea of African freshwater fishes ; their biology and
distribution. /. Zoo/. Lond. 156 : 45-95.
GREENWOOD, P. H. 1959. Quaternary fish fossils. Explor. Pare natn. Miss. J. de Heinzelin
de Braucourt 4 (i) : 1-80.
— 1962. A revision of certain Barbus (Pisces, Cyprinidae) from East, Central and South
Africa. Butt. BY. Mus. nat. Hist. (Zool.) 8 (4) : 151-208.
— 1964. Environmental effects on the pharyneal mill of a cichlid fish Astatoreochromis
alluaudi, and their taxonomic implications. Proc. Linn. Soc. Lond., 176 (i) : i-io.
1966. The Fishes of Uganda. 2nd ed. The Uganda Society, Kampala.
GREGORY, J. W. 1896. The Great Rift Valley, being a Narrative of a Journey to Mount Kenya
and Lake Baringo. John Murray, London.
GROENEWALD, A. A. v. J. 1958. A revision of the genera Barbus and Varicorhinus (Pisces,
Cyprinidae) in Transvaal. Ann. Transv. Mus. 23 (3) : 263-330.
GUNTHER, A. C. 1869. The fish of the Nile. In : PETHERICK, J. & PETHERICK, B. H.,
Travels in central Africa 2 : 197-268. Tinsley Bros., London.
- 1880. An Introduction to the Study of Fishes. Adam & Charles Black, Edinburgh.
— 1894. Report on the collection of reptiles and fishes made by Dr J. W. Gregory during
his expedition to Mount Kenya. Proc. zool. Soc. Lond. : 84-91.
— 1896. Report on a collection of fishes made by Dr A. Donaldson-Smith during his
expedition to Lake Rudolf. Proc. zool. Soc. Lond. : 217-224.
HOLLY, M. 1929. Einige neue afrikanische Fishformen. Anz. Akad. Wiss. Wien 66 :
32-35-
JACKSON, P. B. N. 1961. The Fishes of Northern Rhodesia. Government Printer, Lusaka.
JUBB, R. A. 1961. An Illustrated Guide to the Freshwater Fishes of the Zambesi River, Lake
Kariba, Pungwe, Sabi, Lundi and Limpopo Rivers. Stuart Manning, Bulawayo.
— 1963. A revised list of the freshwater fishes of Southern Africa. Ann. Cape prov. Mus.,
3: 5-39-
— 1965. Freshwater fishes of Cape Province. Ann. Cape prov. Mus. 4 : 1-72.
- i96ya. Freshwater Fishes of South Africa. A. A. Balkema, Cape Town.
— ig67b. Notes on the popular angling yellow fishes and distribution outside their native
range. Piscator 20 : 120-125.
- 1968. The Barbus and Varicorhinus species (Pisces : Cyprinidae) of Transvaal. Ann.
Transv. Mus. 26 (4) : 79-97.
KARAMAN, M. L. 1971. Siisswasserfische der Turkei. 8. Revision der Barben Europas,
Vorderasiens und Nordafrikas. Mitt, hamburg. zool. Mus. Inst. 67 : 175-254.
KENDALL, R. L. 1969. An ecological history of the Lake Victoria basin. Ecol. Monogr.
39 (2) : 121-176.
KLAUSEWITZ, W. & NIELSEN, J. G. 1965. On Forskal's collection of fishes in the zoological
museum of Copenhagen. Spolia zool. Mus. haun. 22 : 1-31.
LADIGES, W., VON WAHLERT, G. & MOHR, E. 1958. Die Typen und Typoide der Fischsamm-
lung des Hamburgischen Zoologischen Staatsinstitut und Zoologischen Museums. Mitt,
zool. Mus. Inst. 56 : 155-167.
138 K. E. BANISTER
MANN, M. J. 1971. Some taxonomical notes on the fish fauna of the Baringo area. Afr.
J. Trop. Hydrobiol. Fish. 1 (i) : 25-34.
MARLIER, G. 1953. Etude biogeographique du bassin de la Ruzizi basee sur la distribution
des poissons. Annls. Soc. r. zool. Belg. (i) 84 : 175-224.
MATTHES, H. 1962. Poissons nouveaux ou interessants du Lac Tanganika et du Ruanda.
Annls. Mus. r. Afr. cent. S6r 8° 111 : 81.
MAYR, E. 1949. Systematics and the Origin of Species. Columbia Biological Series No. 13.
New York.
MENON, A. G. K. 1964. Monograph of the Cyprinid fishes of the genus Garra, Hamilton.
Mem. Indian Mus. 14 (4) : 173-260.
NICHOLS, J. T. & GRISCOM, L. 1917. Fresh-water fishes of the Congo basin obtained by the
American Museum Congo Expedition, 1909-1915. Bull. Am. Mus. nat. Hist. 37 (25) :
653-756.
NORMAN, J. R. 1925. A new siluroid fish of the genus Clarias from South-western Uganda.
Occ. Pap. Boston Soc. nat. Hist. 5 : 189-190.
PAPPENHEIM, P. & BOULENGER, G. A. 1914. Fische. Wiss. Ergebn. dt. ZentAfr. Exped.
Zool. (5) Zoologie 3 : 225-260.
PARENZAN, P. 1939. I pesci del bacino del Lago Regina Margherita nel Galla e Sidama.
Boll. Pesca Piscic. Idrobiol. 15 : 146-173.
PEEL, C. V. A. 1900. Somaliland. Robinson & Co., Gt Russel St, London.
PELLEGRIN, J. 1933. Poissons de la region du Kivu adresses par M. Guy Babault. Bull.
Soc. zool. Fr. 58 : 169-175.
1935- Poissons de la region du Kivu recoltes par M. Guy Babault. Revue Zool. Bot.
afr. 27 (3) : 376-385-
PFEFFER, G. 1889. Uebersicht der von Herrn Dr Stuhlmann in Aegypten, auf Zanzibar
und dem gegeniiberliegenden Festlande gessamelten Reptilien, Amphibien, Fische,
Mollusken und Krebse. Jb. Hamb. wiss. Anst. 6 (2) : 1-36.
POLL, M. 1939. Poissons. Explor. Pare natn. Albert Miss G. F. de Witte 24 : 1-81.
1946. Revision de la faune ichthyologique du Lac Tanganika. Annls. Mus. r. Congo
Beige. Zool. (i) 4 (3) : 141-364.
— 1948. Poissons recueillis au Katanga par H. J. Bredo. Bull. Mus. r. Hist. nat. Belg.
24 (21) : 1-24. •'.
1953. Lac Tanganika (poissons non Cichlidae). Result, scient. Explor. hydrobiol. Lac
Tanganika 3 (5 A) : 1-251.
— 1967. Contribution a la faune ichthyologique de 1'Angola. Publcoes cult. Co. Diam.
Angola 75 : 1-381.
ROBERTS, D. G. 1970. A discussion mainly concerning the contributions by Hutchinson
and by Baker. In : A discussion on the structure and evolution of the Red Sea, Gulf of
Aden and Ethiopian rift junction, organised by FALCON, N. L. et alii. Phil. Trans. Roy.
Soc. Lond. A 267 : 399-407.
RUPPELL, E. 1837. Neuer Nachtrag von Beschreibungen und Abbildungen neuer Fische,
im Nil Entdeckt. Mus. senckenb. 2 : 5-10.
SANDON, H. 1950. An Illustrated Guide to the Freshwater Fishes of Sudan. Sudan Notes Rec.
Khartoum.
STUVIER, M., DEEVEY, E. S. & GRALENSKY, L. J. 1960. Yale natural radiocarbon measure-
ments. Radiocarbon 2 : 56.
TAKING, A. V. 1952. Experimental study of meristic characters in fishes. Biol. Rev. 27 :
169-193.
TEMPLE, P. H. 1969. Some biological implications of a revised geological history for Lake
Victoria. Biol. J. Linn. Soc. 1 (4) : 363-373.
VINCIGUERRA, D. 1898. I pesci dell'ultima spedezione del Cap. Bottego. Annali Mus.
civ. Star. Natn. Giacomo Doria (2) 19 : 240-261.
WHITEHEAD, P. J. P. 1959. The anadromous fishes of Lake Victoria. Revue Zool. Bot.
Afr. 59 (3-4) : 329-363-
A REVISION OF THE LARGE AFRICAN BARBUS 139
WORTHINGTON, E. B. iQ32a. Scientific results of the Cambridge Expedition to the East African
Lakes. 1930-1931. 2. Fishes other than Cichlidae. /. Linn. Soc. zool. 38: 121-134.
i932b. The Lakes of Kenya and Uganda. Geogrl. J. 79 (4) : 275-297.
& RICARDO, C. K. 1937. The fish of Lake Tanganyika (other than Cichlidae). Proc.
zool. Soc. Lond. (4) : 1061-1112.
WORTHINGTON, S. & E. B. 1933. The Inland Waters of Africa. MacMillan & Co. Ltd, London.
WUNDER, W. 1939. Die ' hungerform ' und die ' mastform ' des Karpfens (Cyprinus carpio L.).
Z. Morph. Okol. Tiere 35 (4) : 594-614.
ZOLEZZI, G. 1939. Descrizione di tre nuovi Ciprinidi raccolti dalla Missione Ittologica in
A.O.I. Boll. Pesca Piscicolt. Idrobiol. Roma 15 : 369.
APPENDIX i
An artificial key to the Barbus species described in this paper
Li : > 46
Lake Zwai basin ....... B. ethiopicus
Li : 38-44
a) No anterior barbel ; posterior barbel minute ; ridges of fat on scales of
adults
L. Tanganyika basin . . . . . B. tropidolepis
b) Anterior barbel present
i) IO more than 8 per cent S.L. ; heavy-bodied, bulky appearance ;
pharyngeal bones short and thick, pharyngeal teeth molariform
L. Tanganyika basin . . . . . B. platyrhinus
ii) IO less than 8 per cent S.L. ; compressed, slender body ; pharyngeal
teeth thin and hooked
L. Zwai basin ...... B. microterolepis
Li : < 38
a) Last unbranched dorsal fin ray weak and flexible with persistent
articulations
i) CP1 > CPd ; body shallow ; L24~3O (most often 26-28), striae on
scales parallel
L. Tanganyika basin and upper reaches of Congo . B. caudovittatus
ii) CP1 — CPd ; body deep ; scales with few converging striae (fig. 64)
Lower Malagarasi and Tanzanian coastal rivers . B. macrolepis
b) Last unbranched dorsal fin ray ossified without articulations, at least in
basal half
i) Body typically deep ; CP1 - CPd
t Dorsal spine thin, only ossified basally
L. Mweru region . . . . . . . B. stappersii
If Dorsal spine long, thick and strong . . . B. bynni complex
(except in B. pagenstecheri) (see below)
ii) Body typically shallow ; CP1 > CPd
t Dorsal spine massive, long ; upper jaw curves down to cover ft
the lower jaw laterally giving a 'clupeoid' appearance to
the mouth
Athi-Tana system . . . . . . B. mariae
ft Dorsal spine not as above
% mouth ventral %%
D straight edge to square LJ ; DD
very small barbels
Rwanda . B. microbarbis
140 K. E. BANISTER
DD mouth semicircular
* Dark pigment of distal portions of pectoral, pelvic
and anal fins
Upper Congo, W. of L. Kivu .... B. paucisquamatus
** Dark pigment on proximal parts of pectoral, pelvic
and anal fins
L. Mweru region . . . . . . B. trachypterus
JJ mouth not as above
| striae on scales vary with position of scale on body, Bfl]
typically striae more radiate on belly than on
back ; DIV-g or 10
O Bulky body (fig. 74)
West of rift valley . . . . . . B. mirabilis
O O Body not bulky
Ruwenzori mountains and upper Malagarasi . . B. somereni
•• striae ± parallel and sinuous, no great variation in
pattern with position on body ; DIV-8 or 9
B. intermedius
complex (see below)
Barbus bynni complex
1) DSP, x = 30-0, spine strong .......... 2)
Nile, L. Albert, L. Abaya, L. Rudolf . . B. bynni
2) DSp, x = 25-0-30-0, strong .......... 3)
a) Ab, x = 9 (range 8-10) ; Pb, x = n (range 9-12).
Pet 21-24
Congo W. of L. Kivu . . . . . . B. longifilis
b) Ab, x = 6 (range 3-10) ; Pb x = 7 (range 3-12).
Pet 22-28, x = 24-4
Athi and Tana systems ..... J5. oxyrhynchus
3) DSp 25-0 or less, but weaker than above
Juba river ........ B. gananensis
4) DSp 20-0 or less, shallow body
Kilimanjaro region . . . . . B. pagenstecheri
Barbus intermedius complex
1) Dorsal profile of head concave ; mouth opens antero-dorsally ; nuchal hump
present
Li 27-30 ; DSp, x = 22-3
Rwanda and Burundi . . . . . . B. acuticeps
2) Dorsal profile of head not concave, mouth sub-terminal or terminal
i) Li 25-26 ; pharyngeal teeth tend to be molariform
Rwanda ......... B. ruasae
ii) Li 28-35 (most often 30-34, m = 32)
DSp, x = 16-9 ; Ab, x = 4-3 ; Pb, x = 5-4
Lakes Victoria, Kivu, Edward and George . . B. altianalis
iii) Li 26-34 (most often 28-32, m = 29)
DSp, x = 20-1 ; Ab, x = 5-8 ; Pb, x = 6-9
Blue Nile, Ethiopia, Lakes Rudolf and Baringo, western part
of Juba . . . . . . . B. intermedius
A REVISION OF THE LARGE AFRICAN BARBUS 141
APPENDIX 2
Gazetteer
As I have found it very difficult (and in some cases impossible) to find some of the
localities mentioned by authors I have compiled this gazetteer giving as many
details as I can find of the collecting sites. Grid references are given where possible.
The details of the rivers in Ethiopia come entirely from Zaphiro's notebook.
Different cultures occupying an area have sometimes changed the names of lakes
or rivers. Where possible all the variants are given.
Abaya = Abbaya, the northern lake of a pair of lakes. The name has been loosely applied to
either or both lakes in the southern part of the Ethiopian rift valley. G.R. 6°3o' N 3o°oo' E.
(See Chiamo, Ganjule, Margharita, Ruspoli.)
Akaki, a river, 10-12 m wide, due south of Addis Ababa, it runs from Legadi south to the
Hawash.
Alaba, a river, some 20 m wide, runs from the Kambata plains to Lake Ganjule. Collecting
altitude 6000 feet.
Arba minch, tributary of Lake Ganjule.
Avakubi, Ituri system (Zaire). G.R. i°24' N, 27°4o' E.
Awata, a tributary of the Juba river. G.R. 6°O5' N, 39°2o' E.
Barja, a river, 10 m wide, collecting altitude 4250 feet. It runs from the hills of Sangana and
Bako to the Zuja river (Lake Stephanie basin).
Basso Ebor, an old name for Lake Stephanie.
Basso Narok, an old name for Lake Rudolf.
Bis (s) an Guarrica, a tributary of the Sagan river to the south of Lake Abaya.
Bobandana, Lake Kivu basin. G.R. i°38' S, 29°oo' E.
Bushiame river, Sankuru system. G.R. 6°oo' S, 24°5o' E to 8°oo' S, 23°oo' E.
Burka, an affluent of Lake Ganjule.
Chiamo (= Ciamo), Lake Ganjule.
Chirangobo, an affluent of Lake Kivu.
Didessa, a river, some 60 m wide, that runs from Guma to the Blue Nile.
Elgo, a fast-flowing river, 15 m wide, collecting altitude 3000 feet, it flows from the Gamu hills
eastwards to Lake Abaya.
Ergino, a river, 20 m wide, collecting altitude 3000 feet, flows from the Basketo hills northwards
to the Omo.
Errer, a river in the Webi Shebeli system. G.R. 9°oo' N, 42°2o' E.
G. el Narua, Guasso el Narua. G.R. o°i6' N, 36°i8' E.
G. Nyuki, Guasso Nyuki, mouth at o°28' N, 36°o8' E (probably a swampy locality).
Ganda, a river, 15 m wide, collecting altitude 3000 feet, flows from the hills of Gama to Lake
Ganjule.
Ganjule, the southern lake of the pair of which the northern lake is Abaya (also called Chiamo,
Ruspoli). G.R. 6°50' N, 37°4o' E.
Gato, a fast flowing river, 15 m wide, which rises in the hills of Gandulla and flows west to the
Sagan. Collecting altitude 3700 feet. (Lake Stephanie basin.)
Gibe, a fast-flowing stream, 20 m wide, which flows directly south from the plains of Gorombi
(altitude 2800 feet) (supposed to be the source of the Omo).
Gofa, a river, Hawash system.
Gombela. G.R. io°5o' S, 27°5o' E.
Gudar (= Gudr), a stream flowing north from the Roggle mountains to the Blue Nile.
Collecting altitude 3400 feet.
Hawash, most collections in this river were made by Zaphiro at Zeluka, altitude 4100 feet.
Jerrer, exact locality untraceable. Mount Jerrer is 32 km southwest of Addis Ababa, according
to Zaphiro's notebook, the river is therefore apparently in the Webi Shebeli system.
142 K. E. BANISTER
Juju, a river, 10 m wide, flowing from the Guma hills to the Blue Nile. Collecting altitude
2000 feet.
Kabiashyia, on the Luanza river, an affluent to the northern edge of Lake Mweru.
Kando, near Tenke. G.R. io°25' S, 26°io' E.
Kansihiti, a river, an affluent of the Luhoho Congo, due west of Lake Kivu.
Kassam, a river, Hawash system. G.R. 9°O5' N, 39°35' E.
Kibonoto ( = Kibongoto), an affluent of the Sanya river, Pangani system. G.R. 3°i i' S, 37°o6' E.
Kibwesi, a river in the Athi system. G.R. ± 2°25' S, 37°56' E.
Kiyimbi, a river, Loama-Lualaba Congo. G.R. 5°oo' S, 28°59' E.
Koki, a river, an affluent of Lake Tanganyika. G.R. 6°O3' S, 29°O5' E.
Laikipa (Leikipa), an escarpment. G.R. o°25' N, 36°io' E.
Loama, an affluent of the Luhoho Congo, due west of Lake Kivu.
Luembe, a river = Cashimo river. G.R. 8°oo' S, 2i°35' E.
Lufiro, a river. G.R. 2°4o' S, 29°oo' E.
Luilu, a river, Sankuru system. G.R. 7°3o' S, 23°3o' E.
Lusine, a river, Rwanda, flows from Lake Mohashi into the Akangaru. G.R. 2°3o' S to 6°oo' S,
30°45' E.
Mahembe, on the Nyawarongo river, Kagera system, Lake Victoria basin. G.R. ± i°52' S,
29°54' E.
Maki, a fast-flowing stream that originates in the hills of Goraza and flows into Lake
Zwai.
Makindu, a river. G.R. 2°O9' S, 37°35' E.
Malawa, a river, Lake Victoria basin. G.R. o°4o' N, 35°3o' E.
Margharita, lake = Lake Abaya.
Metti, a fast-flowing river, 20 m wide, altitude 3500 feet. Flows from Tuludimtu northwards
to the Gudar river. (Blue Nile.)
Mkunga (= Mukungwa), a river, near Ruasa, northwest Rwanda, runs from the southeastern
part of Lake Luhondo to the Nyawarongo. G.R. i°35' S, 29°4o' E.
Modjo, a river, Webi Shebeli system.
Mogre, a river, tributary of the Blue Nile.
Mutamphu, a river, a tributary of the River Akangaru, Kagera system. Specimens with this
locality were collected about i 2 km along the road to Shangugu from Butare. G.R.
± 2°43' S, 29043' E.
Mwogo, an affluent of the Kagera.
Narok, a river. G.R. o°32' N, 36°52' E.
Nyabarongo, a river in the Luhoho Congo system.
Nyabugogo, an affluent of the Lusine river.
Nyawarongo, Kagera system.
Nyundeulu. G.R. io°58' S, 25°5o' E.
Rugwero. Lake, Rwanda. G.R. 2°25' S, 3O°2o' E.
Ruspoli = Lake Ganjule.
Sagan, a fast-flowing stream, 15 m wide, at Wondo (altitude 2800 feet) where most collecting
was done. Blue Nile system.
Sang6, an affluent of the Ruzizi. G.R. 3°O4' S, 29°o8' E.
Sir 6 (= Siri), a stream flowing from the Gamu hills eastwards to Lake Abaya, collecting altitude
3000 feet.
Suksuki, a river linking Lake Zwai and Lake Suksuki.
Tchitatu, a river, near Luachimo, Upper Kasai region. G.R. 7°4o' S, 2O°5o' E.
Tokwe, a river, near Bwambe, Uganda.
Tsavo, a river. G.R. 2°59' S, 38°O2' E.
Tshikapa, a river, Upper Kasai (Congo) system. G.R. 6°28' S, 2o°48' E.
Uaso narok, a river flowing from the Laikipa escarpment northeastwards to the Uaso Nyiro.
Urgessa, a river, 15 m wide, flowing northwest to the Wama river (Blue Nile system).
Wondinak, a small stream flowing northwest into the Gibe river (Omo system).
A REVISION OF THE LARGE AFRICAN BARBUS 143
Zeissi, a turbulent stream, about 10 m wide, flowing from the Zeissi hills eastwards to Lake
Ganjule. G.R. i°52' N, 37^9' E.
Zendo, a river, 15 m wide, flowing from the Anko hill eastwards to the Maze river, an affluent
of the Omo. Collecting altitude 4300 feet.
Zuga, a river, 30 m wide, running from the hill Marta to Lake Stephanie. Collecting altitude
4200 feet.
APPENDIX 3
Barbus susanae, a replacement name for Barbus gregorii Norman.
Barbus gregorii Norman (1923) from China is preoccupied by Barbus gregorii
Boulenger (1902) from Africa. Fowler (1958) noticed this and proposed Barbus
yunnanensis to replace Barbus gregorii Norman. Fowler had unfortunately over-
looked the fact that Barbus yunnanensis had already been used by Regan (1904)
for a fish from Yunnan, China, the same area as that from which Barbus gregorii
Norman comes.
I have examined the type of Barbus yunnanensis Regan B.M. (N.H.) Reg. No.
1904.1.26 : 78 and the syntypes of Barbus gregorii Norman B.M. (N.H.) Reg. Nos.
1923.2.21 : 29-36, and they are very distinct.
I therefore propose Barbus susanae to replace Barbus gregorii Norman.
Derivation of name : named after my wife.
BOULENGER, G. A., 1902
Description of new fishes from the collection made by Mr E. Degen in Abyssinia.
Ann. Mag. nat. Hist. (7) 10 : 422.
REGAN, C. T., 1904
A collection of fishes made by Mr J. Graham at Yunan Fu. Ann. Mag. nat. Hist.
(7) 13 : i9i.
NORMAN, J. R., 1923
Three new fishes from Yunan collected by Professor J. W. Gregory, F.R.S. Ann. Mag.
nat. Hist. (9) 2 : 562.
FOWLER, H. W., 1958
Some new taxonomic names for fish-like vertebrates. Not. nat. (310). August 1958 : 12.
APPENDIX 4
A complete list of the registered material examined.
In some cases more specimens were examined than were used in the description
of the species, consequently a complete list was deemed advisable.
The Ethiopian Barbus spp. from the Stazione Idrobiologia, Rome, do not have
any register numbers.
The species are arranged alphabetically and the following code applies :
B.M. (N.H.) British Museum (Natural History)
C.F.M. Chicago Field Museum
M.A.C.T. Musee d'Afrique Centrale, Tervuren
M.H.N.P. Museum National d'Histoire Naturelle, Paris
M.S.N.G. Museo Civico di Storia Naturale G. Doria, Genova
P.A.S. Philadelphia Academy of Natural Sciences, U.S.A.
S.M.F. Senckenberg Museum, Frankfurt
S.M.N.H. Stockholm Museum of Natural History
Z.M.B. Zoologisches Museum an der Humboldt - Universitat zu Berlin
Z.M.H. Zoologisches Staatsinstitut und Zoologisches Museum, Hamburg.
I44 K- E- BANISTER
The first paragraph contains the register numbers of the type specimens of the
species and of the species now in the synonymy. The following paragraph contains
entirely non-typical material.
Barbus acuticeps M.A.C.T. 130313 (holotype) ; 130310-312 ; 130314 (paratypes).
M.A.C.T. 91755-6, 92214.
Barbus alluaudi M.H.N.P. 09-586, 09-587 (syntypes).
Barbus altianalis B.M. (N.H.) 1906.9.6 : 13 (lectotype of Barbus altianalis altianalis)
1906.9.7 : 41 (holotype of Barbus altianalis eduardianus) ; 1904.5.19 : 13 (holotype of Barbus
altianalis radcliffii) ; 1906.5.30 : 117-121 ; 1906.9.6 : 14-15 ; 1906.9.7 : 41 ; 1906.9.7 : 42-43
1911.3.3:5; 1929.1.24:84; 1929.1.24:105-108; 1929.1.24:191-192; M.H.N.P. 30-118
35-153 ; Z.M.H. 19052.
B.M. (N.H.) 1906.5.30:107-115; 1912.10.15:15-19; 1912.10.31:1; 1925.8.7:1-2
1928.1.25 : 12 ; 1928.5.24 : 18-25 '• 1932.6.13 : 280-289 ; I932-6.i3 : 312-320 ; 1932.6.13 : 332
1938.12.6:19; 1957.10.1:1-9; 1962.7.26:1-14; 1965.11.7:7-10; 1966.6.3:192; 1966.9.1
i; 1966.9.1:2-4; 1967.5.17:52-60; 1971.1.5:121-133; 1971.2.19:28-57; 1971.8.31
7-12 ; 1971.8.31 : 28-39 ; M.H.N.P. 35-143, 144 ; M.A.C.T. 66182 ; 66183 ; 92608 ; 91118-
91122; 129096; Z.M.H. 19052.
Barbus bynni B.M. (N.H.) 1907.12.2 : 1230 (neotype) ; 1932.6.13 : 300-303 ; M.S.N.G.
J7333; M.H.N.P. 05-275.
B.M. (N.H.) 1861.9.9 : 39-43 ; 1861.9.9 : 64 ; 1862.6.17 : 117-122 ; 1862.6.17 : 130-131 ;
1905.10.26:11; 1907.12.2:1181-1229; 1907.12.2:1231-1251; 1907.12.2:3721-3728;
1908.1.20:111-116; 1908.11.7:1; 1929.1.24:109-112; 1937.4.20:7-9; 1968.7.24:37;
1970.12.15 : i ; M.S.N.G. 17337 ; P.A.S. 16710.
Barbus caudovittatus M.A.C.T. 1168 (holotype) ; B.M. (N.H.) 1901.12.26 : 26 (paratype) ;
1919.7.24 : 7 ; 1919.7.24 : 8-9.
B.M. (N.H.) 1920.5.26 : 75-76 ; 1936.6.15 : 643-644 ; M.A.C.T. 6785 ; 6786 ; 6992 ; 14551 ;
154!°; 15584; 39456; 43823; 44446-44461; 44482; 44483; 44551-44563; 47341;
48504; 48505; 50061-62; 56416; 56417; 61304-63011; 74754; 774°7; 78927; 78928;
81618; 81619; 81620; 81622-31; 81632-35; 81637-81656; 81661-81984; 81988; 91117;
92561 ; 92562 ; 92563-78 ; 92608 ; 94318 ; 96108-114 ; 99654 ; 102024 ; 121781-826 ;
124937 ; 125774-780 ; 129095 ; 130067 ; 130068 ; 131355 ; 134956 ; 134957 ', ^8957 ',
148829 ; 148830 ; 148839 ; 153485 ; 160152-163 ; 160165-67 ; 164571 ; 166954.
Barbus ethiopicus B.M. (N.H.) 1971.7.12 : 1-3.
Barbus gananensis M.S.N.G. 17525 (holotype, not seen by me) ; M.S.N.G. 17331 ; 17339 ;
I7341 ; 17342.
Barbus intermedius S.M.F. 6778 (holotype of Barbus intermedius intermedius) ; B.M. (N.H.)
No. 18 in 1932.6.13 : 191-200 (holotype of Barbus intermedius australis) ; S.M.F. 2586 ; 2619 ;
6779 ; 6786 ; M.H.N.P. 05-252 ; 05-257 ; B.M. (N.H.) 1893.12.2 : 40-45 ; 1902.12.13
211-212; 1902.12.13:225-228; 1902.12.13:261-270; 1902.12.13:274-275; 1902.12.13
294; 1902.12.13:295-298; 1902.12.13:303-304; 1902.12.13:305-306; 1902.12.13:309
1902.12.13:328-331; 1903.11.16:1-7; 1908.1.20:100; 1908.1.20:103-106; 1908.1.20
107-109; 1908.1.20:110; 1908.1.20:131-133; 1908.1.20:170; 1908.1.20:181-183
1916.1.14 : 7 ; 1937.4.20 : 66.
B.M. (N.H.) 1893.12.2 : 46-47 ; 1901.6.24 : 83-85 ; 1902.12.13 : 229 ; 1902.12.13 : 231-250
1902.12.13:277-283; 1902.12.13:284-290; 1902.12.13:308; 1902.12.13:311; 1902.12.13
312-315 ; 1902.12.13 : 317-319 ; 1902.12.13 : 320-326 ; 1902.12.13 : 332-337 ; 1902.12.13
339; 1902.12.13 : 357 ; 1908.1.20:86; 1908.1.20:91-95; 1908.1.20:97-99; 1908.1.20
101-102 ; 1908.1.20 : 117-130 ; 1908.1.20 : 134-155 ; 1908.1.20 : 157-168 ; 1908.1.20
171-180; 1912.3.22:50-60; 1912.11.11:9-10; 1932.6.13:191-200 (less the holotype o:
Barbus intermedius australis); 1932.11.5:246-60; 1937.4.20:29-37; 1937.4.20:39-60
A REVISION OF THE LARGE AFRICAN BARBUS 145
1937.4.20:65; 1959.12.15:82-86; 1968.7.24:3-7; 1968.7.24:9-17; 1968.7.24:20-48;
1971.8.31:15-16; P.A.S. 14541 ; 14542.
Barbus longifilis M.H.N.P. 35-150 (lectotype) ; 35-145 - 149 ; 35~I5I > 35-^5^-
M.H.N.P. 35-67 ; M.A.C.T. 42934.
Barbus macrolepis Z.M.H. £[331 (lectotype).
B.M. (N.H.) 1909.2.25 : 8 ; 1922.4.18 : 13 ; 1971.6.22 : 131-134. 1972-11.28 : 9-12.
Barbus mariae holotype not seen.
B.M. (N.H.) 1936.12.22 : 35-39 ; iQ37-6-4 : l6-
Barbus microbarbis M.A.C.T. 41847 (holotype).
Barbus microterolepis B.M. (N.H.) 1902.12.13 : 220 (holotype).
Barbus mirabilis Z.M.B. 19059 (holotype).
Barbus oxyrhynchus Z.M.H. H339 (lectotype) ; B.M. (N.H.) 1893.12.2 : 24-29 ; 1893.12.2
32-34 ; 1893.12.2 : 37-39 ; 1902.5.26 : 25-28 ; 1902.5.26 : 35-38 ; 1906.8.25 : 7-9 ; M.H.N.P
26-285 ; S.M.N.H. 9238 ; F.M.C. 6108 ; 6109 ; Z.M.H. 340.
B.M. (N.H.) 1893.12.2 : 30 ; 1893.12.2 : 35 ; 1902.5.26 : 29-34 • I9°4-I-3° : 5-18 ; 1904.10.10
2-4; 1904.12.23:48-49; 1906.8.25:6; 1906.8.25:10-16; 1908.9.17:8-12; 1909.11.15
43-47; 1914.9.21:4; 1915.12.2:2-20; 1936.12.22:2; 1936.12.22:30-34; 1937.12.11
1-4; 1937.6.4:15; 1959.12.15:81; 1965.11.1:7-12; 1966.6.3:252; 1969.3.24:10; 1969.3.24
44-50 ; 1971.5.10 : 55 ; 1971.8.12 : 2-6 ; F.M.C. 6110 ; M.A.C.T. 47341 ; S.M.N.H. 8061
Barbus pagenstecheri Z.M.H. H.342 (lectotype), H.34I (paralectotype) .
Barbus paucisquamatus M.H.N.P. 35-76 (lectotype).
M.H.N.P. 35-77; 35-78; 35-I53: 35-r54 '. 35-IlS (3 specimens); M.A.C.T. 42932;
130145 ; 130146.
Barbus platyrhinus B.M. (N.H.) 1906.9.6 : 12 (holotype).
M.A.C.T. 89789-92 ; 130658-60.
Barbus ruasae Z.M.B. 19053 (lectotype, one of two specimens).
M.A.C.T. 91755 ; 91756 ; 92214.
Barbus somereni B.M. (N.H.) 1911.7.26 : i (holotype) ; M.A.C.T. 46963-65.
B.M. (N.H.) 1969.3.3:13-14; 1971.1.5:96-99; 1971.1.5:100-117; 1971.1.5:120;
1971.1.5:145; 1971.2.19:26-27; 1971.8.31:13-14; 1971.10.18:1-5; M.A.C.T. 46952-62 ;
46966-47339; 47342; 55788-792; 56456-478; 56480-91; 71781-82; 73I57~73i62 ;
87692-695; 87696-698; 94221-232; 92579-89; 92591-92607; 93356-377; 129108.
Barbus stappersii M.A.C.T. 14250 (holotype) ; 14765 ; 14233 ; 14172.
M.A.C.T. 81945 ; 81985 ; 122295 '• !38958-959.
Barbus trachypterus M.A.C.T. 81621 (holotype).
M.A.C.T. 129097-100 ; 165254-256.
Barbus tropidolepis B.M. (N.H.) 1906.9.6: 19-21 (the smallest specimen is the lectotype).
B.M. (N.H.) 1906.9.8 : 50-52 ; 1920.5.25 : 38-46 ; 1936.6.15 : 568-629 ; 1955.12.20 :
720-731; 1955.12.20:733-815; 1955.12.20:873; 1955.12.20:888; 1955.12.20:1169-1171.
INDEX
The generic name Barbus has been omitted from this list. Where the genus is not Barbus it is
Varicorhinus and is symbolized by V.
acuticeps Matthes 1959, 5
affinis Ruppell 1837 = B. intermedius intermedius, 47
affinis brevibarbus Bini 1940 = B. intermedius intermedius, 47
affinis nedgia Bini 1940 = B. intermedius intermedius, 47
ahlselli Lonnberg 1911 = B. oxyrhynchus, 94
146 INDEX
alluaudi Pellegrin 1909, 9
altianalis Blgr. 1900, 9
altianalis altianalis Worthington 1932, 9
altianalis eduardianus Worth. 1932, 9
altianalis radcliffii Worth. 1932, 9
altianalis var. labiosa Pell. 1932 = B. altianalis altianalis, g
altianalis var. lobogenysoides Pell. 1935 = B. paucisquamatus, 102
altianalis var. longifilis Pell. 1935 = B. longifilis, 76
altianalis var. paucisquamata Pell. 1935 = B. paucisquamatus, 102
altianalis urundensis David 1937 = &• somereni, in
alticola Blgr. 1906 = B. intermedius intermedius, 47
a^Ai Hubbs 1918 = B. oxyrhynchus, 94
babaulti Pell. 1926 = B. oxyrhynchus, 94
babaulti Pell. 1935 = B. lapsus, 94
F. babaulti : Bertin and Esteve 1947 = B. oxyrhynchus, 94
bayoni Blgr. 1911 = B. altianalis altianalis, 9
Capoeta bingeri Pell. 1905 = B. intermedius intermedius, 47
bingeri : Blgr. 1911 = .B. intermedius intermedius, 47
F. bingeri : Bertin and Esteve 1947 = B. intermedius intermedius, 47
bottegi Blgr. 1906 = B. intermedius intermedius, 47
F. bredoi Poll 1948 = B. trachypterus, 119
brevibarbus Blgr. 1902 = B. intermedius intermedius, 47
brunellii Bini 1940 = B. intermedius intermedius, 47
brunellii acutirostris Bini 1940 = B. intermedius intermedius, 47
bynni (Forssk.) 1775, 27
bynni rudolfianus Worthington 1932 = B. bynni, 27
caudovittatus Blgr. 1902, 36
F. chapini Nicholls and LaMonte 1950 = B. tropidolepis, 123
chilotes sakaniae Poll 1938 = B. caudovittatus, 36
curtus Blgr. 1915 = B. stappersii, 116
dainellii Bini 1940 = B. intermedius intermedius, 47
dainellii macrocephalus Bini 1940 = B. intermedius intermedius, 47
degeni Blgr. 1902 = B. intermedius intermedius, 47
degeni leptorhinus Bini 1940 = B. intermedius intermedius, 47
donyensis Holly 1929 = B. oxyrhynchus, 94
duchesnii Blgr. 1902 = B. intermedius intermedius, 47
duchesnii ibridus Bini 1940 = B. intermedius intermedius, 47
duchesnii maximus Bini 1940 = B. intermedius intermedius, 47
eduardianus Blgr. 1901 = B. altianalis eduardianus, 9
elongatus Riipp. 1837 = B. intermedius intermedius, 47
erlangeri Blgr. 1903 = B. intermedius intermedius, 47
ethiopicus Zolezzi 1940, 41
euchilus Blgr. 1919 = B. caudovittatus, 36
eumystus Blgr. 1906 = B. intermedius intermedius, 47
fergussonii Blgr. 1901 = B. altianalis eduardianus, 9
gananensis Vincig. 1895, 44
gorguarii Riipp. 1837 = B. intermedius intermedius, 47
gorguarii macroptalmus Bini 1940 = B. intermedius intermedius, 47
INDEX 147
gregorii Blgr. 1902 = B. intermedius australis, 75
gudaricus Blgr. 1906 = B. intermedius intermedius, 47
harringtonii Blgr. 1902 = B. intermedius intermedius, 47
hindii Blgr. 1902 = B. oxyrhynchus, 94
hollyi Lohberger 1929 = B. altianalis radcliffii, g
hursensis Blgr. 1902 = B. intermedius intermedius, 47
ilgi Pell. 1905 = B. intermedius intermedius, 47
intermedius Rtipp. 1837, 47
intermedius australis s. sp. nov., 75
intermedius gorgorensis Bini 1940 = B. intermedius intermedius, 47
intermedius intermedius s. sp. nov., 74
intermedius microstoma Bini 1940 = B. intermedius intermedius, 47
jarsinus Blgr. 1902 = B. intermedius intermedius, 47
kassamensis Blgr. 1902 = B. intermedius intermedius, 47
kiogae Worth. 1929 = B. altianalis radcliffii, 9
kivuensis Pappenheim 1914 = B. altianalis altianalis, 9
krapfi Blgr. 1911 = B. oxyrhynchus, 94
labiatus Blgr. 1902 = B. oxyrhynchus, 94
leptosoma Blgr. 1902 = B. intermedius intermedius, 47
lestradei David 1936 = B. caudovittatus, 36
lobogenys Blgr. 1906 = B. altianalis radcliffii, 9
longifilis : Pellegrin 1935, ?6
longirostris Worth. 1929 = B. altianalis radcliffii, 9
macmillani Blgr. 1906 = B. intermedius intermedius, 47
macrolepis Pfeffer 1889, 79
macronema Blgr. 1902 = B. intermedius intermedius, 47
macronema var. parenzani Zolezzi 1940 = B. intermedius intermedius, 47
margaritae Blgr. 1906 = B. intermedius intermedius, 47
mariae Holly 1929, 83
mathoiae Blgr. 1911 = B. oxyrhynchus, 94
matris Holly 1928, 83
meneliki Pell. 1905 = B. bynni, 27
mento Blgr. 1902 = B. intermedius intermedius, 47
microbarbis David and Poll 1937, 87
microterolepis Blgr. 1902, 89
miochilus Blgr. 1919 = B. caudovittatus, 36
mirabilis Papp. 1914, 91
moeruensis Pell. 1922 = B. stappersii, 116
nairobi Holly 1928 = B. oxyrhynchus, 94
nedgia Riipp. 1837 = B. intermedius intermedius, 47
neuvillii Pell. 1905 = B. intermedius intermedius, 47
obesus Worth. 1929 = B. altianalis radcliffii, 9
oreas Blgr. 1902 = B. intermedius intermedius, 47
oxycephalus Blgr. 1915 = B. stappersii, 116
oxyrhynchus Pfeffer 1889, 94
148 INDEX
pagenstecheri Fischer 1884, 101
paucisquamatus : Pellegrin 1933, 103
perplexicans Blgr. 1902 = B. oxyrhynchus, 94
pietschmanni Lohberger 1929 = B. altianalis vadcliffii, 9
plagiostomus Blgr. 1902 = B. intermedius intermedius, 47
platyrhinus Blgr. 1900, 106
platystomus Blgr. 1902 = B. intermedius intermedius, 47
platystomus daga Bini 1940 = B. intermedius intermedius, 47
platystomus dekkensis Bini 1940 = B. intermedius intermedius, 47
platystomus platystomus Bini 1940 = B. intermedius intermedius, 47
platystomus prognathus Bini 1940 = B. intermedius intermedius, 47
platystomus var. vatovae Zolezzi 1940 = B. intermedius intermedius, 47
pojeri Poll 1944 = B. caudovittatus, 36
procatopus Blgr. 1914 = B. intermedius australis, 75
radcliffii Blgr. 1903 = B. altianalis radcliffii, 9
rhinoceros Copley 1938 = B. mariae, 83
ruasae Pappenheim and Blgr. 1914, 109
ruepelli Blgr. 1902 = B. intermedius intermedius, 47
ruspolii Vincig. 1896 = B. bynni, 27
somereni Blgr. 1911, in
stappersii Blgr. 1915, 116
V. stappersii Blgr. 1917 = B. caudovittatus, 36
surkis Riipp. 1837 = B. intermedius intermedius, 47
tanensis Giinther 1894 = B. oxyrhynchus, 94
trachypterus Blgr. 1915, 119
tropidolepis Blgr. 1900, 123
urundensis : Poll 1946 = B. somereni, in
volpinii Parenzan 1940 = B. intermedius intermedius, 47
zaphiri Blgr. 1906 = B. intermedius intermedius, 47
zuaicus Blgr. 1906 = B. intermedius intermedius, 47
DR K. E. BANISTER
Department of Zoology
BRITISH MUSEUM (NATURAL HISTORY)
CROMWELL ROAD
LONDON SW7 560
A LIST OF SUPPLEMENTS
TO THE ZOOLOGICAL SERIES
OF THE BULLETIN OF
THE BRITISH MUSEUM (NATURAL HISTORY)
1. KAY, E. ALISON. Marine Molluscs in the Cuming Collection British Museum
(Natural History) described by William Harper Pease. Pp. 96 ; 14 Plates.
1965. (Out of Print.) £3.75.*
2. WHITEHEAD, P. J. P. The Clupeoid Fishes described by Lacepede, Cuvier and
Valenciennes. Pp. 180 ; u Plates, 15 Text-figures. 1967. £4.
3. TAYLOR, J. D., KENNEDY, W. J. & HALL, A. The Shell Structure and
Mineralogy of the Bivalvia. Introduction. Nuculacea-Trigonacea. Pp. 125 ;
29 Plates, 77 Text-figures. 1969. £4.50.
4. HAYNES, J. R. Cardigan Bay Recent Foraminifera (Cruises of the R.V. Antur)
1962-1964. Pp. 245 ; 33 Plates, 47 Text-figures. 1973. £10.80.
Printed in Great Britain by John Wright and Sons Ltd. at The Stonebridge Press, Bristol BS4 jNU
SOME DIGENETIC TREMATODES
FISHES FROM THE BAY OF BISCAY
AND NEARBY WATERS
R. A. BRAY
BULLETIN OF
THE BRITISH MUSEUM (NATURAL HISTORY)
ZOOLOGY Vol. 26 No. 2
LONDON: 1973
SOME DIGENETIC TREMATODES IN FISHES
FROM THE BAY OF BISCAY
AND NEARBY WATERS
BY
RODNEY ALAN BRAY
X
Pp 149-183 ; ii Text-figures
BULLETIN OF
THE BRITISH MUSEUM (NATURAL HISTORY)
ZOOLOGY Vol. 26 No. 2
LONDON: 1973
THE BULLETIN OF THE BRITISH MUSEUM
(NATURAL HISTORY), instituted in 1949, is
issued in five series corresponding to the Departments
of the Museum, and an Historical series.
Parts will appear at irregular intervals as they
become ready. Volumes will contain about three or
four hundred pages, and will not necessarily be
completed within one calendar year.
In 1965 a separate supplementary series of longer
papers was instituted, numbered serially for each
Department.
This paper is Vol. 26, No. 2, of the Zoological series.
The abbreviated titles of periodicals cited follow those
of the World List of Scientific Periodicals.
World List abbreviation :
Bull. Br. Mus. nat. Hist. (Zool.)
Trustees of the British Museum (Natural History), 1973
TRUSTEES OF
THE BRITISH MUSEUM (NATURAL HISTORY)
Issued 15 November, 1973 Price £1.60
SOME DIGENETIC TREMATODES IN FISHES
FROM THE BAY OF BISCAY
AND NEARBY WATERS
By RODNEY A. BRAY
CONTENTS
SYNOPSIS ........... 151
INTRODUCTION .......... 151
HOST - PARASITE LIST . . . . . . . . 154
SYSTEMATIC SECTION ......... 155
Family BUCEPHALIDAE ........ 155
Family OPECOELIDAE ......... 159
Family LEPOCREADIIDAE ........ 162
Family AZYGIIDAE ......... 167
Family ZOOGONIDAE ......... 167
Family FELLODISTOMIDAE . . . . . . . .173
Family HEMIURIDAE ........ .178
ACKNOWLEDGMENTS .......... 181
REFERENCES 181
SYNOPSIS
Sixteen species of digenetic trematodes are recorded from seventeen species of fishes caught in
the Bay of Biscay and in waters off the north-west coast of Spain and the Atlantic coast of
Morocco.
The following two species are described as new : Bathycreadium biscayense and Steringophorus
blackeri. Seven of the remaining species are described, these are Dolichoenterum longissimum
Ozaki, Prosorhynchus crucibulum (Rud.), Lepidapedon rachion (Cobbold), L. elongatum (Lebour),
Steganoderma abyssorum (Odhner), Neosteganoderma glandulosum Byrd and (?) Steringotrema
diver gens (Rud.). The following new combinations are made : Bathycreadium [Nicolla] elonga-
tum (Maillard) and Neosteganoderma [Proctophantastes] polymixiae (Yamaguti).
INTRODUCTION
THE author was given the opportunity in 1971 to obtain parasitic worms from marine
fishes, whilst accompanying a cruise of the Ministry of Agriculture, Fisheries and
Food Research Vessel 'Cirolana', the primary purpose of which was to collect blood
from hake for electrophoretic and chromosome studies. The collecting was carried
out between the ist and the i4th of January and was conducted primarily in the
Bay of Biscay, but also in other areas, including the waters off the western coasts of
Spain and Morocco. The locations of the stations mentioned in this report are
shown on Fig. i and further details are given in Table I.
The fish were caught by stern trawl and the alimentary tract removed from them.
The stomach and intestine were separated before being opened up with scissors, and
then the contents were shaken into a tube of sea-water. These tubes were kept for
152
R. A. BRAY
N.I.O.
6429
FIG. i. Map showing positions of stations mentioned in this report.
FISH TREMATODES FROM BAY OF BISCAY 153
ft I"
<D C
ft "~
4
^
ffi
±5 tj HfOOOOOrONfO^-Nt^OOOO
2j >> looooooooooooo
OOOOOOOOOOOOO
R-r £ooroMo»t^ooo^"-it^M'<t-
.•SH ~*NO»o-<*-HroTj-u-)ioH-(O'^-^-
-4->ij -4->ooooooooooooo
rt PH o oo c**1 *«o *O ^o ^^ co oo co o O co co
S~ M O w>
— 58 ?L o
H
w
«
^
itation List
ft'? Ot^OOOOOinOONOOOC
ft ""
(ji
_^4J OMOOOOOOMOOfOMt^O^
'tuO ^ oooooooooooooo
H
CO
u->ioo
RCO OMH
;rt CQ ino>O
C— I d. WWH
I
i
CO
154
R. A. BRAY
periods of up to six hours in a refrigerator at just above freezing point, before being
topped up with a fixative, either 'Dowicil' or 10 per cent formalin. A degree of
contraction resulted from the use of these fixatives.
The material from the 'Cirolana' cruise was supplemented by two trematodes
collected from preserved fish in the collection of the National Institute of Oceano-
graphy (now the Institute of Oceanographic Sciences) at Wormley, Surrey.
The parasites of seventeen species of fish are mentioned in this report . The maj ority
of the sixteen species of digenetic trematodes studied are already well known as
widespread species. It is interesting to note, however, that Neosteganoderma
glandulosum, reported originally from Florida, has now been recovered from off the
north-west coast of Spain and that Steganoderma abyssorum, previously reported
only from Norway, occurs in the eastern Bay of Biscay. Further evidence of the
affinities of species in the area under discussion with Florida on one hand, and the
north-eastern North Atlantic on the other, is indicated by the finding of two new
species whose closest relatives appear to be species from these two areas.
HOST-PARASITE LIST
Alosa fallax (Lacepede) - Hemiurus appendiculatus (Rud.).
Beryx decadactylus Cuvier - Neosteganoderma glandulosum Byrd.
Capros aper (L.) - (?) Steringotrema diver gens (Rud.).
Coelorinchus caelorhinchus (Risso) - Bathycreadium biscayense sp. nov.
Steganoderma abyssorum (Odhner).
Conger conger (L.) - Dolichoenterum longissimum Ozaki.
Prosorhynchus crucibulum (Rud.).
Lecithochirium gravidum Looss.
Dalatias licha (Bonnaterre) - Otodistomum veliporum (Creplin) .
Gaidropsarus vulgaris (Cloquet) - Helicometra fasciata (Rud.).
Lepidion eques (Giinther) - Bathycreadium biscayense sp. nov.
Lepidapedon elongatum (Lebour).
Steganoderma abyssorum (Odhner).
Lophius piscatorius L. - Bucephahpsis gracilescens (Rud.).
Merluccius merluccius (L.) - Derogenes various (Muller).
Micromesistius poutassou (Risso) -Derogenes various (Muller).
Molva macrophthalma (Rafinesque-Schmaltze) -Derogenes various (Muller).
Phycis blennoides (Brunnich) - Lepidapedon rachion (Cobbold) .
Derogenes various (Muller).
Scomber scombrus L. - Lecithocladium excisum (Rud.).
Trachyrhynchus trachyrincus (Risso) - Bathycreadium biscayense sp. nov.
Lepidapedon elongatum (Lebour).
Steganoderma abyssorum (Odhner).
Derogenes various (Muller).
Xenodermichthys copei (Gill) - Steringophorus blackeri sp. nov.
Xenodermichthys sp. indet. - Steringophorus blackeri sp. nov.
FISH TREMATODES FROM BAY OF BISCAY 155
SYSTEMATIC SECTION
Family BUCEPHALIDAE Poche, 1907
Dolichoenterum longissimum Ozaki, 1924
(Fig. 2)
Host and locality : intestine of Conger conger, Stn. 7.
Twenty-three specimens were found, and these vary from 3-5 to 6-5 mm in length.
The longer worms are less contracted and have a long narrow neck, and an elongate
oval hindbody (Fig. 2a), but in the shorter contracted specimens there is little indica-
tion of a neck (Fig. 2b). Whilst, superficially, these two forms may appear distinct
an examination of the internal morphology of the worms leaves no doubt that they
are specifically identical. The cuticle has, imbedded in it, numerous narrow pointed
spines.
The anterior sucker is surmounted by a thick muscular hood, which in an extended
condition, bears seven to eight horn-like conical projections on the dorsal and lateral
rims (Figs. 2c and d). In an extended condition the sucker is 0-53-0-68 mm in
diameter. When contracted the sucker becomes almost globular, and the projections
are less conspicuous, and apparently less numerous in some cases, and point inwards.
The sucker in this condition measures 0-32-0-44 mm in diameter. The aperture of
the sucker is subterminal, with a more or less narrow extension down the ventral
surface.
The contraction or expansion of the worm also affects the anterior portion of the
alimentary canal. In neither whole mounts nor serial sections was evidence of an
oral sucker seen. The specimens lacking a neck bear a simple mouth situated closely
behind the anterior sucker. A short prepharynx runs posteriorly to the somewhat
transversely-oval pharynx measuring 0-3-0-32 mm x 0-21 mm. From the pharynx
the intestine runs anteriorly to a point which may or may not be nearer to the an-
terior sucker than the pharynx, where it turns conversely to reach the hinder end
of the body. In both extended and contracted specimens the intestine follows a
transversely undulating course and reaches nearly to the posterior end of the cirrus-
sac. The posterior limit of the intestine is not easy to make out in whole mounts as
it is often obscured by the eggs in the uterus.
A sinus occurs at the posterior margin of the worm, normally in the median line,
though contraction has, in some cases, displaced the opening to the left of the
median line. Into it open the ducts of the male and female systems and of the excre-
tory vesicle. The opening of the sinus does not appear to be surrounded by a
sphincter. The elongate cirrus-sac, which, due to contraction, lies at an acute angle
to the median line, measures 0-95 x 0-23-0-28 mm and possesses a fairly thick mus-
cular wall. It contains an oval seminal vesicle, 0-38-0-43 mm x 0-23-0-25 mm,
which occupies the anterior portion of the cirrus-sac. The pars prostatica passes
out of the middle region of the seminal vesicle, and runs to near the posterior ex-
tremity of the cirrus-sac where it unites with a short muscular cirrus. The testes are
transversely elongate, well separated one behind the other, and are situated in the
R. A. BRAY
FIG. 2. Dolichoenterum longissimum Ozaki : a) extended specimen (ventral view) ;
b) contracted specimen (ventral view) ; c) anterior sucker (dorso-lateral view) ; d)
longitudinal section of anterior sucker.
middle region of the hinder half of the body. They are 0-56-0-84 mm x 0-22-0-41
mm. The ovary is also transversely elongate measuring 0-53-0-82 mm x 0-23-
0-34 mm, and lies a little to the left of the median line between the testes. Directly
behind it lies a large diffuse Mehlis' gland. Laurer's canal opens dorsally at about
the level of the posterior testis. There appears to be no receptaculum seminis, but
sperm is accumulated in the initial slings of the uterus, which is long and coiled
running to a position some way anterior to the anterior testis before it turns and passes
posteriorly, with many tightly-packed transverse slings. It passes into the posterior
FISH TREMATODES FROM BAY OF BISCAY 157
region of the body, where it opens into the base of the genital sinus near the cirrus-
sac, via a short muscular metraterm. There are numerous small eggs, many of
which are collapsed, but measuring 21-27 /mi x 12-18 /mi when not collapsed.
The vitelline follicles lie in two lateral groups, with about 12-16 follicles in each
group, extending from about the level of the anterior testis to a position anterior to
the foremost extent of the uterus. A few follicles also occur scattered between the
testes.
The excretory vesicle opens into the genital sinus, and runs forward to a position
just posterior to the posterior testis.
Dolichoenterum longissimum was originally described from Conger myriaster in
Japanese waters. It has been found a number of times since in the same waters in
C. myriaster (Ozaki 1928 ; Yamaguti 1938 ; Machida et al. 1970) and in Muraenesox
cinereus (Yamaguti 1934). Prior to the present record, it has also been found in
Conger conger in New Zealand waters (Manter 1954) and in two localities off the
western coast of Scotland (Williams 1960).
Ozaki (1928) and Yamaguti (1934) disagree on the presence of a so-called ' urogeni-
tal ' pore, that is, the common opening of the genital sinus and the excretory vesicle.
Yamaguti states that Ozaki's figure 23 is misleading, and suggests that the 'genital
sinus' is no more than a slight depression. Ozaki's figure, however, seems to
accurately represent the condition in the present material. It would seem that the
depth of the genital sinus is influenced somewhat by the contraction of the body.
Another feature which varies with the condition of the worm is the number and
configuration of the projections surmounting the anterior sucker. Such a difference
is, according to Tendeiro (1955), a major factor in separating D. manteri Tendeiro,
1955 (with only five projections) from D. longissimum. Tendeiro found a single
specimen of D. manteri in C. conger from the Algarve coast of Portugal, and considered
it distinct for a number of other reasons. The only one of these that appears to be
a valid differentiating feature is the length of the intestine, which in D. manteri is
said to extend only a little behind the middle region of the body. As Tendeiro had
but one specimen it may have been that, as is the case in many of my specimens, it
was not possible to trace the full extent of the intestine. If this were found to be so,
then it seems probable that D, manteri should be considered a synonym of D.
longissimum.
Bucephalopsis gracilescens (Rud., 1819) Nicoll, 1914
Host and locality : intestine of Lophius piscatorius, Stn. 54.
This parasite occurred in great numbers. The shape of these worms varies
considerably as indicated by Dawes (1947).
Prosorhynchus crucibulum (Rud., 1819) Odhner, 1905
(Fig. 3)
Monostoma crucibulum Rud., 1819.
Distoma crucibulum : Dujardin, 1845.
Gasterostomum crucibulum : Gervais and van Beneden, 1859.
158
R. A. BRAY
FIG. 3. Prosorhynchus crucibulum (Rud.) : ventral view.
Host and locality : intestine of Conger conger, Stn. 7.
A single specimen was present, and it was examined as a whole mount and then
in serial sections. It is a small worm, truncated anteriorly and pointed posteriorly,
1-5 mm in length and 0-87 mm in maximum width. The cuticle bears numerous
scale-like spines. At the anterior end there is a muscular, funnel-shaped rhynchus,
0-48 mm across at its widest point and extending posteriorly to very near the intes-
tine, that is, about 0-55 mm from the anterior end of the worm. A slit-like transverse
mouth, 52 /im across, is situated at about the anterior third of the body length, and
near the base of the rhynchus. It leads immediately to a nearly globular pharynx,
measuring 0-108 x 0-115 mm, which in turn leads into a muscular oesophagus
FISH TREMATODES FROM BAY OF BISCAY 159
0-07 mm long, thence into the saccular intestine, which lies dorsally to the pharynx
and is about 0-2 mm x 0-16 mm.
The excretory pore is terminal posteriorly, and the excretory vesicle reaches to
about the level of the anterior end of the cirrus-sac.
The genital sinus opens near to the posterior extremity of the body on the ventral
side of the worm. A muscular cirrus-sac, 0-47 mm long by about 0-22 mm wide,
reaches anteriorly to a position just in front of the anterior margin of the posterior
testis. It contains a sigmoid seminal vesicle, which extends from the foremost
extremity of the cirrus-sac to a position near to the posterior extremity, where it
forms a muscular bulb before passing into an elongate pars prostatica. This latter
organ runs nearly to the anterior of the cirrus-sac before passing posteriorly to where
it unites with a relatively short muscular cirrus, which opens into the sinus at the base
of the genital tongue. There is no indication of an external seminal vesicle. The
testes lie one just posterior to the pharynx with the other just behind it. They are
oval with smooth margins and measure 0-27 x 0-19 mm (anterior testis) and 0-21 x
0-2 mm (posterior testis). The ovary, of similar size and shape to the testes (0-2 mm
dia.), lies just antero-dorsal to the anterior testis. The ' shell '-gland lies immediately
posterior to the ovary, and Laurer's canal opens dorsally just posterior to the ovary.
The vitellaria consist of about 25 irregular follicles lying more or less in an arc just
behind the rhynchus and anterior to the pharynx and uterus. This latter organ is
extensive in the region posterior to the pharynx, obscuring parts of the reproductive
system in the whole mount. A short, muscular metraterm enters the genital sinus
dorsally to the cirrus-sac. The eggs, which are very numerous and mostly collapsed,
measure about 24-26 p,m x 16-18 p,m.
Prosorhynchus crucibulum is a common parasite of the conger and is morphologi-
cally very similar to another parasite of this fish, P. aculeatus Odhner, 1905, with
which it has been synonymized by Dawes (1947). Brinkmann (1957) has given a
number of reasons why he considers these species to be distinct, and, having examined
a number of specimens assigned to both species in the collections of the British Museum
(Natural History), I agree that the features selected by Brinkmann appear to be of
value in distinguishing these species. The most noticeable difference is the size and
shape of the rhynchus, which in P. aculeatus is small and rounded, and in crucibulum
is larger and triangular or funnel-shaped. This character appears to be fairly
constant, and according to Matthews (1973) the typical shape of the rhynchus in
P. crucibulum is developed in the metacercaria during its first month in the second
intermediate host. As can be seen, the present specimen fits P. crucibulum in this,
as well as the other less prominent characteristics.
Family OPECOELIDAE Ozaki, 1925
H elicometra fasciata (Rud., 1819) Odhner, 1902
Host and locality : intestine of Gaidropsarus vulgaris, Stn. 22.
One specimen was present.
160 R. A. BRAY
Bathycreadium biscayense sp. nov.
(Fig- 4)
Hosts and localities : caeca and intestine of Trachyrhynchus trachyrincus (type-host),
Stn. 7, Stn. 24 ; caeca and intestine of Coelorinchus caelorhinchus, Stn. i ; intestine
of Lepidion eques, Stri. i.
This species occurred commonly in T. trachyrincus, in smaller numbers in C.
caelorhinchus and only one specimen was found in L. eques.
The worm is elongate oval with lateral margins roughly parallel for much of its
length, which varies between 0-4 mm and 1-7 mm. The maximum width, which
normally occurs at the level of the ventral sucker, is 0-19-0-42 mm. The tegument
is not armed, but bears annular wrinkles, probably due to contraction. Circular and
subterminal, the oral sucker measures 0-065-0-14 mm in diameter and leads, without
an apparent prepharynx, into a globular pharynx 0-032-0-066 mm in diameter.
The oesophagus, up to 0-13 mm in length and lined with a cuticle, bifurcates at about
the anterior margin of the ventral sucker. The intestinal caeca reach to near the
posterior end of the worm, where they unite to form a cyclocoel. The ventral sucker
is situated close behind the oral sucker in the anterior quarter or third of the body.
It is often situated on a slight protuberance, and is transversely elongate, its trans-
verse diameter being 0-19-0-29 mm, thus giving an oral : ventral sucker ratio of
i : 2-2-3-5.
The excretory pore is terminal, and the cylindrical vesicle reaches to the ovary.
The genital pore occurs just to the left of the median line, closely anterior to the
ventral sucker. Into it opens a claviform cirrus-sac measuring 0-25-0-4 mm x 0-05-
0-085 mm, which normally reaches just posteriorly to the ventral sucker, and its
posterior extremity lies amongst the uterine coils. It contains a convoluted internal
seminal vesicle, a relatively short, oval pars prostatica and a long muscular cirrus.
The testes are situated in tandem, close together, in the middle third of the body.
They are usually wider than long with entire or slightly indented margins and measure
in mature specimens 0-08-0-12 mm x 0-12-0-15 mm (anterior testis) and 0-09-
0-12 mm x 0-12-0-15 mm (posterior testis).
The ovary is situated anteriorly and adjacent to the foremost testis, usually in
the median line, but sometimes moved to one side or the other by the development
of the uterus. It is normally wider than long, oval, with smooth margins and
measures 0-025-0-065 x 0-083-0-1 mm. There appears to be a seminal receptacle
lying dorsally to the ovary, but, despite several attempts, it has not been possible to
obtain satisfactory serial sections of the female complex, so whether this is a true
seminal receptacle or part of the uterus has not been determined, nor has the course
of Laurer's canal, if present. The viteilaria consist of numerous oval follicles in
lateral fields reaching from about the posterior border of the ventral sucker to near
the posterior end of the worm. Behind the posterior testis the fields meet in the
ventral plane, and come close together in the dorsal, being separated only by
the excretory vesicle. Between the testes, and between the anterior testis and the
ovary the follicles may extend towards the median line, but do not reach it. The
vitelline fields are sometimes interrupted opposite the ovary and the testes. The
FISH TREMATODES FROM BAY OF BISCAY
161
FIG. 4. Bathycreadium biscayense sp. nov. : specimen from Trachyrhynchus
trachyrincus (ventral view).
uterus runs from the ovary to the posterior margin of the ventral sucker, and con-
tains a few (up to about 10) large eggs. These measure 80-92 /mi x 43-52 /mi.
MATERIAL. British Museum (Nat. Hist.) Reg. no. 1973.4.9.1-20 (syntypes).
Bathycreadium Kabata, 1961, has been reduced to synonymy with Nicolla Wisniew-
ski, 1933, by Slusarski (1971). Kabata did not compare the two genera so it was
left to Pritchard (1966) to list the differences between them. Slusarski dismisses
Pritchard's diagnostic characters by saying that they are too variable in Nicolla,
and he is probably correct in this assumption in respect of most of these characters.
It does seem, however, that at least one character mentioned by Pritchard may be
used in separating these two genera, along with another feature she did not mention.
162 R. A. BRAY
As she noted, it appears that in all the species of Nicolla (with one exception, i.e.
N. elongata) the vitellaria reach anteriorly to the ventral sucker, whilst in Bathy-
creadium they do not reach further forward than the posterior margin of the ventral
sucker. It is also apparent that the normal situation of the posterior testis in
Nicolla is close to, or contiguous with, the cyclocoel with few or no vitelline follicles
between the posterior testis and the intestinal arch (again N. elongata is an exception).
In Bathycreadium there is at least a testis diameter, usually considerably more,
between the posterior testis and the cyclocoel, and the bulk of the vitellaria is in
this region. From its differentiating characteristics it is evident that Nicolla
elongata Maillard, 1970, from Onos tricirratus in the Gulf of Lion should be assigned
to Bathycreadium elongatum (Maillard, 1970) n. comb.
It may also be mentioned here that the concept of Bathycreadium as given above
is close to that of Coitocaecum Nicoll, 1915, as represented by Yamaguti (1971). He
seems to be in error, however, in stating that the cirrus-sac is ' retort-shaped contain-
ing elongate saccular or tubular seminal vesicle, prostatic complex and ejaculatory
duct' in Coitocaecum. The type-species of this latter genus, C. gymnophallum
Nicoll, 1915, has, according to the original author, 'no true cirrus-pouch', while
Crowcroft (1951) who re-examined Nicoll's material states that there is 'a small
membranous cirrus-sac enclosing a short, terminal portion of the male duct'.
B. biscayense differs from the other two species in the genus, B.flexicollis Kabata,
1961, and B. elongatum (Maillard, 1970), in the anterior extent of the vitellaria, in
the proximity of the testes and the ovary and in the sucker-ratio. It also appears
that it normally contains fewer eggs than the other species. The egg-size of B.
biscayense differs greatly from that of B. elongatum, but is similar to that of B.
flexicollis as corrected by Pritchard (1966) and verified by the author from paratype
specimens in the collections of the British Museum (Natural History).
Family LEPOCREADIIDAE Nicoll, 1935
Lepidapedon rachion (Cobbold, 1858) Stafford, 1904
(Fig- 5)
Distomum rachion Cobbold, 1858.
Lepodora rachiaea : Odhner, 1905.
Host and locality : intestine of Phycis blennoides, Stn. 17.
Nine contracted specimens were present. They are small worms o-66-i mm long
and 0-27-0-44 mm wide, being widest behind the ventral sucker, and tapering to-
wards the anterior end. The cuticle bears rows of closely-set narrow pointed spines.
The oral sucker is subterminal, measuring 0-097-0-115 mm in diameter, and leads into
a prepharynx which, when apparent, varies in length up to 44 /xm and is usually,
but not invariably, longer than the oesophagus, which may reach a length of 28 jtrni.
The phary nx is large and pyriform 0-09-0- 13 mm x 0-066-0-106 mm. The intestinal
bifurcation occurs close to the ventral sucker, and the caeca, which reach to the
posterior end of the body, are wide. The rounded ventral sucker lies in the anterior
FISH TREMATODES FROM BAY OF BISCAY
163
FIG. 5. Lepidapedon rachion (Cobbold) : ventral view.
half of the worm, and measures 0-08-0-12 mm in diameter, giving an oral : ventral
sucker ratio of i : 0-75-1-08. Only in the largest specimen was the ventral sucker
slightly larger than the oral, whilst in the other specimens the oral sucker was
distinctly larger than the ventral.
The excretory pore is terminal and leads into a simple elongate excretory vesicle,
which has been traced as far forward as the posterior testis.
The genital pore occurs closely anterior to the ventral sucker and to the left of the
median line ; in fact, it lies ventrally to the left caecum. The muscular cirrus-sac
reaches to near the posterior margin of the ventral sucker, measures about 0-09 mm
in length and contains a coiled, tubular, seminal vesicle and a pars prostatica, which
is slightly longer than the short muscular cirrus. The external seminal vesicle is
164 R. A. BRAY
also tubular and is coiled amidst a mass of gland-cells. This mass of cells extends
behind the ventral sucker and is surrounded by a thin membrane. In the posterior
third of the body lie the testes, in tandem formation. They are transversely elongate,
have smooth or indented margins and measure 0-1-0-14 mm x 0-03-0-05 mm
(anterior) and 0-08-0-14 mm x 0-04-0-06 mm (posterior).
The ovary lies just anteriorly to the foremost testis in the median line and is
separated from the ventral sucker by the uterine coils. It is transversely elongate,
measuring 0-08-0-13 mm x 0-04-0-07 mm, with smooth or slightly indented mar-
gins. Dorsal to the ovary lies a globular receptaculum seminis. The vitelline glands
are numerous and reach anteriorly as far as about the middle level of the ventral
sucker. They lie laterally and ventrally to the intestinal caeca and reach inwardly
to the gonads, and although they intrude between the gonads they do not meet in
the median line. Posteriorly to the testes, however, the vitelline fields are confluent
in the median line. The uterine coils lie between the ovary and the ventral sucker
and contain eggs, all of which, in the present specimens, are collapsed. In this
condition they measure about 57-60 /mi in length.
These worms are recognizable as Lepidapedon rachion despite the fact that, pre-
sumably due to contraction, the worms are on average shorter than is normal for
this species. This parasite is fairly common in Gadidae in the northern Atlantic,
but has not, apparently, been recorded from this host before.
Lepidapedon elongatum (Lebour, 1908) Nicoll, 1910
(Fig. 6)
Lepodora elongatum Lebour, 1908.
Hosts and localities : intestine of Lepidion eques, Stn. i ; intestine of Trachyrhynchus
trachyrincus, Stn. i, Stri. 7, Stn. 24.
These worms were common in the intestine of both hosts. Their normal shape
appears to be elongate oval, but some of the present specimens are contracted to a
more pyriform shape, whilst some, taken from a frozen specimen of L. eques, are
very elongate in the forebody. The length of the worm varies between 0-73 mm and
2-1 mm, and the width between 0-17 mm and 0-42 mm. The cuticle bears rows of
narrow pointed spines. Rounded and subterminal, the oral sucker measures
0-09-0-17 mm in diameter. The ventral sucker, which is situated in the middle
third of the body, is also rounded and measures 0-06-0-13 mm m diameter. The
oral : ventral sucker ratio is i : 0-55-1. The prepharynx in grossly contracted
specimens is not apparent, but in less contracted specimens it varies greatly in length,
measuring up to 0-6 mm in one specimen from a frozen L. eques. Normally, the
variation appears to be between about o-oi mm and o-i mm. The pharynx is longer
than wide, measuring 0-06-0-14 mm x 0-04-0-12 mm and has a tendency to be of a
similar size to or somewhat larger than the ventral sucker. It leads into the oeso-
phagus which, like the prepharynx, varies considerably in length, and is usually
shorter than the prepharynx, being 0-0-06 mm in length. It bifurcates about midway
FISH TREMATODES FROM BAY OF BISCAY
165
FIG. 6. Lepidapedon elongatum (Labour) : three specimens from Trachyrhynchus
trachyrincus (ventral view).
between the suckers or nearer to the ventral sucker. The caeca, lined with a deeply
staining epithelium, reach to near the posterior end of the body.
The excretory pore is terminal and leads into a simple tubular vesicle which has
been traced anteriorly to the level of the hinder margin of the hinder testis.
The genital pore is situated to the left of the median line close to the anterior
margin of the ventral sucker or more further forward. The strongly muscular, oval
cirrus-sac measures about 0-07-0-14 mm x 0-04-0-09 mm, but is not easy to measure
satisfactorily due to the angle at which it lies to the body surface. It contains a
coiled tubular seminal vesicle, a distinctly widened pars prostatica and a muscular
cirrus. The internal seminal vesicle leads, via a narrow canal, into a convoluted
166 R. A. BRAY
external seminal vesicle surrounded by gland-cells. These cells are in turn surrounded
by a very thin membrane, and reach to a position well posterior to the ventral sucker.
The testes have smooth margins and are either transversely elongate or rounded.
They are arranged in tandem at about the middle of the hindbody. The anterior
testis measures 0-08-0-22 mm x 0-04-0-13 mm and the posterior 0-09-0-2 mm x
0-04-0-16 mm. There is a narrow space between the testes, and also between the
anterior testis and the ovary, which lies in the median line, and is separated from the
ventral sucker by the uterine coils. Like the testes it may be either transversely
elongate or rounded, and measures 0-06-0-17 mm x 0-04-0-11 mm. The seminal
receptacle lies dorsally to the ovary and Laurer's canal runs to the dorsal surface at
about the same level. The vitelline glands do not reach anteriorly as far as the
ventral sucker, but terminate on a level with the hinder end of the external seminal
vesicle, or thereabouts. The follicles lie mostly in a plane ventral to the intestinal
caeca, with some lateral to them. The lateral vitelline fields may be interrupted
beside the ovary and testes and may reach to the median line between these organs,
and also ventrally to the uterus. Behind the testes the fields usually merge in the
median line, but sometimes the excretory vesicle appears to form a barrier between
them. The uterus runs between the ovary and the ventral sucker, and opens into
the genital atrium through a muscular metraterm of about the same length as the
cirrus-sac. The uterus contains numerous eggs, the few uncollapsed ones measuring
58-65 />im long by about 30 /u,m wide.
Several authors have used the name Lepidapedon elongatum for specimens de-
scribed from various species of fishes from the west and east coasts of North America
(Manter 1926 ; Linton 1940 ; Ching 1961), the Pacific coast of Panama (Caballero y
C. el al. 1955), Bombay (Gupta and Sehgal 1971), the Black Sea (Skrjabin and Koval
1960) and the Barents Sea (Dogiel 1936), in addition to those originally described
from Gadus morhua on the Northumberland coast of England by Lebour (1908).
The variation represented in these descriptions indicates that more than one species
may be involved. The position is further complicated by the descriptions of very
similar trematodes under other names, for example, Lepidapedon coelorhynchi
Yamaguti, 1938, L. gadi (Yamaguti, 1934) and L. microcotyleum Dollfus, 1953. It
seems that the latter two species may be distinguished from Lebour's form by egg-
size, although their egg-size overlaps some of the later descriptions of ' L. elongatum '.
L. coelorhynchi was considered a synonym of L. elongatum by Hanson (1950) and this
has been accepted by most subsequent authors.
The specimens described above differ from Lebour's description in that most of
them have a ventral sucker distinctly smaller than the oral. She gives measure-
ments of only one specimen, the ventral sucker being 0-12 mm across and the oral,
o-i mm across. It should be noted, however, that in her figure the ventral sucker
appears to be slightly smaller than the oral. The distribution of the vitellaria
between the ovary and the ventral sucker is rather variable and probably not a good
feature to use for differentiating species in this group, as was suggested by Polyanski
(1955). Figure 6 illustrates variation in three specimens from the intestine of the
same specimen of T. trachyrincus, but at the same time they also show a certain
constancy of morphological features.
FISH TREMATODES FROM BAY OF BISCAY 167
Family AZYGIIDAE Odhner, 1911
Otodistomum veliporum (Creplin, 1837) Stafford, 1904
Host and locality : intestine of Dalatias licha, Stn. i.
This trematode has been reported from this host by a number of authors, including
Dollfus (1937) who records it from the Bay of Biscay, off Biarritz, and also gives a
list of previous records. Since then it has been found in this host in Norwegian
waters by Braten (1964) and in the Cook Strait, New Zealand, by Manter (1954).
Family ZOOGONIDAE Odhner, 1911
Steganoderma abyssorum (Odhner, 1911) Manter, 1947
(Fig- 7)
Proctophantastes abyssorum Odhner, 1911.
Deretrema abyssorum : Price, 1934.
Hosts and localities : intestine of Trachyrhynchus trachyrincus , Stn. i, Stn. 7, Stn.
24 ; intestine of Coelorinchus, caelorhinchus , Stn. i, Stn. 37 ; intestine of Lepidion
eques, Stn. i.
The degree of contraction of the fixed worms varies considerably so that the shape
varies from a very elongate pyriform to a short, broadly pyriform. The length is
between 0-62 mm and 1-66 mm and the greatest width is 0-28-0-63 mm, which occurs
invariably at the region of the ventral sucker. The rounded oral sucker is subter-
minal, 0-18-0-32 mm in diameter, whilst the ventral sucker is considerably larger,
situated somewhat posteriorly to the middle of the worm, and is wider than long,
being 0-25-0-6 mm across by 0-18-0-44 mm- Transversely across the inside of the
ventral sucker runs a low muscular ridge. The oral : ventral sucker width ratio is
i : 1-08-2-1 and the sucker length ratio is i : 0-9-1-6. Only in extended specimens
is a prepharynx to be seen, and it reaches to no greater length than 0-04 mm. The
small globular pharynx has a diameter of 0-03-0-08 mm and from it leads an oeso-
phagus which varies greatly in length between 0-08-0-22 mm. In grossly contracted
specimens it has not been observed. It bifurcates about halfway between the
suckers and the caeca are fairly short, terminating dorsally to the testes or vitellaria.
The excretory pore is terminal and leads into a vesicle which reaches forward,
dorsally to the uterus, for an undetermined distance and as far as can be seen is a
simple tube.
Approximately on a level with the intestinal bifurcation, in a submarginal position,
lies the genital pore. The cirrus-sac reaches past the anterior margin of the ventral
sucker, and in contracted specimens nearly reaches to the transverse ridge, measuring
0-2-0-37 mm x 0-06-0-09 mm- It contains a coiled seminal vesicle, the coils of
which are often tightly packed, a well-developed pars prostatica invested with a mass
of gland cells and a cirrus, of about one-third the length of the cirrus-sac. The
proximal half of the cirrus appears to be lined with a cuticle which may be thrown
in wrinkles. The two oval or elongate oval testes are always partly obscured by the
1 68
R. A. BRAY
FIG. 7. Steganoderma abyssorum (Odhner) : two specimens from Trachyrhynchus
trachyrincus (ventral view).
eggs or the vitelline glands and lie symmetrically one on each side of the body,
separated from the posterior border of the ventral sucker by some of the vitellaria.
They measure about 0-16-0-22 mm x 0-13-0-2 mm.
The rounded ovary is smaller than the testes, being about 0-15 mm in diameter.
It is situated more or less on the posterior border of the ventral sucker, but sometimes
entirely posterior or anterior to it. Posteriorly and dorsally to the ovary lies
the rounded or oval seminal receptacle, which is rather smaller than the ovary.
Immediately posterior to the ventral sucker, in two symmetrical lateral groups, are
the vitellaria, which lie somewhat ventrally to the testes, and overlap them in vary-
ing degrees. There are between five and ten follicles in each group. The uterus fills
FISH TREMATODES FROM BAY OF BISCAY 169
most of the area posterior to the ventral sucker, and therefore in the majority of
specimens the details of the reproductive organs are obscured. It contains numerous
operculate eggs measuring 34-41 />tm long by 16-21 /mi. The metraterm is long,
muscular and helical. It enters a small, muscular genital atrium, which is surrounded
by a scattering of glandular cells.
This species was originally described as the type-species of the genus Procto-
phantastes Odhner, 1911, from Macrurus (Coryphaenoides) rupestris and Gadus
aeglefinus from Norway. Price (1934) considered Proctophantastes to be a synonym
of Deretrema Linton, 1910, but this synonymy was questioned by Manter (1947)
who considered it a synonym of Steganoderma Stafford, 1904. Odhner (191 ia) does
not discuss the relationship of Proctophantastes with either of the genera with which
it has been synonymized. Whilst Price (1934) makes his synonymy with little
comment, Manter (1947) discusses the relationship of the three genera in some detail.
As has been mentioned above, the latter author considers Proctophantastes to be a
synonym of Steganoderma and he suggests that the differences between the type-
species of Steganoderma, S. formosum Stafford, 1904, and P. abyssorum are of specific
rather than generic value. These differences are in sucker-size, oesophagus-length
and length of vitelline fields. The present author agrees with Manter's conclusion.
Therefore, the differences Manter (1947) indicates between Steganoderma and Dere-
trema also apply to Proctophantastes when compared with Deretrema. In short,
Price's synonymy is rejected, because the vitellaria in Deretrema do not reach pos-
teriorly to the ventral sucker, whilst in Steganoderma and Proctophantastes the majority
of the vitellaria are situated in this region. Subsequent to Manter (1947), only
Yamaguti (1953, 1958, 1970, 1971) has continued to consider Proctophantastes a
valid genus. The features he uses to distinguish Proctophantastes from Steganoderma
are, I believe, not sufficient to uphold the genus. In 1953 he indicates that in
Proctophantastes the vitellaria are compact, while in Steganoderma they are in longi-
tudinally elongate groups. This type of difference would appear to be of specific
rather than generic importance, as it concerns fairly small details of the vitelline
extent. Yamaguti (1958) distinguishes the genera on the shape of the cirrus-sac
and the position of the ventral sucker. An examination of Miller's (1941) redescrip-
tion of Stafford's specimen of S. formosum, and also of the other descriptions of this
species given by Manter (1925, 1926), Linton (1940), Ching (1960), Ronald (1960)
and Strelkov (1960) shows that the cirrus-sac of this species resembles that of P.
abyssorum in shape. The ventral sucker in this latter species lies just behind the
middle of the body, while in S. formosum it lies just anterior to the centre of the body.
This condition probably varies with the contraction or growth of the specimen, so
that its value as a generic character is rather doubtful. In 1970, Yamaguti suggested
that a 'profusely developed "periatrial gland" may be a generic, not a specific
character, though not mentioned by Odhner '. He is referring to a character present
in a new species he is describing, and this character is, I believe, a feature of Neo-
steganoderma, and is discussed in connection with the following species. The few
scattered gland-cells around the genital atrium in the present specimen do not, it
appears, represent a feature of generic importance, and, moreover, similar glands
have been described in other species of Steganoderma, e.g. S. elongatum Manter, 1947.
iyo R. A. BRAY
Finally, Yamaguti (1971) upholds Proctophantastes on account of the characteristic
structure of the genital atrium. There is, however, no peculiar feature in the genital
atrium of the present specimens, nor is it indicated in Odhner's generic definition
where he merely says ' Genitalsinus ziemlich klein'. As the various attempts to
validate Proctophantastes do not appear convincing, this genus is here considered to
be a synonym of Steganoderma.
The present specimens show no outstanding morphological differences from P.
abyssorum, as described by Odhner (igiia), and are therefore considered to belong
to this species. There are, however, two minor differences, the first being the sparse
gland-cells around the genital atrium and the second the presence of a ridge running
transversely across the inner surface of the ventral sucker. This latter feature was
not easy to detect in many whole mounts and may easily have been overlooked by
Odhner.
Neosteganoderma glandulosum Byrd, 1964, emend
(Fig. 8)
Proctophantastes glandulosa : Yamaguti, 1971.
Host and locality : intestine of Beryx decadactylus , Stn. 37.
All the worms are contracted to a broadly pyriform shape, which tapers towards
the anterior end. The length varies between i mm and 1-38 mm and the maximum
width, which occurs at the level of the testes, between 0-76 mm and 0-96 mm.
Anteriorly, the cuticle bears long, narrow spines which also occur near the posterior
extremity. The subterminal oral sucker is rounded, 0-27-0-38 mm across by
0-26-0-33 mm long. At about the middle of the worm, or just behind, lies the
transversely elongate ventral sucker, measuring 0-69-0-77 mm across by 0-27-0-33
mm long, reaching close to the lateral margins of the body on either side. The
oral : ventral sucker width ratio is i : 2-1-2-8, whilst the length of both suckers is
almost the same. A small prepharynx can be seen in serial section. It enters a
globular pharynx of about 0-11-0-13 mm diameter. The oesophagus is fairly short,
curved dorsally to the pharynx, and bifurcates just posteriorly to the pharynx. The
caeca are quite short, reaching to a position dorsal to the testes.
The excretory pore is terminally situated, or, when the worm is greatly distended
with eggs, it may appear to be slightly dorsal. The saccular vesicle reaches as far
as the posterior border of the ovary, and in this region part of it lies ventrally to the
seminal receptacle.
On the left side of the body, submarginally and just anterior to the ventral sucker
lies the genital pore. The genital atrium has a strong muscular wall, with irregular
outpocketings. Into it leads a large cirrus-sac, the bulk of which lies just anterior
to the ventral sucker and at right angles to the long axis of the worm. The cirrus-sac
reaches transversely across the body to the right intestinal caecum, where it twists
and runs slightly posteriorly towards the centre of the ventral sucker, in a plane dorsal
to the remainder of the cirrus-sac. The relatively small, coiled internal seminal
FISH TREMATODES FROM BAY OF BISCAY
171
I72 R. A. BRAY
vesicle leads into a well-developed pars prostatica, which, with its innumerable as-
sociated glands, fills most of the cirrus-sac. The muscular cirrus is shorter, contain-
ing some long, narrow spines, lying lengthwise, but as they have been seen clearly
only in serial section the overall shape of these spines has not been made out. The
testes are arranged symmetrically, just posterior to the ventral sucker and close to
the lateral margins of the worm. Apparently, normally they are of an oval shape,
but with growth of the uterus certain lobe-like processes appear. In one specimen
the testes measured 0-27 x 0-3 mm and 0-4 x 0-19 mm.
Between the testes, and lying towards the right of the worm, is the ovary. In
one specimen it was seen to be of an irregular quadrilateral shape, 0-27 x 0-27 mm.
The large, oval receptaculum seminis lies dorsally to the ovary and the left testis.
It may measure 0-3 x 0-19 mm. Ventral to the anterior extremity of the recepta-
culum seminis is a well-developed 'shell '-gland. The vitelline follicles are situated
in two lateral fields, ventrally to the testes and reaching to about the posterior margin
of the ventral sucker. Each field contains four to eight follicles. The egg-filled
uterus occupies the great majority of the hindbody, obscuring the reproductive
organs in this region. The metraterm is long and, particularly near the genital
atrium, strongly muscular. It enters the latter from the dorsal side. For most of
its length it is surrounded by an investment of gland-cells (Fig. 8b). Where the
metraterm and the cirrus-sac run adjacent to each other the cells invest both organs
and also surround the genital atrium. The operculate eggs measure 26-31 jum x
16-21 /u,m.
Byrd (1964) described Neosteganoderma glandulosum from a beryciform fish,
Polymixia lowei, in the Straits of Florida. He differentiated his new genus from other
zoogonids with follicular vitellaria situated posteriorly to the ventral sucker by a
number of features. These included the posterior position of the ventral sucker and
the elongate oesophagus and forebody, but as can be seen from the present contracted
specimens these are not reliable features. It should be noted that Byrd's specimens
were treated with chloretone in sea-water before fixation, which probably accounted
for the relaxation of the forebody. The other distinguishing characteristics he
mentions are the very large ventral sucker and the ' large glandular mass surrounding
the genital pore'. The latter feature is very prominent in the present specimens
and appears to be the most reliable feature for distinguishing Neosteganoderma from
the closely-related genus Steganoderma. Yamaguti (1971), however, places Byrd's
species in Proctophantastes , which, as pointed out above, does not possess a large
mass of gland-cells surrounding the genital atrium. It appears that Yamaguti is
basing his definition of Proctophantastes Odhner, 1911, not on its type-species, P.
abyssorum Odhner, 1911, but on his own species P. polymixiae Yamaguti, 1970, from
Polymixia japonica in the Pacific Ocean off Hawaii, which he describes as having a
large 'periatrial gland'. This suggests that this species should become Neostegano-
derma polymixiae (Yamaguti, 1970) n. comb. It seems possible that it may be
found to be conspecific with N. glandulosum.
As stated above, certain features of the morphology of the present specimens are
different from those described by Byrd, but these are considered to be due to the
differences in the fixation technique used. The oesophagus in the present forms is
FISH TREMATODES FROM BAY OF BISCAY 173
particularly reduced by comparison with the original description. Byrd's statement
that the metraterm enters the genital atrium from the ventral side appears to be a
typographical error, as a close examination of his figures shows that in his specimens,
as in mine, the metraterm enters the genital atrium from the dorsal side.
Family FELLODISTOMIDAE Nicoll, 1913
(?) Steringotrema diver gens (Rud, 1809) Odhner, 1911
(Fig- 9)
Distoma diver gens Rud., 1809.
Fasciola diver gens : de Blainville, 1820.
Host and locality : intestine of Capros aper, Stn. 38, Stn. 53.
Seven specimens were recovered from a fish at Stn. 38 and two from a fish at Stn.
53. These specimens are, unfortunately, rather contracted, but enough detail has
been observed to suggest that they probably belong to Steringotrema diver gens.
The worms are oval, often tapering at each end, and measure 0-75-1-2 mm long
by 0-42-0-47 mm in maximum width at the level of the ventral sucker. The cuticle
is unarmed. The subterminal oral sucker is 0-11-0-15 mm m diameter and leads
into a globular pharynx of 0-09-0-11 mm diameter. Because of the contraction
the oesophagus is curved, but it bifurcates just about the level of the genital pore,
and the intestinal caeca extend to about the level of the testes. About halfway
along the body lies the transversely-elongate ventral sucker, measuring 0-21-
0-31 mm x 0-38-0-42 mm, giving an oral : ventral sucker ratio of about i : 2-7-3-6.
The excretory pore is terminal, and the vesicle is V-shaped, the limbs of which
have been traced almost to the ventral sucker in serial sections.
About halfway between the suckers and just to the left of the median line lies the
genital pore. The rather stout cirrus-sac, 0-18-0-19 mm long and 0-11-0-13 mm
wide, contains a bipartite seminal vesicle leading into an elongate pars prostatica
which passes from the anterior end of the seminal vesicle to near the posterior end
of the cirrus-sac before turning and running anteriorly to unite with a short, muscular,
smooth cirrus. Due to contraction of the body, the testes, the ovary and the egg-
filled uterus are pressed closely against the ventral sucker. Thus, as the eggs are
opaque, the testes and the ovary are discernible only in serial section. The testes
lie symmetrically at the lateral margins of the body, with the ovary lying close to the
right testis. The vitelline follicles are arranged in four groups, two on each side of
the worm, one group anterior to the ventral sucker and the other group posterior to
the sucker. The uterus fills almost the whole of the hindbody of the worm and
contains numerous operculate eggs, 46-48 /xm x 33-37 ju.m with shell about 2-3 /xm
thick.
Any slight morphological differences which can be detected between the above
description and those given by Odhner (191 ib), Palombi (1931) and Mathias (1934)
may probably be explained by the contraction of the present worms. The eggs, as
measured when teased from the body as well as in utero, are rather wider than those
174
R. A. BRAY
FIG. 9. (?) Steringotrema divergens (Rud.) : diagrammatic ventral view with intestine
omitted.
previously recorded. A more definite statement on the identity of these specimens
awaits better-preserved material. It should be noticed that the previously-reported
hosts of S. divergens (i.e. gobies and blennies) have a rather different biology from that
of the present host.
Steringophorus blackeri sp. nov.
(Fig. 10)
Hosts and localities : intestine and body-cavity of Xenodermichthys copei (type-host),
Stn. 24 ; intestine of Xenodermichthys sp. indet., National Institute of Oceano-
graphy, Stn. 6429 - 28°07' N., I3°52' W. depth 0-390 m.
Four whole specimens were recovered from X. copei. One whole specimen and
an anterior fragment were found in the body-cavity of the same fish specimen, the
fish itself having itself been preserved in formalin. The remainder of the specimens
from X. copei and the two specimens found in X. sp. indet. were taken from the
intestine. The worms occurred either singly or in pairs in their hosts. The whole
worm from the body-cavity (Fig. lob) was of a shape and size rather different from
FISH TREMATODES FROM BAY OF BISCAY
175
0-1mm
I76
R. A. BRAY
those in the intestine (Fig. loa), two of which were sectioned for further study. The
other specimens, collected by the author from fishes in the collection of the National
Institute of Oceanography, included one which was apparently young (Fig. ice) and
an older worm (Fig. loe). Despite the variation in size, the internal morphology
indicates clearly that all these specimens represent the same species.
The 'intestinal form' from X. copei is elongate, being widest just posteriorly to
the ventral sucker, which is about one-fifth of the body-length from the anterior end.
The forebody tapers sharply, while the tapering of the hindbody is less pronounced,
and it is rounded terminally. The forms from X. sp. indet are much smaller, but
with a similarly shaped forebody, and are widest just posteriorly to the ventral
sucker (at about one-third of the body-length from the anterior end) , the hindbody
having slightly tapering lateral margins and a rather truncate posterior end. The
' body-cavity form ' from X. copei has a similar forebody to the other forms, but the
hindbody is considerably swollen by the large bulbous excretory vesicle, the limbs
of which reach into the forebody, but do not appear to be swollen in the same way.
The dimensions of the various forms are included in Table 2.
TABLE 2
Measurements of Steringophorus blackeri sp. nov.
Host
X. copei
X. copei
X. sp. indet.
X. sp.
indet.
'Form'
Intestinal
Body-cavity
Younger
Older
No. of specimens
3 (2 sectioned)
i + fragment
i
i
Length (mm)
4-6-9
2-7
2-2
3-08
Width (mm)
2-03-2-04
1-45
0-8
1-68
Oral sucker
0-47
0-34-0-35
0-243
0-4
diameter (mm)
'
Ventral sucker
°'43
0-29-0-33
0-240
o-45
diameter (mm)
Oral : ventral
i : 0-92
i : 0-85-0-94
i : 0-98
i : 1-12
sucker ratio
Pharynx (mm)
0-32 x 0-26
0-22-0-26 x
0-125 x o-ii
0-21 x
0-25
Oesophagus (mm)
Cirrus-sac (mm)
Ovary (mm)
Testes (mm)
Eggs (/*m)
0-21-0-19
0-13-0-23 0-14 0-105 0-16
about 0-32-0-35 about 0-21-0-32 0-26 x 0-14 0-48 x 0-16
x 0-19 x 0-12
0-27 dia. 0-21 x 0-16 0-18 x o-u 0-21 x 0-19
about 0-21 x o-n-o-12 x 0-07-0-13 x 0-13-0-14 x
0-18 0-16 0-14 0-17-0-2
57-63 x 32-34 52-58 x 27-30 54 x 27-30 54-57 x 28-30
The cuticle bears no spines, but has numerous transverse wrinkles. The oral and
ventral suckers are rounded, and with one exception the oral is slightly larger than
the ventral. In some specimens from X. copei contraction has caused the opening
of the ventral sucker to be disposed anteriorly. There is no prepharynx, so the
almost globular pharynx abuts directly with the oral sucker. It leads into a fairly
short oesophagus, which bifurcates at about the level of the anterior margin of the
ventral sucker, or slightly further forward. The intestinal caeca are lined with a
FISH TREMATODES FROM BAY OF BISCAY 177
glandular epithelium and run posteriorly to a point normally amidst the uterine
coils where they terminate, about halfway along the hindbody.
The excretory pore is terminal and leads into a very small globular vesicle (of
about 0'03-o-o6 mm in diameter) before entering the main Y-shaped vesicle. The
stem of the vesicle reaches to about the middle of the body, whilst the limbs reach
anteriorly to the oral sucker. The stem may be very wide or narrow, but the limbs
are constantly narrow. The excretory pore itself appears to be on a short
protuberance at the base of a depression and is surrounded by glandular cells.
The genital pore is situated on a level with, or just anterior to, the intestinal bifur-
cation and slightly to the left of the median line. It leads into a muscular genital
atrium. The cirrus-sac overlaps the anterior border of the ventral sucker and con-
tains a relatively small, bipartite seminal vesicle, a long, coiled pars prostatica and
a short muscular cirrus. The testes lie posteriorly to the ventral sucker a little
obliquely to each other, with the left testis being slightly anterior to the right. They
are close to the ovary and to each other, are small by comparison with the body-size
and have more or less deeply indented margins. The multilobate ovary lies anteriorly
and sometimes slightly obliquely to the right testis. It is associated with a large
' shell '-gland, about the same size as the ovary, lying between it and the left testis.
Laurer's canal opens dorsally, level with the ' shell '-gland, and may contain a small
amount of spermatozoa in its proximal region. Sperm may also lie in the proximal
part of the uterus. The vitellaria consists of numerous globular or short trans-
versely-elongate follicles disposed in two symmetrically placed fields near the lateral
margins of the body. These fields reach from about the level of the anterior border
of the ventral sucker to a position some distance posterior to the testes. The uterus
runs in the median line from the ovary to about the middle of the worm, in a series of
lateral coils. It then forms a descending series of loops on each side of the body to
about two-thirds of the way along the hindbody before passing anteriorly again as a
series of loops. It has not been possible to ascertain whether the uterus passes down
one side of the body and then down the other, or whether it divides before passing
down both sides. The later coils of the uterus pass forward in the median line
ventrally to the descending coils and the gonads. In the specimens with the swollen
excretory vesicle the lateral uterine slings are forced to lie close to the margins of
the body. Within the uterus lie numerous eggs, many of which are collapsed, but
which appear to be operculate, with a brownish-yellow shell. This shell is covered
with numerous small bosses giving the surface a granular appearance, which is
rather less marked in the young specimen from X. sp. indet.
MATERIAL. British Museum (Natural History) Reg. no. 1973.4.9.21-24 (syntypes).
Although some recent authors have followed Yamaguti (1953) in considering
Steringophorus Odhner, 1905, a synonym of Fellodistomum Stafford, 1904, it would
seem that there may be one feature differentiating these genera, if the distribution of
the vitellaria is taken into consideration. In Fellodistomum the vitellaria extend
anteriorly well beyond the ventral sucker, whilst in Steringophorus they do not. A
close examination of material of the type-species of both genera from the type-hosts
may well indicate other differences, particularly in the length of the caeca and the
configuration of the fully-developed uterus.
178 R. A. BRAY
This new species would appear to be most closely related to Steringophoms magnus
Manter, 1934, from an unidentified eel-like fish in 300 fathoms at Tortugas, Florida.
It differs, however, in the granular pattern of the egg-shell, in the sucker ratio, in the
distribution of the vitellaria and in the configuration of the uterus in the hindbody.
The sculpturing on the egg-shell seems to be similar to that described in Benthotrema
plenum Manter, 1934, and it may be noted that although Manter (1934) originally
considered this to be a generic character, later (1954) he described a further species
in the genus, B. richardsoni, in which the eggs apparently have smooth shells.
Family HEMIURIDAE Liihe, 1901
Hemiurus appendiculatus (Rud., 1802) Looss, 1899
Host and locality : stomach of Alosa fallax, Stn. 41.
Seven specimens were found. This is a common parasite of shads and has been
recorded in A . fallax [A . finta] in various localities in the Mediterranean and north-
eastern Atlantic. Nevertheless, H. appendiculatus of de Oliveira Rodrigues et al.
(1972) recorded from A . fallax from the coast of Portugal may not belong to this
species, as it has, judging from the authors' figure, a relatively short pars prostatica
and an undivided seminal vesicle. The normal condition appears to be that rep-
resented by the present specimens, in which the pars prostatica is convoluted pos-
teriorly to the ventral sucker, and in which the anterior part of the bipartite seminal
vesicle has a thick muscular wall.
Lecithocladium excisum (Rud., 1819) Liihe, 1901
Host and locality : stomach of Scomber scombrus, Stn. 44.
Lecithochirium gravidum Looss, 1907
Host and locality : stomach of Conger conger, Stn. 7.
In their revision of Lecithochirium Liihe, 1901, Nasir and Diaz (1971) consider
L. gravidum to be a synonym of L. rufoviride (Rud., 1819). These two species are
certainly very similar morphologically and according to Looss (1908) and Guiart
(1938) they are distinguishable only by size and sucker ratio. L. gravidum is small
with an oral : ventral sucker ratio of about i : 2 whilst rufoviride is larger and has
suckers of more or less equal size. The present specimens (six in number) have an
oral sucker diameter of 0-24-0-29 mm and a ventral sucker diameter of 0-47-0-52 mm,
and therefore a sucker ratio of i : 1-8-2-1. In an attempt to discover whether or
not the sucker ratio in this group is influenced by allometric growth 60 specimens,
assigned to one or other of these species, were measured. These specimens, from the
collections of the British Museum (Natural History), were all from Conger conger
collected at Plymouth and Guernsey. The sucker ratio was then plotted against the
body-length as is shown in Fig. n. As can be seen there are two distinct regions,
one representing those specimens with a sucker ratio of more than 1:1-7 and the
other representing those with a ratio of less than i : 1-6. Allometric growth does
FISH TREMATODES FROM BAY OF BISCAY
179
a
bo
O OH
,. <U
-%
-G
3
^ 'I
^£
T3 fl
11
rt CJ
en 0)
c« >— i
8 ^
5-g
s -0
-I
O
£
i8o
R. A. BRAY
O\ ro
M ro
I I
00 O
X X X X X
'rt .
ro N N
6 66
i i i
00 N N CO <~O
N ro N N N
66666
"rt 5
b 'S
o
i
r- N
o o o o o
ro
W
»
H
p
n-
en
ro
^ s
6
i
6
i
6
i
>*
fl H
0
*o
C-1
0
o
> *S- ^
ro
H
ro
r-i
'S
o
O
O
0
O
i
ft
S
o5
<N
O
r^
E
^ "S"
M
M
6
Q
bo S
1
I
i
^^
£ £ "~>
oo
-t-
ro
ro
S £, CTI
M
'O
10
iO
rr\
0
M
O
O
O
S -S
«•* J
o « -S ^> «
^^^S^
FISH TREMATODES FROM BAY OF BISCAY 181
not appear to have a significant effect on the sucker ratio. It can therefore be seen
that, commonly occurring in Conger conger, there are two very similar species
differing in their sucker ratio, if not in any other obvious feature. This informa-
tion seems to uphold the validity of the species L. gravidum, to which the present
specimens are assigned.
Derogenes various (Miiller, 1784) Looss, 1901
Hosts and localities : stomach of Molva macrophthalma, Stn. 43 ; stomach of
Trachyrhynchus trachyrincus , Stn. 43 ; stomach of Micromesistius poutassou,
Stn. 6 ; stomach of Phycis blennoides, Stn. 17 ; stomach of Merluccius merluccius,
Stn. 8.
This parasite has been recorded from numerous fishes in polar and temperate
waters of the world. This appears, however, to be the first record of this worm from
M. macrophthalma, T. trachyrincus and M. poutassou. For this reason a table of
important dimensions is given (Table 3).
ACKNOWLEDGMENTS
My thanks are due to the following : Dr J. P. Harding, late of the British Museum
(Natural History), and Dr A. J. Lee of the M.A.F.F. Fisheries Laboratory, Lowestoft,
for arranging my participation in the cruise ; Mr R. J. Blacker, the naturalist-in-
charge, and other M.A.F.F. staff and crew members of 'Cirolana' for their help and
advice ; Dr N. R. Merrett of the National Institute of Oceanography, Wormley,
Surrey, for allowing me to examine fishes in the N.I.O. collection ; and Mr A.
Wheeler of the British Museum (Natural History) for identifying some of the fish.
Finally, I would like to thank Mr S. Prudhoe for his advice and encouragement
during this study.
REFERENCES
BRATEN, T. 1964. Otodistomum veliporum (Creplin, 1837) (Trematoda, Azygiidae) i Norge.
Fauna, Oslo, 17 : 73-75.
BRINKMAN, A., Jr. 1957. Fish trematodes from Norwegian waters. Ila. The Norwegian
species of the orders Aspidogastrea and Digenea (Gasterostomata) . Univ. Bergen Arb.
4 : 29 pp.
BYRD, M. A. 1964. Neosteganoderma glandulosa gen. n.( sp. n. (Trematoda : Steganodermati-
dae) from an Atlantic fish. Proc. helminth. Soc. Wash. 31 : 105-108.
CABALLERO Y. C., E., BRAVO HOLLIS, M. & GROCOTT, R. G. 1955. Helmintos de la Republica
de Panama. XIV. Trematodos monogeneos y digeneos de peces marines del Oceano
Pacifico del Norte, con descripcion de neuvas formas. An. Inst. Biol. Univ. Mex. 26 : 117-
147.
CHING, H. L. 1960. Some digenetic trematodes of fishes of Friday Harbor, Washington.
/. Parasit. 46 : 241-250.
1961. Redescription of the digenetic trematodes Lepidapedon calli and L. pugetensis
Acena, and new host records for L. calli and Eurycreadium vitellosum Manter from fishes of
Washington State. Can. J. Zool. 39 : 615-621.
CROWCROFT, P. W. 1951. Notes on the taxonomy of the genus Coitocaecum Nicoll, 1915
(Digenea : Opecoelidae) . /. Parasit. 37 : 251-256.
i82 R. A. BRAY
DAWES, B. 1947. The Trematoda of British fishes. Ray Society, London : 364 pp.
DE OLIVEIRA RODRIGUES, H., CARVALHO VARELA, M., SODRE RODRIGUES, S. & CRISTOFARO, R.
1972. Alguns tremat6deos digeneticos de peixes do Oceano Atlantico. - Costa Continental
Portuguesa e Costa Continental da Africa. Atas Soc. Biol. Rio de J. 15 : 87-93.
DOGIEL, V. 1936. [Parasites of cod from Lake Mogilnoe.] Uchen. Zap. leningr. gos. Univ.
3 (7) : 123-133. (In Russian.)
DOLLFUS, R. P. 1937. Les trematodes Digenea des selaciens (Plagiostomes) . Catalogue par
h6tes. Distribution geographique. Annls Parasit. hum. comp. 15 : 57-73.
GUI ART, J. 1938. Trematodes parasites provenant des campagnes scientifiques du Prince
Albert ier de Monaco (1886-1912). Result. Camp, scient. Prince Albert 99 : 84 pp.
GUPTA, N. K. & SEHGAL, S. K. 1971. Some allocreadiid trematodes from marine food fishes of
India. Res. Bull. Panjab Univ. Sci. 21 : 267-275.
HANSON, M. L. 1950. Some digenetic trematodes of marine fishes of Bermuda. Proc.
helminth. Soc. Wash. 17 : 74-89.
KABATA, Z. 1961. A new genus and species of trematode parasitic in Macrurus fabricii
(Sundeval), a deep-sea fish. Proc. Zool. Soc. Lond. 136 : 285-292.
LEBOUR, M. V. 1908. Fish trematodes of the Northumberland coast. Rep. Invest. Northumb.
Sea Fish. Comm. (1907) : 11-57.
LINTON, E. 1940. Trematodes from fishes mainly from the Woods Hole region, Mass. Proc.
U.S. natn. Mus. 88 : 1-172.
Looss, A. 1908. Beitrage zur Systematik der Distomen. Zur Kenntnis der Familie Hemiuri-
dae. Zool. Jb. (Syst.) 26 : 63-180.
MACHIDA, M., ICHIHARA, A. & KAMEGAI, S. 1970. Digenetic trematodes collected from fishes
in the sea north of Tsushima Islands. Mem. natn. Sci. Mus., Tokyo 3 : 101-112.
MAILLARD, C. 1970. Nicolla elongata n. sp. (Digenea, Coitocaecidae) parasite d'Onos tricirratus
(Brunnich, 1768) (Teleostei). Bull. Mus. natn. Hist. nat. Paris 42 : 410-414.
MANTER, H. W. 1925. Some marine fish trematodes of Maine. /. Parasit. 12 : 11-18.
1926. Some North American fish trematodes. Illinois biol. Monogr. 10 : 127-264.
1934. Some digenetic trematodes from deep-sea fish of Tortugas, Florida. Papers
Tortugas Lab., Carnegie Inst. Washington 28 : 257-345.
1947- The digenetic trematodes of marine fishes of Tortugas, Florida. Am. Midi. Nat.
38 : 257-416.
1954- Some digenetic trematodes from fishes of New Zealand. Trans. R. Soc. N.Z.
82 : 475-568.
MATHIAS, P. 1934. Sur quelques Trematodes de poissons marins de la region de Banyuls.
Archs Zool. exp. gen. 75 : 567-581.
MATTHEWS, R. A. 1973. The life-cycle of Prosorhynchus crucibulum (Rudolphi, 1819) Odhner,
1905, and a comparison of its cercaria with that of Prosorhynchus squamatus Odhner, 1905.
Parasitology 66 : 133-164.
MILLER, M. J. 1941. A critical study of Stafford's report on 'Trematodes of Canadian fishes'
based on his trematode collection. Can. J. Res. 19 : 28-52.
NASIR, P. & DIAZ, M. T. 1971. A revision of genus Lecithochirium Luehe, 1901, and redescrip-
tion of L. monticellii (Linton, 1898) Skrjabin and Guschanskaja, 1955. Riv. Parassit.
32 : 27-36.
NICOLL, W. 1915. The trematode parasites of north Queensland. III. Parasites of fishes.
Parasitology 8 : 22-41.
ODHNER, T. 191 la. Zum natiirlichen System der digenen Trematoden. II. Zool. Am.
37 : 237-253.
191 ib. Zum natiirlichen System der digenen Trematoden. III. (Ein weiterer Fall von
sekundarem Anus.) Zool. Anz. 38 : 97-117.
OZAKI, Y. 1928. Some gasterostomatous trematodes of Japan. Jap. J. Zool. 2 : 35-60.
PALOMBI, A. 1931. Per una migliore conoscenza dei Trematodi endoparassiti dei pesci
del golfo di Napoli. i . - Steringotrema divergens (Rud.) e Haploporus benedeni (Stoss.).
Annuar. Mus. zool. Univ. Napoli 6 : 1-15.
FISH TREMATODES FROM BAY OF BISCAY 183
POLYANSKY, Y. I. 1955- Parasites of the fish of the Barents Sea. Trudy zool. Inst. Leningr.
19 : 5-170. (English translation : Israel program for scientific translations, Jerusalem,
1966.)
PRICE, E. W. 1934. New digenetic trematodes from marine fishes. Smithson. misc. Collns
91 (7): i -8.
PRITCHARD, M. H. 1966. Studies on digenetic trematodes of Hawaiian fishes. Family
Opecoelidae Ozaki, 1925. Zool. Jb. (Syst.) 93 : 173-202.
RONALD, K. 1960. The metazoan parasites of the Heterosomata of the Gulf of St. Lawrence.
VI. Digenea. Can. J. Zool. 38 : 923-937.
SKRJABIN, K. I. & KOVAL, V. P. 1960. Suborder Allocreadiata Skrjabin, Petrov and Koval,
1958. Part two. In : Skrjabin, K. I., Ed. Trematodes of animals and man. Principles of
trematodology . Moscow, 18 : 15-377. (English translation : Israel program for scientific
translations, Jerusalem 1965.)
SLUSARSKI, W. 1971. A redescription of Nicolla testiobliqua (L. W. Wisniewski, 1933) R. Ph.
Dollfus, 1958, a fish trematode from Bosnia and some remarks on the present status of
European and other representatives of the family Coitocaecidae Ozaki, 1928 (Podocotyloidea
R. Ph. Dollfus, 1960). Acta parasit. Pol. 19 : 121-132.
STRELKOV, J. A. 1960. Endoparasitic worms of marine fishes of East Kamchatka. Trudy
zool. Inst. Leningr. 28 : 147-196.
TENDEIRO, J. 1955. Sobre um novo trematodeo digenetico, Dolichoenterum manteri n. sp.,
parasita do safio, Conger conger (L.) algumas considera£oes sobre a sistematica da subordem
Gasterostomata Odhner, 1905. Bolm Soc. port. Cienc. nat. 20 : 19-43.
WILLIAMS, H. H. 1960. A list of parasitic worms, including 22 new records, from marine
fishes caught off the British Isles. Ann. Mag. nat. Hist. Ser. 13 2 : 705-715.
YAMAGUTI, S. 1934. Studies on the helminth fauna of Japan. Part 2. Trematodes of
fishes, i. Jap. J. Zool. 5 : 249-541.
- 1938. Studies on the helminth fauna of Japan. Part 21. Trematodes of fishes. IV.
Kyoto, Japan : 139 pp.
1953. Systema Helminthum. Part i. Digenetic trematodes of fishes. Tokyo, Japan :
405 pp.
1958. Systema Helminthum. i. The digenetic trematodes of vertebrates. Interscience
Publishers Inc., New York : 1575 pp.
1970. The digenetic trematodes of Hawaiian fishes. Keigaku Publishing Co., Tokyo :
436 pp.
1971. Synopsis of digenetic trematodes of vertebrates. Keigaku Publishing Co., Tokyo :
1074 pp.
R. A. BRAY
Department of Zoology
BRITISH MUSEUM (NATURAL HISTORY)
CROMWELL ROAD
LONDON SW7 5BD
A LIST OF SUPPLEMENTS
TO THE ZOOLOGICAL SERIES
OF THE BULLETIN OF
THE BRITISH MUSEUM (NATURAL HISTORY)
1. KAY, E. ALISON. Marine Molluscs in the Cuming Collection British Museum
(Natural History) described by William Harper Pease. Pp. 96 ; 14 Plates.
1965. (Out of Print.) £3.75."
2. WHITEHEAD, P. J. P. The Clupeoid Fishes described by Lacepede, Cuvier and
Valenciennes. Pp. 180 ; n Plates, 15 Text-figures. 1967. £4.
3. TAYLOR, J. D., KENNEDY, W. J. & HALL, A. The Shell Structure and Mineralogy
of the Bivalvia. Introduction. Nuculacea-Trigonacea. Pp. 125 ; 29 Plates
77 Text-figures. 1969. £4.50.
4. HAYNES, J. R. Cardigan Bay Recent Foraminifera (Cruises of the R.V. Antur)
1962-1964. Pp. 245 ; 33 Plates, 47 Text-figures. 1973. £10.80.
Printed in Great Britain by John Wright and Sons Ltd. at The Stonebridge Press, Bristol BS4 sNU
OBSERVATIONS ON
LINEARE PENARD (TESTACEA :
PROTOZOA)
R. H. HEDLEY
AND
C. G. OGDEN
BULLETIN OF
THE BRITISH MUSEUM (NATURAL HISTORY)
ZOOLOGY Vol. 26 No. 3
LONDON: 1974
OBSERVATIONS ON TRINEMA LINEAR
PENARD (TESTACEA: PROTOZOA)
BY
RONALD HENDERSON HEDLEY
AND
COLIN GERALD OGDEN /
Pp. 185-199 ; 7 Plates, i Text-figure
BULLETIN OF
THE BRITISH MUSEUM (NATURAL HISTORY)
ZOOLOGY Vol. 26 No. 3
LONDON: 1974
THE BULLETIN OF THE BRITISH MUSEUM
(NATURAL HISTORY), instituted in 1949, is
issued in five series corresponding to the Departments
of the Museum, and an Historical series.
Parts will appear at irregular intervals as they
become ready. Volumes will contain about three or
four hundred pages, and will not necessarily be
completed within one calendar year.
In 1965 a separate supplementary series of longer
papers was instituted, numbered serially for each
Department.
This paper is Vol. 26, No. 3 of the Zoological
series. The abbreviated titles of periodicals cited
follow those of the World List of Scientific Periodicals.
World List abbreviation :
Bull. Br. Mus. nat. Hist. (Zool.)
Trustees of the British Museum (Natural History), 1974
TRUSTEES OF
THE BRITISH MUSEUM (NATURAL HISTORY)
Issued 7 February, 1974 Price £1.50
OBSERVATIONS ON TRINEMA LINEARE
PENARD (TE STAGE A : PROTOZOA)
By R. H. HEDLEY AND C. G. OGDEN
CONTENTS
Page
SYNOPSIS ........... 187
INTRODUCTION ........... 187
Previous work - biology . . . . . . . .188
Previous work - taxonomy . . . . . . . .188
MATERIAL AND METHODS ....... . . 189
DISTRIBUTION ........... 190
MORPHOLOGY AND VARIATION .... ... 191
REPRODUCTION .......... 192
ULTRASTRUCTURE OF VEGETATIVE STAGE ... . 192
ACKNOWLEDGEMENT .......... 197
REFERENCES ........... 197
SYNOPSIS
Trinema lineave, a geographically widespread, fresh-water, moss and soil inhabiting testacean,
has been established in clonal culture with a doubling time of between 72 and 78 hours. Full
descriptions are given of the siliceous shell and the fine structure of the vegetative stage. Ultra-
structural features of special interest include the presence of microbodies, and a microtubule-
organizing-centre associated with the nucleus at prophase.
INTRODUCTION
THE five known species of Trinema Dujardin, 1841, are amongst the most widely
reported rhizopods to be found in soil, sphagnum moss and fresh-water habitats.
Very little is known of their biology and there has been no previous report of the
cytoplasmic ultrastructure. The present paper is an account of the fine structure of
cultured specimens of Trinema lineare Penard, 1890, isolated from a sample of moss
and soil from Cliffe Marshes, Rochester, Kent, in February, 1970. It is the second
of a series of papers devoted to testate amoebae aimed at providing a fuller under-
standing of their biology, and a fuller appreciation of the significance and nature of
the siliceous shell. The first paper in the series was devoted to a detailed account of
Euglypha rotunda - a widely distributed species (Hedley and Ogden, 1973).
Trinema belongs to the family Euglyphidae and the classification adopted here is
that proposed by Loeblich and Tappan (1961) :
Class RHIZOPODEA Von Siebold, 1845
Subclass FILOSIA Leidy, 1879
Order GROMIDA Claparede and Lachmann, 1859
Superfamily EUGLYPHACEA Loeblich and Tappan, 1961
i88 R. H. HEDLEY AND C. G. OGDEN
Family EUGLYPHIDAE Wallich, 1864;
test hyaline, symmetrical, elongate, composed of rounded siliceous
scales, aperture rounded or elongate : one nucleus.
Previous work - biology
Leidy (1879) examined several testate amoebae from North America arid suggested
that the cytoplasmic structures of Trinema were similar to those in Euglypha, in that
each had a single nucleus and two contractile vacuoles. Penard (1902) observed
that when the animals were inactive, or in the vegetative phase of their life cycle,
the granular zone and nucleus appeared to be distinct. Dunkerly (1923) reported
that reserve shell-plates were arranged around the nucleus, and suggested that
certain dark granules were chromidia. Chardez (1960) noted that the cytoplasm
did not fill the shell cavity and that it appeared to be attached only at the apertural
collar.
Both Penard (1902) and Leidy (1879) described the pseudopodia as very fine,
usually two or three and occasionally six in number. According to Leidy (1879)
when the animal moves, the body is inclined so that the aperture faces anterior and
down, while the fundus points backwards and up.
During binary fission, Penard (1902) observed that the animals are diametrically
opposed whereas at conjugation they appear to be directly apposed. Chardez
(1960) observed the part of the reproductive cycle when clear cytoplasm passed
into the daughter shell, until the moment that the daughter cell became packed with
agitated vacuoles. The cytoplasmic volume attained full size within 20 minutes.
Chardez (1960) also reported the formation of a cyst with two nuclei as a phenomenon
of conjugation, in which the cytoplasm of the two conjugants become joined in one
shell, the empty shell remaining provisionally united. Similar united individuals,
with one shell empty and the other containing cytoplasm and two nuclei, were
observed by Dunkerly (1923) who considered this to be a stage of encystation, after
which the cytoplasm contracted, the chromidia disappeared - at least as staining
bodies - and finally the nuclei fused. Dunkerly (1923) also suggested that the
chromidia were used up during encystation as reserve food material.
As a result of an ecological study of a Netherland fen, de Graff (1956) reported
that Trinema enchelys and T. lineare show an optimum distribution in moderate
dry mosses, but were found in most kinds of biotopes, only T. lineare avoiding the
drier mosses. In a review of soil protozoa, Stout and Heal (1967) described T. lineare
as an ubiquitous species, found in both organic and mineral topsoils, and having the
same pH tolerance as the common soil ciliates and flagellates. They also stated
that most testaceans - with some exceptions including T. lineare - reproduce slowly,
have poor encystment mechanisms and poor tolerance to high carbon dioxide and
low oxygen tensions and salinities.
Previous work - taxonomy
T. lineare was described first by Penard (1890), who considered it to be smaller
and more elongate than T. enchelys (Ehrenberg, 1838). The same author (Penard,
TRINEMA LINEARE (PROTOZOA) 189
1902) later redefined the size as being 16-26 /urn and rarely 30 /un. Cash et al. (1915)
redescribed three species of Trinema, including T. lineare, and one variety in a review
of the British Rhizopoda Fauna. They listed T. acinus Leidy, 1879 as a synonym of
r. lineare, but this is probably a doubtful species because Leidy (1879) described only
T. enchelys and figured only T. acinus. Both Volz (1929) and Kufferath (1932)
suggested that T. lineare was a synonym of T. enchelys, whereas subsequent authors
have recognized both species. Hoogenraad and de Groot (1940) redescribed three
species and listed some measurements of previous authors. Chardez (1956) figured
numerous variations in the shape and size of the shell of both T. enchelys and T.
lineare.
Thomas (1958) suggested that there are three types of shell : firstly, those with
large completely overlapping plates ; secondly, those with large incompletely over-
lapping plates with smaller plates filling the gaps ; and thirdly, those with large
separate plates with many small plates ; the first type is seen fairly often, the second
is very common whilst the last type is rare.
Bonnet and Thomas (1960) redescribed five species of Trinema and one variety,
T. complanatum Penard, 1890 ; T. complanatum var. globulosa Chardez, 1959 ; T.
enchelys ; T. galeata (Penard, 1890) ; T. lineare and T. penardi Thomas and Chardez,
1958, with a list of their recorded distribution. In several publications Decloitre
(i96ia, b, I964a, b, I965a) has provided additional measurements and localities,
and discussed variation seen in some testaceans. Decloitre (1962) described a new
variety of T. lineare, T. lineare var. terricola having a ventral aperture, and more
recently Stepanek (1967) described two varieties, T. lineare var. globulosa having a
mouth at. one side and T. lineare, var. pellucida having an elliptical mouth.
MATERIALS AND METHODS
T. lineare was isolated from a sample of moss and soil collected on Cliffe Marshes,
near Rochester, Kent, in February, 1970. Crude cultures were made from small
portions of this material placed in the culture liquid and kept in the laboratory at
room temperature, i8-20°C. Agnotobiotic cultures were kept, in small plastic
containers, on a thin substrate of agar (i per cent agar agar in distilled water) with a
sterilized wheat grain added prior to setting, and covered with a shallow layer of the
culture liquid. This liquid was a 5 per cent (w/v) solution of soil extract, plus
100 mg/l"1 of sodium nitrate and 15 mg/1"1 of sodium dihydrogen orthophosphate,
in distilled water.
Clonal cultures were obtained by isolating single, active animals. One such clone
was subsequently used to produce working cultures. If sub-cultures are made at
intervals of between three or four weeks the animals appear to feed and reproduce
readily. The clone which was used to produce working cultures is now deposited
at the Culture Centre of Algae and Protozoa, the Natural Environment Research
Council, Cambridge, England.
Optical microscopy - The animals were examined by bright-field and phase-
contrast microscopy, either alive or after fixation. Smears fixed in either
Schaudinn's fluid or glutaraldehyde were stained with either borax carmine or iron
haematoxylin.
igo R. H. MEDLEY AND C. G. OGDEN
Scanning electron microscopy - For morphological studies on external morphology
both live animals and empty shells were used. Living specimens were fixed initially
in 3 per cent glutaraldehyde in distilled water for 30 minutes. The empty shells
were washed initially in several changes of distilled water. Single specimens were
then passed through several changes of triple glass distilled water using either a
single-hair brush or a fine-bore pipette. They were then manipulated onto a small
cover-slip, previously cleaned with acetone and lint-free tissue, and allowed to dry.
Dried specimens adhere well to glass but can be moved by use of a moistened single-
hair brush. For the examination of individual siliceous plates, single clean specimens
were placed on a fragment of cleaned cover-slip, covered with a small drop of con-
centrated sulphuric acid and gently heated. On evaporation of the acid the plates
are liberated from the organic cement. The treated cover-slips were attached to
'Stereoscan' stubs by an electrically-conductive paint, 'Silver Dag', and then coated
evenly with 10-15 nm °f g°ld using an Edwards coating-unit with a planetary
specimen holder (Harris et aL, 1972). The stubs were examined on a Cambridge
Stereoscan Mk II at either 15 or 20 kV and the results recorded on Ilford 35 mm H?3
film.
Transmission electron microscopy - Animals were fixed at room temperature for
15 minutes in i per cent glutaraldyhyde in 0*05 M Sorenson's phosphate buffer plus
0-0015 M calcium chloride, followed by 10 minutes in 3 per cent glutaraldehyde in the
same buffer. After several rinses in buffer, they were post-fixed in i per cent osmium
textroxide in o-i M Sorenson's buffer. To facilitate subsequent handling the
specimens were occasionally embedded in 1-5 per cent agar at this stage, prior to
dehydration. The material was dehydrated by passage through a series of graded
water/ethanol mixtures, ending in absolute ethanol and embedded in Epon 812.
Sections were cut on a Porter Blum MT2 ultramicrotome using a Du Pont diamond
knife, stained with alcoholic uranyl acetate and Reynold's lead citrate, and examined
in an A.E.I. 6B electron microscope operating at 60 kV. The results were recorded
on Ilford's 'Special Lantern Contrasty' plates.
DISTRIBUTION
T. lineare is commonly found in damp and wet mosses and various soils. A list of
localities and references from which it has been recorded is given by Bonnet and
Thomas (1960). Additional localities are given here based on records of several
authors, namely, Hoogenraad and de Groot (i952a, b), Bonnet (1966), Decloitre
(i964a, I965a, b), Chardez (1961, 1969) and Golemansky (1971).
The following list illustrates the geographically wide distribution of T. lineare :
EUROPE : Great Britain, Spitzbergen, Iceland, Finland, France, Belgium,
Netherlands, Germany, Italy, Czechoslovakia and Hungary.
NORTH AMERICA : United States of America, Canada and Greenland.
SOUTH AMERICA : Tristan da Cunha, Chile, Bolivia and Peru.
TRINEMA LINEARE (PROTOZOA) 191
AFRICA : Angola, Morocco, Congo, French Cameroons, Algiers, Guinea,
South Africa and the Seychelles.
ASIA : India, China, Siberia and Japan.
AUSTRALASIA : Australia, New Zealand and Tasmania.
ANTARCTICA : South Georgia and Gough Island.
MORPHOLOGY AND VARIATION
The siliceous shell is elliptical in shape through the major axis and circular through
the minor axis ; it varies in length between 25 and 34-5 /mi and in breadth between
14.5 and 19 /mi (PI. i, figs. B and E). The aperture is normally circular, invaginated
and situated sub-terminally, having a diameter of between 6 and 8-5 /mi (PI. i,
figs. A, B and E). Its position may vary greatly through various angles to the
longitudinal axis of the shell, including being terminal. The shell is composed of
three different types of siliceous plates : small, circular apertural plates (PI. i,
figs. C and D), large, circular shell-plates (PI. i, fig. B) and small, oval to circular
shell-plates (PI. i, figs. D and F).
The apertural plates are roughly circular, folded along the median line with a small
central dorsal tooth (PI. i, figs. C and D), and vary between 1-2 and 1-5 /mi in dia-
meter. The number per specimen appears to vary between 18 and 28, but the majority
of animals have between 22 and 26. The apertural plates may occasionally be moved
from their marginal position either to form double rows or to be displaced out of
symmetry. Usually either one or two rows of between 13 and 16, small shell-plates
are arranged adjacent to the apertural plates (PI. i, fig. A), but these are also subject
to some variation, in one instance being replaced by large shell-plates.
The large circular, slightly convex shell-plates have a diameter of between 4-0 and
5-5 /mi and a thickness of between 0-12 and 0-20 /mi ; whereas the small oval to
circular, convex shell-plates (PI. i, fig. F) vary in length between 2-4 and 3-5 /mi
and are between o-n and 0-20 /mi thick. T. lineare has approximately 50 incom-
pletely overlapping large shell-plates with an unknown number of small shell-plates
haphazardly filling the interstices, many of them being completely overlaid by the
large shell-plates. This is the second and most common of the three types of shell-
plate arrangement noted by Thomas (1958). Although the plates are normally
arranged evenly, an occasional shell is seen in which some shell-plates are reversed,
with the concave surface facing outwards (PI. 6, fig. E).
The incidence of abnormally shaped shells (Plate 2) is usually less than i per cent,
although as many as 10 per cent of one culture was observed to have abnormal
forms. Cash et al. (1915) and Chardez (1970) have illustrated previously the variation
in the position of the aperture and included some examples of evaginated apertures.
The most frequently seen abnormal forms are a single shell possessing two apertures -
an aperture being defined as an opening bordered by apertural plates. Openings in the
shell due to inadequate shell-plate coverage are common (PL 2, fig. F), and as many as
six openings have been seen in one individual. It would appear that the number of
shell-plates in abnormal forms is roughly proportional to the number of apertures.
IQ2 R. H. HEDLEY AND C. G. OGDEN
There is also a slight increase in the amount of organic cement binding the plates of
the abnormal forms.
There appear to have been only two previous reports concerning the surface ultra-
structure of T. lineare, both of which are based on carbon and metal-shadowed pre-
parations. Thomas and Hovasse (1962) described two types of plates, large and
small, with the aperture bordered by a collar of biconvex spines. Mercier et al.
(1964) described the aperture as being surrounded by one or two circles of small spiny
plates.
REPRODUCTION
An estimate of the doubling time was obtained by growing three replicate cultures,
and recording the number of animals present at regular intervals. Growth curves
produced from these results show that the doubling time is between 3-0 and 3-4 days.
A similar calculation for Euglypha rotunda gave a doubling time of between 1-4 and
1-9 days (Hedley and Ogden, 1973).
Binary fission is completed in approximately 60 minutes. The initial stages are
difficult to observe in T. lineare due to the small size of the animal and the oblique
position of the aperture. Cytoplasmic division, once the daughter shell has been
formed, proceeds in a manner similar to that previously described (Hedley and
Ogden, 1973) for E. rotunda.
ULTRASTRUCTURE OF VEGETATIVE STAGE
The cytoplasm does not quite fill the shell cavity (PI. 3, fig. A), but appears to be
anchored to the apertural collar. Numerous fine processes can be seen extending
between the cytoplasm and the inner shell wall (PI. 3, fig. A). Unevenly spaced
pellicular microtubules lie beneath the plasmalemma and run in an antero-posterior
direction. The mitochondria are ovoid or spherical in shape, possess tubular cristae,
and appear to be distributed at random throughout the cytoplasm.
Nucleus. The vegetative or interphase nucleus is usually spherical, between
4-4 and 5 -7 /nm in diameter, occupying a central position at the posterior end of the
cytoplasm (PI. 3, figs. A and D). It is bound by a nuclear envelope made up of two
tripartite membranes, the outer membrane being continuous with the granular
endoplasmic reticulum. The nuclear matrix is finely granular, with small con-
centrations of densely-staining chromatin scattered throughout and a dense nucleolus.
Although the nucleus during interphase was spherical in most animals examined,
a number have been observed which are cone-shaped posteriorly (PI. 4, fig. A), and
it is assumed that this is correlated with the early stages of prophase. The cone-
shape is caused by the convergence of numerous microtubules towards a specific
region at the posterior end of the cytoplasm. The microtubules are first seen con-
verging on the nucleus around its equatorial region (Text-fig, i ; PI. 6, fig. A), whilst
at a level in the region of the contractile vacuoles they lie in the endoplasmic reticulum
at some distance from the nucleus. Although it is difficult to estimate the number
of microtubules present because of the close relationship of the dense endoplasmic
reticulum, it is apparent that the numbers increase to approximately 70 in the pos-
terior region of the cytoplasm.
TRINEMA LINEARE (PROTOZOA)
193
---B
C-
E--
MTOC
(a)
r--D
(b)
E-- -
-F
FIG. i. (a) Diagram of a longitudinal section through the nucleus (see PI. 4, fig. A),
(b) Tranverse sections at those levels of the nucleus marked A-B, C-D and E-F in (a),
to illustrate the microtubules converging towards the microtubule-organizing- centre
(MTOC), as discussed in the text.
194 R- H- HEDLEY AND C. G. OGDEN
The microtubules lie close to the nuclear membrane (PI. 4, fig. C), but do not lie
in invaginations similar to those reported by Leadbeater and Dodge (1967) for the
dinoflagellate Woloszynskia micra. They appear to attach to an electron-dense area
close to the distal end of the nucleus (PI. 4, figs. C and D), which is comparable with
previously described and similar regions in the alga, Char a (Pickett-Heaps, 1968),
fungal zoospores (Fuller and Calhoun, 1968), the soil amoeba, Acanthamoeba castel-
lanii (Bowers and Korn, 1968), the marine amoebae, Stereomyxa ramosa and 5.
angulosa (Benwitz and Grell, igyia, b), and the marine protist, Labyrinthula (Perkins,
1970). Such regions of attachment for microtubules are referred to in other cells
under a variety of terms. For example, in dividing plant cells, Pickett-Heaps (1969)
referred to it as a ' microtubule-organizing-centre ' - MTOC, and suggested that such
centres initiate and control the arrangement of microtubules. An MTOC was also
described by McCully and Robinow (1972) in association with the nucleus during
mitosis in yeasts. In Labyrinthula, Perkins (1970) described the granular aggregate
as a 'protocentriole', while the term ' paracentrosome ' was suggested by Manton
et al. (1970) for the precursor material seen to accumulate near each pole at meiotic
division in the marine centric diatom, Lithodesium undulatum. Fuller and Calhoun
(1968) stated that the kinetosome of fungal zoospores is an unlikely 'organizing
centre' but suggested that the electron-opaque material surrounding the proximal
third of the kinetosome could be such a 'centre'. In mouse oocytes electron-dense
fibrillar areas, close to the nucleus, from which microtubules radiate are referred to
by Szollosi et al. (1972) as 'microtubule foci'. Evidence that such regions indicate
the site of microtubule formation in many systems has been provided by Tilney and
Goddard (1970) and Tilney (1971), who conducted experiments on the breakdown and
reformation of microtubules in certain species of Heliozoa.
Contractile vacuoles. Two or three contractile vacuoles occur at the edge of the
granular endoplasmic reticulum (PI. 3, fig. D) in the region of the nucleus. These
vacuoles are often surrounded by numerous vesicles which are associated with the
lumen of the vacuole, and they discharge directly into the shell cavity.
Microbodies. Microbodies appear in all the specimens examined (PI. 3, fig. B).
They are ovoid or spherical in shape, varying in size between 0-30 and 0-55 /mi,
with a dense granular matrix surrounded by a single unit membrane. Tubular
elements, between 18 and 26 nm in diameter, appear within the matrix (PI. 4, fig. E),
and as many as four tubules are apparent as cross-sections or loops in some micro-
bodies.
The microbodies in T. lineare are similar to those reported in various Foraminifera
by Hedley et al. (1967), Hedley and Wakefield (1969) and Febvre-Chevalier (1971),
in possessing tubular-like elements within the matrix. Hedley and Wakefield
(1969) imply that such organelles are possibly a normal component of the cytoplasm
of the Foraminifera. The structure and function of microbodies has been reviewed
recently by Hruban and Rechcigl (1969).
Endoplasmic reticulum. A concentrated mass of granular endoplasmic reticulum
usually surrounds the nucleus, and appears more electron-dense than the surrounding
cytoplasm due to the concentration of ribosomes (PI. 3, figs. A and D). In early
TRINEMA LINEARE (PROTOZOA) 195
prophase, however, the nucleus is drawn out distally from the endoplasmic reticulum
region (PI. 4, fig. A).
Pigment zone. A zone of large vacuoles lies immediately anterior to the endoplas-
mic reticulum region (PL 3, fig. A). They contain electron-dense material and have
previously been equated by Hedley and Ogden (1973) to the 'pigment zone' of earlier
light microscopy workers. A probable developmental sequence in the formation of
the electron-dense material in the vacuoles is illustrated in PI. 3, figs. A and B - the
various stages are labelled i to 4. The initial stage (i) shows the matrix to be
granular with a small electron-dense area in the centre. As this deeply stained area
increases in size (2 and 3) the matrix becomes coarser and uneven rents appear in the
electron-dense material. In the final stage (4) the vacuole is composed mainly of
electron-dense material. The rents in the electron-dense material are possibly
caused by its impermeability to the embedding resin. No explanation can be made
at present regarding the occurrence of this constant structure, or of the occasional
empty vacuoles seen here and reported previously by Hedley and Ogden (1973) for
Euglypha rotunda.
Golgi apparatus. A single Golgi apparatus lies immediately posterior to the nucleus
at the edge of the endoplasmic reticulum region (PI. 4, fig. A). Both smooth and
coated vesicles are associated with the Golgi saccules (PI. 4, fig. B). In addition, the
saccules of the Golgi are often distended by small concentrations of densely staining
fibrillar material (PL 5, fig. C). These concentrations appearing in the outer saccules
of the dictyosome become progressively spherical as the concentration of material
increases and finally are detached at the margins of the saccules as membrane-
bound vesicles. At this stage the fibrillar material is concentrated mainly at the
centre of the vesicle, and has small strands radiating from the centre (PL 3, fig. C).
The vesicles then pass around the outside of the endoplasmic reticulum region and
are distributed randomly throughout the cytoplasm.
It has already been established (Favard, 1969) that the Golgi apparatus appears
to play a role in the packing of secretory products for export and storage. The
chemical constituents of such secretory products in both plants and animals are
usually polysaccharide or protein macromolecules. Hedley and Wakefield (1969)
suggested that the polysaccharide produced by the Golgi apparatus of the marine
protozoon, Gromia oviformis, appeared to be used in the formation of the proteinace-
ous shell-wall. Schwab (1969) also suggested that the fibrillar shell-wall material
in the marine foraminifera, Myxotheca arenilega, is produced by the Golgi apparatus.
More recently, Hedley and Ogden (1973) have suggested that the fine fibrillar, poly-
saccharide material contained in the spherical, membrane-bound vesicles of E.
rotunda might be utilized to form the proteinaceous cement or glue, that lines the
inside of the siliceous shell and also binds the shell-plates of this testate amoeba.
It seems reasonable to suggest that the fibrillar vesicles (PL 3, fig. C) produced by
the Golgi apparatus in T. lineare might be the initial stages in the formation of the
organic cement bodies (PL 5, fig. E).
Another type of vesicle, with a double unit' membrane and usually electron-
transparent contents, often occurs lying in the area of cytoplasm between the distal
ig6 R. H. HEDLEY AND C. G. OGDEN
end of the granular endoplasmic reticulum region and the Golgi apparatus (PI. 5,
fig. D) . Occasionally they occur in the cytoplasm anterior to the granular endoplas-
mic reticulum region, surrounded by a ring of smooth endoplasmic reticulum (PI. 5,
fig. A). The function of these vesicles is at present unknown.
Food particles. Food particles, which are usually gram-negative bacteria, occur
throughout the cytoplasm, whereas in E. rotunda they are confined to the anterior
third of the cytoplasm (Hedley and Ogderi, 1973). Nevertheless, digestion in T.
lineare probably occurs in the enlarged food vacuoles which are found normally in
the anterior cytoplasm.
Reserve shell-plates. Reserve shell-plates are formed in that region of granular
endoplasmic reticulum that surrounds the nucleus and close to the Golgi apparatus.
As the shell-plates are formed they become closely packed together in a region just
anterior to the nucleus (PI. 6, fig. D). Both large and small shell-plates appear to be
formed at the same time. The apertural plates appear to be the last to be formed
and are usually seen only in the posterior region of the cytoplasm, with their dorsal
teeth pointing outwards.
Large inclusions, containing electron-dense material, are present in the granular
endoplasmic reticulum region (PI. 3, fig. A). These inclusions appear to be less
dense than the ' pigment zone ' vacuoles and are often seen to fuse with the unit
membrane of reserve shell-plates (PI. 6, fig. B). It seems probable that these
inclusions are associated with the formation of the siliceous plates as they are
frequently seen in specimens containing reserve shell-plates. It is noted, however,
that whereas in sections of animals fixed and subsequently treated in the absence of
heavy metals the siliceous plates are naturally electron-dense (PI. 7, fig. A), the
inclusions in the endoplasmic reticulum are electron-transparent.
The only additional reference to those previously reported by Hedley and Ogden
(1973) regarding the formation of siliceous material is that of Cachon and Cachon
(1971) on silica metabolism in Radiolaria. They suggest that the siliceous shells of
digested micro-organisms, upon which radiolarians feed, are used to produce their
siliceous skeletons, and the unused siliceous material is rejected in gel form.
Organic cement. The organic cement or glue that holds the plates together is a
fine fibrillar material probably produced by the Golgi apparatus. It is circulated
within the cytoplasm in membrane-bound vesicles (PI. 5, fig. E), and is discharged
mainly in the anterior region where the vesicles fuse with the plasmalemma. In
abnormal forms there is usually more cement at shell-plate junctions than in normal
specimens.
Pseudopodia. The fine structure of the pseudopodia consists of ground plasm
limited by a membrane, and occasionally containing microfilaments. Our observa-
tions are, however, limited to cytoplasmically joined individuals, in which micro-
tubules are often seen in the cytoplasm internal to the aperture, but not externally
(PI. 6, fig. C). Similar observations on microtubules were reported for Difflugiella
sp. by Griffin (1972) and for E. rotunda by Hedley and Ogden (1973).
Rosette groups. Rosette-like groups are frequently seen in clonal cultures which are
four or five weeks old (PI. 2, fig. G and PI. 7, figs. C-E). The groups are usually
TRINEMA LINEARE (PROTOZOA) 197
composed of five individuals or less and are joined by cytoplasmic connections.
These connections contain numerous cement bodies and less frequently mitochondria
(PL 7, fig. C). In contrast to the situation reported by Hedley and Ogden (1973),
for similar formations in E. rotunda, microfilaments are seldom present. Individuals
with reserve shell-plates are seldom seen in rosette groups, suggesting that such
formations may be the result of starvation.
Abnormal forms
The cytoplasm of animals with abnormal shapes (PI. 2) is similar to that of normal
animals with the exception that the number of nuclei appears to be related to the
number of apertures (PI. 7, fig. F). From our observations all attempts by these
forms to divide are abortive.
ACKNOWLEDGEMENT
We would like to acknowledge the technical assistance of Mrs J. E. Dorahy.
REFERENCES
BENWITZ, G. & GRELL, K. G. 1971 (a). Ultrastruktur mariner Amoben. II. Stereomyxa
ramosa. Arch. Protistenk. 113 : 51-67.
1971 (b). Ultrastruktur mariner Amoben. III. Stereomyxa angulosa. Arch. Prostistenk.
113 : 68-79.
BONNET, L. 1966. Le peuplement thecamoebien de quelques sols du Chili (I). Protistologica
2: 113-139.
BONNET, L. & THOMAS, R. 1960. Faune terrestre et d'eau douce des Pyrenees-Orientales.
Thecamoebiens du sol. 103 pp. Hermann, Paris.
BOWERS, B. & KORN, E. D. 1968. The fine structure of Acanthamoeba castellanii. I. The
trophozoite. /. Cell. Biol. 39 : 95-111.
CACHON, J. & CACHON, M. 1971. Recherches sur le metabolisme de la silice chez les Radio-
laires. Absorption et excretion. C. v. Acad. Sci. (Paris), 272 : 1652-1654.
CASH, J., WAILES, G. H. & HOPKINSON, J. 1915. The British Freshwater Rhizopoda and
Heliozoa. Vol. 3 : Rhizopoda 156 pp. 24 pis. The Ray Society, London.
CHARDEZ, D. 1956. Variations morphologiques et teratologie chez quelques Rhizopodes
Testaces. Biol. Jaarb. 23 : 265-276.
1960. Sur quelques Thecamoebiens du genre Trinema Dujardin. Bull. Inst. agron. Stns
Reck. Gembloux 28 : 266-271.
1961. Catalogue de Thecamoebiens de Belgie. Bull. Inst. agron. Stns. Rech. Gembloux
29 : 269-300.
• 1969. Contribution de la faune Thecamoebienne de 1'Islande (Protozoa, Rhizopoda,
Testacea). Bull. Inst. r. Sci. nat. Belg. 45 : 1-16.
1970. Sur Trinema lineare Penard 1890. Protozoa Rhizopoda Testacea. Revue verviet
Hist. nat. 27 : Nos 10 and 12.
DECLOITRE, L. 1961 (a). Materiaux pour une faune thecamoebienne du Maroc. Premiere
note. Butt. Soc. Sci. nat. Maroc. 41 : 117-119.
1961 (b). Materiaux pour une faune thecamoebienne du Maroc. Deuxieme note. The-
camoebiens des sols aeriens des palmiers de Marrakech. Bull. Soc. Sci. nat. Maroc, 41 : 121 -
136-
1962. Trinema lineare var. terricola nov. var. (Thecamoebiens). Int. Revue ges. Hydrobiol.
47 : 163.
ig8 R. H. HEDLEY AND C. G. OGDEN
DECLOITRE, L. 1964 (a). Especes, variet^s et anomalies dans le monde des Thecamoebiens.
Archs. Zool. exp. gen. 104 : 61-64.
— 1964(13). Thecamoebiens de la XI I erne Expedition Antarctique Francaise. Territoire des
Terres Australes et Antarctique Franfaises. Expeditions Polaires Fran9aises (Missions
Paul-Emile Victor). Publ. no. 259 : 1-47.
— I965(a). Amoebida testacea (Rhizopoda). Zoology Iceland 2 (i) : 1-58.
— 1965 (b). Contribution a la faune du Congo (Brazzaville). Mission A. Descarpentries et
A. Villiers. III. Rhizopodes, Thecamoebiens. Bull. Inst. fr. Afr. Noire, 27A : 165-184.
DUNKERLY, J. S. 1923. Encystation and reserve food formation in Trinema lineare, Penard.
Trans. R. Soc. Edinb. 53 : 297-300.
FA YARD, P. 1969. The Golgi apparatus. In Handbook of Molecular Cytology, Vol. 15 : 1130-
1155. North Holland, Amsterdam and London.
FEBVRE-CHEVALIER, C. 1971. Constitution ultrastructurale de Globigerina bulloides d'Orbigny,
1826 (Rhizopoda- Foraminif era) . Protistologica, 7 : 311-324.
FULLER, M. S. & CALHOUN, S. A. 1968. Microtubule-kinetosome relationships in the motile
cells of the Blastocladiales. Z. Zellforsch. mikrosk. Anat. 87 : 526-533.
GOLEMANSKY, V. 1971. Taxonomische und zoogeographische Notizen tiber die thekamobe
Fauna (Rhizopoda, Testacea) der Kiistengrundgewasser der sowjetischen Fernostkiiste
(Jananisches Meer) und der Westkiiste Kanada (Stiller Ozean). Arch. Protistenk. 113 :
235-249.
GRAAF, FR. de. 1956. Studies on Rotatoria and Rhizopoda from the Netherlands. I.
Rotatoria and Rhizopoda from the 'Grote Huisven'. Biol. Jaarb. 23 : 145-217.
GRIFFIN, J. L. 1972. Movement, fine structure, and fusion of pseudopods of an enclosed
amoeba, Difflugiella sp. /. Cell Sci. 10 : 563-583.
HARRIS, R. H., MARTIN, B. S. & OGDEN, C. G. 1972. Notes on the preparation of natural
history specimens for scanning electron microscopy. Bull. Br. Mus. nat. Hist. (Zool.)
24 : 223-228.
HEDLEY, R. H. & OGDEN, C. G. 1973. Biology and fine structure of Euglypha rotunda
(Testacea : Protozoa). Bull. Br. Mus. nat. Hist. (Zool.) 25 : 119-137.
HEDLEY, R. H., PARRY, D. M. & WAKEFIELD, J. ST. J. 1967. Fine structure of Shepheardella
taeniformis (Foraminifera : Protozoa). // R. microsc. Soc. 87 : 445-456.
HEDLEY, R. H. & WAKEFIELD, J. ST. J. 1969. Fine structure of Gromia oviformis (Rhizo-
podea : Protozoa). Bull. Brit. Mus. nat. Hist. (Zool.) 18 : 69-89.
HOOGENRAAD, H. R. & GROOT, A. A. DE 1940. Fauna van Nederland. Vol. 9 : Zoetwater-
rhizopoden en Heliozoen. Leiden. 302 pp.
I952(a). Thekamobe Moosrhizopoden aus Nordamerika. Arch. Hydrobiol. 47 : 229-262.
I952(b). Thekamobe Moosrhizopoden aus Asien. Arch. Hydrobiol. 47 : 263-287.
HRUBAN, Z. & RECHCIGL, M. 1969. Microbodies and related particles : morphology,
biochemistry, and physiology. Int. Rev. Cytol. Suppl. i. Academic Press, London.
296 pp.
KUFFERATH, H. 1932. Rhizopodes du Congo. Rev. Zool. Bot. Afr. 23 : 52-60.
LEADBEATER, B. & DODGE, J. D. 1967. An electron microscope study of nuclear and cell
division in a dinoflagellate. Arch. Mikrobiol. 57 : 239-254.
LEIDY, J. 1879. Freshwater Rhizopods of North America, in Vol. 12 : United States Geological
Survey of the Territories. Washington. 324 pp.
LOEBLICH, A. R. & TAPPAN, H. 1961. Suprageneric classification of the Rhizopodea. /.
Paleontology, 35 : 245-330.
MANTON, I., KOWALLIK, K. & VON STOSCH, H. A. 1970. Observations on the fine structure
and development of the spindle at mitosis and meiosis in a marine centric diatom (Litho-
desmium undulatum). IV. The second meiotic division and conclusion. /. Cell Sci.
7 : 407-443.
McCuLLY, E. K. & ROBINOW, C. F. 1972. Mitosis in heterobasidiomycetous yeasts. II.
Rhodosporidium sp. (Rhodotorula glutinis) and Aessosporon salmonicolor (Sporobolomyces
salmonicolor}. J. Cell Sci. 11: 1-31.
TRINEMA LINEARE (PROTOZOA) 199
MERCIER, M., LE BLANC, M., THOMAS, R. & CAMBAR, R. 1964. Observations en microscopie
61ectronique, sur la constitution de la theque de quelques Euglyphidae (Rhizopodes testaces) .
C.r. Acad. Sci. (Paris). 258: 5967-5968.
PENARD, E. 1890. Etudes sur les Rhizopodes d'eau douce. Mem. Soc. Phys. Hist. not.
Geneve, 31 : 1-230.
1902. Faune Rhizopodique du Bassin du Ldman. Geneva. 700 pp.
PERKINS, F. O. 1970. Formation of centriole and centriole-like structures during meiosis and
mitosis in Labyrinthula sp. (Rhizopodea, Labyrinthulida) . An electron-microscope study.
/. Cell Sci. 6 : 629-653.
PICKETT-HEAPS, J. D. 1968. Ultrastructure and differentiation in Chara (Fibrosa). IV.
Spermatogenesis. Aust. J. biol. Sci. 21 : 255-274.
1969. The evolution of the mitotic apparatus : an attempt at comparative ultrastructural
cytology in dividing plant cells. Cytobios, 3 : 257-280.
SCHWAB, D. 1969. Elektronenmikroskopische Untersuchung an der Foraminifere Myxotheca
arenilega Schaudinn. Z. Zellforsch. Mikrosk. Anat. 96 : 295-324.
STEPANEK, M. 1967. Testacea des Benthos der Talspere Vranov am Thayafluss. Hydro-
biologia, 29 : 1-66.
STOUT, J. D. & HEAL, O. W. 1967. Protozoa. Chap. 6, pp. 149-195, in Soil Biology. Aca-
demic Press, London and New York.
SZOLLOSI, D., CALARCO, P. & DONAHUE, R. P. 1972. Absence of centrioles in the first and
second meiotic spindles of mouse oocytes. /. Cell Sci. 11 : 521-541.
THOMAS, R. 1958. Observations sur le revetement des Trinema. Bull. Microsc. appl.
8 : 105-108.
THOMAS, R. & HOVASSE, R. 1962. Sur la constitution des theques des The"camoebiens. I.
Le genre Trinema et Trinema linear e Penard. Bull. Microsc. appl. 12 : 117-119.
TILNEY, L. G. 1971. How microtubule patterns are generated. The relative importance of
nucleation and bridging of microtubules in the formation of the axoneme of Rhaphidiophrys.
J. Cell. Biol. 51 : 837-854.
TILYEY, L. G. & GODDARD, J. 1970. Nucleating sites for the assembly of cytoplasmic micro-
tubules in the ectodermal cells of blastulae oiArbaciapunctulata. J. Cell. Biol. 46 : 564-575.
VOLZ, P. 1929. Studien zur Biologie der bodenbewohnenden Thekamoben. Arch. Protistenk.
68 : 349-406.
RONALD HENDERSON HEDLEY, D.Sc.
BRITISH MUSEUM (NATURAL HISTORY)
CROMWELL ROAD
LONDON SW7 5BD
COLIN GERALD OGDEN
BRITISH MUSEUM (NATURAL HISTORY)
CROMWELL ROAD
LONDON SW7 5BD
PLATE i
A. Aperture of T. lineare ; note the invagination produced by the curved shell-plates, x 7100
B. Apertural view of T. lineare showing the arrangement of large, circular shell-plates, x 2900
C. Two apertural plates ; the lower one shows a characteristic median fold. x 23 ooo
D. A preparation from which the organic cement has been removed illustrating the arrange-
ment of apertural and small shell-plates. x 8650
E. Lateral view of T. lineare illustrating the oblique position of the aperture. x 2900
F. Individual shell-plates ; note the fold in small shell-plate (arrowed). x n 500
Bull. Br. Mus. nat. Hist. (Zool.) 26, 3
PLATE i
PLATE 2
A-F. Abnormal variants of T. linear e ; note that the apparent imbrications are due possibly
to electrons penetrating the borders of the shell-plates, these are not seen when the
accelerating voltage on the scanning electron microscope is reduced.
A, B, D, E and F x 1000 C x 2000
G. Groups of three animals in 'rosette' formation. x 1400
Bull. Br. Mus. nat. Hist. (Zool.) 26, 3
PLATE
PLATE 3
A. Longitudinal section showing the pigment zone (pz), electron-dense inclusion (i), granular
endoplasmic reticulum (ger), nucleus (n) and Golgi apparatus (G). x 4300
B. Section showing various vacuoles containing bacteria (b), microbodies (m), and the probable
stages in the development of electron-dense particles (stages 1-3) -stage 4 is shown in
fig. A. x 15 950
C. Section showing vesicles containing fibrillar material (fm) concentrated in the centre with
radiating strands. x 23 900
D. Transverse section showing three contractile vacuoles (cv) which are at systole, nucleus (n)
and the concentrated mass of perinuclear endoplasmic reticulum (ger). x 7800
Bull. Br. Mus. nat. Hist. (Zool.) 26, 3
PLATE 3
PLATE 4
A. Longitudinal section showing the position and ovoid shape of the nucleus (n) at prophase,
Golgi apparatus (G), the cytoplasmic attachments to the shell and the concentrated mass
of perinuclear endoplasmic reticulum (ger.). x 7800
B. Cross-section of the posterior region of nucleus (n) showing proximity of numerous micro-
tubules (arrowed), Golgi apparatus (G) and coated vesicles (eves). x 20 750
C. Section showing microtubule (mt) close to nuclear membrane (nm) and the terminal position
of the microtubule-organizing-centre, MTOC (arrowed). x 30 100
D. Section immediately posterior to the nucleus showing nuclear envelope (ne) and numerous
microtubules (mt) converging onto microtubule-organizing-centre, MTOC (arrowed).
x 56 700
E. Section through two microbodies showing tubular elements. x 44 600
Bull. Br. Mus. nat. Hist. (Zool.) 26, 3
PLATE
PLATE 5
A. Vesicle with double-unit membrane and enclosed by smooth endoplasmic reticulum (ser)
(see also fig. D). x 30 100
B. Transverse section through anterior region of cytoplasm showing numerous cement bodies
(cm) and thin cytoplasmic strands in the space between cytoplasm and shell. x 10 400
C. Section showing Golgi apparatus (G) with concentrations of fibrillar material (fm) in the
saccules, nucleus (n) and pellicular microtubules (pmt). x 22 300
D. Probable origin of electron-transparent vesicles in the region between granular endoplasmic
reticulum (ger) and Golgi apparatus (G) (see also fig. A). x 20 050
E. Cement bodies with fibrillar matrix and electron-dense centres. x 40 100
Bull. Br. Mus. nat. Hist. (Zool.) 26, 3
PLATE 5
PLATE 6
A. Section through equatorial region of nucleus (n) showing surrounding microtubules
(arrowed). x 30 100
B. Fusion of electron-dense inclusion with reserve shell-plate vesicle. x 22 300
C. Section through apertural region showing internal microtubules (mt), apertural plate (ap)
and pseudopodial trunk (pt) which is relatively structureless. x 15 ooo
D. Stack of reserve shell-plates each in separate membrane-bound vesicle. x 22 300
E. Portion of shell showing shell-plates (arrowed) with concave surface outwards. x 4300
Bull. Br. Mm. nat. Hist. (Zool.) 26, 3
PLATE 6
IP
PLATE 7
A. Glutaraldehyde-fixed and unstained section showing the arrangement of reserve shell-plates.
x 4300
B. Section of two animals directly apposed ; note the larger individual contains numerous
reserve plates and a well-defined pigment zone (pz). x 2100
C. Apertural region of a 'rosette' group of four individuals ; note the presence of vacuoles
containing bacteria (b) and the numerous cement bodies (cm). x 4300
D. Section through two united adult specimens showing cement bodies (cm) and pseudopodial
extensions ; note the structureless nature of the pseudopodia. x 7800
E. A 'rosette' group of three animals showing the diffuse nature of the cytoplasm. x 2800
F. Abnormal individual with two nuclei. x 5750
Bull. Br. Mus. nat. Hist. (Zool.) 26, 3
PLATE 7
A LIST OF SUPPLEMENTS
TO THE ZOOLOGICAL SERIES
OF THE BULLETIN OF
THE BRITISH MUSEUM (NATURAL HISTORY)
1. KAY, E. ALISON. Marine Molluscs in the Cuming Collection British Museum
(Natural History) described by William Harper Pease. Pp. 96 ; 14 Plates.
1965. (Out of Print.) £3-75."
2. WHITEHEAD, P. J. P. The Clupeoid Fishes described by Lacepede, Cuvier and
Valenciennes. Pp. 180 ; n Plates, 15 Text-figures. 1967. £4.
3. TAYLOR, J. D., KENNEDY, W. J. & HALL, A. The Shell Structure and Mineralogy
of the Bivalvia. Introduction. Nuculacea-Trigonacea. Pp. 125 ; 29 Plates
77 Text-figures. 1969. £4.50.
4. HAYNES, J. R. Cardigan Bay Recent Foraminifera (Cruises of the R.V. Antur)
1962-1964. Pp. 245; 33 Plates, 47 Text-figures. 1973. £10.80.
Printed in Great Britain by John Wright and Sons Ltd. at The Stonebridge Press, Bristol BS4 sNU
22 JULJ
THE INDIGENOUS EARTHWORMS
(MEGASCOLECIDAE : OLIGOCHAETA)
OF TASMANIA
B. G. M. JAMIE SON
BULLETIN OF
THE BRITISH MUSEUM (NATURAL HISTORY)
ZOOLOGY Vol. 26 No. 4
LONDON: 1974
THE INDIGENOUS EARTHWORMS
(MEGASCOLECIDAE : OLIGOCHAETA)
OF TASMANIA
BY
BARRIE GILLEAN MOLYNEUX JAMIESON
University of Queensland
Pp. 201-328 ; 10 Plates ; 32 Text-figures
BULLETIN OF
THE BRITISH MUSEUM (NATURAL HISTORY)
ZOOLOGY Vol. 26 No. 4
LONDON: 1974
THE BULLETIN OF THE BRITISH MUSEUM
(NATURAL HISTORY), instituted in 1949, is
issued in five series corresponding to the Departments
of the Museum, and an Historical series.
Parts will appear at irregular intervals as they
become ready. Volumes will contain about three or
four hundred pages, and will not necessarily be
completed within one calendar year.
In 1965 a separate supplementary series of longer
papers was instituted, numbered serially for each
Department.
This paper is Vol. 26, No. 4 of the Zoological
series. The abbreviated titles of periodicals cited
follow those of the World List of Scientific Periodicals.
World List abbreviation :
Bull. Br. Mus. nat. Hist. (Zool.).
Trustees of the British Museum (Natural History), 1974
TRUSTEES OF
THE BRITISH MUSEUM (NATURAL HISTORY)
Issued 22 May, 1974 Price £6.75
THE INDIGENOUS EARTHWORMS
(MEGASCOLECIDAE : OLIGOCHAETA)
OF TASMANIA
By B. G. M. JAMIESON
CONTENTS
Page
SYNOPSIS. ........... 204
INTRODUCTION ........... 205
EXPLANATION OF DESCRIPTIVE FORMAT AND TERMINOLOGY . . . 207
SYSTEMATICS ........... 212
KEY TO THE GENERA OF TASMANIAN MEGASCOLECIDAE . . . 212
Subfamily ACANTHODRILINAE . . . . . . . .212
Tribe Acanthodrilini
Genus Rhododrilus Beddard 1889 . . . . . . 212
Rhododrilus littoralis sp. n. . . . . . . . 213
Subfamily MEGASCOLECINAE . . . . . . . . 216
Tribe Perionychini .........
Genus Perionychella Michaelsen, 1907 . . . . . 216
Subgenus Perionychella Michaelsen, 1907 .... 220
Subgenus Vesiculodrilus Jamieson, 1973 .... 221
P. (Perionychella) capensis sp. n. . . . . . . 225
P. (Perionychella) hickmani sp. n. . . . . . 226
P. (Perionychella) irregularis (Spencer, 1895) • • • 22^
P. (Perionychella) kershawi sp. n. ..... 229
P. (Perionychella) weldboroughi sp. n. . . . . . 230
P. (Vesiculodrilus) bithecata sp. n. . . . . . 233
P. (Vesiculodrilus) dilwynnia (Spencer, 1895) • • • 234
P. (Vesiculodrilus) evansi sp. n. . . . . . . 236
P. ( Vesiculodrilus) glandifera sp. n. ..... 237
P. (Vesiculodrilus) hellyeri sp. n. . . . . . . 238
P. (Vesiculodrilus) hobartensis (Spencer, 1895) . . . 241
P. (Vesiculodrilus) lacustris (Stephenson, 1924) . . . 245
P. (Vesiculodrilus) montisarthuri sp. n. .... 246
P. (Vesiculodrilus) mortoni (Spencer, 1895) .... 247
P. ( Vesiculodrilus) obliquae sp. n. . . . . . 250
P. (Vesiculodrilus) pedderensis sp. n. . . . . 251
P. (Vesiculodrilus) tunnackensis sp. n. . . . . . 253
P. (?) bassana (Spencer, 1895) ...... 253
P. (?) decatheca (Michaelsen, 1910) ..... 255
P. (?) ellisi (Spencer, 1895) ...... 257
P. (?) moroea (Spencer, 1895) ...... 257
P. (?) richea (Spencer, 1895) ...... 258
P. (?) scolecoidea (Spencer, 1895) ...... 259
Genus Graliophilus Jamieson, 1971 . . . . . . 259
Graliophilus tripapillatus sp. n. . . . . . . 261
Genus Pinguidrilus gen. nov. ....... 263
Pinguidrilus tasmanianus (Fletcher, 1888) .... 263
Genus Woodwardiella Stephenson, 1925 ..... 265
W oodivardiella ? tessellatus (Spencer, 1895) .... 266
204 B. G. M. JAMIESON
Genus Cryptodrilus Fletcher, 1887 ...... 266
Cryptodrilus albertisi (Cognetti, 1910) ..... 270
Cryptodrilus avesiculatus sp. n. . . . . . . 271
Cryptodrilus brunyensis sp. n. ...... 273
Cryptodrilus dartnalli sp. n. . . . . . . 275
Cryptodrilus enter onephricus sp. n. ..... 277
Cryptodrilus ? officeri Spencer, 1895 ..... 280
Cryptodrilus polynephricus Spencer, 1895 .... 282
C. polynephricus polynephricus Spencer, 1895 . . . 284
C. polynephricus urethrae subsp. n. . . . . . 288
C. p. polynephricus ad urethrae . . . . . 291
Cryptodrilus simsoni Spencer, 1895 ..... 293
Genus Pseudocryptodrilus Jamieson, 1972 ..... 296
Pseudocryptodrilus acanthodriloides sp. n. 298
Tribe Dichogastrini
Genus Hickmaniella gen. nov. ....... 300
Hickmaniella opisthogaster sp. n. ..... 301
Tribe Megascolecini
Genus Oreoscolex Jamieson, 1973 ...... 302
Oreoscolex bidiverticulatus sp. n. . . . . . . 305
Oreoscolex campestris (Spencer, 1895) . . . . . 307
Oreoscolex irregularis (Spencer, 1895) ..... 309
Oreoscolex leai (Michaelsen, 1910) . . . . . 310
Oreoscolex longus sp. n. . . . . . . . 311
Oreoscolex peculiaris sp. n. . . . . . . . 313
Oreoscolex sexthecatus sp. n. . . . . . . 315
Oreoscolex wellingtonensis (Spencer, 1895) . . . . 317
Genus Megascolex Templeton, 1844 . . . . . . 318
Megascolex burniensis sp. n. . . . . . . 319
Megascolex montisarthuri sp. n. . . . . . . 321
Megascolex tasmanicus (Spencer, 1895) .... 324
ACKNOWLEDGEMENTS . . . . . . . . . 326
REFERENCES ........... 327
SYNOPSIS
This paper raises the number of indigenous earthworms known from Tasmania to forty-eight,
in ten genera. The genus Perionychella Michaelsen, i9O7a, is re-established and extended to
include lumbricine species with or without nephridial bladders and perichaetine, vesiculate
species. The other species resemble the type-species in being perichaetine with avesiculate
nephridia. All vesiculate species are included in the subgenus Vesiculodrilus Jamieson, I973a,
initially established as a genus for lumbricine, vesiculate species. Perionychella absorbs all
Australian species of Diporochaeta, which is now regarded as endemic in New Zealand, and a
large portion of the residue of Plutellus, a genus which has been revised elsewhere. A check
list of the two subgenera of Perionychella is given which lists twenty-three Tasmanian species.
Perionychella is therefore the dominant genus in numbers of species in Tasmania, as in Victoria.
Twelve of the Tasmanian Perionychella species are new species and a further five species are re-
described, three of these from new material. Only P. scolecoidea (Spencer, 1895) was previously
referred to the genus (Michaelsen, i9O7a).
The remaining earthworm fauna consists of a new species and the first Tasmanian record of
Rhododrilus ; a new species, also the first record, of Graliophilus ; the new genus Pinguidrilus
erected for Notoscolex tasmanianus Fletcher ; a species of Woodwardiella which, like the Gralio-
philus species, may be referable to Perionychella ; eight species of Cryptodrilus, of which four
TASMANIAN EARTHWORMS 205
and one subspecies are new ; a new species of the formerly monotypic Victorian genus Pseudo-
cryptodrilus ; a new monotypic genus Hickmaniella ; eight species of Oreoscolex, of which four
are new, and three species provisionally referred to Megascolex, of which two are new. There
is almost total specific endemicity in Tasmania, a single species, Megascolex tasmanicus, being
questionably a junior synonym of a mainland species. In contrast only two of the ten genera
are endemic, the remaining eight occurring on the mainland and five of these in Victoria, the
region with the highest zoogeographic affinity with Tasmania. Noteworthy features of the
Tasmanian fauna are the high species diversity relative to the area of the island (paralleled,
however, in Victoria), the high proportion of primitive, chiefly perionychin genera and species ;
and radiative evolution in the excretory system within the perionychin genus Cryptodrilus
which has resulted in convergent acquisition of grades of organization characteristic of the tribes
Dichogastrini and Megascolecini, these tribes being poorly represented in Tasmania.
INTRODUCTION
THE earthworms of Tasmania have not been reviewed since the original work of
Spencer (1895). As a result of the kind cooperation of Mr A. J. Dartnall of the
Tasmanian Museum and Art Gallery, Hobart, and of Dr J. L. Hickman of the
Zoology Department, University of Tasmania, in placing their oligochaete collec-
tions at the author's disposal, the taxonomic and zoogeographic relationships of the
Tasmanian earthworm fauna have been investigated and the number of known
species from that island has more than doubled. The present paper is a report of
these investigations.
Before the study was begun, the known oligochaete fauna of Tasmania consisted
of six species of aquatic microdriles, twenty-three indigenous species of megascolecid
earthworms, and introduced earthworms of the family Lumbricidae. The microdriles
are the cosmopolitan and probably anthropochorous Lumbriculus variegatus (Muller,
1774) (Lumbriculidae) ; the endemic Telmatodrilus multiprostatus and T. pectinatus,
both of Brinkhurst 1971 ; Antipodrilus davidis (Benham, 1907) which also occurs in
Australia and New Zealand ; the cosmopolitan Limnodrilus udekemianus Claparede,
1862 (all Tubificidae) ; and the species inquirenda, Tasmaniaedrilus tasmaniaensis
Goddard and Malan, 1913 (Phreodrilidae) .
The previously known indigenous earthworms are listed below under their original
binomina. Generic identifications were rarely correct but the species are reassigned
and all previous generic attributions are listed in the specific synonymies.
Species of Spencer, 1895 :
Cryptodrilus irregularis Megascolides simsoni
polynephricus bassanus
mortoni Perichaeta tasmanica
hobartensis moroea
campestris richea
tesselatus dilwynnia
insularis scolecoidea
ellisi irregularis
wellingtonensis
officeri
144° 145°
146°
147°
148°
145°
146
147°
148°
FIG. i. Tasmanian records of Perionychella species. 0, present; Q. absent; ba, P.
bassana ; bi, P. (V.) bithecata ; c. P. (P.) capensis ; d, P. (V.) dilwynnia ; el, P. ellisi ;
ev, P. (V.) evansi ; g, P. (V.) glandifera ; he, P. (V.) hellyeri ; hi, P. (V.) hickmani ; ho,
P. (V.) hobartensis ; i, P. (P.) irregularis ; k, P. (P.) kershawi ; 1, P. (P.) lacustris ; mon,
P. (V.) montisarthuri ; mor, P. moroea ; mort, P. (F.) mortoni ; o, P. (F.) obliquae ; p,
P. (F.) pedderensis ; r, P. richea ; s, P. scolecoidea ; t, P. (F.) tunnackensis ; w, P. (P.)
weldboroughi. Not mapped : P. decatheca. (Numerals pertain to map references in text.)
TASMANIAN EARTHWORMS 207
Further species :
Notoscolex tasmanianus Fletcher, i888b Notoscolex leai Michaelsen, 1910
Megascolides albertsi Cognetti, 1910 Perionyx lacustris Stephenson, 1924
Plutellus decatheca Michaelsen, 1910
EXPLANATION OF DESCRIPTIVE FORMAT AND TERMINOLOGY
Those unfamiliar with oligochaete morphology and terminology are referred to
Stephenson (1930) for what remains the most comprehensive account of this subject.
It should be noted, however, that the taxonomy employed by Stephenson is now
largely obsolete. In the following explanation of the format used in describing
species in the present work some additional clarification of terminology will be given.
The descriptive format has been kept brief. Some characters, e.g. septal thicken-
ing, which are customarily described but are of limited value are omitted in favour of
fuller treatment of systems of major significance in taxonomy which are frequently
neglected, such as the excretory system. Some information retrievable from the
illustrations, e.g. spermathecal proportions and dimensions, is excluded from the
text.
These preliminary accounts are primarily limited to a holotype and paratype but
variation, particularly in the highly species-specific genital fields, is appended. For
species with eight setae per segment, setal ratios, which are population-specific
(Jamieson and Bradbury 1972), are given for several specimens where these are
available. Specimens are sexual, and usually clitellate, unless otherwise stated.
Characters employed, in sequence are as follows.
Dimensions : 1 = length, w = midclitellar, or greatest width where accounts are
drawn only from the literature, s = number of segments. These and other numerical
data are given in the sequence holotype and paratype I. Prostomium whether
epilobous or tanylobous and canaliculate (with middorsal longitudinal groove or not) .
Body form is mentioned only where the cross section departs from an approximate
circle. The anus is terminal in all cases. Position of the first dorsal pore : deter-
mined by parting the intersegmental furrows with a probe or observing expulsion of
alcohol when the specimen is immersed temporarily in water.
Setal ratios (intersetal distances) : the four setae on each side in a segment are
designated, in the ventral to dorsal direction, a, b, c and d. The distances aa : ab :
be : cd : dd : dc : cb : ba, measured by camera lucida, are expressed in the tables rela-
tive to a constant total (periphery, u) of 100. Direct comparison of ratios between
individuals without recourse to additional computations is also facilitated by stating
the distances as ratios relative to one of the distances (ab) expressed as unity. Where
there are more than eight setae per segment (the perichaetine condition), the longi-
tudinal setal rows are designated, in the ventral to dorsal direction, setal line I, 2,
3, etc. but the ventral two rows are alternatively termed a and b. The dorsalmost
row, so as readily to indicate this position, is termed setal line z and the penultimate
and next ventral rows are y and x respectively. Setae commence on segment II
in all described species.
208
B. G. M. JAMIESON
144° 145"
146°
147°
148°
40
O polynephricus polynephricus
O polynephricus urethrae
p. polynephricus ad urethrae
enteronephricus
avesiculatus,
dartnalli
145°
146°
147°
148°
FIG. 2. Tasmanian records of Cryptodrilus species and subspecies. 0, present ;
Oi absent. (Numerals pertain to map references in text.)
TASMANIAN EARTHWORMS 209
Nephropores in holonephric Tasmanian species are invariably near the anterior
margins of their segments. Their position relative to the setal lines is recorded.
The clitellum may be annular, i.e. completely encircling the body, or saddle-shaped,
i.e. interrupted ventrally throughout its extent. Segmentation indicated as, for
instance, XIII-XVII, indicates that the clitellum extends throughout the mentioned
segments, in this case from the anterior border of the thirteenth to the posterior
border of the seventeenth segment. A fractional indication, such as ^XIII-^XVII,
indicates the position on the segment at which the clitellum begins or ends, in this
case the clitellum would include the posterior two-thirds of XIII and the anterior
third of XVII. The male genital field is described next, including the position of the
porophores bearing the combined pores of the vasa deferentia and the pair of prostate
glands, and the form and configuration of accessory genital markings in their vicinity.
These markings are usually protuberant, and where an elliptical marking has a dis-
tinct central area are described as eye-like. Other accessory genital markings may
be present near the spermathecal pores. Female pores are without exception paired
and anteromedian to setae a of segment XIV in Tasmanian species of Perionychella
but their position is recorded for each species pending description of at least the
entire Australian fauna. Spermathecal pores are intersegmental or segmental and
paired or single in Tasmanian species ; their number, intersegments or segments
occupied and location relative to setal lines are indicated.
The internal anatomy (examined after middorsal longitudinal incision) is described
in the following sequence. Blood vascular system : in all Tasmanian species the
dorsal vessel is continuous onto the pharynx and a subneural vessel is absent ; the
segment occupied by the last hearts is given and the statement that they are latero-
oesophageal refers to their origin from the supra-oesophageal vessel (or the oesophageal
plexus) . A latero-oesophageal heart usually also receives a connective from the dorsal
vessel but this was not always demonstrable in the material examined. The extent
of the supra-oesophageal vessel is indicated where precisely determinable ; the stated
degree of development of the vessel refers to the maximum development in its course.
The account of the alimentary canal gives the segmental location and degree of
development of the gizzard ; modification of the oesophagus, including, where
feasible, dilatation and internal specializations such as rugae and lamellae, and the
position, if present, of extramural oesophageal (calciferous ?) 'glands' ; the segment
of origin of the intestine (determined by intestinal expansion and, where possible,
identification of the oesophageal valve between oesophagus and intestine) and
presence or absence of a typhlosole. Special muscular thickening of the intestine
is present only in Hickmaniella. Intestinal caeca are absent in all Tasmanian
species of the Megascolecidae.
The types of nephridia found in oligochaetes are described by Jamieson (igyia).
Nephridia are holonephridia in all species of Perionychella, Graliophilus, Pinguidrilus,
Woodwardiella and Rhododrilus but in some of these holonephric species anterior
nephridia are tufted, having numerous tubules, with simple or less commonly com-
posite or multiple ducts ; it remains to be ascertained whether the tufted nephridia
always retain the preseptal funnel (the stomate condition) of the simple nephridia.
The most anterior segment in which the nephridia were seen to have preseptal funnels
144° 145°
B. G. M. JAMIESON
146° 147°
148°
145'
146
147°
148°
FIG. 3. Tasmanian records of Oreoscolex species. 0, present ; Q. absent. Not mapped :
Oreoscolex leai. (Numerals pertain to map references in text.)
TASMANIAN EARTHWORMS 211
is stated but it is not to be assumed that funnels were absent from more anterior
nephridia as demonstration of funnels is difficult, particularly where disruption of
other organ systems is undesirable. Presence or absence of ectal bladders, the
vesiculate or avesiculate condition, and their form is recorded, as is the occurrence
of diverticula on the bladders.
The anterior male organs are described next ; whether holandric (testes and funnels
paired in each of X and XI) or metandric (these organs in XI only) and iridescence of
the funnels is noted (an indication, albeit only partly reliable, that the organs are
functional). Location of seminal vesicles is noted.
Female organs in Tasmanian megascolecids are metagynous (ovaries and funnels
in XIII) as in all but a few species of the suborder Lumbricina of Brinkhurst and
Jamieson 1971. This and the presence or absence of ovisacs (dependent from the
anterior wall of XIV) is noted for each species.
The form and extent of the prostate glands are recorded ; variation in the form of
these is discussed in the introductory discussion of Perionychella. Penial setae :
occurrence of these enlarged copulatory setae in segment XVIII, associated with the
male pores, is noted. Lengths of well-developed setae are given and their mor-
phology as revealed by the light microscope and scanning electron microscope is
briefly described. The penial setae in Tasmanian species of Perionychella are un-
usually slender and are at most only weakly ornamented, a contrast with other
genera on the island, and reproduction of the electron microscope photographs is
rarely merited. In several species of Perionychella complete or undamaged setae
were not recoverable and the morphology could not be elucidated. A study of the
degree of inter-specific variation in their ultra-morphology is required.
Finally, the number of spermathecae and the form and number of their diverticula
are noted, detailed information being retrievable from the illustrations.
Material examined : Under this heading are given, in sequence, the locality with
habitat data, the longitude and latitude of the locality, the grid reference on the
map, the collector's name, the date of collection, and the institutions in which the
holotype, paratypes and additional specimens are lodged. Letters and numbers
following the abbreviation for the institution are the registration numbers of the
institution. Lectotypes and paralectotypes are designations of Jensz and Smith
(1969). Abbreviations for the institutions are given below and listed with them
are the names of collectors.
Collectors Institutions in which specimens
are lodged
Mr E. A. Bradbury AM Australian Museum, Sydney
Mr A. Dartnall BJ Author's collections
Dr J. W. Evans BM(NH) British Museum (Natural
Dr J. L. Hickman History)
Professor V. V. Hickman NMV National Museum of Victoria
Mr W. Jackson TM Tasmanian Museum,
Dr B. G. Jamieson Hobart
Mr R. W. Kerr
B. G. M. JAMIESON
Collectors (cont.)
Mr R. C. Kershaw
Mr C. D. King
Mr W. Radford
Tasmanian Biological Survey
Mr D. Tyler
SYSTEMATICS
The generic and specific descriptions are arranged under the relevant subfamilies
and tribes but a key to all Tasmanian genera, ignoring suprageneric ranks, will be
given first. This key is intended only for known Tasmanian species of the genera
concerned and should not be used as a general diagnostic key for these genera in
other regions.
KEY TO THE GENERA OF TASMANIAN MEGASCOLECIDAE
1 Combined male and prostatic pores a pair on XVII (i6th setigerous segment).
(Holonephric ; bladders absent) ... . RHODODRILUS
- Combined male and prostatic pores a pair on XVIII (lyth setigerous segment).
(Holonephric or meronephric ; bladders present or absent) .... 2
2 Nephridia one pair per segment (holonephridia) in midbody ..... 3
- Nephridia more than one pair per segment (meronephridia) in midbody ... 6
3 Nephridial bladders present .......... 4
- Nephridial bladders absent .......... 5
4 Nephridia discharging in irregular rows, from c to near middorsum. Spermathecae
adiverticulate PINGUIDRILUS
- Nephridia discharging in a straight line, in the vicinity of c lines. Spermathecae
diverticulate . .... PERIONYCHELLA (VESICULODRILUS)
5 Intestinal typhlosole present GRALIOPHILUS (part ; see also WOODWARDIELLA)
- Intestinal typhlosole absent . . . PERIONYCHELLA (PERIONYCHELLA)
6 Nephridia 2 or 3 on each side except at the caudal extremity which is holonephric
PSEUDOCR YPTODRILUS
- Nephridia more than 2 per segment throughout ....... 7
7 Setae more than 8 per segment ..........
Setae 8 per segment .........••• 9
8 A gizzard present at the beginning of the intestine. Oesophageal gizzard also
present HICKMANIELLA
- Oesophageal gizzard present. Intestinal gizzard absent . . MEGASCOLEX
9 Caudal nephridia with numerous funnels on each side on the anterior face of each
septum. Bladders absent ....... OREOSCOLEX
- Caudal nephridia with at most a single funnel on each side on the anterior face of each
septum. Bladders often present CRYPT ODRILUS
Subfamily ACANTHODRILINAE
Tribe ACANTHODRILINI s. Jamieson, igjiz
Genus RHODODRILUS Beddard, 1889
DIAGNOSIS. Setae 8 per segment. Prostatic pores i pair, on XVII (in one species
on XVI) ; male pores combined with the prostatic pores or i pair, usually on XVII
TASMANIAN EARTHWORMS 213
close to them, sometimes anterior on XVIII, occasionally associated with the pos-
terior ends of seminal grooves. Female pores i pair, on XIV. Spermathecal pores
1-5 pairs, the last pair anterior to 9/10. Nephropores in a single series on each
side. Last hearts in XII or XIII. Gizzard single, occupying i or 2 segments in
the region of V-VIII, usually strongly developed, occasionally small. Holonephric
throughout or exceptionally with anterior meronephridia. Nephridial vesicles of
various forms or absent. Calciferous glands present or absent. Holandric ;
gymnorchous. Prostates tubular ; penial setae usually present. Spermathecae
diverticulate.
TYPE-SPECIES. Rhododrilus minutus Beddard, 1889.
DISTRIBUTION. New Zealand and neighbouring islands. Two species in Austra-
lia : one in Queensland, the other in Tasmania.
TASMANIAN SPECIES. Rhododrilus littoralis sp. n.
REMARKS. The above generic definition is derived from that of Jamieson (igjib).
In the latter work attention was drawn to the unsatisfactory distinction between
Rhododrilus and Microscolex. Although the majority of species of Microscolex differ
from Rhododrilus in having two pairs of prostates, the type-species, M. phosphoreus
(Duges), like M. dubius (Fletcher), has a single pair of prostates, as in Rhododrilus,
and has been shown by Lee (1970) to have strong phenetic affinities with the latter
genus (including the type-species, R. minutus) over which it has chronological
priority. It is possible that Rhododrilus will have to be synonymized with Micro-
scolex and new genera erected for some of the species of the present Microscolex-
Rhododrilus complex which fall outside a redefined Microscolex. The following new
Tasmanian species is provisionally placed in Rhododrilus because, like some other
species of this genus, including the type-species, it lacks nephridial bladders whereas
these are invariably present in Microscolex. In location of nephropores shortly
below c lines it resembles Microscolex dubius, however, and differs from Rhododrilus
in which nephropores are in b or c.
Rhododrilus littoralis sp. n.
Fig. 4, 26C, 32E ; Plate 60. Table i
1 = 65, 74 mm, w = 2 mm, s = 103, 105.
Setal annuli strongly protuberant, excepting the anterior preclitellar and extreme
caudal regions, and demarcated posteriorly by a groove equalling the intersegmental
furrows in distinctness. Prostomium epilobous f, \, open, margins posteriorly
convergent but indistinct. Dorsal pores weakly developed behind the clitellum,
possibly present as rudiments in the immediate preclitellar segments. Setae 8 per
segment, commencing on II, in regular longitudinal lines throughout ; setae a and b
absent (replaced by penial setae) in XVII. Nephropores not externally visible.
Clitellum externally indistinct, annular, £XIII-f XVIII (= 5| segments) from
dorsal incision (holotype) ; weakly developed but limits indeterminable in paratype i.
Male pores in XVII, in b on small whitish papillae which fill ab transversely and
2I4
144° 145°
B. G. M. JAMIESON
146° 147°
148°
145°
146°
147°
148°
FIG. 4. Records of Tasmanian megascolecid earthworms excluding Perionychella, Crypto-
drilus and Oreoscolex species. •, present ; Q, absent. (Numerals pertain to map
references in text.)
TASMANIAN EARTHWORMS 215
occupy the posterior two-thirds of the segment. Accessory genital marking a
transverse pad in 19/20, extending laterally shortly beyond b lines, and longitudinally
abutting on the setal annuli of XIX and XX ; strongly tumid and transversely
bisected by a furrow corresponding with but not peripherally continuous with inter-
segmental furrow 19/20 (holotype) ; genital marking not certainly detectable in
paratype I. Female pores a pair on XIV, midway between the setal arc and an-
terior margin, very slightly lateral of a lines (paratype i) ; not certainly demonstrable
but apparently concealed in the anterior furrow delimiting the setal annulus in
holotype. Spermathecal pores minute, concealed (holotype) or visible with difficulty
(paratype i) in intersegments 7/8 and 8/9, in b lines.
TABLE i
Intersetal distances in segment XII in Rhododrilus littoralis
mm
aa ab be cd dd dc cb ba
holotype 0-7 0-4 0-9 0-8 1-8 0-9 0-8 0-4
paratype I 0-8 0-5 0-9 0-8 1-8 0-8 0-9 0-5
standardized (u — 100)
aa ab be
holotype
paratype i
mean
interval/aft
Last hearts in XIII (those in X-XIII latero-oesophageal ; with connectives to
the dorsal and supra-oesophageal vessel) . Supra-oesophageal vessel in VIII ? -XIV ;
moderately developed. Gizzard small and easily compressible, though muscular,
in V (holotype) ; rudimentary in paratype i. Oesophagus lacking calciferous
glands ; especially vascularized and with low internal rugae in VIII-XVI ; slender
and less vascular in XVII and XVIII ; intestine commencing, with dilatation, in
XIX ? (holotype, paratype i) ; the alimentary canal in XVII and XVIII possibly
intestinal, however ; typhlosole absent. Nephridia st ornate holonephridia, com-
mencing in II, preseptal funnel (demonstrated in intestinal region) large, in b line ;
avesiculate duct entering parietes one-third be below c. Holandric ; gymnorchous ;
testes large, much branched ; sperm funnels small, faintly iridescent ; seminal
vesicles racemose, in IX, XI and XII increasing in size posteriorly. Metagynous
(ovaries with many egg-strings) ; ovisacs present (paratype i) ; not detectable in
holotype. Prostates one pair, very slenderly tubular, in XVII-XXIII (holotype) -
XXIV (paratype i) ; vas deferens joining the muscular duct well ectal of the gland.
Penisetal follicles conspicuous, as wide as the prostates and almost as extensive,
extending from XVII-XXII. Penial setae filiform, sinuous ; the ectal extremity
depressed spatulate ; ornamentation absent ; length = 2-8-3-7 mm, general width
of shaft = ii /mi (paratype i). Spermathecae 3 pairs, in VII, VIII and IX, each
with ovoid ampulla and shorter digitiform (uninseminated) diverticulum on the
aa
ab
be
cd
dd
dc
Cb
ba
10-00
5'79
13-16
12-63
27-36
13-16
12-11
5'79
11-22
6-63
13-27
n-73
26-02
11-22
12-76
7-14
10-61
6-21
13-22
12-18
26-69
12-19
12-44
6-47
1-71
I-OO
2-13
1-96
4-30
I-96
2-00
1-04
216 B. G. M. JAMIESON
short, poorly demarcated duct ; the spermatheca and diverticulum almost sessile
on the body wall.
Genital markings. Of 13 paratypes, n have the genital field described for the
holotype, with a single accessory genital marking, sometimes only faintly indicated,
at 19/20 ; 2 lack the marking but are less mature than the others.
MATERIAL EXAMINED.
Eagle Hawk Neck, in kelp on rocks, I47°55'E. 43°oo'S. Map. 19/2, Dr J. L. Hickman, Apr
1956 - holotype TM : K39& ; paratypes I and 4 BM(NH) : 1973.2.53-54 ; paratypes 2 and 8
BJ : T77~78 ; paratypes 3, 5-7 TM : 1^397-400. Hobart, and 'Straham' [possibly Strahan,
west coast of Tasmania?] i47°2o'E. 42°5o'S. Map, 14/5 collector ? AM: 62628 - paratypes
9-13, AM : W5339-5343.
REMARKS. R. liitoralis differs from all other species of the genus in the con-
figuration of the male genital field. The other Australian species, R. queenslandica,
differs among other respects in having only one pair of spermathecae.
Subfamily MEGASGOLEGINAE
Tribe PERIONYCHINI Jamieson,
Genus PERIONYCHELLA Michaelsen, i907a emend.
DIAGNOSIS. Setae 8 or more per segment. Combined pores of vasa deferentia
and tubular or tubuloracemose prostates on XVIII. Gizzard in V, rarely VI or VII
(well developed to vestigial) ; typhlosole absent. Extramural calciferous glands
usually absent, and never paired, but the oesophagus often vascularized and intern-
ally rugose, especially in XIV and XV. Nephridia stomate, exonephric holo-
nephridia with or without bladders ; the anterior nephridia sometimes tufted.
Spermathecae I to 5 pairs.
DESCRIPTION. Terrestrial or rarely aquatic worms. Form circular in cross
section or depressed ; anus terminal ; with or without pigmentation. Dorsal pores
present or absent, usually commencing in 3/4 or 4/5. Setae 8 per segment through-
out (lumbricine condition) or becoming more numerous (perichaetine condition)
after a varying number of anterior segments, or perichaetine throughout. Nephro-
pores in a single series throughout in the vicinity of c lines (where setae are lumbricine)
or their equivalent in perichaetine species. Posterior limit of clitellum shortly
anterior to the male pores or (P. (P.) canaliculatus and phalacrus) including these.
A pair of pores of the combined vasa deferentia and a pair of tubular, tubuloracemose
or racemose prostates on XVIII. Female pores paired or exceptionally single mid-
ventral, in XIV. Spermathecal pores I to (usually) 5 pairs in front of or usually
including intersegment 8/9. Accessory genital markings present in the vicinity of
the male pores and sometimes of the spermathecal pores.
Last hearts in XII or (P. (V.) grandis) XIII. Supra-oesophageal vessel (always ?)
restricted to the post-gizzard oesophagus, moderately to well developed, sometimes
not differentiated from the oesophageal plexus. Subneural vessel absent. Gizzard
in V or exceptionally in VI, frequently only moderately developed or vestigial (in
TASMANIAN EARTHWORMS 217
some northern Queensland species in VII). Oesophagus commonly with internal
rugae, these often best developed in XIV and XV ; sometimes with internal laminae
which exceptionally line one or more annular, unpaired, axial calciferous glands.
Intestine commencing in XVI-XX, usually XVII ; typhlosole, muscular thicken-
ing and caeca absent or exceptionally, a rudimentary typhlosole-like dorsal ridge
present. Nephridia simple, exonephric holonephridia throughout, or rarely preceded
in anterior segments by tufted nephridia with simple or (P. (P.) davallid) composite
ducts. Ectal nephridial bladders present (V esiculodrilus] or absent (Perionychella) ;
tufted nephridia retaining bladders in vesiculate species. Holandric (testes in X
and XI), with seminal vesicles in IX and XII or rarely also in X, or X and XI, or in
XI and XII ; or rarely metandric, with anterior testes and seminal vesicles reduced
or absent. Metagynous (ovaries in XIII) ; ovisacs present or absent. Prostates
tubular, with central lumen, though often flattened, or tubuloracemose (i.e. lobulated
to externally racemose but with axial lumen with (or without ?) side branches).
Vasa deferentia usually joining the ectal end of the prostate gland, occasionally the
junction of gland and duct. Penial setae usually present. Spermathecae with
single, clavate or digitiform, uniloculate diverticulum.
DISTRIBUTION. Eastern Subregion of Australia : mostly in Victoria but extending
northwards to the Atherton Tableland in North Queensland and with a morphologic-
ally isolated species near Cooktown. The dominant genus in Tasmania.
REMARKS. The genus Perionychella was established by Michaelsen (i907a) for
four eastern Australian species : Perichaeta dendyi Spencer, i8g2b, and Megascolex
pritchardi Spencer, 1900, both from Victoria ; Perichaeta canaliculata Fletcher,
i888a, from north Queensland ; and P. scolecoidea Spencer, 1895, from Tasmania.
Michaelsen's diagnosis of Perionychella was : setae at least in the mid- and hind-body
many (more than 8) per segment. Spermathecal pores 2-4 (-5 ?) pairs between
segments IV and IX. i gizzard, in segment 5 or 6. Completely meganephric
(holonephric). Prostates with branched canal system.
The only expressed difference from the typically southwestern Australian Wood'
wardia (now Woodwardietta Stephenson, 1925) was the increase in setal numbers
beyond 8 per segment. The status of Woodwardiella is discussed on p. 219.
In 1916 Michaelsen emended the definition of the genus Perionyx Perrier, 1872, to
agree in all essentials with that of Perionychella, which was therefore suppressed, and
included Diporochaeta Beddard in Perionyx as a subgenus. The subgenus Perionyx
was distinguished by branching of the prostate duct on entering the gland (Pheretima
prostate) while in Diporochaeta the gland possessed a central axial lumen with or
without branched or unbranched lateral canals. Perionychella dendyi (like P.
canaliculata) was shown to have the Diporochaeta condition, with lateral canals, and
was placed in that subgenus. Stephenson (1923) reinstated Diporochaeta as a
separate genus while suggesting a close relationship with Perionyx. Perionychella
remained suppressed in Diporochaeta. Gates (1959) placed Diporochaeta and Perionyx
in different families, the redefined Acanthodrilidae and Megascolecidae respectively.
Jamieson (i97ia) demonstrated that the grounds for distinguishing the two families
were invalid and, having redefined and reconstituted the two entities, reduced them
13
218 B. G. M. JAMIESON
to subfamilial rank in the Megascolecidae. Restriction of Perionyx to oriental
species (Gates, 1960) was accepted and 40 Australian species were provisionally
referred to the genus Diporochaeta. It was suggested that revision of Diporochaeta
required consideration of certain species placed in Plutellus which were distinguished
solely by the numbers of setae and that it would necessitate establishment of several
new genera. It was also proposed that Perionychella should be resurrected for
Perichaeta dendyi as it was clearly not congeneric with the New Zealand type-species
of Diporochaeta (see p. 219). Relationship of Perionyx and Diporochaeta s. lat. was
acknowledged by placing the two genera, with other holonephric megascolecine
genera, in the new tribe Perionychini.
Revision of Plutellus is now far advanced and has involved resurrection of Argilo-
philus Eisen for American species, and division of the Australian species, after re-
instatement of Fletcher odrilus Michaelsen, into the genera Plutellus s. strict., Simsia
Jamieson, ig72a ; Graliophilus Jamieson, 19716, and V esiculodrilus Jamieson,
I973a. In the latter paper attention was drawn to the close morphological affinity
of V esiculodrilus with similarly vesiculate Australian species of Diporochaeta s. lat.
The Tasmanian fauna provides the key to revision of Diporochaeta in Australia,
and to the relationships of V esiculodrilus , as vesiculate perichaetine species which
would formerly have been assignable to Diporochaeta s. Stephenson are seen to have
morphological (phenetic) affinities, which are unequivocally intrageneric, with species
attributable to V esiculodrilus. Transference of these perichaetine species to
V esiculodrilus greatly reduces the numbers of species in Diporochaeta but leaves a
large residue of species either lacking nephridial bladders (avesiculate species) or
with the condition of the nephridial duct undescribed. The Tasmanian fauna,
again, allows determination of the affinities of the avesiculate species as most of
these on the island show high morphological similarities with V esiculodrilus. They
are also indistinguishable from Perionychella, of which the Victorian type-species is
avesiculate. (Variation in form of the prostates from tubular to racemose is con-
sidered unimportant as all intermediates are seen, author's examination.) It thus
appears that V esiculodrilus must be included in Perionychella, a prior genus which
has been suppressed for nearly sixty years.
It is here proposed that V esiculodrilus be retained as a subgenus of Perionychella
distinguished from the nominate subgenus (P. (Perionychella}} in possessing nephridial
vesicles. It is possible that some species of V esiculodrilus may be morphologically
closer to avesiculate Perionychella species than to species in their own subgenus (and
vice versa}, but recognition of the two subgenera achieves one of the aims of practical
taxonomy, convenience. Separation of V esiculodrilus from Perionychella also finds
some support in setal ratios (Table i) though whether the distinction by ratios will
be upheld should additional lumbricine species of the latter subgenus be found
remains to be seen. Thus, for Tasmanian species at least, dd = 2-o-^iab>O'iS-o-22u
in V esiculodrilus (with the exception of the morphologically very isolated P. (V.)
glandifera in which dd = $-6ab, o-2ju) while in the two lumbricine species of Periony-
chella dd = 6-5-7-8^6, O'33-o-39w. Furthermore, while recognition of V esiculodrilus
species presents little difficulty, there may be some difficulty in ascertaining whether
an avesiculate specimen should be referred to the subgenus Perionychella or to other
TASMANIAN EARTHWORMS 219
similarly holonephric genera (particularly the Western Australia avesiculate Gralio-
philus) or merits recognition of a new genus. (It should be noted that the difficulty
in distinguishing between avesiculate Perionychella species and Graliophilus stems
largely from paucity of knowledge of some species-groups of Graliophilus and that it
is not proposed that Graliophilus should be united with Perionychella. Rather, as
suggested by Jamieson (19716), Graliophilus requires splitting into further genera.)
Reinstatement of Perionychella is necessitated by the rules of priority but Michael-
sen's definition of Perionychella does not diagnose the majority of species of the genus
as emended and enlarged on the basis of consideration of large numbers of characters,
i.e. overall phenetic resemblance of constituent species, and an emended definition
is therefore required. It will be shown that the only characters which are constant
throughout the emended genus are general characters of suprageneric, often ordinal
or subordinal significance. The genus is thus polythetic, that is to say characters of
generic significance are not present in all constituent species but jointly associate
these species at a high level of similarity. Furthermore, several generic characters,
such as nephridial bladders and five pairs of spermathecae, are constant for the
majority of species and give striking morphological homogeneity to these species
which can often only be distinguished with certainty by the configuration of genital
markings.
Something more must be said of the status of Woodwardiella and of Diporochaeta.
The fact that in the present work Perionychella has been extended to include forms
with eight setae per segment removes all justification for separating Woodwardiella
as defined by Michaelsen from this genus. However, the Western Australian species,
including the type-species, of Woodwardiella invariably have two pairs of spermathe-
cae and are characterized by anterior tufted nephridia. These conditions do occur
exceptionally in Perionychella but it seems inadvisable at present to reduce Wood-
wardiella to the status of a junior synonym in the former. The refractory state of
material of Woodwardiella in the Western Australian Museum (Jamieson 1970)
precludes thorough revision of the genus. The status of these Eastern Australian
species tentatively assigned, from poor descriptions, to Woodwardiella requires
mention. One of these, Cryptodrilus mortoni Spencer, 1895, is shown to be a
Perionychella (V esiculodrilus] in the present work. A second C. tesselatus Spencer,
1895, is almost certainly also a Perionychella. It has not been rediscovered in this
survey. The third, Woodwardia healesi Michaelsen, 1923, cannot be placed with
certainty in any known genus.
The type-species of Diporochaeta, D. intermedia Beddard, 1888, is exceptional in
the Megascolecinae in having its last hearts in segment XI rather than in XII or
XIII. This seems a normal condition in the species as it has been observed in
material from several localities (Lee 1959). The fact that the ovaries are in the nor-
mal megadrile location in XIII and that spermathecae may persist in IX indicates
that the unusual distribution of hearts and spermathecae is not simply due to deletion
of a prespermathecal metamere. Loss of hearts from XII would be contrary to
evolutionary trends in the Megascolecinae. If last hearts were in XII, D. intermedia
might be considered congeneric with the extended Perionychella of the present work
and, as Diporochaeta is the prior genus, all Perionychella species would have to be
220 B. G. M. JAMIESON
placed in it. Other Diporochaeta species in New Zealand appear from the literature
to qualify for inclusion in Perionychella s. lat. but must be re-examined before this
can be confirmed. It is here proposed to retain Perionychella for Australian forms
conforming with the Australian type-species Perionychella dendyi, and to allow the
heart-location in Diporochaeta intermedia to merit regarding this species as con-
stituting a monotypic genus Diporochaeta. This restriction of Diporochaeta will not,
however, be formally made until the status of the other New Zealand species
currently in Diporochaeta is decided, when they may be referred to Perionychella or
considered to warrant one or more genera distinct from both Perionychella and a
monotypic Diporochaeta.
Perionychella accounts for almost half of the native Tasmanian earthworm
species. The very profound emendation of the genus after a long period of suppres-
sion makes it desirable to list those non-Tasmanian species assignable to the genus
and it will be seen that these comprise substantial components of the fauna of neigh-
bouring Victoria.
Subgenus PERIONYCHELLA Michaelsen, I9o;a emend.
DEFINITION. As for the genus with nephridia lacking bladders.
DISTRIBUTION. As for the genus but the generic identity of the Queensland species
is uncertain.
TYPE-SPECIES. Perionychella dendyi (Spencer, i8g2b) Victoria. Syntype (?)
examined, AM : W.I2Q4.
CHECK LIST
, Tasmanian species
P. (P.) capensis sp. n.
P. (P.) hickmani sp. n.
P. (P.) kershawi sp. n.
*Perichaeta irregularis Spencer, 1895
P. (P.) weldboroughi sp. n.
Other species
Unless otherwise stated, material has been examined by the author as indicated.
Victoria
*Perichaeta copelandi Spencer, i8g2b Lectotype NMV : Gi86 ; paralectotype
Gi443
^Diporochaeta davallia Spencer, 1900, vide Jamieson (1970)
*Perichaeta dubia Spencer, i892b Syntype (?) AM : Wi293
Megascolex pritchardi Spencer, 1900, vide Michaelsen (i907a) and lectotype
NMV : Gio8
*Perichaeta tanjilensis Spencer, i892b Syntype (?) AM : Wi298
New South Wales
None known
TASMANIAN EARTHWORMS 221
Queensland
*Perionyx (Diporochaeta) athertonensis Michaelsen, 1916
Perichaeta canaliculata Fletcher, i888a, vide Michaelsen (igoyb)
*Perionyx (Diporochaeta) erici Michaelsen, 1916
*Perionyx (Diporochaeta) phalacrus Michaelsen, 1916
*Perionyx (?) terrareginae Fletcher, iSSgb
* Comb. nov.
REMARKS. The condition of the nephridial ducts is not described for the Queens-
land species but they presumably lacked bladders as all were described by Michaelsen
(1916) who in the same account described nephridial bladders for a further Perionyx
species (P. (D.) sigillatus). The types are not available for examination. That
these species, from the Cairns-Atherton Tableland area, should be referred to the
genus Perionychella is nevertheless questionable. The gizzard is located more
posteriorly than in Victorian and Tasmanian species, the number of setae per segment
is more numerous and, although it is deduced that nephridial bladders are absent,
they show affinities with Perionyx (Diporochaeta) sigillatus Michaelsen, 1916, from
the same area, which has bladders and appears to be related to the similarly vesiculate
genus Fletcherodrilus, rather than with Perionychella. Exclusion of the five north
Queensland species from Perionyx as emended by Gates (1960) is necessitated by a
set of characters of the latter genus, viz. female pore unpaired, midventral ; sub-
neural vessel present ; spermathecal diverticula absent.
Subgenus VESICULODRILUS Jamieson,
DEFINITION. As for the genus with nephridia possessing ectal bladders. (One
morph of P. (V.) mortoni lacks bladders and their presence is questionable in the
type-specimens of P. (V.) lacustris.)
DISTRIBUTION. As for the genus.
TYPE-SPECIES. Cryptodrilus frenchi Spencer, i892a. Victoria.
CHECK LIST
Tasmanian species
P. (V.) bithecata sp. n.
*Perichaeta dilwynnia Spencer, 1895
P. (V.) evansi sp. n.
P. (V.) glandifera sp. n.
P. (V.) hellyeri sp. n.
Cryptodrilus hobartensis Spencer, 1895 (new synonym C. insularis Spencer, 1895)
*Perionyx lacustris Stephenson, 1924
P. (V.) montisarthuri sp. n.
*Cryptodrilus mortoni Spencer, 1895
P. (V.) obliquae sp. n.
P. (V.) pedderensis sp. n.
P. (V.) tunnackensis sp. n.
* Comb. nov.
222 B. G. M. JAMIESON
Other species
Unless otherwise stated, material has been examined by the author as indicated.
Victoria
*Perichaeta alsophila Spencer, i892b Syntype (?) AM : Wi297
*Diporochaeta arnoldi Spencer, 1900 Lectotype NMV : G203
*Diporochaeta euzona Spencer, 1900 Lectotype NMV : Gi2O
Cryptodrilus frenchi Spencer, 18920, vide Jamieson (ig73a)
*Diporochaeta frosti Spencer, 1900 Lectotype NMV : 0117
Cryptodrilus gippslandicus Spencer, i892a, vide Jamieson (1970)
*Perichaeta lochensis Spencer, i892a, q.v.
*Diporochaeta manni Spencer, 1900 Lectotype NMV : Gn8
*Diporochaeta mediocincta Spencer, 1900 Lectotype NMV : Gi24
*Diporochaeta richardi Spencer, 1900 Lectotype NMV : G22O
Cryptodrilus tanjilensis Spencer, i892a, vide Jamieson (i973a) and paralectotype
NMV : 055
Megascolides tisdalli Spencer, 1900, vide Jamieson (1972) and lectotype NMV : 0178
Plutellus uncinatus Stephenson, 1933, q.v.
Cryptodrilus victoriae Spencer, 18920, vide Jamieson (i973a) and paralectotype
NMV : 01410
Megascolides volvens Spencer, 1900, vide Jamieson (ig73a) and syntype (?)
AM : Wi292
*Cryptodrilus willsiensis Spencer, i892a Syntype (?) AM : Wi279
New South Wales
V esiculodrilus frenchi (Spencer), vide Jamieson (ig73a)
*Diporochaeta pheretima Jamieson, ig73a
V esiculodrilus purpureus Jamieson, i973a
Queensland
*Perichaeta barronensis Fletcher, i887b Syntypes (?) AM : W2387
*Diporochaeta grandis Spencer, 1900 Lectotype NMV : 0184
Perionychella species of uncertain subgeneric status
The following species which are not available for re-examination accord with the
above-emended definition of the genus Perionychella but are of uncertain subgeneric
status as their authors did not record presence or absence of nephridial bladders.
They are known customarily to have omitted reference to these in accounts of species
since shown to possess bladders. All except Perionychella scolecoides (Spencer,
1895) are new combinations in Perionychella.
Tasmania
Megascolides bassanus Spencer, 1895 (King Island)
Plutellus decatheca Michaelsen, 1910
Cryptodrilus ellisi Spencer, 1895
TASMANIAN EARTHWORMS 223
Tasmania (cont.)
Perichaeta moroea Spencer, 1895
Perichaeta richae Spencer, 1895 (this supposedly has gizzards in III and IV, a
most doubtful observation)
Perichaeta scolecoidea Spencer, 1895
Victoria
Diporochaeta apiocystis Stephenson, 1933
Perichaeta bakeri Fletcher, i888b
Perichaeta dicksonia Spencer, i892b
Diporochaeta faucium Michaelsen,
Diporochaeta lindti Spencer, 1900
Diporochaeta nemoralis Spencer, 1900
Perichaeta obscura Spencer, i8g2b
Diporochaeta sedecimalis Michaelsen,
Diporochaeta spenceri Michaelsen, i
Diporochaeta telopea Spencer, 1900
Perichaeta walhallae Spencer, i8g2b
New South Wales and Queensland
No additional species
KEY TO TASMANIAN SUBGENERA AND SPECIES OF PERIONYCHELLA
For species of uncertain subgeneric status see specific descriptions (p. 253 et seq.).
1 Nephridial bladders absent subgenus PERIONYCHELLA, 2
- Nephridial bladders present .... subgenus VESICULODRILUS, 6
2 Setae 8 per segment throughout.* Accessory genital markings unpaired mid ventral
or paired medianly contiguous respectively in 17/18 and 18/19. Fig. yA
P. (P.) weldboroughi sp. n.
Setae more than 8 per segment throughout or behind the clitellum. Postclitellar
genital markings not median .......... 3
3 Setae 8 per segment anteriorly ; numerous posteriorly. Fig. 56 P. (P.) kershawi sp. n.
- Setae more than 8 per segment throughout ....... 4
4 Spermathecae 2 pairs. Fig. 5A ..... P. (P.) capensis sp. n.
- Spermathecae 3-5 pairs ........... 5
5 Spermathecae 3 pairs, f Fig. 6C P. (P.) irregularis (Spencer, 1895)
Spermathecae 5 pairs. Fig. 6A, B P. (P.) hickmanni sp. n.
6 Setae 8 per segment throughout ......... 7
- Setae more than 8 per segment throughout or at least caudally . . . . 13
7 Spermathecal pores 5 pairs .......... 8
- Spermathecal pores less than 5 pairs . . . . . . . . 10
8 Dorsal blood vessel bifid in the oesophageal region. A median oesophageal gland in
XV. Seminal vesicles in XI and XII. Pericardiac testis-sacs present. Fig.
loA, B P. (V.) glandifera sp. n.
* An avesiculate morph of P. (V.) mortoni keys out here but differs from weldboroughi in having a
genital field much as Fig. I2B.
f Typical specimens of P. (V.) lacustris with rudimentary or no bladders key out here but differ from
irregularis in having a genital field as in Fig. yB.
224
B. G. M. JAMIESON
VI
sp.p
sp.p5
Imin
B
FIG. 5. Genital fields. A, Perionychella (P.) capensis, holotype. B, P. (P.) kershawi,
holotype.
Abbreviations used in figures 5-32. $, female pore; g.m, accessory genital marking; £, male pore;
p.s, penial seta; sp.p., spermathecal pore. Roman numerals indicate respective segments. Clitellum
shaded. All figures by camera lucida.
TASMANIAN EARTHWORMS 225
- Dorsal blood vessel single. No extramural oesophageal glands present in XV
though internal rugae may be present here and elsehwere. Seminal vesicles in
IX and XII. Testis-sacs absent ......... 9
9 Paired accessory genital markings of the male field at the posterior borders of their
segments but not intersegmental. Fig. 126 . . P. (V.) mortoni Spencer, 1895)
- Paired accessory genital markings of the male field intersegmental. Fig. 1 1
P. (V.) hobartensis (Spencer, 1895), part
10 Spermathecal pores i pair. Fig. 8A ..... P. (V.) bithecata sp. n.
- Spermathecal pores 2 to 4 pairs ......... 1 1
11 Spermathecal pores 2 pairs. Fig. gA. ..... P. (V.) evansi sp. n.
- Spermathecal pores 3 or 4 pairs ......... 12
12 Spermathecal pores 3 pairs. Fig. 14. . . . P. (V.) tunnackensis sp. n.
- Spermathecal pores 4 pairs. Fig. n. . P. (V.) hobartensis (Spencer, 1895), part
13 Setae 8 per segment in some anterior segments ; numerous posteriorly. Spermathe-
cal pores 3 pairs ............ 14
- Setae more than 8 per segment throughout. Spermathecal pores 3-5 pairs . . 15
14 Genital markings present in aa in XVII and XVIII. Fig. 96. P. (V.) hellyeri sp. n.
- Genital markings absent in aa. Fig. I2A . . . P. (V.) tnontisarthuri sp. n.
15 Spermathecal pores 5 pairs .......... 16
- Spermathecal pores 4 pairs or less ....... . . 17
1 6 Accessory genital markings including a mid ventral unpaired papilla posteriorly in
each of VI and VII and anteriorly in XVIII. Fig. 8B P. (V.) dilwynnia (Spencer, 1895)
- Accessory genital markings may include a mid ventral papilla in VII but not else-
where. Fig. 1 3 A, B P. (V.) obliquae sp. n.
17 Spermathecal pores 3 pairs. Dorsal pores present. Fig. 76
P. (V.) lacustris (Stephenson, 1924)
- Spermathecal pores 4 pairs. Dorsal pores absent. Fig. 126 P. (V.) pedderensis sp. n.
Perionychella (Perionychella) capensis sp. n.
Fig. i, 5A, i6A, B ; Plates 91 and 92
1 = 50 mm, w = 2-4, 2-5 mm, s = 108 (holotype ; paratype i is a posterior
amputee).
Prostomium epilobous f or proepilobous. First dorsal pore 5/6. Setae in XII 22,
24 ; caudally 31, 29 or more in some segments ; ventral gap (aa), approximately
constant, = 2-^ab throughout ; dorsal gap (zz) = i-2yz, narrowing posteriorly but
still distinct. Nephropores : preclitellar (XIII) in setal lines 6 ; intestinal in s.l. 8.
Clitellum annular ^XIII-XVII. Male pores on XVIII on b, each with a pair of
protruding, very long penial setae, on dome-shaped papillae extending from a to
slightly median of c. Paired eye-like genital markings from s.l. i to s.l. 3 at 17/18
and from median to s.l. i to s.l. 2 at 19/20, that in 19/20 present on the left side only
in paratype i. Female pores paired anterior and slightly median of a on XIV.
Spermathecal pores 2 pairs on small papillae in s.l. 3 at 7/8 and 8/9.
Last hearts in XII (latero-oesophageal) . Supra-oesophageal in VIII-XIII, well
developed. Gizzard vestigial, in V, an equally large but thinner walled proventricu-
lus in IV. Extramural calciferous glands absent but oesophagus with circumferential
vascular striae in VII-XV, especially dilated in XIV and XV in which, however,
longitudinal rugae remain few and low. Intestinal origin |XVI ; a very low dorsal
typhlosole present (hardly appreciable).
226 B. G. M. JAMIESON
Nephridia : simple stomate, avesiculate, exonephric holonephridia ; commencing
segment indeterminable ; preseptal funnels demonstrated for those of XV pos-
teriorly. Holandric (funnels iridescent in X and XI) ; gymnorchous ; seminal
vesicles racemose, in IX and XII. Metagynous ; ovisacs present (paratype i) or
not developed (holotype). Prostates tubuloracemose, in XVIII, elongate, tongue-
shaped, deeply incised, almost sessile as the duct is short ; a narrow central lumen
present in the gland. Penial setae (holotype) filiform and strongly sinuous ; ectal
end in profile with a group of a few small anteriorly directed spines repeated at
intervals longitudinally ; tip with long, horn-shaped projections in two sequential
sets of two, slightly below the pointed, asymmetrical extremity, and probably
homologous with the spines ; length (right setae) 1-95, 2-74 mm. Spermathecae 2
pairs ; diverticulum (inseminated) single, digitiform, uniloculate, longer than the
ampulla.
MATERIAL EXAMINED.
Table Cape, i45°45'E. 4i°oo'S. Map, 7/1, Dr J. L. Hickman, 24 Aug 1954 ~ holotype
TM : K259 ; paratype i BM(NH) : 1972.8.3 ; paratype 2 BJ : Ti.
REMARKS. P. (P.) hickmani is morphologically the closest species to P. (P.)
capensis in the subgenus. Among other differences from hickmani are the two pairs
of spermathecal pores and the genital markings in 19/20. It is also unique in the
genus, in Tasmania, in possessing a typhylosole-like dorsal thickening of the intestinal
wall (questionably to be considered a true typhlosole) and in location of the first
dorsal pore in 5/6.
Perionychella '(Perionychella) hickmani sp. n.
Fig. i, 6A, B, i6C, D
1 = 94> 75 mm> w = 4 mm, s = 125, 112.
Prostomium epilobous f , open. First dorsal pore 4/5 (rudiment in 3/4 in holo-
type ?). Setae in XII 24, 21 ; caudally 31, 33 ; dorsal and ventral rows (a and z)
straight throughout ; ventral gap approximately %ab throughout ; dorsal gap
approximately 1-5 or 2yz anterior to clitellum ; equal toyz caudally. Nephropores :
preclitellar pores in setal lines 6-7 ; postclitellar in setal lines 7. Clitellum annular
^XIII-|XVII, XVII. Males pores equatorial in XVIII in ab-b, on prominent
papillae. Paired eye-like genital markings in 16/17, 17/18 and 18/19 in ab.
Female pores paired on XIV, anterior and slightly median of a. Spermathecal
pores 5 pairs, in 4/5-8/9, on small papillae in ab.
Last hearts in XII (latero-oesophageal) . Supra-oesophageal in VIII-XIII, well
developed. Gizzard large but compressible, in V. Oesophagus simple ; extramural
calciferous glands absent. Intestinal origin XVIII (though pushing septum 17/18
forwards) ; typhlosole absent. Nephridia simple stomate, avesiculate, exonephric
holonephridia commencing in II (preseptal funnel demonstrated in XIV but pre-
sumably present further anteriorly as nephridial morphology similar) ; tufts absent ;
TASMANIAN EARTHWORMS
227
sp.pl
sp.p 5
XVI
XVII
XVIII
1mm
1mm
FIG. 6. Genital fields. A & B, Perionychella (P.) hickmani : A, holotype ; B, paratype 3.
C, P. (P.) irregularis, lectotype.
ducts tubular, seminal vesicles racemose, in IX and XII. Metagynous ; ovisacs
large, in XIV. Holandric (funnels iridescent in X and XI) ; gymnorchous. Pros-
tates : gland almost circular, appearing to form a simple flattened lobe but resolvable
into a zig-zag with narrow central lumen, the arms of which are expanded and not
tubular (holotype) or depressed tubular, zig-zag in XVII -XIX (paratype i) ; vas
deferens joining ectal end of gland. Penial setae slender, almost filiform ; tip
rounded, possibly worn, and with no notable sculpturing (holotype, paratype 2) or
with a delicate tip hooked through about sixty degrees (paratype 3); length (right
setae), 0-96, 0-85 and 1-22 mm, respectively. Spermathecae 5 pairs ; diverticulum
single, clavate, uniloculate.
Genital markings. 7/3 - paratype 2 agreeing exactly with holotype and paratype
1 '> I3/5 ~ paratypes 3 and 4, mature specimens as holotype but lacking accessory
genital markings in 18/19.
228 B. G. M. JAMIESON
Other variation. The two mature paratypes from Lake St Clair agree closely in
morphology with the holotype, differing in the following unimportant respects :
1 = 47-49 mm ; s = 105-108 ; the prostomium is closed ; the setal count rises to
46 caudally ; the nephridia of II are astomate rudiments, those of III having demon-
strable funnels (see also Genital markings}.
MATERIAL EXAMINED.
Fern Glade, Emu River, Burnie, I45°55'E. 4i°O5'S. Map, 7/2, Dr J. L. Hickman, 24 Aug
1954 - holotype TM : K26o. Hellyer Gorge, i45°35'E. 4i°2o'S. Map, 7/3, Dr J. L. Hickman,
28 May 1954 - paratype J BM(NH) : 1972.8.12. Parrawe, I45°35'E. 4i°2o'S. Map, 7/3, Dr
J. L. Hickman, 25 Aug 1954 - paratype 2 TM : K26i. Lake St Clair, i46°io'E. 42°O5'S.
Map, 13/5, Dr J. W. Evans, Feb 1941 -paratypes 3 and 4 BM(NH) : 1972.8.13-14; paratype 5
AM: W5i85.
REMARKS. Differences between this species and P. (P.) capensis, which is closest
to it morphologically, are given under the latter species.
Perionychella (Perionychella) irregularis (Spencer, 1895)
Fig. i, 6C
Perichaeta irregularis Spencer, 1895 '• 53~54» PI- V. fig. 52-54.
Diporochaeta irregularis ; Michaelsen, 1900 : 206 ; Jamieson, 19715 : 83.
1 = 87 mm, w = 4-7 mm, s = 104.
Prostomium canaliculate, epilobous |. First dorsal pore in 4/5. Setae n on each
side in V, 15-16 on each side in XII and onto the midbody ; 20 on each side caudally ;
ventral and dorsal setal gaps conspicuous throughout ; in XII aa = 2ab, zz = ^zy ;
in the midbody and caudally aar = $ab, zz = ^zy. Nephropores between setal lines
8 and 9. Clitellum annular, |XIII-XVII and, dorsally, anterior XVIII ; setae,
intersegments (weaker) and dorsal pores retained. Male pores on prominent elliptical
papillae between setal lines 2 and 3, nearer 2, on XVIII, the intervening setal rows
not demonstrably bearing setae in this segment. Two pairs of eye-like accessory
genital markings in (or, lectotype, immediately median to) setal lines 3, in 19/20
and 20/21. Female pores in XIV, anteromedian of a, near the ventral midline.
Spermathecal pores 3 pairs, in 6/7, 7/8 and 8/9, in setal lines 3, 4 and 5 respectively
(Spencer) but only slightly divergent posteriorly.
Dorsal blood vessel single, continuous onto the pharynx. Last hearts in XII
(latero-oesophageal but lacking dorsal connectives). Supra-oesophageal weakly
developed, in X-XII. Gizzard small, with pronounced muscular sheen but easily
compressed, appearing to lie in VI but the very attenuated septum 5/6 adherent to
it slightly behind its anterior limit, suggesting that the gizzard has moved from V
into VI by adhesion of the attenuated septum. Calciferous glands absent but oeso-
phagus vascularized in VIII-XVI and swollen in some of these. Intestinal origin
XVII ; typhlosole absent. Nephridia st ornate avesiculate holonephridia, commenc-
ing in II (funnels demonstrated from XVIII) ; tufted nephridia absent. Holandric
(funnels iridescent in X and XI) ; seminal vesicles 4 pairs, small and simple in IX
and X, large and racemose in XI and XII. Metagynous ; small ovisacs in XIV.
TASMANIAN EARTHWORMS 229
Prostates depressed, linear, laterally deeply incised, in XVII -XXII, resolvable
with difficulty into a tubular form ; external muscular duct very long and tortuous.
Penial setae absent. Spermathecae 3 pairs, not notably departing from a line
parallel with the body axis ; diverticulum (inseminated) single, clavate, uniloculate.
TYPE-LOCALITY. King River Valley, approximately i45°25'E. 42°io'S. Map, 12/1.
MATERIAL EXAMINED.
The previously dissected lectotype, NMV : Ga88.
REMARKS. The above account considerably augments that of Spencer and
demonstrates the avesiculate condition of the nephridial ducts for the first time.
The dimensions accord with the specimen described by Spencer.
Closest to this species are the similarly perichaetine species P. (P.) hickmani and
P. (P.) capensis. It differs from both in having three pairs of spermathecal pores
and in lacking genital markings anterior to the male pores.
Perionychella (Perionychella) kershawi sp. n.
Fig. i, 5B, i6E. Table 2
1 = 171 mm, w = 6-1 mm, s = 119.
Prostomium epilobous £, closed with a grooved peristomium giving a tanylobous
appearance. First dorsal pore 4/5. Setae 8 per segment anterior to clitellum ;
posterior to clitellum occasionally with an extra seta in cd ; in the posterior half,
setae at first added in the dorsal gap, caudally the rows becoming irregular and
approximately 21 seta per segment. Nephropores in c lines anteriorly ; visible
posteriorly but setal lines irregular. Clitellum annular, XIII-XVII. Male pores
on small dome-shaped papillae in a on XVIII. Small eye-like unpaired midventral
accessory genital markings in 11/12 and 12/13 ; paired transversely conjoined mark-
ings, extending lateral of mid be at 17/18, extending shortly lateral of b at 18/19 and
occupying bb at 19/20. Female pores a pair in XIV, anteromedian of a. Spermathe-
cal pores 5 pairs in 4/5-8/9, on small papillae in a.
Last hearts in XII, latero-oesophageal, connectives to dorsal vessels very slender,
thread-like. Supra-oesophageal weak in VII, well developed in VIII-XII. Gizzard
very large but easily compressible in V. Extramural calciferous glands absent.
Oesophagus with engorged circumferential vascular striae in VIII -XV ; in (XIII ?),
XIV and XV strongly dilated and with longitudinal striations which correspond
with low internal ridges (severely macerated in this region) which suggest some
(calciferous ?) modification. Intestinal origin XVII ; typhlosole absent. Nephridia
tufted holonephridia in II-IX, with posteriorly decreasing numbers of coiled loops,
discharging presetally in c lines ; stomate at least from X in which coiling in insuf-
ficient to be termed tufted, but anterior tufted nephridia probably are also stomate
as a slender neck is demonstrable to the anterior septum. Bladders absent unless
terminal dilatations of the ducts of the anterior tufts be considered vestigial bladders.
Holandric (funnels iridescent in X and XI) ; gymnorchous ; seminal vesicles
racemose, in IX and XII. Presence of ovaries or ovisacs indeterminable owing to
23°
B. G. M. JAMIESON
maceration. Prostates slender coiled only slightly depressed tubes, winding through
XVII-XXIV ; vas deferens joining the gland near the junction of the sinuous
duct. Penial setae present, form not elucidated ; length (holotype) 2-25, 2-37,
2-39 mm. Spermat-hecae 5 pairs ; diverticulum (inseminated) single, clavate,
uniloculate.
MATERIAL EXAMINED.
Under a rotting log near the old Breiseis Tin Mine dam, Derby, i47°5o'E. 4i°io'S. Map, 9/4,
Mr R. C. Kershaw, 19 Sep 1971 - holotype TM : 262.
REMARKS. This species differs from other Tasmanian species of the sugbenus in
having midventral genital markings in 11/12 and 12/13 and in transition of the setae
from 8 per segment anteriorly to many posteriorly.
TABLE 2
Mean intersetal distances in segment XII in lumbricine species
of Perionychella (Perionychella and Vesiculodrilus)
P. (P.) kershawi
(i)
12-65
P. (P.) weldboroughi
(i)
"•57
P. (F.) bithecata
(5)
I2-I
P. (F.) evansi
(2)
10-35
P. (F.) glandifera
(2)
9-89
P. (F.) hellyeri
(5)
12-42
P. (F.) hobartensis
(12)
n-59
P. (F.) montisarthuri
(I)
12-5
P. (F.) mortoni
(3)
11-83
P. (F.) tunnackensis
(!)
11-74
no. of
specimens
aa
P. (P.) kershawi
(i)
2-44
P. (P.) weldboroughi
(i)
2-33
P. (F.) bithecata
(5)
1-54
P. (F.) evansi
(2)
1-34
P. (F.) glandifera
(2)
2-OO
P. (F.) hellyeri
(5)
1-52
P. (F.) hobartensis
(12)
1-62
P. (F.) montisarthuri
(I)
I-42
P. (F.) mortoni
(3)
i-73
P. (F.) tunnackensis
(I)
1-67
no. of
standardized to a periphery of 100
specimens
aa
ab
be
cd
dd
dc
cb
ba
(i)
12-65
5-14
11-46
10-28
33-2
9-88
12-25
5-14
(i)
u-57
4-96
9-92
9-92
38-84
9-92
9-92
4-96
(5)
I2-I
7-87
12-45
12-76
21-53
13-25
12-12
7-87
(2)
10-35
7-71
14-28
12-98
20-09
12-98
13-9
7-71
(2)
9-89
4-94
18-39
8-34
27-44
8-18
I7-37
5-46
(5)
12-42
8-17
12-82
12-51
20-70
12-62
12-76
8-00
(12)
n-59
7-16
13-22
13-0
21-98
12-87
I3-2
6-94
(I)
12-5
8-8
ii-n
14-81
17-59
15-74
n-57
7-87
(3)
11-83
6-83
15-89
n-4
19-87
11-74
15-79
6-64
(J)
11-74
7-04
I3-I5
14-08
19-72
14-08
I3'I5
7-04
ab
i-oo
i-oo
I-OO
I-OO
I-OO
I-OO
I-OO
I-OO
I-OO
I-OO
be
2-23
2-OO
1-58
1-85
1-85
1-26
2-33
1-87
intervals/aft
dd
6-46
7-83
2-74
2-61
5-55
2-53
3-07
2-OO
cd
2-OO
2-OO
1-62
1-68
1-69
1-82
1-68
1-67
2-OO
2-9I
2-80
dc
1-92
2-OO
1-68
1-68
1-66
1-8
1-79
1-72
2-OO
cb
2-38
2-OO
1-54
i -80
3-52
1-56
1-84
2-31
1-87
ba
i-oo
I-OO
I-OO
I-OO
I-IO
0-98
o-97
0-89
0-97
I-OO
Perionychella (Perionychella) weldboroughi sp. n.
Fig. i, 7A, i6F. Table 2
1 = 48 mm, w = 1-6 mm, s = 71.
Prostomium tanylobous. First dorsal pore 4/5. Setae 8 per segment in regular
longitudinal rows throughout. Nephropores faint in c lines. Clitellum annular
TASMANIAN EARTHWORMS
231
XIII-
XIV-
XVI-
sp.p1
XVIII
1mm
B
FIG. 7. Genital fields. A, Perionychella (P.) weldboroughi, holotype. B, P. (V.) lacustris,
TM : K265.
XIII-XVI. Male pores equatorial in XVIII, on pronounced papillae in ab with
centres nearer b than a ; preceded by a pair of medianly contiguous small papillae
in aa at 17/18, with a further midventral unpaired papilla at 18/19. Faint traces
of presetal midventral unpaired markings present in VII ?, VIII ? and IX. Female
pores a pair on XIV anterior to and slightly median of a. Spermathecal pores 5
pairs in 4/5-8/9 in or slightly lateral of a.
Last hearts in XII. Supra-oesophageal very weakly developed. Gizzard
moderate, but easily compressible, in V. Calciferous glands absent but oesophagus
greatly dilated in XIII-XV in which vascularization is increased and internal
longitudinal rugae, though low, are conspicuous. Intestinal origin XVII ; typhlosole
absent but intestinal wall slightly thicker middorsally. Nephridia simple stomate
avesiculate holonephridia commencing in II (funnels demonstrated in the intestinal
region) ; ectal ducts lacking bladders. Holandric (funnels in X and XI iridescent) ;
gymnorchous ; seminal vesicles very large, racemose, in IX and XII. Metagynous ;
232
B. G. M. JAMIESON
1mm
A
sp.p1
g.m
B
FIG. 8. Genital fields. A, Perionychella (Vesiculodrilus) bithecata, holotype.
B, P. (V.) dilwynnia, TM : Ka66.
ovisacs small, in XIV. Prostates flattened, somewhat incised, tubular, in XVIII-
XX ; vas deferens joining gland near the tortuous duct. Penial setae (holotype)
filiform and sinuous, lacking ornamentation but tip apparently missing ; length
(incomplete ?) = 0-42 mm. Spermathecae 5 pairs, diverticulum single, slenderly
clavate, uniloculate.
MATERIAL EXAMINED.
1-6 miles from the eastern end of Weldborough Pass, I47°55'E. 4i°io'S. Map, 9/3, Dr J. L.
Hickman, 26 Aug 1953 - holotype TM : K2&3-
REMARKS. The midventral genital marking in 18/19 diagnoses this species.
TASMANIAN EARTHWORMS 233
Perionychella (Vesiculodrilus) bithecata sp. n.
Fig. i, 8A, i6H ; Plate 93. Table 2
1 = 41, 35 mm, w = 1-5 mm, s = 93, 98.
Prostomium epilobous ^ open. First dorsal pore 4/5. Setae 8 per segment in
regular longitudinal rows throughout. Nephropores in c lines. Clitellum annular
^XIII-XVI. Male pores equatorial in XVIII in ab on small medianly conjoined
papillae ; 2 transverse tumid bands spanning the space between the porophores, one
anteriorly the other posteriorly in XVIII, each band traversed by a groove which
is parallel to but distinct from the intersegmental furrows. Genital markings ; 3
pairs of eye-like markings anterior to a in VII, VIII and XVII and on the right side
only in XIX (holotype) ; these replaced in paratype I by paired markings in ab,
posteriorly in VI and XVII. Female pores a pair on XIV anteromedian to a sur-
rounded by a nonpigmented oval marking filling aa. Spermathecal pores I pair, in
6/7, in b.
Last hearts in XII (latero-oesophageal). Supra-oesophageal moderately de-
veloped. Gizzard small to moderate, easily compressible, in V. Calciferous glands
absent ; oesophagus widest in XIV and XV in which internally it has large,
regular longitudinal folds ; intestinal origin XVII ; typhlosole absent. Nephridia
simple stomate, vesiculate, exonephric holonephridia commencing in II (funnels
demonstrated at least as far forward as the nephridia of III) ; bladder transversely
ellipsoid and projecting median of the pore (in c line) through not truly diverticulate.
Holandric (funnels iridescent in X and XI) ; gymnorchous ; seminal vesicles race-
mose in IX and XII. Metagynous ; ovisacs small in XIV. Prostates depressed
tubular, compacted coils in XVI I -XXI ; Penial setae (paratypes i, 3) filiform and
sinuous ; tip inrolled, hooked and with a pair of 'dorsal', short horns visible only
under the electron microscope ; length (paratype i) 0-39 mm. Spermathecae i pair,
with pores in 6/7 ; the ampulla extending into IX or VIII ; diverticulum (insemi-
nated) single, uniloculate, bent clavate.
Genital markings in the holotype and 3 paratypes
Total number
Specimen of specimens
Paired posterior in VI in ab - Pi. P3(R) 2
Paired presetal in VII in a H -
Paired presetal in VIII in a H -
Paired presetal in XVII in a H -
Paired posterior in XVII in ab - Pi, 2. P3(R)
Unilateral (R) in 18/19 in ab - P2
Unilateral presetal in XIX in a (R) H —
or ab (L) - P3
H = Holotype; P = Paratypes. R = right. L = left.
MATERIAL EXAMINED.
Hellyer Gorge, I45°35'E. 4i°2o'S. Map, 7/3, Dr J. L. Hickman, 28 May 1954 - holotype
TM 1X264; paratype i BM(NH) : 1972.8.1 ; paratypes 2, 3 AM : W$ 186-7 : paratype 4
BJ : T2.
14
234 B- G- M- JAMIESON
REMARKS. The configuration of genital markings in the paratypes differs
markedly from that in the holotype but similarity of general morphology, including
the bithecal condition (which is unique in the genus), indicates conspecificity.
Perionychella (Vesiculodrilus) dilwynnia (Spencer, 1895)
Fig. i, 8B, i6J ; Plates 94-96
Perichaeta dilwynnia Spencer, 1895 : 50-51, PI. IV, fig. 46-48.
Diporochaeta dilwynnia ; Michaelsen, 1900 : 204 ; Jamieson igyib : 83.
1 = 71, 52 mm, w = 3-0, 2-6 mm, s = 104, 118 (Specimens i and 2 respectively,
throughout account).
Prostomium epilobous £, open, but deep dorsal canalicula extending to inter-
segment 1/2. Peristomium similarly grooved midventrally. First dorsal pore 4/5.
Caudal end dorsoventrally depressed. Setae in XII, 16, 14 ; caudally 24 per seg-
ment ; dorsal and ventral rows (a and z) straight throughout but caudally some setae
of z line are omitted ; ventral gap approximately 2ab throughout ; dorsal gap in-
constant even over a few segments, usually a definite interruption of the setal circlet
but not more than twice the general zy width. Nephropores : preclitellar in setal
lines 4-5 ; postclitellar in 4 ; caudally in 6-7 ; in a straight lateral row on each side.
Clitellum annular, XIII, £XIII-XVII. Male pores equatorial in XVIII in mid ab,
on large prominent papillae ; a transversely elongate elliptical papilla filling the
space between them in the anterior half of the segment. Paired eye-like genital
markings in 19/20 and 20/21 in ab and an unpaired postsetal midventral marking in
each of VI and VII. Female pores shortly anteromedian to setae a of XIV. Sperma-
thecal pores 5 pairs, on small papillae in 4/5-8/9, in ab.
Last hearts in XII (latero-oesophageal). Supra-oesophageal well developed in
IX-XII, weak in VIII and XIII. Gizzard small, almost rudimentary, in V.
Oesophagus simple, extramural calciferous glands absent. Intestinal origin XVII ;
typhlosole absent. Nephridia simple stomate, exonephric, vesiculate holonephridia
throughout, commencing in II ; each with large, single preseptal funnel and long,
wide, tortuous tubular bladder. Holandric (funnels iridescent) ; gymnorchous ;
seminal vesicles racemose, in IX and XII. Metagynous ; ovisacs in XI V. Prostates
depressed, lobulated tubular, with compressed coils, in XVII-XX (i), XVIII only
(2) ; vas deferens joining junction of gland and duct. Penial setae (specimens i
and 2) slender, needle-like, not filiform ; tip pointed, asymmetrical, convex on one
side, straight or slightly concave on the other ; ornamentation in profile consisting
of alternating V-shaped depressions the border of which, at the angle, projects
towards the apex of the seta as an irregular scale ; length (specimen i) = 0-77 mm.
Spermathecae 5 pairs ; diverticulum (inseminated) single, clavate, uniloculate.
Genital markings. 13/4 - (specimens 1-9) median genital marking in VI absent in
specimens 3-6, 8 ; otherwise agreeing in all respects with specimen i.
TYPE-LOCALITY. Dee Bridge, i46°35'E. 42°i5'S. Map, 13/7.
TASMANIAN EARTHWORMS
235
g.m
VI
XVI
1mm
B
FIG. 9. Genital fields. A, Perionychella (Vesiculodyilus) evansi, holotype.
B, F. (V.) hellyeri, holotype.
236 B. G. M. JAMIESON
MATERIAL EXAMINED.
Tarraleah, i46°25'E. 42°2o'S. Map, 13/4, over pipeline, Dr J. L. Hickman, 27 May 1954 ~
specimens 1-9 TM : K266-274 ; specimens 10-13 BM(NH) : 1972.8.4-7 ; Butlers Gorge Road,
I -4 miles from Tarraleah, Dr J. L. Hickman, 27 May 1954 ~ specimens 14-20 TM : 1^275-281 ;
specimens 21-32 BJ : T3- 14 ; Tarraleah, Lyell Highway, Dr J. L. Hickman, 22 May 1954-
specimens 33-36 AM : W5i88~5i9i.
REMARKS. The new material was collected a few miles from the Dee Bridge
type-locality. The type-specimens are lost (Jensz and Smith 1969). It appears
from Spencer's description that they differed from the Tarraleah specimens in clitel-
lar extent (XIV-XVII) ; in location of the paired genital markings in 18/19 and 19/20
and absence of the midventral markings in VI and VII ; the median marking in
XVIII was present and setal counts agreed closely. Nephridial bladders were not
mentioned.
Perionychella (Vesiculodrilus) evansi sp. n.
Fig. i, gA, 156, i6K. Table 2
1 = 51 (?) mm (the paratype damaged anteriorly), w = 2-6, 2-5 mm, s = 102, 117.
Prostomium epilobous §, closed ; first dorsal pore (2/3 ?), 3/4 (holotype). Setae in
8 regular longitudinal rows throughout. Nephropores in c lines. Clitellum annular,
XIII -XVI. Male pores on small papillae in a on XVIII. Five transversely ellip-
tical, unpaired, midventral genital markings in the anterior half of VII, VIII, IX,
XVIII and XIX (holotype), that in IX absent in paratype, and a pair of eye-like
markings at 18/19 with centres in b (both specimens). Female pores a pair in XIV
anteromedian of a. Spermathecal pores 2 pairs in 7/8 and 8/9, in a.
Last hearts in XII (laterb-oesophageal). Supra-oesophageal well developed.
Gizzard in V (damaged and not recognizable in paratype), moderately large but fairly
easily compressible. Calciferous glands absent but oesophagus internally rugose in
VI -XV, especially so in XIV and XV in which it is conspicuously dilated, but lacking
specialized internal lamellae. Intestinal origin XVI ; typhlosole absent. Nephridia
simple stomate vesiculate exonephric holonephridia ; first seen in III but pores
apparently commencing anteriorly in II ; funnels present at least as far forward as
VII ; the bladders at first pyriform but by IX becoming distinctly bilobed, the ectal
nephridial duct joining the median lobe. Holandric (funnels very large, only those
in X iridescent in the holotype, iridescent in XI also in the paratype) ; seminal
vesicles racemose, in IX and XII. Metagynous ; ovisacs very large, multiloculate,
in XIV (holotype ; not demonstrable in paratype) . Prostates appearing racemose,
in XVII-XX, XXI but resolvable into a compressed and minutely lobulated tubular
form. Penial setae (holotype) needle like, not filiform ; tip smooth and rounded ;
ornamentation absent, length = 0-71 mm. Spermathecae 2 pairs ; diverticulum
(inseminated) single, clavate, uniloculate.
MATERIAL EXAMINED.
Lake St Clair, i46°io'E. 42°O5'S. Map, 13/5, Dr J. W. Evans, Feb 1941 - holotype TM :
K282. Same locality, Tasmanian Biological Survey : J 20, Feb 1941 -paratype BM(NH) :
1972.8.8.
TASMANIAN EARTHWORMS
237
sp.p 5
1mm
FIG. 10.
1mm D
Genital fields. Perionychella ( Vesiculodrilus) glandifera : A, holotype ;
B, paratype.
REMARKS. This species appears to be closest morphologically to P. (V.) lacustris.
Some differences from the latter are the lumbricine setae, the two pairs of spermathe-
cal pores, the presetal rather than postsetal location of the anterior genital markings
and the configuration of those of the male genital field.
Perionychella (Vesiculodrilus) glandifera sp. n.
Fig. i, loA, B, 150, i6L, M. Table 2
1 = 75, 73 mm, w = 6-5, 7-5 mm, s = 98.
Prostomium canaliculate, broadly tanylobous. Body rectangular in cross section.
First dorsal pore 3/4 (imperf orate ? holotype), 4/5 perforate. Setae 8 per segment in
regular longitudinal rows throughout. Nephropores in c lines. Clitellum annular,
XIII- £X VIII (holotype ; the paratype is aclitellate) . Male pores equatorial in a lines
of XVIII on weakly developed papillae on the setal ridge ; 2 pairs of large sucker-like
genital markings with pore-like centres in intersegments 17/18 and 18/19, the centres
lateral to b lines, each marking spanning the distance between the setal arcs of ad-
jacent segments ; a transverse glandular depression linking each marking with that
238 B. G. M. JAMIESON
of the other side (holotype). Two pairs of eye-like transversely conjoined markings
shortly lateral of b at 17/18 and 18/19 m paratype. Female pores shortly anterior
to the setal arc of XIV at 1/3 aa. Spermathecal pores minute, 5 pairs in 4/5-8/9 in a
lines.
Dorsal vessel segmentally bifid in VIII (paratype), IX (holotype) -XVIII ; last
hearts in XI I (latero-oesophageal). Supra-oesophageal weakly developed. Gizzard
large, in V but extending well posteriorly. Oesophagus in XV expanded to form a
subspherical unpaired calciferous gland with many closely packed internal radial
laminae which fill the oesophageal lumen but do not unite centrally. Intestinal
origin ^XVII (paratype) or posterior XVII (holotype) ; typhlosole absent. Nephri-
dia vesiculate, exonephric holonephridia discharging by large bladders throughout,
commencing in II (holotype) ; those in II forming small tufts and lacking detectable
funnels (no nephridia detectable in II in paratype) ; the remaining nephridia simple
and each with a preseptal funnel ; bladders in the anterior segments very long, wide
tortuous tubes ; in VI (holotype) or somewhat more posteriorly (paratype) the blad-
der is bent near its ectal end and the bend is drawn out slightly as a diverticulum ;
succeeding bladders are increasingly diverticulate ; full development of the diverti-
culum as a long lateral digitiform process is attained by segment XIX and is main-
tained to the caudal extremity although caudal nephridia are smaller than those in
the anterior region. Holandric (funnels iridescent) ; cleistorchous, pericardiac
testis-sacs formed by longitudinal fusion of septa 9/10, 10/11 and 11/12 ; seminal
vesicles racemose in XI (in the testis-sac) and XII (free). Metagynous. Ovisacs
absent (holotype) or present in XIV (paratype). Prostates depressed tubular,
compactly coiled, in XVIII-XX (holotype) ; restricted to XVIII and not fully
developed in paratype ; vas deferens joining ectal end of gland (holotype) ; penial
setae absent. Spermathecae 5 'pairs ; diverticulum (inseminated only in holotype)
single, clavate, uniloculate.
MATERIAL EXAMINED.
St Helens, on road to Launceston via Scottsdale between 94 and 96 mile posts, I48°I5'E.
4i°2o'S. Map, 10/2, Dr J. L. Hickman, 26 Aug 1953 - holotype TM : K283. St Columba Falls,
I47°55'E. 4i°2o'S. Map, 9/2, Dr J. L. Hickman, 17 Apr 1954 - paratype BM(NH) : 1972.8.9.
REMARKS. Perionychella (V.) glandifera is unique among the known Tasmanian
species of the genus in the bifid dorsal vessel, the extramural oesophageal gland in
XV, in possessing testis-sacs and in location of the seminal vesicles in XI and XII.
Perionychella (Vesiculodrilus) hellyeri sp. n.
Fig. i, 96, i6N, O. Table 2
1 = 54, 62 mm, w = 2-6, 3-0 mm, s = 104, 97.
Prostomium indistinctly epilobous |, open. First dorsal pore 4/5. Setae 8 per
segment, the rows becoming irregular in the posterior 12 or 40 segments in which
there are 20-24 setae per segment. Nephropores in c lines. Clitellum annular,
£XIII-XVI. Male pores on prominent rounded papillae, equatorial in XVIII, in b ;
TASMANIAN EARTHWORMS 239
the papillae transversely conjoined by a midventral boss which bears presetally two
widely conjoined pore-like markings, a similar boss present with conjoined presetal
markings on XVII ; a pair of eye-like genital markings present in 17/18 in b lines
(holotype) ; presetal markings absent from XVII in paratype I and the median
boss in XVIII replaced with 2 small circular markings in aa joined to the male
papillae by slight ridges ; eye-like markings present in 17/18. Two pairs of eye-like
markings presetally in b, on IX and X in paratype I ; absent in holotype. Female
pores paired, on XIV, anteromedian of a. Spermathecal pores 3 pairs, in 6/7-8/9,
in b, on small papillae each of which is preceded by a lip-like prominence (holotype,
paratypes 1-5).
Last hearts in XII (latero-oesophageal). Supra-oesophageal in VIII-XIII, well
developed. Gizzard vestigial (questionably present), in V. Extramural calciferous
glands absent but oesophagus enlarged and vascularized with lamellar internal rugae
in XIV and XV. Intestinal origin XVII ; typhlosole absent. Nephridia simple,
vesiculate exonephric holonephridia ; preseptal funnel demonstrated in those of
III ; possibly rudimentary nephridia in II ; bladder large, subspherical to trans-
versely elliptical, the lateral extremity protuberant and almost forming a diverti-
culum. Holandric (gymnorchous) or probably incipiently metandric ; only posterior
funnels iridescent and seminal vesicles large in XII but vestigial in IX (holotype) ;
male organs of IX and X completely absent in paratypes I and 2 (metandric).
Metagynous ; ovisacs in XIV. Prostates flattened tubular, with lobulated, com-
pacted coils in XVIII-XX, XXI. Penial setae (paratype i) moderately stout, not
filiform, but ectal region sigmoid ; tip simple ; ectally weakly ornamented with
anteriorly directed ^-shaped scales flush with the surface and commonly in triads ;
the tip of a reserve seta (paratype 5) has a delicate, hooked extremity ; length of a
well-developed seta (paratypes 2 and 5) = 1-18, i-oo mm. Spermathecae 3 pairs ;
diverticulum (inseminated) single, clavate, uniloculate.
Genital markings of the holotype and 5 paratypes
Total number
Specimen of specimens
Paired presetal in a of VII — Pi, 5 2
Paired presetal in a of VIII - PS i
Paired presetal in a of IX - PS, 5 2
Paired presetal in b of IX - Pi i
Paired presetal in i of X - Pi I
Paired presetal in aa of XVII H Pa, 3, 5 4
Paired presetal in aa of XVIII H Pi -5 6
Paired in b in 17/18 H Pi -5 6
H = Holotype ; P = Paratypes .
Spermathecal pores and male porophores invariable in 6 specimens.
MATERIAL EXAMINED.
Hellyer Gorge, I45°35'E. 4i°2o'S. Map, 7/3, Dr J. L. Hickman, 28 May 1954 - holotype
TM : Ka84 ; paratypes i and 2 BM(NH) : 1972.8.10-11 ; paratypes 3 and 4 AM : W5I92-
5193 ; paratype 5 BJ : 115. Mt Arthur, in rainforest, i47°2o'E. 4i°i5'S. Map, 9/1, Mr A. J.
Dartnall, 13 Mar 1971 - 2 specimens TM : 1^285-286.
240
B. G. M. JAMIESON
XIV
XVI-
sp.pl
sp.p 5
sp.p 1
sp.p 5
sp.p
sp.p 5
1mm
1mm
TASMANIAN EARTHWORMS 241
REMARKS. This species is close morphologically to P. (V.) montisarthuri but is
readily distinguished from it by the presence of postclitellar genital markings in aa.
Perionychella (Vesiculodrilus) hobartensis (Spencer, 1895)
Fig. i, nA-F, I5A, B, i6P-R ; Plates 97, 98. Table 2
Cryptodrilus hobartensis Spencer, 1895 : 37-38, PL I, fig. 10, u, 12.
Plutellus hobartensis ; Michaelsen, 1900 : 175-176 ; Jamieson, 19710 : 87.
Cryptodrilus insularis Spencer, 1895 : 41-42. PL 2 ; fig. 19-21 (new synonym).
Plutellus insularis ; Michaelsen, 1900 : 176 ; Jamieson, 19710 : 87.
Vesiculodrilus hobartensis ; Jamieson, I973a : 225.
Vesiculodrilus insularis ; Jamieson, i973a : 225.
1 = 28 mm+ (posterior regenerate), w = 2-7 mm, s = 91 + .
Form moderately stout ; approximately circular in cross section but flattened
between adjacent setae. Prostomium epilobous f, open but apparently with a
transverse groove at about half peristomium ; a middorsal groove commencing on
the prostomium and continuous throughout the body. First dorsal pore 4/5. Setae
readily visible, in 8 regular longitudinal rows throughout, commencing on II ; ab
absent, cd present, on XVIII.
Nephropores in c lines at the anterior borders of their segments (clitellum and
posteriorly). Clitellum annular, strongly protuberant on XIV-XVI but XVII with
clitellar modification, especially dorsally over the anterior third ; dorsal pores present
but imperforate ; setae and intersegmental furrows retained but less distinct ;
nephropores visible. Small indistinct (imperforate) male porophores in ab nearer
b, of XVIII ; paired eye-like accessory genital markings in a lines in 16/17, 17/18 and
18/19, the anterior pair large and more conspicuous than the second pair, the pos-
terior pair rudimentary. Female pores minute, shortly anteromedian of a in XIV,
each with a distinctly visible white 'halo'. Spermathecal pores not visible.
Last hearts in XII (latero-oesophageal) ; connections of the 4 pairs of hearts with
the dorsal vessel presumably present but not verified with certainty as hearts were
brittle and were separate from the dorsal vessel on examination. Supra-oesophageal
well developed in VIII-XII, and with a slender posterior continuation into XIII and
apparently intramurally to |XV.
Gizzard in V, almost cylindrical and only slightly wider than the oesophagus but
recognizable by its muscular sheen ; easily compressible ; preceded in IV by a less
muscular proventriculus of the same width, from which it is not separated by any
appreciable constriction. Oesophagus narrow (though not much narrower than the
gizzard) and not evidently vascularized in VI and VII ; still narrow but with intra-
mural circumferential vascular striae, which form the supra-oesophageal vessel in
VIII-XI ; forming a segmental unpaired annular dilatation in each of segments XII,
FIG. u. Genital fields. Perionychella (Vesiculodrilus) hobartensis : A, lectotype ; B,
13/6, TM : K293 ; C, 14/2, TM : K299 ; D, 14/1, TM : K294 ; E, 19/2, AM :
F, 14/6, TM : K3o6.
242 B. G. M. JAMIESON
XIII and XIV, the dilatations increasing in size posteriad ; the oesophagus also
widened, but less so, to ^XV but narrow and chloragenous looking in the posterior
half of this segment ; the internal walls of the dilatations with rounded longitudinal
ridges too low to be considered lamellae. Intestinal origin XVII ; typhlosole
absent.
Genital markings
Specimen Paired genital markings
15/16 16/17 17/18 18/19 19/20 20/21
Spencer, 1895 + + + +
Lectotype + + + ? +
7/2 TM : K288 + + + +
8/2 TM : K28g + R
8/2 TM : K290 + +
10/1 TM : K2Qi + + +
10/1 TM : K292 + + +
13/6 TM : K293 +
14/1 TM : K294 R +
14/1 TM : K295 +
14/1 TM : K296 +
14/1 TM : K297 +
14/2 TM : K299 + + +
14/3 TM : K3oo + + + ?
14/4 TM : K302 + L + +
14/5 Lenah V TM : K304 + + + +
14/5 Sandy Bay TM : K305 + + + + R
14/6 Risdon TM : K3o6 + +
14/6 E. Risdon BM(NH)
14/6 BM(NH)
14/6 BM(NH)
1972.8.15 + + +
1972-8.16 + + +
1972.8.17 + + +
19/2 AM : W5I94 + + R
19/2 AM : W5I95 + +
19/3 TM : K3o8 + +
L = left only ; R = right only.
Nephridia simple vesiculate holonephridia throughout ; the first discharging at
intersegment 2/3, each with a wide elongate bladder bent into a U or J shape, all
bladders discharging in front of setae c ; 'necks' to septa seen but no funnels demon-
trated with certainty. Holandric (funnels iridescent in X and XI) ; seminal vesicles
not recognizable. Prostates represented only by thick muscular ducts each with the
form of a short, straight, entally narrowing cone. Penial setae not detectable in the
lectotype and not reported by Spencer (1895) ; demonstrated in one or (loc. 19/2) two
specimens from each of localities 7/2, 8/2, 10/1, 14/1, 14/5, 14/6 and 19/2 ; filiform
and sinuous, tip frequently worn or damaged but intact in 14/5 and 14/6 in which it
is widely bifid, while simple and hooked in 7/2 ; ornamentation a few faint longi-
tudinal ridges (14/5, 14/6) or groups of several scarcely protuberant, approximately
palmately arranged, apically directed teeth at intervals near the tip (7/2) or a few
very faint triangular marks and longitudinal ridges (14/6) ; length = 0-97, 1-56 mm
(19/2), 1-71 mm (14/6). Ovaries (laminate with several strings of large oocytes) and
TASMANIAN EARTHWORMS 243
stout funnels in XIII ; ovisacs absent. Spermathecae 5 pairs, the last in IX ; with
ovoid ampulla sessile on the body wall and a lateral clavate diverticulum as long as
or longer than the ampulla. In one spermatheca the diverticulum is very short
and there is a further diverticulum on the median aspect of the spermatheca further
entally. Length of the right spermatheca of IX = 0-75 mm ; ratio total length :
length diverticulum = 0-8.
Other variations. Spermathecal pores are 4 pairs in the material listed above
from localities 7/2 and 14/1, but 5 pairs in all other specimens examined. The gizzard
is very large in specimens from locality 10/1, small to vestigial in all others. In-
testinal origin is XVII (Spencer 1895 ; lectotype ; localities 8/2, 10/1, 13/6, 14/2,
14/3, 14/4, 14/5, 14/6) ; JXVII (locality 14/1) ; XVI (locality 14/6, 19/2) or |XVI
(locality 7/2) . Penial setae are present in all specimens examined with the exception
of the lectotype. The latter specimen lacks seminal vesicles, has prostate ducts
but no glands and has abnormal Spermathecae. These features suggest that it is a
parthenogenetic, uniparental morph and that absence of penial setae, in a species
normally possessing these, is further evidence of parthenogenesis.
Ovisacs were demonstrated for specimens from all localities excepting 13/6 and
14/5 and the lectotype.
TYPE-LOCALITIES. Parattah and Mt Wellington.
MATERIAL EXAMINED.
Parattah, I47°25'E. 42°2o'S. Map, 14/1, Baldwin Spencer collection, Jan 1893 -the pre-
viously undissected, clitellate lectotype, in poor condition, NMV : 050. The two paralecto-
types, 051, were examined but were not used for the above account as one is broken in two at the
clitellum and the other is imperfectly clitellate.
The following new material has been identified on external and internal anatomy : Fern Glade,
Emu River, I45°55'E. 4i°o5'S. Map, 7/2, Dr J. L. Hickman, 24 Aug 1954 -TM : K28y-288.
Great Lake, near stones around base of gum tree, slopes of lake, i46°45'E. 4i°55'S. Map, 8/2,
Dr J. L. Hickman, 26 May 1954 ~ TM : 1^289-290. Goulds County, near Lottah, fireplace at
creek, I48°05'E. 4i°i5'S. Map 10/1, Dr J. L. Hickman, 16 Apr 1954 -TM : K2gi -292.
Marlborough Highway, near Bronte, i46°3o'E. 42°io'S. Map, 13/6, Dr J. L. Hickman, 26
May I954-TM : K293. Parattah (type-locality) I47°25'E. 42°2o'S. Map, 14/1, under moss,
in earth at base of cliff, also in earth along sides of logs. Map, 14/1, Professor V. V. and Dr J. L.
Hickman, 18 Aug 1954 -TM : 1^294-298. Tunnack, under logs and stones, damp conditions,
i47°3o'E. 42°25'S. Map, 14/2, Dr J. L. Hickman, 18 Aug 1954 - ™ : K299. Collinsvale near
Hobart, Myrtle Forest, I47°O5'E. 42°5o'S. Map, 14/3, Dr J. L. Hickman, 8-9 Nov 1955 -
TM : K300-30I. Mt Wellington (type-locality) I47°I5'E. 42°55'S. Map, 14/4 : under stones,
creek, Bett's Vale, Dr J. L. Hickman, 4 Mar 1954 - ™ : K3°2 > Shoobridge Bend Track,
approx. 580 m altitude, in loam and clay in Eucalypt-fern woodland, B. G. M. Jamieson and
E. Bradbury, 19 Aug 1971 -TM : 1X303. Lenah Valley, near Newton Falls, i47°2o'E. 42°5o'S.
Map, 14/5, Dr J. L. Hickman, 24 Jun 1957 - TM : 1X304. Mt Nelson, Sandy Bay, i47°2o'E.
42°55'S. Dr J. L. Hickman, n Sep 1953 - TM : 1^305. Risdon, i47°2o'E. 42°5o'S. Map,
14/6, Professor V. V. Hickman, 26 Jun 1947 - TM : 1^306-307 ; East Risdon, from under
stones on hill and in valley, Dr J. L. Hickman, 14 Aug 1954 - BM(NH) : 1972.8.15-17. Eagle-
hawk Neck, under bark, i47°55'E. 43°oo'S. Map, 19/2, Dr J. L. Hickman 1954 - AM : W5I94-
5198 ; BJ : Ti6-i8. Tinderbox, under fallen eucalyptus leaves, i47°2o'E. 43°O5'S. Map,
19/3, Dr J. L. Hickman, 4 Aug 1957 - TM : KsoS.
REMARKS. The above account of the lectotype confirms and considerably aug-
ments Spencer's description. As shown in the table of field-variation, the accessory
244
B. G. M. JAMIESON
PQ
*
o
O
1
S
N 0,'
IH '
6
TASMANIAN EARTHWORMS 245
genital markings were more numerous in the material described by Spencer but the
evidence offered by Jensz and Smith (1969) for regarding the lectotype which they
selected as part of the type-series is satisfactory. Location of paired genital markings
in intersegmental furrow 16/17 is almost constant for the species, the sole exceptions
in the material examined being those from Goulds County (10/1). The latter speci-
mens are also exceptional in having genital markings in 20/21 (seen, however, in the
right side in a Sandy Bay specimen (14/5)) and have a very large gizzard. The
possibility that these, and perhaps specimens from other localities, are subspecifically
or specifically distinct from the type-population deserves further investigation but
on present evidence recognition of distinct taxa is not justified.
Perionychella (Vesiculodrilus) lacustris (Stephenson, 1924)
Fig. i, 76, 16!
Perionyx lacustris Stephenson, 1924 : 546-547.
Diporochaeta lacustris : Jamieson, I97ib : 83.
1 = 45-60 mm, w = 2-4-3-0 mm, s = 95-107.
Prostomium epilobous f, wedge-shaped (type) or tanylobous (Bronte). First
dorsal pore 4/5. Setae on XII 18 (type), 22 (Stephenson), 14 and 16 (Bronte) ;
caudally (Bronte) 16 and 23 ; 20 on V and XIX, 22 on IX (Stephenson). The dorsal
setal gap regular anteriorly and in the midbody equalling 2-$yz, but irregular and
indistinct caudally ; ventral gap regular throughout, equalling or > 2ab anteriorly.
Clitellum annular |XIII-|XVII. Male pores in b on prominent ovoid papillae
occupying ac on XVIII. Paired eye-like markings in be in 16/17, I7/1^ and 18/19,
those of 17/18 slightly median of the others ; a median sucker-like pad postsetal
ventrally in each of VII and VIII ; an indistinct glandular tumescence present on
each side in ac on XVII and XIX. Female pores paired in XIV anteromedian of a.
Spermathecal pores 3 pairs, in 6/7 to 8/9, on small papillae in cd, closer to c.
Last hearts in XII (latero-oesophageal) . Supra-oesophageal in VIII-XII, well
developed. Gizzard small, almost vestigeal, in V, hidden in septal glands. Calci-
ferous glands absent but in the type the oesophagus is swollen and vascularized, and
has low internal rugae, in XIV and XV. Intestinal origin |XVII ; typhlosole absent.
Nephridia simple, stomate, vesiculate, exonephric holonephridia commencing (but
rudimentary ?) in II (preseptal funnel demonstrated in XIV) ; bladders large, sub-
spherical, tapering ectally or, in the type, represented at most by slight intraparietal
dilatations of the terminal duct. Holandric (funnels iridescent) ; gymnorchous ;
seminal vesicles racemose, in IX and XII. Metagynous ; ovisacs large in XIV.
Prostates tubuloracemose, in XVII to XX (type) or to XXI, XXII (Bronte) ; only
partially resolvable into a simple linear arrangement ; depressed, deeply incised and
tabulated ; vas deferens (see in Bronte specimens only) joining gland near its
muscular duct. Penial setae delicate, filiform, sharply pointed ; apical extremity
rounded ; ornamentation absent ; length ? Penial setae present in the type but not
examined. Spermathecae 3 pairs ; diverticulum (inseminated) single, clavate,
uniloculate.
246 B. G. M. JAMIESON
TYPE-LOCALITY. Great Lake, under stones in water at margin of the lake,
i46°45'E. 4i°55'S. Map, 8/2.
MATERIAL EXAMINED.
Syntypes, BM(NH) : 1924.10.21.1-5, of which one, referred to as 'type' in the above account,
was closely examined.
Lyell Highway, 5 miles from Bronte towards Hobart, i46°35'E. 42°i5'S. Map, 13/7, Dr J. L.
Hickman, 24 May 1954 ~ tw° specimens TM : K265 and BM(NH) : 1972.8.2.
REMARKS. The new description of a type-specimen in the above account con-
siderably augments Stephenson's description. In this type-specimen, which had not
been previously dissected, and in the single one of the type-specimens which had
previously been dissected, the following observations of Stephenson are not confirmed
and are considered incorrect : clitellum in XIV-XVI ; gizzard in VI ; intestinal
origin (about) XVIII ; penial setae absent. Genital markings are similar in the
two specimens and include midventral markings in VII and VIII.
P. (V.) montisarthuri is morphologically close to this species but shows sufficient
differences to be considered a distinct species. These differences include the setal
counts, the bilobed nephridial vesicles, the sessile spermathecae and the precise
configuration of the genital field.
Perionychella (Vesiculodrilus) montisarthuri sp. n.
Fig. i, I2A, 15!!, 168. Table 2
1 = 79 mm, w = 3-6 mm, s = 86.
Prostomium epilobous \, open. First dorsal pore 3/4. Setae on XII 8 (10 on XI) ;
caudally, 14 or 16 per segment " the dorsal gap regular anteriorly and, in the mid-
body, equalling 2,yz, irregular caudaUy ; ventral gap regular throughout, equalling
2ab anteriorly. Nephropores in a single straight series on each side in setal line 3
anteriorly and in the mid-body ; caudally in s.l. 4-5. Clitellum, annular £XIII-
XVI. Male pores in b on small, flat, ovoid markings on XVIII ; two ovoid genital
markings occupy the space from a to a little lateral of b at 17/18 and 18/19. Female
pores paired on XIV, anterior and slightly median to a. Spermathecal pores 3 pairs
in b in 6/7-8/9, on small papillae each of which is succeeded by a small ovoid marking
on the posterior segment.
Last hearts in XII (latero-oesophageal) . Supra-oesophageal in VIII-XII, well
developed. Gizzard in V, moderately large but only weakly muscular. Calciferous
glands absent. Oesophagus with circumferential vascular striae in VII-XIV,
widest in XIV and XV in which internally there are numerous radial lamellae which
almost occlude the lumen ; narrow in XVI in which there are a few, low radial folds.
Intestinal origin |XVII ; typhlosole absent. Nephridia simple stomate, vesiculate,
exonephric holonephridia commencing in II (funnels demonstrated in the forebody) ;
bladders at first elongate avoid but by VI with a suggestion of a lateral basal lobe ;
thereafter the lateral lobe increases in size relative to the median portion (the latter
receiving the nephridial duct) until, by about XIX, the bladder is broader than long
and consists of two approximately equal lobes, the lateral lobe constituting a broadly
TASMANIAN EARTHWORMS 247
sessile diverticulum ; caudally the bladders are less distinctly bilobed. Holandric
(funnels iridescent in X and XI), gymnorchnous ; seminal vesicles racemose, in IX
and XII. Metagynous ; ovisacs small in XIV. Prostates depressed very tortuous
tubes, in XVII-XXI ; double vas deferens joining the gland shortly ental of the
slender, sinuous muscular duct. Penial setae (holotype) slender, strongly curved,
almost filiform with delicate thinner, elongate tip ; ensheathed in setal follicle and
not suitable for electron microscopy ; length (?), incomplete. Spermathecae 3
pairs, diverticulum (inseminated) single, clavate, uniloculate, exceptional in joining
the ectal end of the broad subspherical ampulla ; ampulla almost sessile on the
parietes.
MATERIAL EXAMINED.
Mt Arthur, from rainforest, i47°2o'E. 4i°i5'S. Map, 9/1, Mr A. J. Dartnall and Mr R. C.
Kershaw, 15 Oct 1971 - holotype TM : Ksog.
REMARKS. This species is morphologically close to P. (V.) hellyeri (q.v.)
Perionychella (Vesiculodrilus) mortoni (Spencer, 1895)
Fig. i, I2B, I5E-G, i6T ; Plates 99-102. Table 2
Cryptodrilus mortoni Spencer, 1895 : 36-37, PI. i, fig. 7-9.
Plutellus mortoni ; Michaelsen, 1900 : 176.
Woodwardiella mortoni ; Jamieson, 1970 : 104, 105 ; 19710 : 92.
Dimensions (see Variation). Anatomy (specimen i, locality 14/4, 19 Aug 1971) :
Prostomium canaliculate, tanylobous. First dorsal pore 4/5. Setae in 8 regular
longitudinal rows throughout. Nephropores conspicuous, commencing anteriorly in
II in c lines but shortly below c lines behind the midclitellum. Clitellum annular,
XIII-XVII, strongly developed and pigmented in XIV-XVII, unpigmented and
weakly developed in XIII ; possibly some slight clitellar modification in XVIII.
Male pores on small papillae in ab of XVIII. Accessory genital markings : paired
elliptical pads with depressed centres in ab in 17/18, 19/20, 20/21 and 21/22, those in
17/18 with centres slightly lateral of those of the other markings and truly inter-
segmental ; the markings in 19/20-21/22, though intersegmental actually postsetal
in XIX-XXI, respectively ; a median elliptical pad present in 17/18 bridging the
paired markings of the intersegment. Median markings each with the form of a
depressed transverse intersegmental band with anterior rim in aa in 5/6-8/9. Female
pores anteromedian of setae a of XIV, midway between the setal arc and anterior
border of the segment in a common oval field. Spermathecal pores 5 pairs on
minute circular papillae immediately in front of intersegmental furrows 4/5-8/9, in
a lines.
Last hearts in XII (latero-oesophageal) . Supra-oesophageal well developed, in
VIII-|XIII. Gizzard very large and firm, in V but projecting posteriorly to the
level of XL Calciferous glands absent. Oesophagus strongly vascularized and
increasingly dilated in XIII-XVI with well-developed internal vascular rugae,
especially in XV and XVI. Intestinal origin XVIII ; typhlosole absent. Neph-
ridia stomate, vesiculate exonephric holonephridia commencing in II (preseptal
248 B. G. M. JAMIESON
funnel demonstrated for those of II) ; tortuously coiled but simple, none tufted ;
bladders at first elongate, wide tubes, each of slightly irregular diameter ; by XII
bent midway at a right angle ; thereafter with a lateral diverticulum extending from
the bend, the portion ectal to the diverticulum being eliminated after a few segments.
Holandric (funnels iridescent in X and XI) ; gymnorchous ; seminal vesicles race-
mose in IX and XII. Metagynous ; ovisacs present. Prostates much coiled
depressed tubes, in XVIII-XXI ; vas deferens joining the duct shortly ectal of the
gland. Penial setae short and moderately stout, the tip irregular, widened, roughly
spatulate and bent a little or through a right angle, this apical modification visible
under the light microscope ; shaft ornamented ectally with a few groups of anteriorly
directed scarcely protuberant teeth regularly spaced along it ; the surface of the seta
undercut beneath them ; length 0-5 mm (3 setae, Mt Wellington and East Risdon
specimens, as illustrated). Spermathecae 5 pairs ; diverticulum (inseminated)
single, clavate, uniloculate.
Variation. A minimum of two specimens from each of localities 7/2, 9/2, 14/3,
14/4, 14/5, 14/6 and the single specimen from 19/2 were examined for variation in
external characters. Variation in internal anatomy is described for at least one
specimen from each locality.
Genital markings. Postsetal genital markings were present in aa in each of
segments IV- VIII in all specimens but in those from Fern Glade (locality 7/2) these
were exceptional in being paired. Paired markings were present in 17/18 and post-
setally in each of segments XIX, XX and XXI, with the exception that those in XXI
were absent in one of the two specimens from each of localities 9/2, 14/3 and 14/5
and from most East Risdon specimens (14/6). The median genital marking in 17/18,
though characteristic of the species, is absent from most but not all of the East
Risdon specimens.
Spencer did not observe genital markings in IV- IX (these are often faint in the new
material) but he depicted median and paired markings in 17/18, as here described,
and paired markings in intersegments 18/19 and 19/20. The latter two pairs were
presumably postsetal rather than intersegmental and it appears probable from the
present study that they lay in segments XIX and XX.
Other variation. 1 = 56-212 mm, w = 3-5-6-5 mm, s = 113-274. Prostomium
epilobous |-f, epitanylobous or tanylobous ; often canaliculate. First dorsal pore
usually in 4/5, rarely in 3/4 or 5/6. Body strongly canaliculate to not canaliculate.
Variation from the description in internal characters occurs in the intestinal origin
which is XVII in 9/2 and 19/2 ; in the presence of a low dorsal intestinal ridge,
though no true typhlosole is present, in 9/2 and 14/6 ; and in the absence of diverti-
cula of the nephridial bladders in 9/2. Absence of nephridial diverticula is especially
noteworthy and is a variation hitherto unknown intraspecifically in otherwise diverti-
culate species. At present there is insufficient justification for taxonomic recognition
of the variant.
MATERIAL EXAMINED.
Approximately 100 sexual specimens were examined from the following localities : Fern
Glade, Emu River, Burnie, I45°55'E. 4i°o5'S. Map, 7/2, Dr J. L. Hickman, 24 Aug 1954.
Mt Arthur (east), wet sclerophyll forest, i47°2o'E. 4i°i5'S. Map, 9/1, Mr A. J. Dartnall and
TASMANIAN EARTHWORMS
249
.m
1mm
sp.p 3
sp.p 5
1mm
FIG. 13. Genital fields.
B
Perionychella (Vesiculodrilus) obliquae : A, holotype ; B, 17/1,
AM : W52O2.
Mr R. C. Kershaw, 15 Nov 1971. St Columba Falls, I47°55'E. 4i°2o'S. Map, 9/2, Dr J. L.
Hickman, 17 Apr 1954. Collinsvale, near Hobart, in Myrtle Forest, i47°io'E. 42°5o'S. Map,
14/3, Dr J. L. Hickman, 8-9 Nov 1955. Mt Wellington, I47°I5'E. 42°55'S. Map, 14/4,
Dr J. L. Hickman, 13 Jan 1954 • Dr J- L. Hickman, 27 Oct 1955 ; Mr W. Radford, Apr 1952 ;
Professor V. V. Hickman, 28 Jun 1947 ; 13 Sep 1951 - B J : Tig-20 ; Betts Vale, Dr J. L.
Hickman, 4 Mar 1954 -AM '• W5 199-5201 ; Shoobridge Bend track, approx. 580 m, Dr B. G.
M. Jamieson and Mr E. A. Bradbury, 19 Aug 1971 - BM(NH) : 1972.8.18-25. Hobart, Water-
works Road, under stones, i47°2o'E. 42°5o'S. Map, 14/5, Dr J. L. Hickman, 17 Aug 1954.
Lenah Valley, track along Newtown Creek, Dr J. L. Hickman, 15 Sep 1953 ; Lenah Valley,
under stones near Newtown Falls, Dr J. L. Hickman, 24 Jun 1957 - BM(NH) : 1972.8.26-30;
15
250 B. G. M. JAMIESON
Sandy Bay, Dr J. L. Hickman, Sep 1954 ; Domain, 14 Aug 1954. East Risdon, under stones
on hill and in valley, i47°2o'E. 42°5o'S. Map, 14/6, Dr J. L. Hickman, 14 Aug 1954 - BJ : T2I -
22 ; Risdon, Professor V. V. Hickman, 26 Jun 1947. Eagle Hawk Neck, i47°55'E. 43°oo'S.
Map, 19/2, Dr J. L. Hickman, 13 May 1954. Tasmania : T.M. I5527/K97, Dr J. L. Hickman.
All specimens lodged in the Tasmanian Museum (K97, K4O2-4I9) except where indicated above.
REMARKS. This species and P. (V.) hobartensis are the most widespread and
commonly found megascolecid earthworms in Tasmania and are morphologically
close. A notable difference is location of the paired genital markings at the posterior
borders of their segments in mortoni whereas in hobartensis they are intersegmental.
Perionychella (Vesiculodrilus) obliquae sp. n.
Fig. i, I3A, B, i6U, V
1 = 115 mm (holotype ; paratype i is a posterior amputee), w = 6-6, 6-0 mm,
s = 118.
Prostomium epilobous £, open. First dorsal pore 3/4. Setae on XII 27, 25 ;
caudally 40, 38 ; ventral break distinct throughout, equal to 2ab in forebody and $ab
posteriorly ; dorsal gap not distinct caudally being equal to dorsal couple (yz) ;
distinct anteriorly, equalling 2yz. Nephropores difficult to observe, anteriorly in
setal lines 6 and posteriorly in s.l. 8 (holotype) or 8-9 (paratype i). Clitellum
annular, |XII-XVII. Male pores on small papillae in b lines. Five pairs of eye-like
markings on the anterior margins of segments XX and XXI in b lines and on segments
XVII, XVIII and XIX in ab. Female pores paired anterior and a little median to
a on XIV. Spermathecal pores 5 pairs, in 4/5-8/9 in ab on small papillae.
Last hearts in XII (latero-oesophageal). Supra-oesophageal in VII-XIII, very
well developed. Gizzard small, almost vestigial, in V. Extramural calciferous
glands absent ; intestinal origin |XVII ; typhlosole absent. Nephridia simple
exonephric, vesiculate holonephridia commencing in II ; preseptal funnels present
at least from III posteriorly ; bladders long, tubular or somewhat dilated, bent into
a U or V shape. Holandric (funnels iridescent) ; gymnorchous ; seminal vesicles
racemose, in IX and XII. Metagynous ; large ovisacs in XIV. Prostates depressed
tubular, somewhat lobulated, with several bends compacted into a circular outline
and spuriously appearing racemose, in XVII-XIX. A small follicle but no seta
present in a of XVIII in the holotype ; penial seta from AM : W5202 very slender,
straight, unornamented (under light microscope), length = 0-58 mm ; a vestigial
seta from paratype under the electron microscope displays no ornamentation ; tip
pointed, aquiline, length = 0-34 mm. Spermathecae 5 pairs ; diverticulum (in-
seminated) single, clavate, uniloculate.
MATERIAL EXAMINED.
Obliqua-iorest, 2 miles inland, south of Interview River, I44°55'E. 4i°35'S. Map, 6/1, Mr
W. Jackson, 31 Dec 1953 - holotype TM : KSIO ; paratype i BM(NH) : 1972.8.31. Port
Davey, Kelly's Beach, I45°55'E. 43°2o'S. Map, 17/1, Tasmanian Biological Survey : J 17,
Jan 1940 - AM : W5202.
REMARKS. Differences from the Interview River type-specimens shown by the
Port Davey specimen are as follows : clitellar width 2-7 mm ; first dorsal pore 4/5 ;
TASMANIAN EARTHWORMS 251
setae 22 in segment XII ; accessory genital markings absent at 17/18 and 18/19 »
a midventral circular genital marking present posteriorly in VII (possibly an arte-
fact) ; gizzard moderate in size ; intestinal origin XVII, not |XVII ; nephridial
bladders subspherical though ectally tapering. These differences do not appear
sufficient for recognition of a distinct species for the Port Davey material, especially
as only a single specimen is available. Larger series from both localities are needed.
The genital field of this species is reminiscent, though substantially different from,
that of P. (V.) hobartensis which obliquae closely resembles. It differs conspicuously
from hobartensis and from P. (V.) tunnackensis, to which it is also morphologically
close, in the perichaetine arrangement of its setae.
Perionychella (Vesiculodrilus) pedderensis sp. n.
Fig. i, I2C, i6W
1 = 50 mm, w = 1-6 mm, s = 123.
Prostomium canaliculate, indistinctly epilobous £, open. Dorsal pores absent.
Forebody dorsoventrally depressed ; hindbody with wide dorsal gutter ; anus
terminal but deeply incising the pygidium dorsally and ventrally. Setae 10 per
segment anteriorly, increasing to 12 per segment caudally, rows regular ; dorsal and
ventral gaps large, the dorsal larger. Nephropores ? Clitellum annular, XIII-XVI.
Male pores in ab of XVIII on small papillae. Accessory genital markings : a mid-
ventral tumid pad traversing bb in each of intersegments 17/18 and 18/19, eacn Pad
with a deep groove corresponding with the intersegment ; paired eye-like genital
markings centred in a lines in 19/20 and 20/21 (that on the left in 20/21 barely per-
ceptible) and posteriorly in VIII shortly lateral of a lines. Female pores at | aa,
midway between the setal arc and the anterior border of XIV. Spermathecal pores
4 pairs, minute, in 5/6 shortly lateral of a, in 6/7 in ab, in 7/8 and 8/9 in b.
Last hearts XII. Suboesophageal VII I -XIII. Gizzard moderate, in VI.
Oesophagus simple ; intestinal origin XVII ; typhylosole absent. Holonephridia
commencing in II but appearing rudimentary to VII ; each with a thick-walled
subspherical bladder discharging presetally in c line ; preseptal funnel well developed.
Holandric (funnels iridescent) ; gymnorchous ; seminal vesicles racemose, in IX
and XII. Metagynous ; large ovisacs in XIV. Prostates thickly tubular, i pair,
in XVIII-XIX, or XX ; vas deferens joining ectal limit of gland. Penial setae
present (holotype) but structure not elucidated ; no visible ornamentation ; length
(incomplete ?) = 0-41 mm. Spermathecae 4 pairs ; diverticulum (inseminated)
single, clavate, uniloculate.
MATERIAL EXAMINED.
Lake Pedder, sorted from interstitial fauna, i46°io'E. 42°55'S. Map, 13/2, Mr D. Tyler,
25 Feb 1971 -holotype TM : K$II.
REMARKS. The absence of dorsal pores, which is unique in the genus in Tasmania,
and the reduction of the anterior nephridia are elsewhere correlated with an aquatic
existence in megadriles.
252
B. G. M. JAMIESON
XII
XVII
VIII
IX
1mm
sp.p
B
1mm
FIG. 14. Genital fields. Perionychella ( Vesi culodrilus) t unnackensis, holotype : A,
male field ; B, spermathecal field.
TASMANIAN EARTHWORMS 253
Perionychella (Vesiculodrilus) tunnackensis sp. n.
Fig. i, I4A, B, i6X. Table 2
1 = 33. 35 mm> w = 1-4, 1-3 mm, s = 82, 81.
Prostomium faintly canaliculate, epilobous £. First dorsal pore 4/5. Setae 8
per segment in regular longitudinal rows throughout. Nephropores in c. Clitellum
annular, 13-16. Male pores in ab on small papillae on 18, the two papillae joined by
a low median ridge, intersegmental ridges at 17/18 and 18/19 flank the male
papillae and correspond in extent to the ridge between the papillae. Paired eye-like
transversely conjoined genital markings in ab at 19/20 (holotype, paratype) and 20/21
(holotype) ; post-setal eye-like markings occur with centres in a in VII (paired in
paratype ; left only in holotype) ; VIII (paired) and IX (paired in paratype I ;
right only in holotype). Female pores a pair anteromedian of a on XIV. Sperma-
thecal pores 3 pairs in a at 6/7, 7/8 and 8/9.
Last hearts in XII (latero-oesophageal). Supra-oesophageal not distinguishable
from the well-developed oesophageal plexus. Gizzard in V, moderate to small,
muscular but easily compressible. Calciferous glands absent. Oesophagus with
circumferential vascular striae in (VII ?), VIII-XV ; in X-XV with at first few and
small, posteriorly increasingly numerous and larger longitudinal rugae which in XIV
and XV approach the appearance of laminae. Intestinal origin XVII ; typhlosole
absent. Nephridia simple stomate, vesiculate holonephridia, commencing in II
(funnels traced at least from IV) ; bladders subspherical, tapering to the pore ; large
in all but a few anteriormost segments. Holandric (funnels iridescent in X and XI) ;
gymnorchous ; seminal vesicles racemose in IX (?) and XII. Metagynous ; ovisacs
large, in XIV. Prostates depressed tubular with incised adpressed coils, in XVIII-
XX ; vas deferens joining the gland shortly ental to the muscular duct. Penial
setae (holotype) filiform ; tip pointed and slightly upturned ; no ornamentation ;
length = 0-76 mm. Spermathecae 3 pairs, decreasing in size anteriorly ; diverti-
culum (uninseminated) single, clavate uniloculate.
MATERIAL EXAMINED.
Tunnack, under logs and stones, damp conditions, i47°3o'E. 42°25'S. Map, 14/2, Dr J. L.
Hickman, 18 Aug 1954 - holotype TM : K3I2 ; paratype BM(NH) : 1972.8.32.
REMARKS. As in all Tasmanian species of the genus, the genital markings of this
species are distinctive. It is closest to P. (V.) hobartensis from which the 3 pairs of
spermathecal pores are an additional distinction.
Perionychella (subgenus ?) bassana (Spencer, 1895)
Megascolides bassanus Spencer, 1895 : 46-47, PI. Ill, fig. 34-36.
Plutellus bassanus ; Michaelsen, 1900 : 169 ; Jamieson, ig7ic : 87.
1 = 88 mm, w > 3 mm, s ?.
Prostomium prolobous. First dorsal pore in 4/5 (?). Setae 8 per segment.
Nephropores in c lines. Clitellum saddle-shaped, embracing XIV-XIX, ventral
254
B. G. M. JAMIESON
FIG. 15. Nephridial bladders in Perionychella. A & B, Perionychella (Vesiculodrilus)
hobartensis : A, anterior segments ; B, an intestinal segment. C, P. (V.) evansi, holoytpe,
D, P. (V.) glandifera, holotype. E-G, P. (V.) mortoni, 14/4, BM(NH) : 1972.8.18.
H, P. (V.) montisarthuri, holotype, in situ, showing setae c. (A & B, left bladders ; others,
right bladders.)
margins in ab excepting the anterior part of XIV, the whole of which is included.
Male pores on papillae in XVIII in a lines. Accessory genital markings : a median
ventral patch on each of XVII and XVIII ; and 2 papillae in XIX in a lines, con-
joined by a median transverse ridge. Female pores on XIV within the tumid annular
part of the clitellum near the midventral line. Spermathecal pores 2 pairs, in 7/8
and 8/9, in ab.
Dorsal blood vessel single ; last hearts in XIII ; no continuous supra-oesophageal.
Gizzard in V. True calciferous glands absent but vascular swellings in XIII and
XIV ; intestinal origin in XIX. Nephridia stomate holonephridia. Holandric ?,
sperm funnels in X and XL Seminal vesicles on anterior walls of X (?), XI and XII.
Metagynous ; ovisacs ?. Prostates tubular, small and coiled, in XVIII. Sperma-
thecae 2 pairs, in VIII and IX ; the diverticulum simple and less than half the length
of the sac.
TYPE-LOCALITY. King Island in Bass Strait.
TASMANIAN EARTHWORMS 255
MATERIAL EXAMINED.
None. The types are missing from the National Museum of Victoria.
REMARKS. It is not entirely certain that this species should be assigned to
Perionychella. The saddle-shaped clitellum is exceptional, and presence or absence
of a typhlosole is not recorded, but otherwise nothing in the brief description ex-
cludes it from the genus.
Perionychella (subgenus ?) decatheca (Michaelsen, 1910)
Plutellus decatheca Michaelsen, 1910 : 81-83, Fig- XIV-XVI ; Jamieson, 19710 : 87.
1 = 100 mm, w = 10 mm, s = 164.
Prostomium epilobous \, closed. First dorsal pore in 4/5. Setae widely paired ;
in the forebody aa : ab : be : cd : dd = 1-5:1:2:1-5: 4-5 ; dd = o-6w ; in the
hindbody 1-5:1:1-3:1: 2-5 ; dd = o-25«. Clitellum annular, XIV-XVIII. Male
pores on small papillae in XVIII, approximately in ab. Accessory genital markings :
paired transverse almost linear papillae anterior and posterior to the male papillae,
in XVII, XIX, XX and XXI extending medianly over a lines and laterally over b
lines ; the markings on XVII much larger than those of XIX-XXI, almost contigu-
ous medianly and nearly obscuring the male papillae. Female pores indicated by a
long transverse furrow anteriorly in XIV. Spermathecal pores 5 pairs, in 4/5-8/9,
in a lines.
Dorsal blood vessel single ; last hearts in XIII. A large gizzard in V. Calciferous
glands absent ; oesophagus in XIV and XV appearing more vascular than elsewhere,
but only slightly swollen. Holonephric. Holandric ; seminal vesicles racemose in
IX and XII. Prostates tubular ; duct short and thin. Penial setae ca. 1-2 mm
long, and 18 p,m thick, almost straight, proximally slightly curved, distally only
slightly tapered. Distal tip rather stout, bent into a small hook. Distal end of the
seta ornamented with wide deep teeth, which project very steeply from the seta ;
the teeth lying in 4 fairly regular spirally arranged lines. Spermathecae with single,
clavate, uniloculate diverticulum.
TYPE-LOCALITY. The vicinity of Hobart.
MATERIAL EXAMINED.
None.
REMARKS. Michaelsen considered P. decatheca to be close to P. insularis (Spencer,
1895) which is now a junior synonym of P. hobartensis. Possibly decatheca is merely
a variant of the sympatric hobartensis but differences from the latter, viz. setal rather
than intersetal accessory genital markings and the form of the penial setae preclude
union of the two taxa on present information. Location of the genital markings
anteriorly rather than posteriorly in their segments and apparently the form of the
penial setae exclude decatheca from the similarly sympatric and not dissimilar P.
mortoni. Known only from the holotype.
256
B. G. M. JAMIESON
IffllR
w
FIG. 16. Spermathecae in Perionychella. A & B, Perionychella (P.) capensis : A, holotype
(left IX) ; B, paratype i (left IX). C & D, P. (P). hickmani : C, holotype (right IX) ;
D, paratype 3 (right VIII). E, P. (P.) kershawi, holotype (right VIII). F, P. (P.) weld-
boroughi, holotype (right IX). G, P. (P.) irregularis, lectotype (left IX). H, P. (7.)
bithecata, holotype (left IX). I, P. (V.) lacustris, TM : K2&5 (left IX). J, P. (V.) dil-
wynnia, 13/4, TM : Ka66 (right IX). K, P. (V.) evansi, holotype (right IX). L & M,
P. (7.) glandifera : L, holotype (left IX) ; M, paratype (right IX). N & O, P. (V.)
hellyeri : N, paratype 2 (right IX) ; O, holotype (right VIII). P-R, P. (V.) hobartensis :
TASMANIAN EARTHWORMS 257
Perionychella (subgenus ?) ellisi (Spencer, 1895)
Fig. i.
Cryptodrilus ellisii (sic) Spencer, 1895 : 42-43, PL 2, fig. 22-24.
Plutellus ellisi ; Michaelsen, 1900 : 172 ; Jamieson, 19710 : 87.
1 = 25-38 mm, w = 3 mm, s ?.
Prostomium epilobous J, it and the body canaliculate. First dorsal pore in 5/6.
Setae 8 per segment throughout ; be = cd = dd. Clitellum annular, XIV-XVI or
posterior XIII also. Male pores in XVIII in b or perhaps slightly ventral of this, on
papillae. Accessory genital markings : paired elliptical patches in ab anteriorly in
X and XI ; in 17/18 and 18/19 and anteriorly in XX and XXI. Female pores
anteromedian of setae a of XIV. Spermathecal pores 3 pairs, in 6/7-8/9, in b lines.
Last hearts in XII. Gizzard in V. Sessile paired dorsolateral calciferous glands in
XIV and XV. Intestinal origin in XVII. Holonephric. Holandric ; seminal
vesicles racemose, in IX and XII. Prostates tubular, coiled, in XVII-XX. Meta-
gynous ; ovisacs in XIV. Spermathecae 3 pairs ; diverticulum short, simple.
TYPE-LOCALITY. Dee Bridge, under logs and stones, I46°35'E. 42°i5'S. Map,
13/7-
MATERIAL EXAMINED.
The four syntypes (NMV : 032) were found to be in poor condition and no longer yield useful
information.
Perionychella (subgenus ?) moroea (Spencer, 1895)
Fig. i.
Perichaeta moroea Spencer, 1895 : 49, PL IV, fig. 40-42.
Diporochaeta moroea ; Michaelsen, 1900 : 207 ; Jamieson, 19710 : 83.
1 = zoo mm, w ?, s ?.
Prostomium very slightly epilobous. First dorsal pore in 3/4. Preclitellar setae
11-18 on each side ; intraclitellar and postclitellar setae 21 per side ; dorsal and
ventral breaks very small. Nephropores in setal lines 9 or 10 in the midbody.
Clitellum ? Male pores on papillae between setal lines 3 and 4. Accessory genital
markings : median ventral ridge on XVIII between the male papillae ; a pair of
elliptical patches in ab, in 19/20 (or 18/19 ?)• Female pores on XIV. Spermathecal
pores 2 pairs, in 7/8 and 8/9, in ab.
Last hearts in XII ; supra-oesophageal present. Gizzard in V. True calciferous
glands absent. Intestinal origin XVII. Holandric ; seminal vesicles racemose, in
P, lectotype (right IX) ; Q, 14/1, TM : K2g4 (right IX) ; R, 14/6, BM(NH) : 1972.8.15
(right IX). S, P. (V.) montisarthuri, holotype (right IX). T, P. (V.) mortoni, 14/4,
BM(NH) : 1972.8.18 (right IX). U & V, P. (V.) obliquae : U, holotype (right IX) ; V,
17/1, AM: W5202 (right IX). W, P. (V.) pedderensis, holotype (right VIII). X, P.
(V.) tunnackensis, holotype (right IX).
258 B. G. M. JAMIESON
IX and XII. Metagynous. Ovisacs absent ? Prostates in XVII-XX. Sper-
mathecae 2 pairs ; diverticulum simple.
TYPE-LOCALITY. Lake St Clair, i46°io'E. 42°o5'S. Map, 13/5.
MATERIAL EXAMINED.
The single type-specimen (NMV : Gaga) is in a very refractory condition and yields no useful
information.
REMARKS. There is a discrepancy in Spencer's account : genital markings are
paired in 19/20 according to the text but in 18/19 m the illustration. If the markings
were in 19/20 there is a possibility that P. (P.) capensis (from Table Cape) is a junior
synonym. Both species are perichaetine. Differences from capensis indicated by
Spencer's account are location of the spermathecal pores in db (not c), absence of
accessory genital markings in 17/18 and intestinal origin in XVII (not £XVI). These
differences, if real, appear to warrant specific distinction of capensis but confirmation
of the distinctness of the two entities must await collection of new material of moroea.
P. (F.) evansi from the same locality as moroea has, like the latter and capensis,
2 pairs of spermathecae. It differs from moroea in having lumbricine setae, in the
genital fields and in other respects and conspecificity is not indicated.
Perionychella (subgenus ?) richea (Spencer, 1895)
Fig. i.
Perichaeta richea Spencer, 1895 : 49-53, PI. V, fig. 49-51.
Diporochaeta richea ; Jamieson, igyic : 83.
1 = 75 mm, w = 3 mm, s ?.
Epilobous \. First dorsal pore in 3/4. Setae 12 on each side anterior to clitellum,
24 half-way along the body. Clitellum annular, XIV- XVI I. Male pores on papillae
in ab. No accessory genital markings developed. Female pores on XIV. Sper-
mathecal pores 5 pairs, on small elliptical patches at the posterior margins of IV- IX,
in a.
Last hearts in XII ; supra-oesophageal in IX-XII. Gizzards in III and IV (?).
No true calciferous glands but oesophagus in XI and XII white and swollen. In-
testinal origin XVII. Holonephric. Holandric ; seminal vesicles racemose, in
IX and XII. Metagynous ; ovisacs in XIV. Prostates wide, tubular, in XVII-
XIX. Spermathecae 5 pairs ; the diverticulum very small.
TYPE-LOCALITY. Mount Olympus, under logs in the Beech forest, i46°io'E.
42°05'S. Map, 13/5.
MATERIAL EXAMINED.
None available.
REMARKS. Reported location of gizzards in III and IV, if accurate, would be
highly distinctive but this position is very questionable as it is unknown elsewhere in
the suborder Lumbricina. This species must be regarded as a species dubium as,
TASMANIAN EARTHWORMS 259
apart from the dubious location of the gizzards, there is nothing distinctive in the
description which would permit identification. The types are no longer traceable
in the Spencer collection (Jensz and Smith 1969).
Perionychella (subgenus ?) scolecoidea (Spencer, 1895)
Fig. i.
Perichaeta scolecoidea Spencer, 1895 : 51-52, PI. V, fig. 47-51.
Diporochaeta scolecoidea ; Michaelsen, 1900 : 207 ; Jamieson, 19710 : 83.
1 = 25-27 mm, w = 5 mm, s = 72-77 (lectotype and 2 paralectotypes).
Form very wide relative to the short length. Prostomium epilobous < £. First
dorsal pore in 3/4 (Spencer). Setae 35 on each side in IV and caudally ; 40-44 on
each side in XX ; no dorsal gap throughout ; some ventral interruption posteriorly
but irregular and slight (paralectotype i). Nephropores and clitellum not visible.
Male pores on minute papillae in setal lines 4 in XVIII. Accessory genital markings
an indistinct transverse pad extending laterally as far as setal line 6 in each of inter-
segments 13/14 and 14/15 (lectotype and paralectotype i ; none recorded in the text
by Spencer but a median pad illustrated in 19/20). Female pores just in front of setae
b of XIV. Spermathecal pores 2 pairs, in 7/8 and 8/9, in or slightly lateral of setal
lines 3 (lectotype) ; or in s.l. 4 (Spencer).
Dorsal vessel single ; last hearts in XII (confirmation). Supra-oesophageal vessel
in VIII-XII (Spencer). Gizzard very large, but not very thick walled, in V (con-
firmation) ; true calciferous glands absent ; intestinal origin XVIII (Spencer).
Holonephric ; nephridia more coiled in II-V than elsewhere. Holandric ; seminal
vesicles racemose, in XII. Metagynous (Spencer). Prostates racemose, depressed
and lobed, in XVIII (confirmation from lectotype). Penial setae absent (lectotype).
Spermathecae 2 pairs, in VIII and IX, each with 2 very small diverticula at the base
of a fair-sized sac (Spencer).
TYPE-LOCALITY. King River Valley, approximately i45°25'E. 42°io'S. Map,
I2/I.
MATERIAL EXAMINED.
Lectotype NMV : 0290 and 2 paralectotypes, 01422.
REMARKS. The material examined was refractory and yielded little information.
The condition of the nephridial ducts was indeterminable.
Genus GRALIOPHILUS Jamieson, 19716
DIAGNOSIS. Combined male and prostatic pores a pair on XVIII. Prostates
tubular. Setae 8 per segment ; cd = o^-i-^bc. Nephridia exonephric, stomate,
avesiculate holonephridia forming a single series on each side discharging in c or d
260 B. G. M. JAMIESON
or above these ; sometimes replaced in a varying number of anterior segments by
coiled or tufted nephridia with anteriorly directed (always buccal ?) ducts. Calci-
ferous glands and intestinal typhlosole present or absent. Spermathecae 2 or less
commonly 3 pairs with extramural diverticula, sometimes with sperm chambers in
the walls but never multiloculate.
TYPE-SPECIES. Graliophilus georgei Jamieson, 19716.
DISTRIBUTION. Western Subregion of Australia, South Western faunal province.
Eastern Subregion, New South Wales (Mt Kosciusko) ?, Tasmania ?.
REMARKS. The above generic diagnosis is based on Jamieson, 19716. Gralio-
philus was erected for Western Australian species which previously would have been
ascribable to Plutellus but clearly were not congeneric with the type-species of that
genus nor with species of any other genera erected during progressive revision of
Plutellus. The two species from New South Wales (G. montiskosciuskoi and G.
woodi Jamieson, I973a), were referred tentatively to Graliophilus but the widening of
Perionychella to include avesiculate and vesiculate, lumbricine and perichaetine
species, permits G. montiskosciuskoi to be placed unequivocally in Perionychella as a
lumbricine species of the nominate subgenus. G. woodi and the Tasmanian species
G. tripapillatus sp. n. differ from Perionychella as currently denned in possessing
true typhlosoles. Typhlosoles are present in some Western Australian species of
Graliophilus and woodi and tripapillatus have therefore been referred to this genus,
with which they conform in general morphology, though the identification is made
with considerable hesitation. It must be noted that in another Tasmanian genus,
Cryptodrilus, typhlosoles may be present or absent in closely related species and that
separation of woodi and tripapillatus from Perionychella may, therefore, be unnatural
as no other significant features 'are known to exclude them from the latter genus.
Furthermore, P. (P.) capensis has a slight dorsal intestinal ridge which might be
considered a rudimentary typhlosole.
This difficulty in generic identification of species reflects that fact that distinction
of Graliophilus from Perionychella is rendered uncertain by the emendment of
Perionychella. While some Western Australian species groups, notably the strelitzi-
group of Graliophilus, are unquestionably distinct from Perionychella, the georgei-
group, containing the type-species of Graliophilus, conforms closely in morphology
with the emended Perionychella. However, in Western Australian Graliophilus
species the number of spermathecae is less (2 or 3 pairs) than the 5 pairs usual in
Perionychella (though some species of the latter have 1-4 pairs) and the avesiculate
and totally lumbricine condition is constant whereas the combination of these charac-
ters is known in no Perionychella species from Victoria or New South Wales and in
only one, P. (P.) weldborough.' , from Tasmania. Wallace (1972) has shown that species
since placed by the author in Perionychella show only a low phenetic resemblance to
Graliophilus species when affinity is assessed from large character-sets using computer
techniques. It is hoped that analysis of the data will allow diagnosis of each genus
from the other without recourse to numerical methods.
In short, the following Tasmanian species referred to Graliophilus may be a
typhlosolate species of Perionychella (Perionychella) .
TASMANIAN EARTHWORMS
261
sp.p
sp.p
sp.p
XII
XIII
XIV—
1mm
1mm
FIG. 17. Genital fields. Graliophilus tripapillatus , holotype : A, spermathecal field ;
B, male field.
Graliophilus (?) tripapillatus sp. n.
Fig. 4, 17, 32 A ; Plate i. Table 3
1 = 60, 59 mm, w = 3-2, 3-0 mm, s = 130, 135.
Prostomium epilobous £, closed. Dorsal pores not visible. Setae 8 per segment,
in regular longitudinal rows throughout. Nephropores in c lines. Clitellum annular,
^XIII-^XVIII. Male pores on XVIII in a on dome-shaped papillae reaching from
b to an approximately equal distance median to the pores. Accessory genital mark-
ings midventral unpaired transverse pads in 15/16, 16/17 and 19/20, extending later-
ally into ab or beyond b and longitudinally to or over the setal rows of adjacent
segments ; the pads increasing in size and prominence posteriorly ; faintly indicated
glandular and slightly tumid interruptions of the intersegments, in 17/18 and 18/19.
Female pores paired on XIV, anterior and only slightly median to setae a. Sper-
mathecal pores 3 pairs, in 6/7-8/9, on small papillae in a.
262
B. G. M. JAMIESON
TABLE 3
Intersetal distances in segment XII in Gmliophilus tripapillatus
mm
aa
ab
be
cd
dd
dc
cb
ba
0-9
0-4
I-O
0-6
3'3
0-6
I-O
0-4
0-7
0-4
0-9
0-7
2-7
0-6
0-8
0-4
i-i
0-4
I-O
0-6
3-6
0-6
I-O
0-4
i'3
0-4
I-O
0-6
3-8
0-6
I-O
0-4
i-o
o-5
0-9
0-6
2-8
0-6
0-9
0-4
standardized (u = 100)
aa
ab
be
cd
dd
dc
cb
ba
"'35
4-80
11-79
7-42
49-61
7-42
11-79
4-80
10-40
4'95
12-87
9-41
38-61
7.92
10-89
4-95
12-55
4-86
ii-34
6-88
41-30
6-88
n-34
4-86
13-96
4'53
11-32
6-79
40-75
6-79
11-32
4'53
12-44
5'99
11-98
8-29
36-87
7-37
11-98
5-07
12-14
5-03
11-86
7.76
39-63
7-28
11-46
4-84
2-41
I-OO
2-36
1-54
7-88
i-45
2-28
0-96
holotype
paratype i
paratype 2
paratype 3
paratype 4
holotype
paratype i
paratype 2
paratype 3
paratype 4
mean
interval/afe
Last hearts in XII (latero-oesophageal) ; supra-oesophageal moderately developed
in VIII-XIII. Gizzard large and firm in V. Calciferous glands absent ; oesopha-
gus internally rugose and with circumferential vascular striae in VIII-XVI. In-
testinal origin XVIII ; deep dorsal typhlosole commencing in XX. Nephridia
simple stomate, avesculate, exonephric holonephridia (funnels demonstrated in fore-
and mid-body) ; ducts entering the parietes presetally in c lines, in the forebody
forming wide-tubes but not sufficiently dilated to be termed bladders ; slender in
the midbody.
Holandric (funnels iridescent in X and XI) ; gymnorchous ; seminal vesicles
racemose, in IX and XII. Metagynous ; ovisacs in XIV. Prostates much coiled,
depressed tubular ; vas deferens joining the sinuous ectally widening duct shortly
ectal of the gland. Penial setae very slender, tip bifid, chelate, the two branches
terminally contiguous, a coarse tooth, almost large enough to be considered a bifurca-
tion, on one branch ; ornamentation absent ; length ? mm, general width of shaft
= 3 /mi (holotype). Spermathecae 3 pairs ; diverticulum (inseminated) single,
clavate, uniloculate.
Genital markings. Of the 15 type-specimens (holotype, paratypes 1-14) only the
holotype and paratype i have the genital marking in 15/16. Otherwise the genital
fields are invariable in all specimens, with median genital marking in each of inter-
segments 16/17 and 19/20 and, less strongly indicated, in 17/18 and 18/19.
MATERIAL EXAMINED.
Tarraleah, over pipeline, I46°25'E. 42°2o'S. Map, 13/4, Dr J. L. Hickman, 27 May 1954 -
holotype TM : K3I3 ; paratypes 2-6 TM : 1^314-318 ; paratypes i, 7-9 BM(NH) : 1973.2.1-4 ;
paratypes 10-12 AM : W5203-52O5 ; paratypes 13 and 14 BJ : T23-24-
REMARKS. The genital field, with its arrangement of three genital markings,
diagnoses this species.
TASMANIAN EARTHWORMS 263
Genus PINGUIDRILUS gen. nov.
DIAGNOSIS. Setae 8 per segment. Combined male and prostatic pores a pair on
XVIII. Prostates a single pair, tubular. Gizzard in V and VI ; extramural calci-
ferous glands and intestinal typhlosole absent. Nephridia vesiculate holonephridia
discharging irregularly in c to near middorsum. Spermathecae adiverticulate.
DESCRIPTION. Large ; terrestrial. Dorsal pores present. Setae 8 per segment.
Nephropores erratically in c to near the middorsal line. Clitellum long (> 9
segments), including the male pores. Male pores (combined pores of the vasa
deferentia and the single pair of prostates) on XVIII. Female pores paired on XIV.
Spermathecal pores 5 pairs, the last in 8/9. Accessory genital markings present in
the vicinity of the male pores.
Last hearts in XIII, latero-oesophageal hearts, in X-XIII, with connectives to
dorsal and the supra-oesophageal vessel. Gizzard large, in V-VI. Extramural
calciferous glands absent ; intestinal origin XIX ; typhlosole absent. Nephridia
simple, exonephric holonephridia with large ectal bladders. Holandric ; gymnor-
chous ; seminal vesicles in IX and XII. Metagynous ; ovisacs absent. Prostates
slightly lobulated tubes. Penial setae absent. Spermathecae lacking diverticula.
DISTRIBUTION. Tasmania. (Monotypic.)
TYPE-SPECIES. Notoscolex tasmanianus Fletcher, i888b.
REMARKS. Pinguidrilus retains the primitive holonephric condition (two neph-
ridia per segment) seen elsewhere in Tasmania only in Rhododrilus (Acanthodrilinae)
and in Perionychella and a doubtful member of each of the genera Graliophilus and
Woodwardiella (Megascolecinae). The italicized features in the above diagnosis
distinguish it from all these genera. Its closest phyletic relationship is presumably
with the similarly vesiculate Perionychella (V esiculodrilus] but the absence of sper-
mathecal diverticula is a very rare condition in the Megascolecidae (with the excep-
tion of the Ocnerodrilinae, in which diverticula are only exceptionally present) and,
with the unusual arrangement of the nephropores, suggests a rather isolated position
for the genus. This irregular disposition of the nephropores is also seen in the
Nearctic plutelloid genus Argilophilus, which shows strong morphological affinities
with Pinguidrilus, though avesiculate and possessing spermathecal diverticula
albeit little more than sperm chambers in the wall of the spermathecal duct.
Pinguidrilus tasmanianus (Fletcher, i888b)
Fig. 4, 256, 326
Notoscolex tasmanianus Fletcher, i888b : 607-611.
Megascolides tasmanianus ; Spencer, 1895 : 33-
Plutellus tasmanianus ; Michaelsen, 1900 : 175 ; Jamieson, 19710 : 88.
1 = ?, w = 23 mm, s = ?.
Circular in cross section, segments I and II simple ; thereafter, in the forebody,
strongly triannulate. Prostomium epilobous, closed, \ ; it and the peristomium
264 B. G. M. JAMIESON
with numerous longitudinal grooves. First dorsal pore in 10/11, minute. Setae
8 per segment but c and d only sporadically visible ; setae a and b absent in XVIII.
Nephropores (see nephridia, below). Clitellum annular, strongly protuberant,
embracing XIV-^XXIII ; dorsal pores absent, intersegmental furrows retained
only ventrally ; ventral setal couples visible ; nephropores not visible. Male pores
distinct but minute apertures in ab on small white papillae which occupy the setal
annulus of XVIII. Genital markings 7 median transverse pads, in intersegmental
furrows 14/15-20/21, each bisected by the furrow, longitudinally occupying the pos-
terior and anterior annuli of the adjacent segments, and extending laterally beyond
setal lines b ; those in 17/18 and 18/19 not as extensive transversely as the others and
only shortly overreaching b lines. Female pores a pair of minute orifices on small,
low papillae, on XIV, very shortly anteromedian of setae a, on the setal annulus.
Spermathecal pores on 5 pairs of small, round papillae, in 4/5-8/9, in a lines.
Septa progressively thickened from 3/4 to 12/13 ; 3/4 and 4/5 obscured by tendons
running to the next posterior septa ; 5/6 strong ; 12/13 immensely thick ; 13/14
moderately thick ; the remainder thin. Dorsal blood vessel single ; continuous onto
the pharynx. Dorsoventral commissural vessels in V-XIII ; those in X-XIII
forming 4 pairs of latero-oesophageal hearts ; those in IX anteriorly progressively
more slender and dorsoventral only. Each latero-oesophageal heart receiving a
slender connective from the dorsal vessel and a thicker connective from the median
supra-oesophageal vessel immediately behind the junction with the latter of a pair of
vessels from the oesophageal wall. Supra-oesophageal beginning (very thin)
anteriorly in VIII and ending posteriorly as the connectives to the hearts in XIII, the
vessel well defined and separate from the oesophagus throughout and posteriorly
almost as thick as the dorsal vessel ; receiving vessels from the oesophagus throughout
its course. Gizzard large and muscular but fairly easily compressible, in V and VI,
septum 5/6 being inserted on its equator where the commissures of V join the dorsal
vessel. Oesophagus in VII long; coiled and muscular ; in VIII-XI vascular, though
not conspicuously, and not dilated ; in XII and XIII segmentally dilated and with
numerous circumferential vascular striae ; in XIV- XVI 1 1 very wide, extremely
thin walled and vascular, giving paired vessels directly to the dorsal vessel.
Intestine commencing in XIX ; no typhylosole detectable.
Nephridia : nephropores erratically positioned (described by Fletcher as an
irregularly sinuous series varying from c lines to near the middorsal line) ; their
position sometimes corresponds with setal loci but these also are erratic. In the
intestinal region nephridia are stomate, vesiculate holonephridia without supple-
mentary loops ; vesicles are large, in two wide sections with a narrow section be-
tween ; a slight angular projection which could be regarded as a diverticulum extends
laterally beyond the position of opening ; no change in structure occurs at the hind
end ; anteriorly nephridial structure is similar but the vesicle is not distinctly divided
into two parts (Wallace, personal communication). Testes ? sperm masses and large
funnels free in X and XI. Seminal vesicles large and racemose in IX and XII ;
approximately equisized. Ovaries paddle-shaped, with many rows of minute
oocytes, and small funnels in XIII. Prostates restricted to XVIII, slender tor-
tuous tubes with slightly lobulated surfaces ; the slender ducts receiving the vasa
TASMANIAN EARTHWORMS 265
deferentia near the junction with the glandular portions. Penial setae absent.
Spermathecae 5 pairs of simple pouches tapering to the pores but lacking distinctly
differentiated ducts ; diver ticula absent.
TYPE-LOCALITY. Thomas' Plains, N.E. Tasmania.
MATERIAL EXAMINED.
A single fragmenting clitellate specimen labelled ' Megascolides tasmanicus [sic] Fletcher jar
2', Baldwin Spencer collection, ex Melbourne University Zoology Department, May 1968.
REMARKS. The above description agrees closely with that of Fletcher and, as the
material of Fletcher species in the Baldwin Spencer collection must be considered
type material, the specimen examined is here designated the lectotype NMV : G2H2.
The following additional data are given by Fletcher : lengths (contracted) 200-250
mm, width 19-24 mm ; about 200 segments. Setal lines d sinuous ; aa : ab : be : cd
= 4:2:4: 3-8. The first dorsal pore was observed in 12/13, and the clitellum on
^XIII-XXII ; the gizzard in V ; the oesophagus dilated and vascular in VIII-XVI
and especially in XIII-XVI, showing effervesence in acid in all but the first segment.
Genus WOODWARDIELLA Stephenson, 1925
DIAGNOSIS. Combined male and prostatic pores a pair on XVIII. Prostates
racemose. Setae 8 per segment ; cd = or slightly > be. Nephridia exonephric,
stomate, avesiculate holonephridia forming a single series on each side in c lines but
(always ?) replaced in anterior segments by tufts discharging into the buccal cavity
or exonephrically in d lines. Calcif erous glands absent ; typhlosole weakly developed
or absent. Spermathecae 2 pairs, diverticulum simple, not multiloculate. (Based
on Jamieson (1970, 19716).)
TYPE-SPECIES. Woodwardia callichaeta Michaelsen, i9O7a.
DISTRIBUTION. South-western Australia. Tasmania ? Victoria ?
REMARKS. Differences between Woodwardiella and the older genus Perionychella
require augmentation. The prostates in Perionychella are tubular or, if externally
racemose, retain an axial lumen whereas those of Woodwardiella have a typical
Pheretima-type prostate with branching internal ducts (Michaelsen, 1916). Al-
though this difference is allowed in Cryptodrilus, a genus in which species with the
two alternative conditions of the prostates are assignable to the same genus by virtue
of the peculiar excretory system which they share, it warrants separation of Wood-
wardiella and Perionychella pending further investigation of the former little known
genus. Tubular prostates are all that is certainly known to distinguish Graliophilus
from the older Woodwardiella and further investigation of the justification of retain-
ing Graliophilus as a separate genus from Woodwardiella or Perionychella is also
required.
The following Tasmanian species was transferred to Woodwardiella by Jamieson
(i97ic) from the artificial congeries Plutellus, in which Michaelsen (1900) had placed
it on the then justified grounds that it had 8 setae per segment, holonephridia and
apparently racemose prostates. As no material of it is available for the present
16
266 B. G. M. JAMIESON
study it is retained in Woodwardiella though it appears likely that its prostates will be
found to be tubulo-racemose and that it will be assignable to Perionychella.
Woodwardiella ? tessellatus (Spencer, 1895)
Fig. 4
Cryptodrilus tesselatus (sic) Spencer, 1896 : 40-41, PI. II, fig. 16-18.
Plutellus tessellatus ; Michaelsen, 1900 : 170-171.
Woodwardiella ? tesselatus ; Jamieson, I97ic : 92.
1 = 25 mm, s = 65.
Prostomium slightly epilobous. First dorsal pore in 5/6. Setae 8 per segment,
the dorsal row very irregular and setae d absent in some segments ; c regular except
caudally. Clitellum annular, XIII -XVII, with midventral continuation in bb in
XVIII and XIX. Male pores on small papillae in XVIII, in ab. Accessory genital
markings 2 pairs of small elliptical patches in 12/13 and 13/14 in ab. Female pores
anteromedian to setae a, at \aa, in XIV. Spermathecal pores 2 pairs, in 7/8 and
8/9, in ab.
Last hearts in XII. Gizzard in V ; true calciferous glands absent ; intestinal
origin XVII. Holonephric. Holandric ; seminal vesicles saccular, in XII. Meta-
gynous. Prostates small flattened lobes, in XVIII. Spermathecae 2 pairs, in VIII
and IX ; diverticulum simple, less than half the length of the ampulla.
TYPE-LOCALITY. Mount Olympus, Lake St Clair, in damp soil under logs, and
amongst decaying leaves in beech forest. Map, 13/5.
MATERIAL EXAMINED.
None available.
Genus CRYPTODRILUS Fletcher, i887a emend.
DIAGNOSIS. A pair of combined male and prostatic pores on the seventeenth
setigerous segment, normally XVIII. Setae 8 per segment. Meronephric ; neph-
ridia typically vesiculate, frequently avesiculate, throughout, not very numerous on
each side, some or all discharging in the setal lines ; the medianmost nephridium
usually (always ? ) with preseptal funnel. Typhlosole present or absent. Prostates
tubular, tubuloracemose, or racemose.
DESCRIPTION. Moderate-sized earthworms with less than 200 segments. Pro-
stomium zygolobous to tanylobous. Dorsal pores present or absent. Setae in 8
regular longitudinal lines, commencing on II (or, C. mediocris, on the product of
union of the first and second metameres). Ventral setal couples widely paired
(aa = i-5~3a6) ; setae of the dorsal couples (cd) widely separated, always further
apart than those of the ventral couple (cd = 2-4^6) and usually a greater interval,
sometimes smaller, than the interval between the couples of a side (be) ; dorsal
median intersetal distance (dd) == 0-2-0-4 of the circumference (u). Setae a and b
absent in the segment bearing the male pores ; usually replaced by penial setae.
Nephropores (conspicuous or not) at least 3 on each side in some segments and
TASMANIAN EARTHWORMS 267
reaching (C. polynephricus) as many as 10 per side ; the pores associated with some or
all four of the setal lines where the number per side is 4 or less ; increase in number
intersetal and setal. Clitellum annular, commencing in XI I -XIV and including or
shortly preceding the segment bearing the male pores. A pair of combined male and
prostatic pores on XVIII (or XVII where intersegment 1/2 is suppressed), in b or
median of this. Accessory genital markings present. Female pores a pair, or un-
paired midventral, in XIV, inconspicuous, though sometimes in a common glandular
field. Spermathecal pores 2 or 3 pairs, the last pair in intersegmental furrow 8/9.
Some preclitellar septa strongly thickened. Gizzard rudimentary to well
developed, in V. Definite calciferous glands absent but oesophagus in some seg-
ments vascularized, often dilated, and internally rugose or lamellate. Intestinal
origin XVI, XVII or XVIII ; typhlosole present or absent. Dorsal blood vessel
(always ?) continued onto the pharynx ; dorso ventral commissurals commencing in
V or further posteriorly ; last hearts in XII or XIII, those in X-XIII forming latero-
oesophageal hearts with connectives from the supra-oesophageal vessel and in some
species from the dorsal vessel. Supra-oesophageal vessel not extending more than a
segment or two in front of the latero-oesophageal hearts. Subneural vessel absent.
Nephridia 3~ca. 15 exonephric nephridia on each side, typically vesiculate but fre-
quently avesiculate, some or all discharging in the setal lines ; the medianmost
nephridium in the intestinal or at least caudal region stomate in most if not all species
and frequently enlarged as a megameronephridium ; posterior nephridia sometimes
sending ducts posteriorly through several segments or sometimes with the stomate
nephridia contributing to a longitudinal excretory duct (ureter) which appears to
discharge at junction of rectum and body wall ; rarely (C. enteronephricus) with duct
of each stomate megameronephridium entering the intestine. Avesiculate exo-
nephric or enteronephric tufts present or absent in anterior segments. Holandric ;
gymnorchous ; seminal vesicles in 2 or more of segments IX, X, XI and XII.
Metagynous ; ovisacs present or absent. Prostates one pair, racemose, with branched
internal and external ducts, or tubuloracemose (lobulated but with single central
lumen which (always ?) has side branches) ; vasa deferentia joining the ectal end of
the gland or the prostate duct shortly ectal of this. Spermathecae discharging
anteriorly in their segments ; the duct with one or two uniloculate diverticula.
TYPE-SPECIES. Cryptodrilus rusticus Fletcher, iSSya.
DISTRIBUTION. Eastern Subregion of Australia, Southern faunal province :
New South Wales and Victoria. Tasmania.
REMARKS. The tribe Perionychini was established by Jamieson (iQ7ia) for those
Megascolecinae which were holonephric, or had meronephridia in a varying number
of segments anterior to holonephridia, and which lacked intestinal enteronephry.
Demonstration of more than two vesiculate nephridia per segment in Cryptodrilus
(vide Jamieson iQ72b) posed the problem that this genus could not be placed in any
one of the three megascolecine tribes (Perionychini, Dichogastrini and Megascolecini)
recognized by the author but studies of general morphology, and a numerical (taxono-
metric) investigation of setal ratios, suggested that the affinities of Cryptodrilus lay
with the Perionychini. An evolutionary pathway for origin of the multiple-bladder
268 B. G. M. JAMIESON
condition from the holonephric vesiculate condition exemplified by Plutellus hetero-
porus, the 'heteropore' condition, was suggested. It was pointed out, and is still
maintained, that recognition of a tribe Cryptodrilini to include Cryptodrilus deserves
consideration. In the present work Cryptodrilus is included in the Perionychini
because of its close morphological and presumed evolutionary links with this tribe,
the nephridial bladders, which are unknown in the Dichogastrini and Megascolecini,
being an especially significant indication of affinity. Pending the possibility of
erection of a tribe Cryptodrilini, then, the definition of the Perionychini must be
enlarged to include species with multiple bladders in a segment or, in other words, all
megascolecine species with nephridial bladders, irrespective of whether holonephric
or meronephric, in addition to holonephric avesiculate species.
Some species of Cryptodrilus, which have clear affinities (sometimes at the infra-
specific level) with vesiculate species, lack nephridial bladders. They have median
stomate nephridia in posterior segments, as do some vesiculate species, and are
presumed to represent an acquisition of the dichogastrin condition of nephridia
independently of the true Dichogastrini. This view is preferred to the alternative
that they represent a survival of populations transitional between the Perionychini
and Dichogastrini as the numerical study indicated that, at least in setal ratios, the
Dichogastrini are distinct from the Perionychini. This distinctness might have been
developed subsequent to origin of Dichogastrini from populations of which Tas-
manian perionychines are representative but variation from the vesiculate to the
avesiculate condition infraspecifically, or at least in very closely related subgeneric
taxa (vide C. polynephricus), indicates that loss of vesicles is currently occurring in
Tasmanian perionychines. The definition of the Perionychini must, therefore, be
further extended to allow inclusion of forms which have acquired the dichogastrin
condition but have clear relationships with perionychin species. In addition, an
avesiculate species, C. enteronephricus, with intestinal enteronephry (a condition
formerly restricted to the Megascolecini) which is unquestionably related to vesiculate
species of Cryptodrilus, must be included within this genus in the Perionychini. Thus
the definition of the Perionychini embraces species which would be referable to the
Dichogastrini or Megascolecini on the anatomy of the excretory system but which
show evident relationships with Perionychini.
Such widening of the definition of discrete groups so that they are no longer totally
interdistinguishable has become accepted in modern taxonomy and is an expected
extension of the concept of polythetic groups, definable by no one single character.
The genus Cryptodrilus, as emended above to include additional evidence from
Tasmanian species, has now become a polythetic grouping as the single formerly
diagnostic character, multiple nephridial bladders in a segment, no longer applies to
all species. Inclusion of avesiculate species appears justified from a consideration of
general anatomy. Such a variation, from vesiculate to avesiculate species is in-
significant relative to the remarkable variation further discussed in the account of C.
polynephricus, from the exonephric to the enteronephric condition in posterior seg-
ments.
Jamieson (igyaa) recognized three species groups in Cryptodrilus. The Tasmanian
species C. avesiculatus, C. brunyensis, C. enteronephricus and C. polynephricus (with
TASMANIAN EARTHWORMS 269
C. albertisi) show clear affinities with the mediocris group and C. simsoni and less
certainly C. dartnalli probably also are placeable in it. It is not proposed, however,
to redefine the group as the emended definition would be so wide as to be poorly
defined from the rusticus and fastigatus groups and would have little operational
value. This is not to deny that at least fastigatus, with its externally branched
prostates, is a very unusual species. If further species with fastigatus prostates were
found it would not be unreasonable to recognize a separate genus for such species,
the name Trinephrus being available for it.
CHECK LIST OF SPECIES OF Cryptodrilus
Tasmanian species
C. albertisi (Cognetti, 1910) New combination (junior synonym of C. polynephri-
cus ?)
C. avesiculatus sp. n.
C. brunyensis sp. n.
C. dartnalli sp. n.
C. enteronephricus sp. n.
C. officeri Spencer, 1895
C. polynephricus Spencer, 1895
C. simsoni Spencer, 1895
Other species
New South Wales
C. fastigatus Fletcher, iSSga
C. mediocris Fletcher, i88ga
C. rusticus Fletcher,
C. tenuis Fletcher,
Victoria
C. dubius Spencer, i892a (junior synonym of C. fastigatus)
KEY TO TASMANIAN SPECIES OF CRYPTODRILUS*
1 3 pairs of spermathecae ....... C. officeri Spencer, 1895
- 2 pairs of spermathecae ........... 2
2 Gizzard large, firm. Intestine commencing in XVI ...... 3
- Gizzard small, flaccid. Intestine commencing in XVII or XVIII .... 6
3 Caudally with median stomate nephridium on each side discharging into the intestine.
Fig. 20 ......... C. enteronephricus sp. n.
- No caudal nephridia discharging into the intestine unless anally .... 4
4 2 nephridia present caudally in a line on each side : i astomate and discharging
exonephrically, the other stomate and contributing to a longitudinal collecting
duct (ureter) ............ 5
- Only one nephridium present caudally in a line on each side, astomate and exonephric.
No ureter present. Fig. i8B C. avesiculatus sp. n.
- Excluding C. albertisi (q.v.)
ayo
B. G. M. JAMIESON
1mm
sp.p
sp.p
g.m
FIG. 1 8. Genital fields. A, Hickmaniella opisthogaster, holotype. B, Cryptodrilus
avesiculatus, holotype.
5 No postsetal genital markings in preclitellar segments. Fig. igA . C. brunyensis sp. n.
- Postsetal genital markings present in preclitellar segments. Fig. 2iA, B, C
C. polynephricus Spencer, 1895
6 Dorsal pores present. Intestine commencing in XVIII. Fig. 25 A C. sitnsoni Spencer, 1895
- Dorsal pores absent. Intestine commencing in XVII. Fig. 196 . C. dartnalli sp. n.
Cryptodrilus albertisi (Cognetti, 1910)
Fig. 2
Megascolides albertisi Cognetti, 1910 : 329-331, Fig. 3-6.
1 = 68 mm, w = 4 mm, s = 107.
Prostomium tanylobous. Dorsal pores not observed. Ventral setae closely
paired, dorsal setae distant, aa < be < ^throughout. InV-XVIII, aa = ab + be ;
TASMANIAN EARTHWORMS 271
ab > be ; cd = 2bc ; dd = I'^cd = o-^u. Behind XVIII setal lines, with the excep-
tion of a, displaced dorsally so that, in XXX, aa slightly > ab, but appreciably
< ab + be ; ab slightly < be ; cd > 2bc ; dd = aa = 0-14^. Clitellum not deve-
loped. Male pores not externally visible. Accessory genital markings 3 pairs of
whitish, slightly tumid papillae posteriorly in segments IX, X and XI in b lines.
Female pores a pair anteriorly in XIV, midway between setal lines a and the ventral
midline. Spermathecal pores 2 pairs of short transverse slits in 7/8 and 8/9, shortly
lateral of a lines.
Last hearts in XIII. Gizzard extending from mid V to mid VI ; calciferous
glands absent ; intestinal origin XVI. Nephridia 'diffuse' ; from approximately
segment XXX, those of the most ventral pair, in b lines, are much larger than the
others. Holandric ; gymnorchous ; seminal vesicles small, in IX, larger in XII,
non-racemose. Prostates one pair, in XVIII-XX, the glandular part oblong with
corrugated surface ; the S-shaped muscular duct penetrating the parietes in,
apparently, b. Penial setae 800 £im long, 30 /u,m wide, slightly curved, ending in a
point which is not very fine ; the ectal half ornamented with small scale-like projec-
tions with minutely dentate margins. Metagynous. Spermathecae 2 pairs, in
VIII and IX ; ampulla transversely compressed with slightly lobulated margin with
stout duct of equal length into which, a short distance from the external pore, opens
a digitiform diverticulum.
TYPE-LOCALITY. Mt Wellington. Map, 14/4.
MATERIAL EXAMINED.
None available.
REMARKS. Cognetti's description is questionably adequate for identification of
any material with this species. Although nephridial bladders and distribution of
nephridia are not described, the distribution of accessory genital markings strongly
suggests that C. albertisi is identical with, and therefore, a junior synonym of,
C. polynephricus Spencer, 1895.
Cryptodrilus avesiculatus sp. n.
Fig. 2, i8B, 3iA ; Plates 2 and 3. Table 4
1 = 81, 69 mm, w = 3-8, 4-0 mm, s = 116, 118.
Prostomium epilobous f , open. Dorsal pores absent. Setae 8 per segment, in
regular longitudinal rows throughout. Nephropores not visible. Clitellum annular,
JXIII-XVII. Male and separate penisetal pores on dome-shaped papillae in b of
XVIII. Accessory genital markings 2 or 3 pairs of eyelike markings in 17/18
(holotype), 18/19 and 19/20 (holotype, paratype i) in a lines ; postsetal paired oval
markings extending over the posterior borders of their segments in VII and VIII
containing the spermathecal pores (holotype) >and in IX (holotype, paratype i) ;
also a faintly indicated paired postsetal patch in and ventral to a in VII and VIII
(holotype) . Female pores shortly anteromedian to setae a of XIV. Spermathecal
pores 2 pairs, in 7/8 and 8/9, on minute papillae in ab.
272
B. G. M. JAMIESON
TABLE 4
Intersetal distances in segment XII in Cryptodrilus avesiculatus
holotype
paratype 3
paratype 4
paratype 5
holotype
paratype 3
paratype 4
paratype 5
mean
interval/aft
aa ab be cd dd dc cb ba
1-6 0-7 1-4 1-5 3-5 1-5 1-4 0-7
i-i 0-6 1-2 1-4 2-4 1-4 i-i 0-6
i-i 0-6 i-i 1-3 2-3 1-3 i-o 0-6
i '4 0-6 i-i 1-3 2-1 1-3 i-o 0-6
standardized (u = 100)
aa
06
be
cd
dd
dc
cfc
ba
13-07
5-68
11-36
12-50
28-69
11-93
11-36
5-40
n-33
6-37
12-04
14-51
24-43
13-81
"•33
6-02
11-97
6'73
11-22
14-21
24-68
I3-46
10-85
6'73
14-96
6-36
"•59
13-84
22-44
13-84
10-47
6-36
12-83
6-29
"•55
I3-77
25-06
13-26
11-00
6-13
2-04
i-oo
1-84
2-19
3-98
2-II
i-75
0-97
Last hearts in XII (latero-oesophageal). Supra-oesophageal present. Gizzard
large, in V. Extramural calciferous glands absent but oesophagus vascularized and
moniliform in IX-XIII (holotype, paratype i) and slightly in VIII (paratype i).
Intestinal origin XVI ; typhlosole absent. Nephridia (holotype) : large (exone-
phric ?) tufts in III and IV ; succeeding segments with a few large avesiculate integu-
mentary micromeronephridia on each side. Caudally 4 exonephric, avesiculate
nephridia on each side discharging in front of setae a, b, c and d ; the dorsal three
astomate, the medianmost nephridium, discharging in a with a large preseptal funnel
and greatly enlarged to form a megameronephridium ; one nephridium only at the a
locus, the megameronephridium ; no ureter. Holandric (funnels iridescent) ;
gymnorchous ; seminal vesicles compact, racemose in IX and XII. Metagynous ;
ovisacs absent. Prostates flattened, lobulated but linear winding through XVII-
XX ; with narrow central lumen. Vasa deferentia joining the ectal end of the gland.
Penial setae moderately stout, strongly curved, widened basally, the smooth delicate
pointed tip upturned ; the ectal region of the shaft, with the exception of the vicinity
of the tip, ornamented by a series of coarsely serrated, scarcely protuberant trans-
verse but appreciably oblique jagged markings each of which occupies approximately
one-quarter to three-quarters of the width of the seta ; length of functional seta =
1-3 mm, general width of shaft = 23 /urn (holotype). Spermathecae 2 pairs ;
diverticulum single, subspherical, in the axis of the duct ; ampulla a lateral appen-
dage.
Genital markings. (Holotype, paratypes 1-7.) Genital markings in IX are
restricted to the Port Davey specimens. In the new Harbour specimens a distinct
oval genital marking is developed in front of each spermathecal pore, on VII and
VIII ; there are only suggestions of these in the Port Davey specimens. All
specimens, from both localities, have paired eye-like marking is 18/19 and 19/20
but only the holotype and paratype 2 have these markings in 17/18 and in only
paratype 3 are they present (median to a] in 20/21.
TASMANIAN EARTHWORMS
273
sp.p
sp.p
VIII
XVIII
XIX-
XX-
XXI-
linfn
FIG. 19. Genital fields. A, Cryptodrilus brunyensis, holotype. B, C. dartnalli, holotype.
C, C. simsoni, 8/1, specimen 2.
MATERIAL EXAMINED.
Kelly's Basin, Port Davey, I45°55'E. 43°2o'S. Map, 17/1, Tasmanian Biological Survey :
Jiy, Mr C. D. King, Jan 1940 - holotype TM : K3ig ; paratype i BM(NH) : 1973.2.5 ; paratype
2 TM : K320. New Harbour, i46°7o'E. 43°3o'S. Map, 18/5, Mr C. D. King - paratypes 3
and 4 TM : K32I-322 ; paratypes 5 and 6 AM : W5206-5207 ; paratype 7 BJ : T25.
REMARKS. C. avesiculatus is closely similar to C. enteronephricus, C. polynephricus
or C. brunyensis. It differs from C. enteronephricus in lacking caudal enteronephry,
and from C. polynephricus and C. brunyensis in lacking a ureter.
Cryptodrilus brunyensis sp. n.
Fig. 2, igA, 316, C ; Plates 4-7. Table 5
1 = 48 mm, 71 mm, w = 3-0 mm, s = 83, 124.
Prostomium epilobous \, open. Dorsal pores absent. Setae 8 per segment in
regular longitudinal rows throughout. Nephropores not visible. Clitellum annular
XIII-XVII. Male pores in ab, nearer a, in XVIII on subcircular papillae which
extend posteriorly over 18/19 slightly into XIX and anteriorly reach the hind margins
of a pair of oval genital markings in ab at 17/18. Two penial setae, projecting from
separate follicles, median to each pore. Two additional pairs of genital markings
aa
ab
be
cd
dd
dc
cb
ba
ii-45
6-17
12-33
15-86
18-94
15-86
13-66
5-73
13-03
6-90
12-64
I5-7I
16-48
I5-7I
12-64
6-90
12-24
6-54
12-49
15-79
17-71
15-79
13-15
6-32
1-87
I -00
1-91
2-41
2-71
2-41
2-OI
0-97
274 B- G- M- JAMIESON
present in 19/20 and 20/21 median to a, the anterior pair fused medianly (holotype) ;
only the conjoined pair in 19/20 present in paratype i. Female pore unpaired,
midventral, between the setal arc and anterior margin of XIV. Spermathecal
pores 2 pairs, in 7/8 and 8/9, in a lines, each surrounded by a glandular field ; the
fields in 7/8 (holotype) or in 7/8 and 8/9 (paratype) medianly united and suggestions
of presetal markings present in b of IX in paratype i.
TABLE 5
Intersetal distances in segment XII in Cryptodrilus brunyensis
mm
aa ab be cd dd dc cb ba
holotype 0-9 0-5 i-o 1-3 1-5 1-3 i-i 0-5
paratype I 1-2 0-6 1-2 1-4 1-5 1-4 1-2 0-6
standardized (u = 100)
aa ab be
holotype
paratype i
mean
interval/a&
Dorsal blood vessel continuous onto the pharynx. Last hearts in XII ; hearts in
X-XII latero-oesophageal, each with a connective to the supra-oesophageal vessel
but none to the dorsal vessel, supra-oesophageal ending at |XIII, well developed.
Gizzard firm and very large, in -V. Calciferous glands absent but oesophagus intern-
ally lamellate in VIII-XIII. Intestinal origin XVI ; dorsal typhlosole a low but
distinctly developed lamina from XVIII but continued as a rudiment to XVI
(holotype, paratype). Nephridia : loose aggregations of nephridia in II and III
send several ducts anteriorly to discharge at the anterior margin of the segment ;
pharyngeal tufts in IV- VI, accompanied by exonephric meronephridia ; in VII and
VIII with a meronephridium discharging above d, at d, cd, c, b and several, of which
some form a small tuft, at a ; this condition persisting in succeeding segments
(holotype) ; in the hindbody with a micromeronephridium discharging in the 4 setal
lines a, b, c and d and in addition a stomate megameronephridium contributing its
duct to a poorly developed longitudinal collecting duct (ureter) (holotype, paratype)
which ends posteriorly at the junction of rectum and body wall and therefore appears
to discharge into the anal aperture on each side ; all nephridia avesiculate and only
the megameronephridia with funnels (holotype). In paratype the ureter is well
developed but is not certainly detectable at the caudal extremity where in a there is
only a single (astomate ?) nephridium ; those in b, c and d persisting.
Holandric (funnels iridescent in X and XI) ; gymnorchous ; seminal vesicles
racemose in XI and XII. Metagynous ; ovisacs present. Prostates broad, flat-
tened, linear, lobulated (tubuloracemose) , in XVIII-XXI (holotype, paratype),
dissected to demonstrate main central internal duct with numerous more slender
lateral branches (holotype). Vasa deferentia on each side winding before joining the
TASMANIAN EARTHWORMS 275
prostate duct at its junction with the gland. Penial seta moderately stout, strongly
curved, tapering to a narrow blunt or pointed tip ; the ectal region of the shaft, with
the exception of the vicinity of the tip, ornamented by numerous groups of small,
irregular, conjoined teeth, the groups alternating along the seta, each containing only
a few teeth but considerably notching the setal surface ; length of a functional seta =
1-9 mm ; general width of shaft = 20 pm (paratype i). Spermathecae (not insemi-
nated) 2 pairs, each with long convoluted duct and long clavate-digitiform diver-
ticulum.
MATERIAL EXAMINED.
Rainforest at summit of Mt Mangana, Bruny Is., 147°I5'E. 43°25'S. Map, 19/1, Mr A. J.
Dartnall, 10 Apr 1971 - holotype TM : 1X323 ; paratype BM(NH) : 1973.2.6.
REMARKS. C. brunyensis is morphologically close to C. enteronephricus, C.
avesiculatus and C. polynephricus. It differs from C. enteronephricus in lacking
caudal enteronephry, from C. avesiculatus in possessing a ureter and from C. poly-
nephricus in lacking preclitellar genital markings.
Cryptodrilus dartnalli sp. n.
Fig. 2, 196, 3iD ; Plates 8-12. Table 6
1 = 85 + , 109 mm, w = 3-7, 3-5 mm, s = 107 + , 143 (holotype, posterior amputee ;
paratype 2).
Prostomium epilobous \, closed, continued posteriorly as parallel grooves, almost
giving a tanylobous condition. Dorsal pores absent. Setae 8 per segment, in
regular longitudinal rows throughout. Nephropores sporadically visible presetally.
Clitellum not developed (holotype) or strongly protuberant, though weak in aa, in
XIV-XVII with some dorsal clitellar modification to |XVIII and perhaps |XIII
(paratype 2). Male pores small slits on XVIII in a on ovoid papillae which extend
laterally as far as b and well median of a. Glandular tumescence extending from setal
arc of XVII to shortly behind 18/19, reaching mid be at 17/18, and including the male
porophores (holotype), similar in paratype I in which a pair of faint eye-like genital
markings is present on the tumescence in 17/18 in b ; indistinctly developed in
paratype 2 and other New Harbour specimens and obscured in these by insinking of
the male field. Female pores a pair very shortly anteromedian of setae a on XIV.
Spermathecal pores 2 pairs of small slits on small but distinct papillae, in 7/8 and
8/9, in a.
TABLE 6
Intersetal distances in segment XII in Cryptodrilus dartnalli
mm
aa ab be cd dd dc cb ba
i-i 0-6 1-3 i-2 4-5 1-2 1-3 0-6
i-i 0-7 1-5 1-4 4-8 1-5 1-6 0-7
1-2 0-8 1-6 1-4 5-8 1-3 1-6 0-7
276
Table 6 (cont.)
B. G. M. JAMIESON
standardized (u = 100)
aa
ab
be
cd
dd
dc
cb
ba
holotype
9-15
4-88
10-98
10-37
38-41
10-37
10-98
4-88
paratype i
8-59
5-37
11-28
IO-2O
35-98
11-28
11-81
5'37
paratype 2
8-16
5-28
11-04
9-60
39-60
9-12
11-04
4-80
mean
8-63
5-18
II-IO
10-06
38-00
10-26
11-28
5-02
interval/a&
1-67
I -00
2-14
1-94
7-34
1-98
2-18
0-97
sp.p
1mm
FIG. 20. Genital fields. Cryptodrilus enteronephricus : A, holotype ; B, paratype 18 ;
C, paratype 3.
TASMANIAN EARTHWORMS 277
Last hearts in XIII (latero-oesophageal ; connectives to dorsal vessel not demon-
strable). Supra-oesophageal in VII-XIII, well developed. Gizzard vestigial, in V,
but musculature thicker than elsewhere. Extramural calciferous glands absent ;
oesophagus with circumferential vascular striae and internal papilliform rugae, in
VII-XIII (holotype), XIV (paratype) ; chloragogenous and narrower in XIV
(holotype), XV (paratype 2) -XVI, though with internal rugae in XIV in holotype ;
intestinal origin XVII ; typhlosole absent. Nephridia : in the forebody, astomate,
avesiculate, exonephric meronephridia commencing in II, few in number, all (?)
associated with setal lines a, b, c and d ; tufts absent. Caudally in the four setal
lines, or in a-c and above d, the median-most nephridium with a preseptal funnel.
Holandric (funnels iridescent in X and XI) ; gymnorchous ; seminal vesicles spheroi-
dal, in IX and XII. Metagynous ; ovisacs absent (holotype), but small sacs
(ovisacs ?) in XIV in paratype 2. Prostates flattened tubular, tortuous, in XVIII
and XIX ; two vasa deferentia joining the junction of each gland with its sinuous
muscular duct (holotype) or joining the ectal region of the gland (paratype 2). Penial
setae moderately stout, gently curved, tapering to a narrow point ; the ectal region
of the shaft, with the exception of the vicinity of the tip, ornamented by oblique
laterally sometimes interrupted circlets of irregular coarse teeth which are directed
anteriorly but may be considerably elevated at their apices from the setal surface ;
the circlets approximately 7 /zm apart (holotype, paratype 8) ; length of functional
seta o-8-i-o mm ; general width of shaft 22, 17 ju,m (holotype, paratype 8). Sperma-
thecae 2 pairs. Diverticulum single, digitiform uniloculate ; shorter relative to
the ampulla in paratype 2 compared with the holotype.
MATERIAL EXAMINED.
Melaleuca Inlet, Port Davey, I45°55'E. 43°2o'S. Map, 17/1, Mr A. J. Dartnall, Dec 1967 -
holotype TM : Kaao, paratype i TM : 1^324. New Harbour, i46°io'E. 43°3o'S. Map, 18/5,
Mr C. D. King, 27 Jan 1938 - paratypes 2 and 3 BM(NH) : 1973.2.7-8 ; paratypes 4-6
AM : W52o8-52io ; paratypes 7-8 BJ : T26-27.
REMARKS. The genital field of C. dartnalli is diagnostic but the species shows
close affinities with C. simsoni. It differs from the latter, among other respects,
in lacking dorsal pores and in origin of the intestine one segment further forward,
in XVII.
Cryptodrilus enteronephricus sp. n.
Fig. 2, 20, 3iE-G ; Plates 13-19. Table 7
1 = 124, no mm, w = 4-6, 4-0 mm, s = 162, 168.
Prostomium epilobous f with narrow open dorsal tongue. Dorsal pores absent.
Setae 8 per segment in regular longitudinal rows throughout. Nephropores not, or
only sporadically, visible. Clitellum annular, |XIII-XVIII (paratype i) or XVII,
with weak extension to |XVIII (holotype). Male pores small transverse slits, not
on porophores, in ab of XVIII, nearer a than b (holotype) or (paratype i) nearer b
than a, with two penisetal pores median to each pore, the three pores of a side on a
small indefinite papilla. Paired accessory genital markings at the posterior borders
278 B. G. M. JAMIESON
of IX and X in ab, intersegmental in 16/17 m a and I7/I8 and 18/19 m a^ (holotype,
paratype i), and intersegmental in 19/20 in and median to ab (holotype) ; median
markings occupying aa in 14/15 and 15/16 (holotype, paratype i). Female pores
anteromedian to setae a of XIV, at approximately \aa. Spermathecal pores 2 pairs,
in 7/8 and 8/9, on small distinctly protuberant papillae in ab, shortly lateral of a.
TABLE 7
Intersetal distances in segment XII in Cryptodrilus enter onephricus
holotype
paratype i
paratype 3
paratype 4
paratype 6
paratype 14
paratype 15
aa
ab j
be i
:d
dd
dc <
cb
ba
•o
o-5
•3
•3
2-0
i-3
•3
o-5
•3
0-6
•5
•8
2-7
1-7
•5
0-6
•2
o-5
•3
•3
1-8
1-3
•3
o-5
•2
0-6
•9
•6
2-6
1-6
•o
0-6
•3
0-6
•7
•8
2-5
1-8
•7
0-6
•0
0-6
•3
•6
2-0
i'5
:4
0-6
•7
0-8
•5
•9
2-6
1-9
•5
I-O
standardized (u = 100)
aa
ab
be
cd
dd
dc
cb
ba
10-89
5-06
14-40
14-01
21-79
14-01
14-79
5-06
11-03
5-07
13-11
I5-50
22-65
14-30
13-11
5-07
12-55
4'94
I4-45
14-45
19-77
M'45
14-45
4-94
10-09
4-90
I5-56
13-26
21-61
13-26
16-43
4-90
10-88
5-00
13-82
15-29
20-88
15-29
13-82
5-00
10-15
5-60
12-60
I5-75
19-95
I5-40
13-65
5-60
12-69
6-48
11-61
14-85
19-98
I4-58
n-34
7-29
11-18
5-29
I3-65
I4-73
20-95
14-47
13-94
5-4i
2-II
I'OO
2-58
2-78
3-96
2-74
2-64
1-02
holotype
paratype i
paratype 3
paratype 4
paratype 6
paratype 14
paratype 15
mean
interval/a&
Last hearts in XIII (latero-oesophageal with supraoesophageal but no dorsal
connectives in X-XIII). Commissurals in IX anteriorly, slender and dorsoventral,
lacking supra-oesophageal connectives. Supra-oesophageal vessel in VIII-XIV,
well developed.
Gizzard large and firm, with anterior rim, in V. Calciferous glands absent.
Oesophagus vascularized in VIII-XIV. Intestinal origin XVI ; a deep laminar
dorsal typhlosole present, rapidly increasing in depth from XVIII, continued as a
very low ridge or thickening into XVI ; muscular thickening and caeca absent.
Nephridia : large tufts in II-V send sheaves of forwardly directed ducts to the
buccal cavity and pharynx, those in II discharging at the junction of buccal cavity
and body wall, those in III entering the buccal cavity and those in IV and V joining
the pharynx behind- the brain. Succeeding segments with several astomate,
avesiculate, exonephric parietal micromeronephridia on each side. Caudally with
4 exonephric, astomate, micromeronephridia on each side discharging in a, b, c and
d lines and in addition a very large stomate megameronephridium which has a large
preseptal funnel near the ventral nerve cord and loops far laterally (to approximately
c line) before turning medianly to send a duct to the ventrolateral surface of the
TASMANIAN EARTHWORMS
279
intestine, i.e. enteronephric. All 5 nephridia on each side persisting to the caudal
extremity. No ureter present.
Holandric (funnels iridescent in X and XI) ; gymnorchous ; seminal vesicles
racemose, in IX, X and XII. Metagynous ; small ovisacs in XIV. Prostates linear,
depressed, winding in XVIII-XXI ; much lobulated but with a narrow axial lumen
(tubuloracemose) but that on the right side in the holotype a broad flattened lobe
in XVI 1 1- XX and not resolvable into a linear form ; vasa deferentia joining gland
near its junction with the sinuous muscular duct. Penial setae moderately stout,
gently curved, tapering ectally to a simple, moderately sharp (paratype i) or (arte-
fact ?) truncated tip (paratype 14) ; the ectal region of the shaft, with the exception
of the vicinity of the tip ornamented by about 20 irregular serrated circlets approxi-
mately 6 )u,m apart ; the serrations forming jagged lines with, here and there,
individual narrow pointed teeth recognizable. The circlets complete or interrupted
along the lateral aspect of the seta, if interrupted the two halves tending to alternate
along the seta. Length 1-05, 0-81 mm ; general width of shaft = 28 /*, 22 (para-
types i, 14). Spermathecae 2 pairs, with ovoid ampulla and approximately equally
long duct which is joined at midlength by an ovoid unstalked diverticulum.
Genital markings. The following conclusions are drawn from examination
of the holotype and eight fully clitellate paratypes : paired genital markings are
always present posteriorly in X but may be absent (paratypes 4, 5) in IX ; the greatest
extent of genital markings occurs in Pio in which, in addition to the markings in IX
and X, there are median markings in 13/14 and 14/15 and paired markings in each of
intersegmental furrows 15/16-20/21 ; markings are present in all specimens in each
of intersegments 15/16-18/19. Those in 15/16 are usually median as are those in
14/15 if present ; those in 13/14 (seen in only two specimens) are median or paired ;
all other markings are paired though there may be some tendency to unite medianly.
Paired markings in IX and X are virtually in line with the spermathecal pores, in ab
usually nearer a.
Distribution of accessory genital markings
in Cryptodrihis enteronephricus
H
Pi
P2
P5
P6
Pio
Total
IX
P
P P
P
P
P
P 7
X
P
P P
P P
P
P
P
P 9
13/14
P
m 2
14/15
m
m
P
m
m
m 6
15/16
m
m m
P Pm
pm
m
m
P 9
16/17
pm
P P
P Pm
P
P
pm
P 9
17/18
P
P P
P P
P
P
P
P 9
18/19
P
P P
P P
P
P
P
P 9
19/20
P
P
P P
P
P 6
2O/2I
P
P P
P
P 5
m = median ;
p = paired ; pm = paired tending to fuse medianly.
H = holotype ;
P = paratype.
Other variations. Seven specimens, including the holotype, were examined
internally. Constant features included : location of gizzard in V ; intestinal origin
28o B. G. M. JAMIESON
in XVI ; large typhlosole ; enteronephric megameronephridia caudally and absence
of bladders ; seminal vesicles in IX, X and XII ; presence of penial setae and 2 pairs
of spermathecae. A major variation was location of last hearts in XII in paratypes
3, 4 and 5, in contrast with location in XIII in the holotype and paratypes i, 6 and
10. Such variation is highly exceptional intraspecifically but there appear to be no
grounds for regarding the individuals with hearts in XII as specifically distinct.
Ovisacs were seen only in paratypes i and 5.
MATERIAL EXAMINED.
Dee Bridge, Lyell Highway, I46°35'E. 42°i5'S. Map, 13/7, Dr J. L. Hickman, 24 May 1954 ~
holotype TM : 1^325. Tarraleah, Lyell Highway, I46°25'E. 42°2o'S. Map, 13/4, Dr J. L.
Hickman, 22 May 1954 - paratypes 4 and 5 AM : ^5211-5212 ; paratypes 18 and 19 BJ : T34-
35. Tarraleah, over pipeline, Dr J. L. Hickman, 27 May 1954 ~ paratypes 11-15 BJ : T28-32.
Marlborough Highway (near Bronte), i46°3o'E. 42°io'S. Map, 13/6, Dr J. L. Hickman,
26 May 1954 - paratype J BM(NH) : 1973.2.9 ; paratype 2 BJ : T33 ; paratypes 6-10 TM :
K326-33O. Lyell Highway, 5 miles from Bronte towards Hobart, I46°35'E. 42°i5'S. Map,
13/7, Dr J. L. Hickman, 24 May 1954 -paratype 3 BM(NH) : 1973.2.10.
REMARKS. C. enteronephricus appears to be related to C. polynephricus, and
especially to the subspecies urethrae, but differs notably from that species (q.v.) in
having enteronephric intestinal nephridia. The development of enteronephry
represents a major departure from perionychin organization and the attainment of a
condition previously known only in the distinct tribe Megascolecini which, in sens,
strict., is otherwise represented in Tasmania only by Oreoscolex.
Cryptodrilus ? officeri Spencer, 1895
Fig. 2
Cryptodrilus officeri Spencer, 1895 : 44-45, PI. Ill, fig. 28-30.
Trinephrus officeri ; Michaelsen, 1900 : 186-187.
Notoscolex officeri ; Jamieson, i97ic : 79.
Cryptodrilus officeri ; Jamieson, 19725 : 166.
1 = 44 mm, w = 6 mm, s ?
Prostomium epilobous f . First dorsal pore 4/5. Setae 8 per segment, irregular
caudally. Clitellum annular, XIV-XVII. Male pores on papillae on XVIII, in b.
Accessory genital markings paired elliptical patches in 15/16 and 16/17 in b, and in
19/20 in ab. Female pores presetal and close together in XIV. Spermathecal pores
3 pairs, in 6/7-8/9, shortly above b.
Last hearts in XII, no continuous supra-oesophageal. Gizzard well marked,
segment ? (no distinct septa in front of 7/8) ; true calciferous glands absent but
vascular swellings in XIII-XV ; intestinal origin XVII. Nephridia ; 3 tufts on
each side. Holandric ; seminal vesicles racemose in IX and XII. Metagynous.
Spermathecae 3 pairs, in VII, VIII and IX ; diverticulum a group of finger-like
processes, the ampulla long and irregular in outline.
TYPE-LOCALITY. King River Valley. Map, 12/1.
TASMANIAN EARTHWORMS
281
17
282 B. G. M. JAMIESON
MATERIAL EXAMINED.
None available.
REMARKS. Spencer's statement that the nephridia are like those of C. fastigatus
and C. dubius necessitates placing officer i in Cryptodrilus but inclusion in this genus is
questionable. The multiloculate spermathecal diverticulum raises the possibility
that it is an Oreoscolex, the 3 pairs of spermathecae and the configuration of the
genital field being reminiscent of 0. sexthecatus. The latter species is distinguished,
so far as the inadequate description of C. officer i indicates, by the more ventral
position of the spermathecal pores (in b lines). New material of C. officeri is required
if its status is to be settled.
Cryptodrilus polynephricus Spencer, 1895
Cryptodrilus polynephricus Spencer, 1895 : 35~36 PI- I> ng- 4. 5. 6.
1 = 98-330 mm, w (midclitellar) = 4-5-12 mm, s = 127-175.
Prostomium epilobous |, f to tanylobous, sometimes canaliculate. Dorsal pores
absent. Setae 8 per segment, in regular longitudinal rows throughout ; a and b
absent on XVIII. Nephropores minute, only sporadically if at all visible. Clitellum
annular, XIII-XVII. Male pores on XVIII in ab, not usually on papillae ; 2
separate penisetal follicles usually visible median to each male pore. Accessory
genital markings ; an anterior series of 3 pairs of elliptical markings posteriorly in IX,
X and XI, or at the posterior intersegment of each of these, in ab, or median of a, or
some or all of these markings absent, or additional markings present at the posterior
margins of VII, VIII and XII. A posterior series of paired or median transverse
markings present in or immediately anterior to some of intersegments 12/13-21/22,
more commonly in 17/18-19/20. Female pores in XIV, paired anteromedian of a, or
unpaired midventral and presetal. Spermathecal pores 2 pairs, in 7/8 and 8/9, in or
slightly lateral of a.
Last hearts in XIII (exceptionally, in some specimens of urethrae, in XII) those in
X-XIII latero-oesophageal, with connectives from the supra-oesophageal vessel
only. Gizzard large, strongly muscular, with anterior rim, in V. Extramural
calciferous glands absent. Intestinal origin XVI ; typhlosole absent, rudimentary
or deep laminar ; muscular thickening and caeca absent. Nephridia : several pairs
of meronephric tufts discharging into the foregut ; exonephric micromeronephridia
present in most segments on each side in the four setal rows, usually with additional
intervening nephridia, caudally with the medianmost nephridium enlarged as a
megameronephridium, with preseptal funnel, the duct of which contributes to a
longitudinal excretory duct (ureter) which apparently discharges into the ureter ;
nephridial bladders present or absent, if present number in a segment ranging from
4 to 10 per side. Testes and iridescent sperm funnels in X and XI ; gymnorchous ;
seminal vesicles in IX and XII, or XI and XII or rarely IX, XI and XII. Prostates
tubuloracemose, extending through 2 or more segments. Penial setae long and stout,
when functional 0-7-4 mm long, with the general width of the shaft 23-45 //,m.
TASMANIAN EARTHWORMS 283
Metagynous ; ovisacs present or absent. Spermathecae 2 pairs, in VIII and IX,
diverticulum single, digitiform, uniloculate with or without a basal swelling.
TYPE-LOCALITY. Mt Wellington. Map, 14/4. Parattah. Map, 14/1.
DISTRIBUTION. See subspecies.
REMARKS. Two subspecies are here recognized and described below. Where a
subspecific character does not vary from the specific description it is not repeated in
the subspecific account.
sp.p
VI
sp.p
1mm
FIG. 22. Genital fields. Cryptodrilus polynephrictts polynephricus : A, specimen 7
B, specimen 24.
284 B. G. M. JAMIESON
Cryptodrilus polynephricus polynephricus Spencer, 1895
Fig. 2, 2iA, 22, 24A, 3iH ; Plates 20-31. Table 8
Cryptodrilus polynephricus Spencer, 1895 : 35-36, PL i, fig. 4, 5, 6.
Trinephrus polynephricus ; Michaelsen, 1900 : 185-186.
Cryptodrilus polynephricus ; Jamieson, 19720 : 169-172, Fig. 6D, 7E, F, G.
1 = 98-200 mm (44 specimens), the extreme specimens with midclitellar widths of
4-5 and 8 mm, and segmental counts of 127 and 158, respectively.
Male pores not usually on papillae, sometimes on oval papillae which extend beyond
a and b and occupy the posterior three-fourths of the segment. Accessory genital
markings : characteristically with 3 pairs of anterior genital markings (raised or
depressed glandular ellipses) posteriorly in IX, X and XI, in ab, in IX in or shortly
median to b ; in X slightly median to b or at approximately mid be ; in XI at mid
ab or nearer but not in a, i.e. lateral of the line of the spermathecal pores and conver-
gent posteriorly.
TABLE 8
Intersetal distances in segment XII in 14 specimens* of
Cryptodrilus polynephricus polynephricus
mm
aa ab be cd dd dc cb ba
maximum 2-3 0-9 2-0 4-1 4-4 4-0 1-9 0-9
minimum i-i 0-6 1-3 1-5 2-1 1-4 i-i 0-3
mean 1-8 0-7 1-6 2-8 3-2 2-7 1-6 0-7
standard deviation 0-3 o-i 0-2 0-9 0-7 0-8 0-2 o-i
standardized (u = 100)
aa ab be cd dd dc cb ba
maximum I3'45 6'9° J4'94 20-86 24-41 20-50 13*64 0-36
minimum 10-63 3'54 9-24 13-03 18-16 13-22 8-43 3-54
mean 12-19 5-10 11-15 17-89 20-79 17-27 10-69 4'93
standard deviation 1-20 0-99 1-63 2-43 1-69 2-35 1-73 0-88
interval/afc 2-39 i-oo 2-19 3-51 4-08 3-39 2-10 0-97
* Specimens i, 4, 5, 7, 10, 12, 14, 18, 24, 26, 32, 34, 37 and 38.
Some individuals (14/4, 27 Oct 1955 ; 13 Nov 1954) lacking markings in IX and
with those in X and XI in or median of a. Posterior genital markings in or immedi-
ately anterior to intersegmental furrows : median at 14/15, occasionally ; paired or
median at 15/16, commonly ; paired or median at 16/17, 17/18 and 18/19, almost in-
variably ; paired but usually medianly conjoined in 19/20, usually, in 20/21,
commonly, and in 21/22 rarely ; the paired markings centred from ab to median of a.
Female pores are an inconspicuous pair in XIV, anteromedian of a. Spermathecal
pores on small papillae.
Internal anatomy (paralectotype i ; gross anatomy confirmed from all localities
and details from 14/4, specimen 12) : supra-oesophageal vessel in £VIII— |XIII,
moderately developed. Oesophagus somewhat dilated in VIII ; in IX-XIII
TASMANIAN EARTHWORMS 285
moderately and uniformly dilated and slightly moniriform ; in XIV and XV more
slender. Intestinal typhlosole absent though in specimen 12 a low diffuse dorsal
ridge is perceptible from XXIV posteriorly.
Nephridia : vesiculate stomate and astomate nephridia and several pairs of
meronephric buccal tufts. The ovoid-subspherical bladders each receive a single
duct from a simple or completely coiled nephridium. The arrangement of the
bladders is as follows : at the posterior end of the body there are 5 (presetal) bladders
on each side, one in each of the four setal rows and a supernumerary bladder in cd,
giving 10 bladders per segment ; the nephridia supplying these posterior bladders
are small and compact but there is in addition an enlarged nephridium (megamero-
nephridium), with a preseptal funnel, the duct of which contributes to a longitudinal
excretory duct (ureter), passes through many caudal segments and appears to dis-
charge at junction of rectum and body wall. In the forebody a bladder persists in
each of the four setal lines of a side but, in addition 1-3 supernumerary bladders are
present in cd and I or 2, usually 2, between d line and the middorsal line ; very
occasionally there is a single supernumerary bladder in be. The number of bladders
on each side in the forebody thus varies from 4 to 10. No bladders are detectable in
segments I and II (paralectotype i) but one is present in d in specimen 12, and the a
bladder appears to be absent in III and is not observable in paralectotype i, though
present in specimen 12 in the anterior spermathecal segment, VIII. Where two
bladders are present above d, one is anterior in the segment and the other is equatorial
and less lateral. Sporadically in paralectotype, but not in specimen 12, there are
two bladders in c, one anterior and one shortly presetal. Numbers of bladders
counted on one side in paralectotype i are : I, o ; II, o ; III, 4 ; IV, 6 ; V, 9 ;
VI, 9 ; VII, 8 ; VIII, 7 ; IX, 9 ; X, 10 ; XI, 10 ; XII, 10 ; XIII, 9 ; post-pros-
tatic, 8 ; caudal, 5 ; these numbers agreeing well though not always exactly with
those in specimen 12. Large paired tufted nephridia, each consisting of numerous
spiral loops, lie in segments II-IX ; each tuft in III-IX contributes to a composite
duct which is continuous, on each side of the nerve cord, from segment IX to the
lower lip of the mouth (paralectotype i) or the buccal cavity (specimen 12). Each
of the pair of tufts in II sends a separate composite duct to the lower lip, lateral of the
other longitudinal duct and sends a few very slender ducts laterally to the body wall
presetally in c although no bladder is detectable.
Simple large saccular seminal vesicles in IX and XII. Prostates irregular, lobu-
lated to racemose, linear structures, extending through XVIII-XXI (XVIII-XIX in
paralectotype 3) ; each with a sinuous narrow muscular duct which is joined entally
near its junction with the gland, by the vas deferens. In transverse section (paralec-
totype i) each gland is seen to have a narrow central lumen around which are many
distinct groups of glandular cells, each group with the appearance of a liver lobule.
The cells of the lobule radiate from a central axis which in some cases can be seen to
form a definite lumen although this is not lined by a recognizable epithelium. The
epithelium of the central lumen of the gland is sporadically interrupted to receive
the lumina or axial cores of the lobuli. Width of the gland : width of the central
lumen = 1-61 : 0-18 mm = 9:1. Penial setae stout, curved, tapering ectally to a
simple blunt tip ; the ectal tapered region strongly convex ' dorsally ', less convex, to
286 B. G. M. JAMIESON
concave, 'ventrally'. The ectal region typically (paralectotype 3) ornamented with
approximately 16 circlets spaced at regular intervals averaging 10 /zm, each circlet
serrated by very numerous simple minute contiguous spines ; the spines anteriorly
directed, adpressed to the surface of the seta, only slightly notching the seta in profile.
Specimens from other localities are similar in general form and may have similar
ornamentation, or the circlets may be interrupted in such a way that the different
portions are not always in the same transverse line, and in some specimens (e.g. 7/2,
specimen i) there is no regular transverse arrangement of spines but merely scattered
short transverse groups. Sometimes (13/4, specimen 5) there are short spine-like
ridges adherent longitudinally to the setal surface immediately ectal to the spines
(and also a tendency to longitudinal furrowing of the seta) ; the presence of ridges in
this specimen approaches the condition characteristic of C. polynephricus poly-
nephricus ad urethrae (q.v.) ; lengths of well developed setae 073-1-5 mm, general
width of shaft 23-37 /um (paralectotype 3 ; specimens I, 5, 10, 14, 18, 24, 34, 42).
Metagynous ; ovisacs not developed (paralectotype i) or present (specimen 12 and
other, but not all, specimens). Spermathecal diverticulum (inseminated) digitiform.
TYPE-LOCALITY. Mt Wellington ; Hobart. Map, 14/4. Parattah. Map, 14/1.
MATERIAL EXAMINED.
Mt Wellington, I47°I5'E. 42°55'S. Map, 14/4, NMV : 01436, 3 paralectotypes.* Fern
Glade, Emu R., Burnie, under logs, I45°55'E. 4i°o5'S. Map, 7/2, Dr J. L. Hickman, 17 May
1954 ~ specimens 1-4 BM(NH) : 1973.2.11-14. Tarraleah, over pipeline, I46°25'E. 42°2o'S.
Map 13/4, Dr J. L. Hickman, 27 May 1954 - specimens 5-6 BM(NH) : 1973.2.15-16. Collins-
vale, near Hobart, myrtle forest, I47°I5'E. 42°5o'S. Map, 14/3, Dr J. L. Hickman, 8-9 Nov
1955 - specimens 7-11 BJ : T.38-42. Mt Wellington, I47°I5'E. 42°55'S. Map, 14/4, Dr J. L.
Hickman, 27 Oct 1955 - specimens 12 and 13 BJ : 7.39-40 ; under stones on top of mountain,
Dr J. L. Hickman, 13 Nov 1954 - specimen 14 BJ : 741 ; specimen 15 AM : W52I3 ; specimens
16 and 17 BM(NH) : 1973.2.17-18. Shoobridge Bend track, altitude approximately 580 m,
in loam and clay in eucalypt - fern woodland, Dr B. G. M. Jamieson and Mr E. A. Bradbury,
19 Aug 1971 - specimens 18-23 BJ : 742-47. Hobart, Waterworks Road, under stones,
i47°2o'E. 42°5o'S. Map, 14/5, Dr J. L. Hickman, 17 Aug 1954 - specimen 24 BM(NH) :
1973.2.19 ; specimen 25 BJ : 748. Lenah Valley, near Newton Falls, under stones, Dr J. L.
Hickman, 24 Jun 1957 - specimen 26 AM : W52I4 ; specimens 27-33 AM : 53o5-53ri '• Lenah
Valley, New Town Creek, Dr J. L. Hickman, 15 Sep 1953 - specimens 34-36 TM : 1^331-333 ;
31 Aug 1953 - specimen 37 TM : 1^334. 'Tasmania', Dr J. L. Hickman, specimens 38-41
TM : K335-338.
REMARKS. In the above account the occurrence of a pair of longitudinal excretory
ducts and associated stomate megameronephridia is demonstrated for the first time.
C. polynephricus polynephricus is distinguishable from C. p. urethrae and from the
closely related C. enteronephricus in location of the anterior genital markings, or at
least those in IX, significantly lateral of the spermathecal pores. The similarity,
nevertheless, of the genital fields and general anatomy of C. polynephricus and C,
enteronephricus, and the intermediate position of C. p. urethrae (q.v.) indicate specia-
tion from a common origin within the genus.
* The specimen designated the lectotype by Jensz and Smith (1969), 041, is here shown to be a
specimen of Perionychella (V esiculodrilus) mortoni. This specimen is clearly not a syntype as it did not
contribute to the type-description of C. polynephricus. Smith has agreed that its designation as the
lectotype was invalid according to Article 74 (a) (i) of the International Code of Zoological Nomenclature.
TASMANIAN EARTHWORMS
287
sp.p
1mm
g.m
B
FIG. 23. Genital fields. Cryptodrilus polynephricus urethrae : A, paratype 8 ; B, paratype 20.
288 B. G. M. JAMIESON
Specimens 14-17, from the summit of Mt Wellington, and specimens 24 and 25,
from Hobart, display a distinct genital field, lacking genital markings in IX, and
having those in X and XI further median than is usual. They are also charac-
terized by a longitudinal reduplication of the nephridial bladders in c lines but the
latter condition occurs sporadically, in some segments, in the paralectotypes of
which, though they are incompletely mature, one has rudiments of a normal field,
with genital markings in IX and X well lateral of the spermathecal pores. As the
nephridial condition of the atypical specimens is echoed in the types, and in view of
the great variation in the genital fields reported for another species of Cryptodrilus
(C . fastigatus v. Jamieson, iQ72b) erection of a separate taxon for specimens 14-17,
24 and 25 does not seem warranted.
Cryptodrilus polynephricus urethrae subsp. n,
Fig. 2, 2iC, 23, 246, 3iJ-M ; Plates 44-59. Table 9
1 = 330, 160 mm ; w = 12, 6 mm ; s = 175, 156.
Prostomium epilobous |, § ; dorsal tongue moderately narrow, closed and (in
holotype) weakly canaliculate. Paired elliptical accessory genital markings with
centres in a and with wide, whitish glandular borders, in intersegments 9/10 and 10/11
(holotype, paratype i) and 11/12 (paratype i) or unilateral, on the right, in holotype.
Similar markings centred in a in 17/18 ; in ab, nearer a, in 18/19 and median to a in
19/20. Female pore in XIV, paired anteromedian of a in holotype ; unpaired, mid-
ventral, in paratype i. Spermathecal pores very slightly lateral of a.
TABLE 9
Intersetal distances in segment XII in 13 specimens* of
Cryptodrilus polynephricus urethrae
holotype
maximum
minimum
mean
standard deviation
aa
ab
be
cd
dd
dc
cb
ba
4-0
i-i
2-4
2-3
14-0
2-6
3-o
i-i
4-0
i-i
2-9
3'3
14-0
3-i
3-0
i-i
1-6
0-6
1-6
1-9
2-7
1-9
1-6
0-6
2-5
0-9
2-1
2'3
6-7
2-3
2-1
0-9
0-8
0-2
0-4
0-4
3'4
0-3
0-4
O-2
standardized (u = 100)
aa
ab
be
cd
dd
dc
cb
ba
13-02
3-72
7-91
7'44
45-58
8-37
9-77
4-19
16-14
5-83
14-29
15-92
45-58
15-92
14-29
5-83
9-30
3-45
7-91
7'44
20-11
8-37
9-09
3-45
12-61
4-71
10-76
12-20
3I-9I
12-19
10-88
4-75
i-75
0-72
1-81
2-74
7-66
2-43
i-53
0-69
2-68
I-OO
2-28
2-59
6-77
2-59
2-31
I-OI
holotype
maximum
minimum
mean
standard deviation
interval/afe
* H, paratypes, i, 7, 8, 9, 10, 12, 13, 16, 18, 20, 21 and 22.
Supra-oesophageal vessel recognizable in |VIII, IX-XIII, XIV well developed.
Oesophagus vascularized and dilated segmentally in VIII-XIII, XIV ; intestinal
TASMANIAN EARTHWORMS 289
typhlosole absent or a rudimentary, very low ridge present from approximately
XXIV (paratype i and holotype respectively). Nephridia : 5 pairs of large tufts,
each with multiple ducts, in II -VI ; the ducts of those of II discharging at the junc-
tion of mouth and body wall ; those of III-VI sending composite ducts to a simple,
though still composite, common longitudinal duct which passes forwards to the
anterior end of the pharynx (i.e. enteronephric) ; sparse exonephric micromero-
nephridia present on the body wall lateral to each tuft. Nephridia in VII forming
smaller tufts which discharge exonephrically in this segment (and holotype, in VI ?)
and appear to send some ducts to the common duct of the anterior tufts ; a tendency
to ventral tufting continues but decreases to the anterior intestinal region where there
are approximately 10 (paratype i) to 15 (holotype) exonephric micromeronephridia
on each side ; the nephridia discharging presetally in a, b, be, c, cd and above d.
The median nephridium of the first intestinal segment (XVI) is the first to have a
preseptal funnel (lying in XV) (demonstrated in paratype i). This arrangement of
nephridia persists posteriorly but the stomate nephridium now enlarges to become a
megameronephridium, the duct of which discharges in a line together with one or
two ast ornate micromeronephridia. Caudalty there is an astomate micromero-
nephridium in each of setal lines a, b, c and d and 4 above d on each side (holotype,
there being few if any above d, but additional intersetal nephridia in paratype i) ;
the megameronephridium contributes its duct to a slender longitudinal excretory
duct (ureter) of which there is one on each side of the ventral nerve cord ; the
posterior end of this ureter was traced into the extreme caudal segments but its
termination was not elucidated. Nephridial bladders totally absent.
Racemose seminal vesicles in XI and XII (holotype, paratype i) with a unilateral
rudiment in IX in holotype. Prostates depressed tubuloracemose, in XVIII-XIX
(holotype) -XX (paratype i), irregular in width and lobulated, with very narrow
central lumen. Penial setae moderately stout, ectally strongly curved, the tip,
which may or may not be reflexed dorsally or bent ventrally, strongly flattened and
almost parallel sided, only slightly tapering ectally, and sometimes ventrally some-
what inrolled ; the end almost straight or convex and simple or denticulate ; orna-
mentation ental to the depressed apical region (holotype) consisting of numerous
transverse rows, spaced at approximately 6 /mi intervals, each row minutely serrate,
the teeth of the serrations very numerous, closely continguous, in a single series,
minute but attenuated in the axis of the seta, short ridges like a fringe present ectal
to each tooth row ; the setal surface also with numerous longitudinal furrows. In
other specimens the tooth rows may be broken up into short transverse or oblique
series of several to many teeth and these groups may lie in the same transverse or
oblique line as adjacent groups or may be scattered apparently randomly, length =
2-1-3-9 mm, general width of shaft = 39-45 /im (holotype ; paratypes i, 6, 20).
Spermathecal diverticulum simple, uniloculate but with a basal swelling.
VARIATION.
Genital markings. The genital field in this subspecies is distinguished from that
of C. p. polynephricus by location of the post-spermathecal, preclitellar accessory
genital markings in or median to the longitudinal lines of the spermathecal pores ;
2QO B. G. M. JAMIESON
the latter pores are usually in or very slightly lateral of a lines. In paratypes 18 and
19 the genital markings are further lateral in ab, but are not significantly lateral of the
spermathecal pores as these are similarly displaced. It can be seen in the following
table that the genital fields in the holotype and paratype I are representative of the
most common conditions in the remaining paratypes. Paratypes 2, 6 and 22 are
shown to have atypical fields but at present there appears to be no justification for
giving them separate taxonomic status.
Variation in the genital markings of 26 specimens of
Cryptodrilus polynephricus urethrae
Total number
Genital markings Specimen of specimens
Small elliptical paired in ab to median of
a at hind margin of VII P6, 22 2
VIII P2, 22 2
IX P2 i
X P2 i
Paired intersegmental in line
with or median to the spermathecal
pores in 9/10 H Pi, 3, 4, 5, 10, n, 16, 16
17, 18, 19, 20, 21, 23,
25, 26
lo/n H Pi-5, 10, n, 12, 13, 14, 21
16, 17, 18, 19, 20-26
11/12 H(R) Pi-5, 8, 12-14, 18-21, 18
23-26
12/13 P4. 5. 8, 12 4
16/17 Pg i
17/18 H Pi-5, 10-14, I6-i9, 21-26 21
18/19 H Pi-6, 8, 10-12, 14, 16-19, 22
21-26
19/20 H Pi-6, 8, 10-13, 16-19, 21 17
20/21 P8 I
21/22 P8 I
H = holotype ; P = paratype.
Other variation. Anatomical variation in this subspecies in some systems exceeds
what would normally be expected within a species. Thus though a typhlosole is
absent or rudimentary in the holotype and paratype, respectively, there is a definite
laminar typhlosole in paratypes 2, 10, n and 20 and appreciable typhlosoles are
present in a considerable proportion of the remaining type-specimens. In specimens
from the same sample (e.g. 13/3, paratypes 10-12) there may be variation from a deep,
to a moderate, to a very rudimentary typhlosole. Other morphological variation does
not appear to correlate with this typhlosole variation and all specimens are distingui-
shed from the typhlosolate C. enteronephricus in lacking segmentally enteronephric
nephridia and in having a ureter. Variation in the distribution of seminal vesicles
also occurs from IX and XII, to IX, XI and XII, to XI and XII and, most remark-
ably, the female pore is paired or (paratypes i, 3, 4, 8, 12, 26) unpaired, median.
TASMANIAN EARTHWORMS 291
Variation in setal ratios is considerable (Table 9) and one of the more extreme
variants (paratype 9) has the last hearts in XII.
REMARKS. The location of the preclitellar genital markings approximately in
line with or median to the spermathecal pores, the absence of nephridial bladders
and the widened ends of the penial setae distinguish this subspecies from the
nominate subspecies.
MATERIAL EXAMINED.
Arne Road, E. side of Styx River bridge, approx. 1000 ft into rainforest, I46°35'E. 42°5o'S.
Map, 13/8, Mr D. Frankombe, 24 Sep 1958 - holotype TM : K34I ; paratype 2 TM : K/342.
Arne Valley, 10 miles from Greeveston, i46°5o'E. 43°io'S. Map, 18/3, Dr J. L. Hickman,
ii Nov 1955 - paratype i BM(NH) : 1973.2.24 ; paratypes 3-5 TM : 1^343-345 ; paratypes
22-25 TM : K346-349. Florentine Valley, surface soil under litter in E. regnans forest, under
stones, i46°25'E. 42°35'S. Map, 13/3, Mr M. Gilbert, 7 Mar 1956 - paratype 6 BM(NH) :
1973.2.25: Florentine, Mr J. M. Gilbert, 21 Jul 1955 - paratype 7 TM : K35O ; Florentine,
rainforest soil, Mr J. M. Gilbert, 4 Aug 1955 - paratypes 8 and 9 AM : ^^5313-5314 ; Florentine
Valley, under pieces of wood and bark, Mr J. M. Gilbert, 14-19 ? Aug 1955 -paratypes 10-12
AM : W53i5~53i7. Tarraleah, i46°25" E. 42°2o'S. Map, 13/4, Professor V. V. Hickman,
Dec 1953 - Jan 1954 ~ paratypes 13 and 14 BJ : T/52-53. Dee Bridge, Lyell Highway, i46°35'E.
42°i5'S. Map, 13/7, Dr J. L. Hickman, 24 May 1954 -paratype J5 BJ : T54. Maydena,
i46°4o'E. 42°45'S. Map, 13/9, Dr J. L. Hickman, 15 Oct 1960 - paratypes 16 and 17 BM(NH) :
1973.2.26-27. Russel River, Mr C. D. King, 16 Feb 1939 - paratypes 18 and 19 BJ : T/55-56.
Kallista, near Junee, i46°35'E. 42°45'S. Map, 13/11, Mrs Rodway, Tasmanian Biological
Survey : Dei3, 19 Apr 1938 — paratype 20 BJ : T57. Snowy Mts, 4000 ft, i46°4o'E. 43°oo'S.
Map, 18/2, Tasmanian Biological Survey: J8, Mr C. D. King, 19 Jan 1939 -paratype 21
TM :
REMARKS. The separate subspecies, C. polynephricus urethrae (and possibly it
should be a species) is recognized for specimens which have anterior genital markings
in or median to the spermathecal lines and which lack bladders while possessing
ureters. It appears to be linked to the nominate subspecies by intermediate forms
which are described below.
Whether polynephricus and urethrae are to be retained as subspecies or should be
regarded as sibling species cannot be unequivocally decided at present as knowledge
of them is limited to their morphology and inadequate geographical evidence. The
map (Fig. 2) shows that they are largely allopatric. Their morphological similarity,
which exceeds that between most species, and the occurrence of morphological inter-
mediates where their distributions overlap, are taken for the time being to indicate
that they are subspecies.
Cryptodrilus polynephricus polynephricus ad urethrae
Fig. 2, 2iB, 246, 31! ; Plates 32-43. Table 10
Some specimens (42-45, 47, 49), referable to C. polynephricus polynephricus from
the appearance of the preclitellar genital field, differ from paralectotypes and other
specimens described above in having the nephridial bladders limited to the hindbody
where they are accompanied, as usual, by a ureter. These specimens are here in-
cluded in the nominate subspecies but are distinguished as Cryptodrilus polynephricus
292
B. G. M. JAMIESON
1
sp.p
FIG
1mm
24. Genital fields. A, Cryptodrilus polynephricus polynephricus, specimen
B, C. p. p. ad urethrae, specimen 49. C, C. p. urethrae, paratype i.
12.
polynephricus ad urethrae. With these are included two specimens (46 and 48) in
which genital fields are undeveloped but which have the same nephridial condition.
This 'polynephricus ad urethrae' morph, in addition to general morphological re-
semblance to C. p. polynephricus, further resembles the latter in having seminal
vesicles in IX and XII and simple-pointed penial setae although in specimen 44
this shows signs of the flattening characteristics of the penial setae of C. p. urethrae.
(The latter specimen is exceptional in having last hearts in XII.) The ornamentation
of some penial setae also seems to be intermediate between that in the two subspecies.
As shown in the map (Fig. 2), the distribution of polynephricus ad urethrae is inter-
mediate between, though partly overlapping, the distributions of the two subspecies,
on present evidence. Details of the penial setae follow : setae stout, curved, taper-
ing ectally to a simple blunt or flattened chisel-like tip ; where simple tipped, the
ectal tapered region strongly convex dorsally, less convex, to concave, ventrally.
The ectal region ornamented with numerous transverse or oblique circlets, at inter-
vals of approximately 10 /urn, each circlet serrated by very numerous spines which in
some specimens (e.g. 13/3, specimen 42) are two deep and are immediately preceded
TASMANIAN EARTHWORMS 293
ectally by short spine-like ridges adherent longitudinally to the setal surface. Spines
anteriorly directed, adpressed to the setal surface, only slightly notching the seta in
profile. The circlets in some (e.g. 13/3, specimen 44) broken up and a regular trans-
verse arrangement scarcely apparent. In 14/4, specimen 48, the setal surface bears
numerous longitudinal furrows. In specimen 44 the spines are simple, and ridges
and significant furrowing are lacking but the tip differs from the typical polynephricus
condition in being flattened ; 1 = 0-55-1-75 mm, general width of shaft = 19-30 /mi
(specimens 42, 44, 48, 49).
TABLE 10
Intersetal distances in segment XII in Cryptodrilus polynephricus ad urethroe
specimen 42 1-3
specimen 44 i-o
specimen 45 i-o
specimen 47 1-5
specimen 48 1-3
specimen 49 2-0
mm
ab
be
cd
dd
dc
cb
ba
0-6
i-i
1-6
2-6
1-8
i-i
0-6
0-4
I-O
I-O
1-6
i-o
i-o
0-4
o-4
0-6
0-9
i-o
1-8
2-3
I-O
1-7
i-o
0-4
0-6
0-7
i-i
1*5
2'5
J"5
I-I
0-7
0-8
1-5
2-3
3-4
2-3
i«5
0-8
standardized (u = 100)
ab be cd dd dc cb ba
I2-OO
5-33
10-67
15-33
24-00
16-67
10-67
5-33
I4'I5
5'37
13-17
I3-I7
21-95
13-66
I3-I7
5-37
14-21
5-58
13-20
14-72
18-78
14-72
13-20
5-58
13-25
5-36
11-99
16-09
20-82
I5'I4
11-99
5-36
12-46
6-40
10-44
14-81
24-24
14-81
10-44
6-40
13-73
5-39
10-29
15-69
23-53
15-69
10-29
5-39
I3-30
5-57
11-63
14-97
22-22
15-12
11-63
5-57
2-39
i-oo
2-09
2-69
3-99
2-71
2-09
I-OO
specimen 42
specimen 44
specimen 45
specimen 47
specimen 48
specimen 49
mean
interval/aft
MATERIAL EXAMINED.
Florentine Valley, surface soil under litter in E. regnans forest, under stones, i46°25'E. 42°35'S.
Map, 13/3, Mr M. Gilbert, 7 Mar 1956 - specimens 42 and 43 TM : 1^339-340 ; specimens 44 and
45 BJ : T49-50. Russell River, i46°4o'E. 42°45'S. Map, 13/9, Mr C. D. King, 16 Feb 1939-
specimen 46 BM(NH) : 1973.2.20. Mt Wellington, i47°i5'E. 42°55'S. Map, 14/4, Dr J. L.
Hickman, 13 Jan 1954 -specimen 47 AM : W53I2 ; Betts Vale at creek, in moss, under logs
and stones, Dr J. L. Hickman, 4 Mar 1954 - specimen 48 BJ : T$i. Eagle Hawk Neck, in kelp
on rocks, I47°55'E. 43°oo'S. Map, 19/2, Dr J. L. Hickman, Apr 1956 - specimen 49 BJ : T5iA.
Cryptodrilus simsoni Spencer, 1895
Fig. 2, 25 A, 3iN, O ; Table n
1 = 61 mm, w = 4, 4-1 mm, 5 = 95+ (posterior amputee), 125 (lectotype,
specimen 2).
Prostomium canaliculate, epilobous f , closed. First dorsal pore 3/4. Setae 8 per
segment in longitudinal rows, d irregular posteriorly. Nephropores not visible
294
B. G. M. JAMIESON
VIII
B
FIG. 25. Genital fields. A, Cryptodrilus simsoni, lectotype. B, Pinguidrilus tasmanicus,
lectotype.
(specimen 2) or sporadically visible throughout (lectotype) and especially distinct on
and shortly behind the clitellum where in each segment there are 6 presetal pores on
each side ; i median to a, i in each of a, b, c and d and i near the dorsal midline
(lectotype) . In some Fern Dene specimens there are as many as 3 presetal pores in
the vicinity of each of setae a and b on the clitellum. Clitellum annular, XIII-
XVIII, weak in XIII and XVIII but well developed between. Male pores on dome-
shaped papillae with diameter equalling ab and central pore in a, on XVIII. Paired
eye-like genital markings with centre in b at 19/20 and in ab at 20/21 ; a median
transverse genital marking occupying bb in 11/12 (specimen 2) or 10/11 (lectotype)
bearing a row of approximately 10 small circular markings corresponding with the
intersegmental furrow. Female pores a pair of small slits just median and slightly
anterior to a on XIV. Spermathecal pores 2 pairs, in 7/8 and 8/9, on small papillae
in a (specimen 2) or slightly lateral of a (lectotype).
Last hearts in XIII (latero-oesophageal) ; connectives to dorsal and supra-
oesophageal vessels present. Supra-oesophageal in IX-XIII (specimen 2) ;
indistinct in lectotype. Gizzard small, easily compressible, in V. Extramural
calciferous glands absent. Oesophagus with circumferential vascular striae and
internally papilliform rugae in X-XVI (specimen 2), the rugae becoming definite
lamellae in XIV-XV and less distinct in XVI in lectotype. Intestinal origin XVIII ;
typhlosole absent.
TASMANIAN EARTHWORMS
295
TABLE n
Intersetal distances in segment XII in Cryptodrilus simsoni
mm
aa
J'4
i-i
i-o
i-i
i-i
i-o
ab
0-4
0-6
0-4
be
i-i
1-2
1-2
I-O
1-2
cd
i-i
dd
4-8
4-7
4'3
4'4
4'3
3-8
standardized (u = 100)
dc
i-i
cb
i-i
1-2
1-2
I-I
1-2
ba
0-4
0-4
aa
ab
be
cd
dd
dc
cb
ba
11-48
3-36
10-08
II-2O
38-64
10-64
10-08
3-o8
9-73
5-oi
9-37
11-21
39-23
11-50
9-44
4-13
9-15
4-25
n-ii
10-13
39-54
10-46
ii-ii
4-25
9-38
3'75
10-63
11-56
38-75
11-56
10-94
3'44
9-87
4-14
9-24
11-78
39-17
11-78
9-87
4-14
9-00
4-82
10-93
12-86
34-4I
12-54
10-61
4-82
9'77
4-22
10-23
11-46
38-29
11-41
io-34
3-98
2-32
I-OO
2-42
2-72
9-07
2-70
2-45
0-94
lectotype
specimen i
specimen 2
specimen 3
specimen 4
specimen 5
lectotype
specimen i
specimen 2
specimen 3
specimen 4
specimen 5
mean
interval/aft
Nephridia : in the forebody and anterior intestinal region meronephridia dis-
charge presetally in each of a, b, c and d lines and between d line and the middorsum.
Those discharging in c and d lie far below these loci and have very long ducts. It was
not possible to demonstrate unequivocally that there were two nephridia at a as the
external pores indicate. In the most anterior segment the ventral nephridia form
small (exonephric ?) tufts and do not appear to be accompanied by the full com-
plement of more dorsal nephridia ; in the intestinal region there is a preseptal funnel
on each side near the nerve cord, this condition persisting to the caudal extremity ;
median nephridia are not enlarged relative to others though they are moderately
large. In the posterior region the dorsalmost nephridia discharge in an approximately
straight line above, in or below d lines which are there irregular. In the midbody a
small bladder-like dilatation (parasitic artefact ?) of a nephridial duct is sporadically
present though most ducts do not reveal even an intramural bladder.
Holandric (funnels iridescent in X and XI) ; gymnorchous ; elongate racemose
seminal vesicles in XI and XII, on the anterior septa ; similar structures of unknown
function are present on the anterior septum of X (specimen 2) and of preceding seg-
ments, to at least V (lectotype) . Metagynous ; ovisacs absent. Prostates restricted
to XVIII, tongue shaped racemose but with a short narrower ectal portion prior to
the duct and, in serial sections, revealing a single narrow central lumen with cuboidal
epithelium. Vas deferens joining the ental end of the sinuous duct. Penial setae
moderately slender, curved, needle like ; the tip tapering narrowly, simply but
bluntly pointed ; ornamentation absent ; length = 1-2-1-4 mm ; general width =
13-14 /am (specimen 2). Spermathecae 2 pairs. Diverticulum single, clavate,
uniloculate (specimen 2) or circumscribed by a furrow so that it has a slight tendency
to be bifid (lectotype).
296 B. G. M. JAMIESON
Genital markings. The location of genital markings in 11/12 (in contrast with 10/11
in the lectotype), 19/20 and 20/21 is constant in the 9 Fern Dene specimens ; the
markings in 20/21 are always more median than those in 19/20.
TYPE-LOCALITY. Emu Bay. Map, 7/2. Launceston. Map, 9/5.
MATERIAL EXAMINED.
Launceston, i47°io'E. 4i°25'S. Map, 9/5, A. Simson, Feb 1892, NMV : Gi 82 -lectotype.
Fern Dene, Ironcliflf Road, near Penguin, I46°O5'E. 4i°o5'S. Map, 8/1, Collector ? 13 Oct
1954 - specimens 1-3 TM : 1^352-354 ; 4 BJ : T58 ; specimens 5 and 8 BM(NH) : 1973.2.29-
30 ; specimens 6 and 7 AM : W53i8-53i9 ; specimen 9 BJ : T59.
REMARKS. The transversely papillate or punctate ventral genital marking in
10/11 or 11/12 is distinctive of this species. The presence of dorsal pores also dis-
tinguishes it from all Tasmanian species.
Genus PSEUDOCRYPTODRILUS Jamieson, 1972!}
DIAGNOSIS. A pair of combined male and prostatic pores on XVIII. Setae 8 per
segment. Nephridia 2 to 3 on each side excepting caudally where there is a single pair
of stomate holonephridia in each segment. Typhlosole weak or absent. Prostates
depressed tubular.
DESCRIPTION. Terrestrial. Dorsal pores present. Setae commencing on II,
in 8 longitudinal rows which may become irregular posteriorly. Ventral setal
couples widely paired (aa < zab) ; setae of the dorsal couples widely separated,
cd > 2ab and approximately as far apart as the two setal couples of a side (cd = 0-8-
i-ibc) ; dorsal median intersetal distance approximately one-third of the circum-
ference (dd = o-3«). Setae a and b of XVIII replaced by penial setae. Nephro-
pores inconspicuous. A pair of combined male and prostatic pores on XVIII ;
postclitellar. Spermathecal pores 2 pairs, in 7/8 and 8/9 or anteriorly in VIII and IX.
Dorsal blood vessel single, continuous onto the pharynx ; last hearts in XII (latero-
oesophageal). Subneural absent. Gizzard moderate to well developed, in V.
Calciferous glands absent but oesophagus modified (dilated or internally rugose) in
XV-XVI, XVII ; intestinal origin XVIII or XIX ; a low typhlosole present or
absent. Anterior nephridia tuft-like or simple. Nephridia exonephric ; in the
anterior intestinal region 2 on each side ; in the forebody 2 (P. diaphanus) or 3 (P.
acanthodriloides, in which the median one is stomate) ; caudally holonephric, with
a single stomate, avesiculate nephridium on each side. Holandric ; gymnorchous ;
seminal vesicles in IX and XII. Prostates depressed tubular. Penial setae present.
Metagynous ; ovisacs present. Spermathecae with i or 2 clavate, uniloculate
diverticula.
DISTRIBUTION. Eastern Subregion of Australia, southern faunal province :
Victoria. Tasmania. (2 species.)
TYPE-SPECIES. Megascolides diaphanus Spencer, 1900.
TASMANIAN SPECIES. P. acanthodriloides sp. n.
REMARKS. The new species is included in the formerly monotypic Pseudocrypto-
drilus chiefly because it is a megascolecine which is meronephric anteriorly but
TASMANIAN EARTHWORMS
297
1mm
A
sp.p
1mm
FIG. 26. Genital fields. A & B, Pseudocryptodrilus acanthodriloides, holotype : A, entire
male field ; B, detail of same. C, Rhododrilus Httoralis, holotype.
18
298 B. G. M. JAMIESON
holonephric caudally. This condition has presumably evolved from a fully holo-
nephric condition and may have been convergently acquired in the two species but
the genus as redefined above to include acanthodriloides is satisfactorily homo-
geneous, there being no variation which is not acceptable as intrageneric.
The longitudinal orientation of the accessory genital markings, constituting
tubercula pubertatis, is highly unusual in the Megascolecidae and it is, therefore,
remarkable that another Tasmanian species Oreoscolex bidiverticulatus (q.v.) has
tubercula and an almost identical genital field. Either tubercula, so rare elsewhere,
have been acquired independently in the two species or a special relationship exists
between the two (i.e. common descent from a species with similar tubercula). A
decision between the two alternatives cannot be made, at least on present evidence,
but the morphology of the Oreoscolex species is not inconsistent with relationship.
Origin of Oreoscolex (Megascolecini) in south-eastern Australia from local holonephric
species or from species which, like P. acanthodriloides, were partly holonephric and
partly meronephric is quite conceivable. Irrespective of its relationships, Pseudo-
cryptodrilus presents a most interesting transition from the holonephric to the mero-
nephric condition.
Pseudocryptodrilus acanthodriloides sp. n.
Fig. 4, 26A, B, 328 ; Plates 61-63. Table 12
1 = 70, 67 mm, w = 3-4, 3-2 mm, s = 127, 126.
Prostomium epitanylobous with grooved peristomium. First dorsal pore 4/5.
Setae 8 per segment, in regular longitudinal rows throughout. Nephropores
sporadically visible in a, b and d lines. Clitellum annular, ^XIII-XVII (holotype ;
not developed in paratype i). Male pores on XVIII in a, A pair of longitudinal
ridges (tubercula pubertatis) present, each occupying ab and extending from just
behind each male pore to the setal arc of XIX ; the pair of tubercula, with the male
pores, lying on a whitish glandular tumescence ; a similar additional pair of tubercula
extending from the male pores to the setal arc of XVII in some specimens (see field
variation). Female pores a pair shortly anteromedian to a of XIV. Spermathecal
pores 2 pairs, at the anterior margins of VIII and IX, but not in the intersegments,
each on a hemispherical papilla ; the ventral surface of segments VII-IX and
posterior VI tumid.
TABLE 12
Intersetal distances in segment XII in Pseudocryptodrilus acanthodriloides
mm
aa ab be cd dd dc cb ba
holotype
paratype i
paratype 2
paratype 3
paratype 4
0-9
o-5
0-9
0-9
2-2
0-9
0-9
o-5
0-9
0-4
0-8
0-9
1-8
0-9
0-8
0-4
0-9
0-8
0-4
0-4
0-8
0-8
0-8
I-O
1-7
1-8
0-8
i-o
0-8
0-8
0-4
0-4
0-9
0-4
1*0
0-9
2-O
0-9
i-o
0-4
aa
ab
be
cd
dd
dc
cb
ba
12-04
6-02
"•57
12-04
28-70
12-04
ii-57
6-02
12-56
6-03
12-06
12-56
26-13
12-56
12-06
6-03
I3-83
6-38
11-70
12-23
25-53
12-23
11-70
6-38
11-44
5'97
11-94
I3-43
25-87
I3-43
11-94
5-97
11-90
5'7i
12-86
11-90
27-14
11-90
12-86
5-7i
12-35
6-02
12-03
12-43
26-67
12-43
12-03
6-02
2-05
I -00
2-OO
2-06
4-43
2-06
2-OO
I-OO
TASMANIAN EARTHWORMS 299
Table 12 (cont.)
standardized (u = 100)
aa ab be
holotype
paratype I
paratype 2
paratype 3
paratype 4
mean
interval/aft
Last hearts in XII (latero-oesophageal) . Supra-oesophageal in VII-XII, weakly
developed. Gizzard moderate, muscular but easily compressible in V. Extramural
calciferous glands absent, longitudinal rugae well developed, though not forming
laminae, in XV and XVI. Intestinal origin |XVIII ; typhlosole absent.
Nephridia : exonephric much coiled, tuft-like nephridia in II, III and IV discharg-
ing by simple ducts presetally in c and or d lines. Further posteriorly simple avesi-
culate nephridia discharging presetally in association with setal lines, apparently not
all 4 setal lines in any one segment (a, b and d in the holotype in which the median-
most nephridium is seen to have a preseptal funnel). In the anterior intestinal
region 2 nephridia on each side, discharging presetally in b and d lines. Caudally
with a pair of stomate, avesiculate holonephridia only, in each segment, discharging
presetally in d lines (holotype and paratype i). Holandric (funnels iridescent in X
and XI) ; gymnorchous ; seminal vesicles racemose, in IX and XII. Metagynous ;
small ovisacs in XIV. Prostates flattened, tubular, tortuous, in XVIII and XIX ;
vas deferens joining junction of gland and duct. Penial setae moderately slender
but relatively short, only slightly curved, the ectal extremity flattened like the blade
of an oar; ornamentation a few deep half-encircling jagged scars ; length = 0-6 mm;
general width of shaft = 10 /zm (paratype 3) . Spermathecae, 2 pairs ; diverticulum
single, clavate, uniloculate.
Genital markings. Of 13 paratypes, 3 (paratypes i, 7, 12) have male fields as the
holotype, with genital markings restricted to a pair of tubercula extending from
|XVIII into XIX ; 6 paratypes (2, 3, 4, 5, n and 13) have these tubercula augmented
solely by a pair (unilateral only in paratypes 3, n (R) and 13 (L)) extending from
approximately |XVIII-|XVII which lie in the same tumid area as the posterior
pair ; in i paratype (8) the posterior tubercula are present but there is also a mid-
ventral circular presetal glandular patch in XVIII and a similar postsetal patch in
XVII ; i paratype (10) has the posterior tubercula and only the midventral presetal
patch in XVIII ; while i (paratype 6) has this condition, and also the anterior paired
tubercula, and i (paratype 9) is similar to paratype 6 but lacks the left anterior tuber-
cula. Most specimens have ventral tumescence in VII-IX and some in VI also.
MATERIAL EXAMINED.
Great Lake, shore under decaying gum leaves, I46°45'E. 4i°55'S. Map, 8/2, Dr J. L. Hickman,
26 May 1954 - holotype TM : K355 ; paratypes i, n and 13 BM(NH) : 1973.2.31-33 ; paratypes
2-5 TM : W356-359 ; paratypes 6, 8-10 BJ : T6o, 61-63 ; paratypes 7 and 12 AM : W532O-
5321.
REMARKS. The male genital field distinguishes this species from P. diaphanus.
300 B. G. M. JAMIESON
Tribe DIGHOGASTRINI Jamieson,
Genus HICKMANIELLA gen. nov.
DIAGNOSIS. Combined male and prostatic pores a pair on XVIII. An oeso-
phageal and an intestinal gizzard present. Nephridia few exonephric avesiculate
meronephridia per segment ; the medium nephridium with preseptal funnel.
Prostates tubuloracemose. Spermathecae 2 pairs.
DESCRIPTION. Terrestrial. Form circular in cross section ; anus terminal.
Dorsal pores present. Setae numerous in each segment (> 30 in posterior segments) ;
ventral gap large ; dorsal gap recognizable anteriorly, inappreciable caudally
(zz = zy). Nephropores not externally visible. Clitellum annular ; anterior to the
male porophores. Combined pores of the vasa deferentia and a pair of tubulorace-
mose prostates one pair on XVIII. Female pores a pair on XIV. Spermathecal
pores 2 pairs, in intersegments 7/8 and 8/9.
Last hearts in XII (latero-oesophageal). Subneural vessel absent. Oesophageal
gizzard in V. Extramural calciferous glands absent. Intestine commencing in the
vicinity of the prostate glands ; with a very thick-walled intestinal gizzard preceded
by a thin-walled crop ; typhlosole and caeca absent. Few, exonephric, avesiculate
meronephridia per segment, the median nephridium in oesophageal and intestinal
regions with preseptal funnel ; tufting and enteronephry absent. Holandric ;
gymnorchous ; seminal vesicles in IX and XII. Metagynous ; ovisacs absent.
Penial setae present. Spermathecae with clavate but internally multiloculate
diverticulum.
TYPE-SPECIES. Hickmaniella ,opisthogaster sp. n.
DISTRIBUTION. Northwest Tasmania. (Monotypic.)
REMARKS. Hickmaniella is almost unique in the known Megascolecidae in possess-
ing an intestinal gizzard. (Pleionogaster has 3 or 4 such gizzards.) This feature is
characteristic of the Oriental family Moniligastridae (in which, however, Gates (1962)
states the gizzards are actually oesophageal though post-ovarian), the holarctic family
Lumbricidae and the palaearctic, neotropical and Ethiopian subfamily Alminae of
the non-Australian family Glossoscolecidae. The Moniligastridae (Order Moniligas-
trida) differ from Megascolecids in many significant respects, including the single-
layered clitellum, large-yolked eggs, presence of intra-septal testis sacs, location of
one or two pairs of male pores in the next segment or segments respectively and
absence of an anterior gizzard. The Lumbricidae are distinguished by the absence
of an oesophageal gizzard, location of the male pores in or in front of segment XV,
location of the clitellum far behind these pores, the exclusively lumbricine setae and
in other respects. In contrast, the Glossoscolecidae, while showing their closest affini-
ties with the Lumbricidae (vide Jamieson 197 id), show notable resemblances to the
Megascolecidae which may indicate that they are more closely related to the latter
than is any other family with the exception of the Eudrilidae. The intestinal gizzard
of Hickmaniella constitutes a further link between the Megascolecidae and the Glossos-
colecidae. Like many other common features of the two families the intestinal gizzard
may have been independently acquired in each but sufficient similarities exist to
TASMANIAN EARTHWORMS 301
suggest that if this be the case we have here a case of parallelism in evolution stemming
from a common, though remote, ancestry rather than convergence in relatively unre-
lated lineages of the order Haplotaxida. These similarities include the following (in
addition to general characteristics of the non-Alluroidid Lumbricina of Brinkhurst and
Jamieson 1971) : progression from the lumbricine to the perichaetine arrangement of
setae ; presence of prostate-like glands in some Glossoscolecidae which resemble the
prostates of the Megascolecidae ; frequent presence in both families of oesophageal
gizzards and the mutual possession of intestinal gizzards ; location of the male
pores in XVIII in some Glossoscolecidae, as in most Megascolecidae ; location of
these pores on the clitellum in some Megascolecidae as in many Glossoscolecidae and
finally replication of the nephridia, in Tritogenia, as in many Megascolecids. While
inclusion, on the one hand, of the Glossoscolecidae and Lumbricidae in the super-
family Lumbricoidea and, on the other hand, of the Megascolecidae and Eudrilidae
in the superfamily Megascolecidae is here upheld, the above-mentioned similarities
between the parent families Megascolecidae and Glossoscolecidae bring the two
families closer together and endorse recognition of the suborder Lumbricina as a
natural grouping. The Alluroididae included in it by Brinkhurst and Jamieson
(1971) relate the suborder to other Haplotaxid suborders, especially the Haplotaxina.
Hickmaniella opisthogaster sp. n.
Fig. 4, i8A, 326, D ; Plates 64-66
1 = 61, 59 mm, w = 5-5, 5-0 mm, s = 87, 82.
Prostomium canaliculate, tanylobous. First dorsal pore 4/5. Setae on XII 32
(holotype) or 49 (paratype i) ; caudally 31 to 38 per segment ; aa regular through-
out (holotype) or becoming irregular posteriorly (paratype i) ; anteriorly aa =• $ab ;
caudally aa = $ab ; zz varying from 2-2-^zy, anteriorly, to i-i-$zy caudally.
Nephropores not externally visible. Clitellum annular, XIII-|XVII. Male pores
approximately in a lines of XVIII, each median to a penisetal pore, on prominent
dome-shaped medianly almost contiguous papillae which expand the segment longi-
tudinally. Accessory genital markings paired, medianly conjoined oval glandular
prominences with papillate or pore-like centres in approximately b lines in inter-
segments 11/12, 15/16 and 20/21, resembling the male porophores in appearance ; an
additional similar marking unilateral on the left, in 14/15, and on the right, in 21/22
in paratype i. Female pores paired, anterior to setae a on XIV. Spermathecal
pores 2 inconspicuous pairs, in 7/8 and 8/9, in a lines.
Last hearts in XII (hearts in X-XII latero-oesophageal, with connective from
dorsal and supra-oesophageal vessel ; supra-oesophageal vessel recognizable in
|VII-|XIV, moderately developed). Oesophageal gizzard moderate, in V. Extra-
mural calciferous glands absent. Intestine commencing in XVIII in which it is
thin-walled and crop-like (holotype, paratype i) ; a very large, extremely thick-
walled intestinal gizzard in XIX (holotype) or XX (paratype i) ; typhlosole absent.
Few, exonephric, avesiculate meronephridia per segment, commencing in II ; 3 on
302 B. G. M. JAMIESON
each side in the intestinal region ; median nephridium with preseptal funnel in
intestinal and at least the posterior oesophageal segments ; tufting and enteronephry
absent.
Holandric ; gymnorchous ; seminal vesicles in IX and XII. Metagynous ;
ovisacs absent. Prostates tubuloracemose, extending laterally in, and restricted to,
XVIII ; with very narrow central lumen wihch bears narrower epithelium-lined side
branches. Penial setae (holotype) 2 mature and i completely developed on each
side ; when functional 1-6-17 mm l°ng> general width of shaft 31 ^m ; gently or
strongly curved, widened basally, the ectal tip flattened, usually curved, and with
margins slightly inrolled ; the ectal region of the shaft, but not the inrolled region,
ornamented by numerous short, transverse minutely serrated linear markings of
which there are 4 or more across the width of the seta ; this ornamentation readily
visible under the higher powers of the light microscope without recourse to oil
immersion. Spermathecae 2 pairs, in VIII and IX ; a single clavate but internally
multiloculate diverticulum ; the duct with a large dorsal swelling.
Genital markings. The four mature paratypes (i, 3-5) resemble the holotype in
having paired intersegmental genital markings in 11/12, 15/16 and 20/21. In
addition paratype i has unilateral genital markings in 14/15 (left) and 21/22 (right)
resembling the other markings.
MATERIAL EXAMINED.
Parrawe, I45°35'E. 4i°2o'S. Map, 7/3, Dr J. L. Hickman, 25 Aug 1954 - holotype TM :
paratype i BM(NH) : 1973.2.34 ; paratype 2 TM : K36i. Table Cape, i45°45'E. 4i°oo'S.
Map, 7/1, Dr J. L. Hickman, 24 Aug 1954 - paratypes 3 and 4 AM : W5322-5323. Hellyer
Gorge, i45°35'E. 4i°2o'S. Map, 7/3, Dr J. L. Hickman, 28 May 1954 - paratype 5 BJ : T64.
Tribe MEGASCOLEGINI s. Jamieson, i97ia
Genus OREOSCOLEX Jamieson, ig73a
DIAGNOSIS. Setae 8 per segment. A pair of combined pores of vasa deferentia
and racemose or tubuloracemose prostates on XVIII. Gizzard in V. Extramural
calciferous glands and typhlosole present or absent. Meronephric ; anterior neph-
ridia astomate, avesiculate and exonephric, or enteronephric, usually tufted ; caudal
nephridia avesiculate, exonephric (or enteronephric ?), with numerous preseptal
funnels on each side in each segment. Spermathecae in 2 or 3 segments, paired or
median ; diverticula uniloculate or multiloculate.
DESCRIPTION. Terrestrial worms. Form circular in cross section ; anus ter-
minal. Prostomium prolobous to tanylobous. Dorsal pores present, commencing
in 3/4-20/21. Setae 8 per segment, in regular longitudinal rows throughout or
becoming irregular posteriorly. Setal ratios very variable aa : ab : be : cd : dd =
1-9-3-4 : i : 1-5-5-9 : I'6-3'4 : 3-1-12-9 ; dd:u = 0-11-0-58 ; cd = or < never sig-
nificantly > be. Clitellum annular or less commonly saddle shaped, occupying 4-6^
segments, its posterior limit in XVII-XIX. Combined pores of the vasa deferentia
and a pair of prostates in XVIII ; the prostates externally racemose or lobulated
TASMANIAN EARTHWORMS 303
tubular, in the former case with or without, in the latter always with an axial lumen ;
this lumen possessing side branches. Female pores a pair, anteromedian of setae a,
in XIV. Spermathecal pores unpaired posteriorly in VII and VIII (0. imparicystis)
or paired in 6/7, 7/8-8/9. Last hearts in XII or (0. bidiverticulatus) XIII ; latero-
oesophageal hearts (always ?) with connectives from dorsal and supra-oesophageal
vessels. Subneural vessel absent. Gizzard large, in V. Oesophagus with internal
rugae, or laminae, or extramural calciferous glands, intestine commencing in XVI,
XVII or XVIII ; typhlosole absent (0. peculiaris), rudimentary or well developed,
commencing i|-8 segments behind the intestinal origin. Meronephric ; anterior
nephridia astomate, avesiculate and exonephric, or enteronephric (entering pharynx
and buccal cavity), usually tufted ; caudal nephridia avesiculate, with numerous
preseptal funnels on each side in each segment and exonephric or (0. imparicystis)
possibly enteronephric. Holandric with seminal vesicles in IX and XII or rarely in
XI and XII or (0. imparicystis, leai) XII only ; or (peculiaris) with anterior male
organs reduced (incipiently metandric) gymnorchous (testis-sacs absent). Metagy-
nous (ovaries in XIII) ; ovisacs present or absent. Vasa deferentia joining the mus-
cular duct of the prostate at midlength or entally, or the ectal extremity of the gland.
Penial setae present or (imparicystis) absent. Spermathecae with one or two uni-
loculate or multiloculate diverticula.
TYPE-SPECIES. Oreoscolex imparicystis Jamieson I973a. New South Wales.
DISTRIBUTION. Eastern Subregion of Australia : New South Wales and Tasmania.
Tasmanian species
0. bidiverticulatus sp. n.
*0. campestris (Spencer, 1895)
*0. irregularis (Spencer, 1895)
*0. leai (Michaelsen, 1910)
0. longus sp. n.
0. peculiaris sp. n.
0. sexthecatus sp. n.
*0. wellingtonensis (Spencer, 1895)
Other species
Four species from New South Wales (examined by the author as indicated) in
addition to the type-species. The additional species accord exactly with the generic
diagnosis but detailed generic characterization is not attempted here.
*Notoscolex grandis Fletcher, i887a 'Cotypes' AM : W.I494-
*Cryptodrilus illawarrae Fletcher, iSSga Mt Kembla AM : W.I3H.
*Cryptodrilus saccarius Fletcher, i887b 'Cotypes' AM : W.I372.
*Cryptodrilus simulans Fletcher, i88gb Syntypes AM : W.I499-
* Comb. nov.
B. G. M. JAMIESON
XIX
B
. km t "
FIG. 27. Genital j&elds. A, Oreoscolex bidiverticulatus, holotype. B, O. longus, holotype.
TASMANIAN EARTHWORMS 305
KEY TO TASMANIAN SPECIES OF OREOSCOLEX
Spermathecal diverticula uniloculate ......... 2
Spermathecal diverticula multiloculate ........ 5
Male pores in ab on XVIII .......... 3
Male pores median to a on XVIII ......... 4
Reniform extramural calciferous glands in XV and XVI. Fig. 27 A O. bidiverticulatus sp. n.
Extramural calciferous glands absent. Fig. 276 . . . . . O. longus sp. n.
Male pores almost contiguous medianly, on a circular porophore median to b lines.
A median genital marking in 20/21. Fig. 28C . . O. irregularis (Spencer, 1895)
Male pores shortly median to a lines on a transverse oval porophore which extends
lateral of b lines and impinges on adjacent segments. Fig. 28B . O. peculiar is sp. n.
Spermathecae 3 pairs. Fig. 2gA O. sexthecatus sp. n.
Spermathecae 2 pairs ........... 6
A pair of eye-like genital markings present in ab in 17/18 or XVII and in each of
several intersegments behind XVIII. Fig. 28A . O. campestris (Spencer, 1895)
Eye-like or other paired genital markings absent ....... 7
A large median transverse genital marking in each of intersegments 15/16-18/19,
extending approximately to b lines. That in 18/19 indistinct ; that in 17/18
indistinct posteriorly O. leai (Michaelsen, 1910)
A large mid ventral glandular pad in 19/20 or 20/21 extending laterally beyond b
lines. Fig. 296 ...... O. wellingtonensis (Spencer, 1895)
Oreoscolex bidiverticulatus sp. n.
Fig. 3, 27A, 32 J ; Plate 67. Table 13
1 = 150, 106 mm, w = 4-9, 4-2 mm, s = 118, 141.
Prostomium epilobous J, open. First dorsal pore 4/5. Setae 8 per segment in
regular longitudinal rows throughout. Nephropores not visible. Clitellum annular
(£XIII dorsally) XI V-XVII ; not developed in paratype i. Male pores on XVIII,
in ab, on long tubercula pubertatis which extend from the setal arc of XVII to that of
XIX ; a glandular depression occupying the space between the tubercula. Female
pores paired anteromedian of a on XIV. Spermathecal pores 2 pairs, in 7/8 and 8/9,
on small papillae in a.
TABLE 13
Intersetal distances in segment XII in Oreoscolex bidiverticulatus
mm
aa ab be cd dd dc cb ba
holotype 1-8 0-6 1-7 i-i 4-4 i-i 1-8 0-5
paratype i 1-5 0-5 1-4 i-o 4-0 i-o 1-4 0-5
standardized (u = 100)
aa ab be cd dd dc cb ba
holotype
paratype i
mean
interval/oft
14-02
4-58
12-94
8-63
33'42
8-63
13-75
4-04
13-54
4-31
12-31
8-92
35-38
8-92
12-31
4-31
13.78
4-45
12-63
8-78
34-40
8-78
13-03
4-18
3-10
I-OO
2-84
1-97
7-73
1-97
2-93
0-94
B. G. M. JAMIESON
sp.p
1mm
FIG. 28. Genital fields. A, Oreoscolex campestris, specimen i. B, O. peculiaris, holotype.
C, O. irregular is, specimen i.
Last hearts in XIII (latero-oesophageal) ; supra-oesophageal in VII-|XIV,
moderately developed. Gizzard large and firm, in V, with anterior rim. A pair of
large reniform extramural calciferous glands in each of segments XV and XVI, each
joined over most of its height to the lateral wall of the oesophagus, free for only a
short extent dorsally, and ventrally, and with a wide communication with the
central lumen of the oesophagus ; sessile, lacking a duct. Each gland with numerous
radial longitudinal laminae which span the lumen of the free part of the gland.
TASMANIAN EARTHWORMS 307
Intestinal origin XVIII but no constricted oesophageal valve at 17/18 ; deep dorsal
typhlosole commencing in XXIII, continued anteriorly as a slight ridge to XIX. In
II-VII with a pair of exonephric tufted nephridia with numerous anteriorly running
ducts (always ?) opening in the same segment ; the tufts largest in V and VI and very
large in V. Few exonephric meronephridia on each side in succeeding segments until
in XV and XVI there are very deep lateral bands of exonephric nephridia, each with
several tubules, with the appearance of transverse tufts ; meronephridia again few
in more posterior segments ; caudally numerous exonephric meronephridia on each
side dependent from the anterior septum and with many preseptal funnels in a
transverse row. Holandric (funnels iridescent in X and XI) ; seminal vesicles race-
mose, in IX and XII. Metagynous ; ovisacs present. Prostates racemose ;
restricted to an enlarged segment XVIII : with (paratype 2) branched internal ducts
and lacking a central lumen. Penial seta almost straight, filiform, tapering ectally,
lacking ornamentation ; the tip broadened, irregular and bent relative to the axis ;
length left seta = 1*1 mm, general width of shaft = 12 pm (holotype). Sperma-
thecae 2 pairs ; each with I (paratype i) or 2 (holotype) (inseminated) clavate,
uniloculate diverticula.
MATERIAL EXAMINED.
Fern Glade, Emu River, Burnie, i45°55'E. 4i°o5'S. Map, 7/2, Dr J. L. Hickman, 24 Aug
1954 ~ holotype TM : K362 ; paratype i BM(NH) : 1973.2.35 ; paratype 2BJ : T65.
REMARKS. Intraspecific variation from one to two diverticula on a spermatheca
is highly unusual in the Oligochaeta but there is no doubt that the holotype and para-
types are conspecific. Location of last hearts in XIII and of calciferous glands in XV
and XVI, together with the configuration of the genital markings, is diagnostic of
this species.
Oreoscolex campestris (Spencer, 1895)
Fig. 3, 28A, 32K ; Plates 68-72. Table 14
Cryptodrilus campestris Spencer, 1895 : 39, PL II, fig. 13-15.
Notoscolex campestris ; Michaelsen, 1900 : 192 ; Jamieson, 19710 : 78.
1 = 87, 82 mm, w = 3-5 mm, s = 109, 114 (specimens i and 2). (1 = 50-75 mm,
w = 3 mm, Spencer.)
TABLE 14
Intersetal distances in segment XII in Oreoscolex campestris
mm
aa ab be cd dd dc cb ba
specimen i 1-4 0-4 2-1 i-i 1-6 i-i 2-2 0-4
specimen 2 1-5 0-5 2-5 i-o i-i i-o 2-5 0-4
specimen 3 i-i 0-3 1-4 0-8 i-o 0-9 1-6 0-3
specimen 4 1-6 0-5 2-4 i-i 1-6 1-2 2-3 0-5
specimen 5 1-4 0-5 2-3 i-o 1-4 i-o 2-5 0-4
13-79
4-14
20-00
II-O3
!5-i7
10-34
21-38
4-14
14-19
4-73
23-65
9-46
10-81
9-46
23-65
4-05
15-00
4-00
19-00
11-00
14-00
I2-OO
22-00
4-00
14-19
4-52
21-29
9-68
14-19
10-97
20-64
4-52
13-50
4-73
21-60
9-45
13-50
9-45
23-63
4-°5
14-13
4-42
21-11
IO-I2
13-53
10-44
22-26
4-15
3-20
I-OO
4-78
2-29
3-06
2-36
5-04
0-94
308 B. G. M. JAMIESON
Table 14 (cow/.)
standardized (u = 100)
aa ab be cd dd dc cb ba
specimen i
specimen 2
specimen 3
specimen 4
specimen 5
mean
interval/aft
Prostomium indistinctly tanylobous, canaliculate (epilobous £, Spencer). First
drosal pore 4/5 (3/4 Spencer). Setae 8 per segment in regular longitudinal rows
throughout, dd narrowing caudally. Nephropores not externally visible. Clitellum
annular, XIII-|XVIII (XIII-XVII, saddle shaped in XVII, Spencer). Male pores
in or median to a on XVIII on a pair of low circular porophores ; each pore preceded
and succeeded, on the porophore, by the pore of a penisetal follicle. Paired eye-like
genital markings occupying ab at 17/18, 19/20, 20/21, 21/22 (specimens, I, 2) and
22/23 (specimen i). (Anterior XVII, 18/19 and 19/20, Spencer.) Female pores,
paired, anteromedian of a on XIV. Spermathecal pores two pairs on hemispheroidal
papillae at the anterior margins of VIII and IX, just lateral of a. Last hearts in XII
(latero-oesophageal). Supra-oesophageal in VIII-|XIII, well developed. Gizzard
firm and very large, in V. Extramural calciferous glands absent ; oesophagus
internally rugose in ?, X-XV. Intestinal origin £XVI (XVI, Spencer) ; a very
deep j_-shaped dorsal typhlosole commencing in XIX. Nephridia : very large
pharyngeal tufts in IV and V ; smaller anterior tufts apparently exonephric ;
caudally with a transverse row of numerous preseptal funnels on each side, the
nephridia exonephric. Holandric (funnels iridescent in X and XI) ; gymnorchous ;
seminal vesicles racemose, in IX and XII. Metagynous ; ovisacs present in XIV (?).
Genital markings in 18 specimens of Oreoscolex campestris
Genital markings Specimen Total number
of specimens
16/17 J6 i
17/18 1-18 18
18/19 5, 13. M (L). 18 4
19/20 1-18 1 8
20/21 1-18 18
21/22 1-3, 7-II, 12 (L), 14, I6-I8 13
22/23 I I
Prostates restricted to an enlarged segment XVIII ; broad flattened smooth lobes,
each with 3 serial portions joined together by narrow alternating bands and probably
originating from a tubular type ; the two glands completely encircling and adpressed
to the intestine and to each other middorsally ; central lumen very narrow, with few
lateral branches ; vas deferens joining the muscular prostate duct at its junction
with the gland. Penial setae gently curved, moderately stout, slightly tapering
ectally, the tip truncated and fluted by longitudinal grooves ; the ectal region
TASMANIAN EARTHWORMS 309
ornamented with the exception of the vicinity of the tip ; ornamentation consisting
of numerous regularly spaced circlets of large acute flattened spines the bases of which
are in contact with adjacent spines in the circlet, some spines bifid, most simple :
length of a functional seta = 0-34 mm ; general width of shaft = 13 /mi (2 specimens) .
Spermathecae 2 pairs ; diverticulum single, sessile, multiloculate.
TYPE-LOCALITY. Parattah. Map, 14/1.
MATERIAL EXAMINED.
Parattah, under moss, in earth at base of cliff, also in earth along edges of logs, i47°25'E.
42°2o'S. Map, 14/1, Professor V. V. Hickman and Dr J. L. Hickman, 18 Aug 1954 ~ specimens
i, 3-6 TM : 1^363-367 ; specimens 2, 7-11 BM(NH) : 1973.2.36-41 ; specimens 12-16 AM :
W5324-5328 ; specimens 17 and 18 BJ : T66-&7.
REMARKS. The syntypes, in the National Museum of Victoria, 048, are desic-
cated, and yield no significant taxonomic information.
Oreoscolex irregularis (Spencer, 1895)
Fig. 3, 28C, 32L. Table 15
Cryptodrilus irregularis Spencer, 1895 : 34~35. PI- I ng- I-3-
Notoscolex irregularis ; Michaelsen, 1900 : 191 ; Jamieson, 19710 : 79.
1 = 194, 195 mm, w = 9 mm, s = 120, 136 mm (specimens i and 2 throughout
this account).
Prostomium epilobous f (ca. \, Spencer) canaliculate, open, parallel-sided. First
dorsal pore 4/5. Setae 8 per segment ; all rows becoming irregular near the caudal
extremity (or from XIV, Spencer) ; setae a and b absent in XVIII although their loci
are lateral of the male pores. Clitellum annular XIV-XVIII ; some clitellar modi-
fication dorsally throughout XIII ; dorsal pores occluded, setae and intersegmental
furrows retained though fainter than elsewhere. Male pores minute, almost con-
tiguous medianly, in a circular depression (with, in specimen i, posterior extension,
i.e. pyriform), on a low median circular porophore which almost fills segment XVIII
longitudinally and extends to a lines laterally. A transversely oval median accessory
genital marking with crater-like centre, in 20/21, extending slightly lateral of a lines
and almost reaching the setal arcs of XX and XXI. Female pores a pair, antero-
median to setae a of XIV. Spermathecal pores not demonstrable externally ; from
internal dissection, well median of a lines if not medianly united, but said by Spencer
to be in a lines.
TABLE 15
Intersetal distances in segment XII in Oreoscolex irregularis
mm
act ab be cd dd dc cb ba
specimen i 2-2 i-i 1-9 1-7 14-4 1-7 1-9 I-i
specimen 2 2-6 1-3 1-7 1.2 15-0 1-2 1-7 1-3
specimen 3 2-6 1-3 2-1 2-6 15-7 2-6 2-0 1-3
specimen 4 2-8 0-9 1-5 2-1 13-9 1-9 1-5 0-9
aa
ab
be
cd
dd
dc
cb
ba
8-49
4-n
7-40
5-58
55-35
6-58
7-12
4-11
9-86
4'93
6-58
4-66
57-54
4-66
6-58
4-93
8-53
4-27
6-87
8-53
52-14
8-53
6-64
4-27
10-88
3-63
5-86
8-37
54HI
7-25
5-86
3-63
9'44
4-24
6-68
7-04
54-86
6-76
6-55
4-24
2-23
I-OO
1-58
1-66
12-94
i-59
J-54
I -00
3io B. G. M. JAMIESON
Table 15 (cow/.)
standardized (u = 100)
aa ab be
specimen i
specimen 2
specimen 3
specimen 4
mean
interval/aft
Dorsal vessel single, continuous onto the pharynx. Last hearts in XII (hearts in
X-XII latero-oesophageal, with dorsal and larger supra-oesophageal connectives) ;
supra-oesophageal in VIII-|XIII, well developed. Gizzard large but easily com-
pressible in V ; oesophagus dilated and vascularized and internally rugose in IX-XV
(-XIII, Spencer), but calciferous glands absent. Intestinal origin posterior in XVII
(in XVIII, Spencer) ; typhlosole rudimentary, commencing in XXV ; muscular
thickening and caeca absent. Nephridia : large tufts in II-IX, especially large in II
and III ; those in V-IX exonephric by composite ducts in their own segments ; those
in IV sending composite ducts forwards, to the vicinity of the pharynx or, II and III,
of the buccal cavity ; some tufting of median nephridia persists to XIII and in II-
XIII there are no nephridia lateral to the median nephridium. In XIV the nephridia
begin to spread laterally to form a parietal band of astomate, avesiculate, exonephric
micromeronephridia and by XIV these bands are well developed and so dense as to
have the appearance of transversely expanded tufts, though with numerous separate
parietal ducts in the same segment. Caudally with transverse bands of 30-40
avesiculate exonephric parietal micromeronephridia on each side, each with a
minute, preseptal funnel.
Holandric (funnels iridescent in X and XI) ; gymnorchous ; seminal vesicles
racemose in IX and XII (XII only, Spencer). Metagynous ; ovisacs present.
Prostates each a broad tongue-shaped lobe, with surface lobulations, racemose in
appearance but internally with single axial duct which gives off lateral branches into
the gland ; double vas deferens joining the middle of the long muscular external duct.
Penial setae absent. Spermathecae 2 pairs, duct about as long as and clearly de-
marcated from the ampulla, joined near its ectal end within the body wall by a
single clavate (inseminated) diverticulum ; the duct entering the body wall at the
lateral border of the ventral nerve cord, near the midventral line.
Genital markings. Constant though not always equally developed. Crater on
male porophore is pyriform in specimen 4 as in specimen i.
TYPE-LOCALITY. Table Cape. (Types in NMV, not seen.)
MATERIAL EXAMINED.
Table Cape, i45°45'E. 4i°oo'S. Map, 7/1, Dr J. L. Hickman, 24 Aug 1954 - specimens I
and 2 TM : 1^368-369 ; specimen 3 BJ : T68 ; specimen 4 BM(NH) : 1973.2.42.
Oreoscolex leai (Michaelsen, 1910)
Notoscolex leai Michaelsen, 1910 : 99-102, Fig. 18, 19.
1 = 310 mm, w = 7 mm, s = 152.
TASMANIAN EARTHWORMS 311
Prostomium prolobous or possibly slightly and broadly pro-epilobous. Setae
unusually delicate, in the mid- and hindbody, with the exception of the hind fourth,
regularly arranged and widely paired, aa : ab : be : cd : dd = 1-6:1:2:1:8. Fur-
ther anteriorly, near the male pores each ventral pair is much narrowed but the
lateral pair is not (in XIX aa = $ab). Clitellum midventrally insunk, though not
saddle-shaped, in |XIII-XIX. Male pores on small, almost circular papillae on
XVIII in ab, about one-twelfth of the circumference apart. Large unpaired, mid-
ventral rounded hexagonal genital markings, extending laterally approximately to
setal lines b and leaving only a small region of the segment between them in 15/16,
16/17, 17/18 and 18/19 ; the most posterior, in 18/19, on^Y indistinct, that in 17/18
posteriorly indistinctly demarcated. Female pores indicated by a rather long, trans-
verse furrow anteriorly in XIV. Spermathecal pores 2 pairs, in 7/8 and 8/9, those of a
pair about one-tenth of the circumference apart, in line with the male pores (in setal
lines ab ?).
Gizzard large in V ; oesophagus only slightly swollen segmentally ; calciferous
glands absent. Intestine, at least in the first 24 segments, spiral. Dorsal vessel
single. Last hearts in XII. Meronephric ; a large number of nephridial tubules in
each segment ; in the f orebody, in the region ' of thickened septa, nephridia less
numerous and aggregated ventrally to give a spurious appearance of holonephridia.
Holandric ; 2 pairs of sperm funnels, in X and XI ; seminal vesicles racemose in XII.
Prostates slenderly tongue-shaped, long, narrow and flattened ; tubuloracemose,
with a central lumen extending almost the whole length of the glandular part ; the
mass of the glandular part composed of numerous adpressed parts with apparently
epithelial ducts opening into the main duct. Vasa deferentia of a side entering the
internal duct of the prostate separately at its distal fourth. Penial setae approx.
2 mm long and, almost throughout, 20 /um wide ; gently and simply curved, ectally
simply and slenderly pointed. Distinct ornamentation absent.
Spermathecae 2 pairs, duct short, conical, bearing a multiloculate diverticulum
with 4 or 5 sperm chambers.
TYPE-LOCALITY. Tasmania, vicinity of Hobart.
MATERIAL EXAMINED.
None available.
Oreoscolex longus sp. n.
Fig. 3, 276, 32N, O ; Plates 73-80. Table 16
1 = 325, 215 mm, w = 6-0, 5-2 mm, s = 210, 231.
Prostomium epilobous |, open. Dorsal pores indistinct ; the first in 20/21.
Setae 8 per segment, in regular longitudinal rows throughout. Nephropores not
visible. Clitellum annular, |XIII-XVIII, weaker ventrally in bb (holotype), not
developed in paratype i. Male pores at the centres of prominent transversely con-
joined papillae occupying ab on XVIII ; two long conspicuous penial setae projecting
in the vicinity of each pore. Accessory genital markings : large oval paired pads in
16/17 and 19/20 extending both laterally and medianly of ab, a smaller pair in 9/10
312 B. G. M. JAMIESON
in ab, and a median ventral pad in 20/21 (holotype) ; or median pads similarly
extending laterally of b in each of intersegments 12/13-16/17 and paired markings
in 19/20 and 20/21 (paratype i). Female pores a pair anteromedian of a, on XIV.
Spermathecal pores 2 pairs, in 7/8 and 8/9, on small papillae in a.
TABLE 16
Intersetal distances in segment XII in Oreoscolex longus
mm
aa ab be cd dd dc cb ba
paratype i 1-3 0-6 1-7 1-2 4-5 1-3 1-7 0-5
paratype 2 1-4 0-6 1-9 1-2 5-4 1-4 1-8 0-6
paratype 3 1-9 0-7 2-1 1-8 4-5 1-8 2-1 0-7
paratype 4 1-7 0-6 2-1 1-4 5-0 1-4 1-6 0-6
paratype 5 1-8 0-6 2-0 1-3 5-0 1-3 1-9 0-6
standardized (u = 100)
aa ab be
paratype i
paratype 2
paratype 3
paratype 4
paratype 5
mean
interval/aft
Last hearts in XII (latero-oesophageal ; connective to dorsal vessel not recogniz-
able). Supra-oesophageal in (?)-• |-XIII ; well developed. Gizzard in V but septa
5/6 and 6/7 very delicate and with more posterior septa deflected far posteriorly by it ;
very strong and firm, fusiform with anterior rim. Extramural calciferous glands
absent. Oesophagus especially vascular in XI-XIII ; in IX-XIII with higher
internal lamellae which are most numerous and most slender in XI and XII in which
they occlude the oesophageal lumen though they do not fuse centrally. Intestinal
origin XVII (holotype), or XVIII (paratype i) ; a deep laminar dorsally typhlosole
(at first rudimentary) beginning in XXIV. Nephridia all avesiculate : anterior
nephridia indistinct ; no enteronephry detected. From approximately XIV pos-
teriorly there is a plumose median nephridium on each side which has a single pre-
septal funnel but which discharges exonephrically by several ducts which converge
to the body wall presetally in a ; other astomate meronephridia discharge presetally
in b, c and in d or above d. Caudally there are several (> 10) preseptal funnels on
each side in each segment, located in ab ; the meronephridia being exonephric
(holotype).
Holandric (funnels iridescent in X and XI) ; gymnorchous ; seminal vesicles
small, racemose, in XI and XII (holotype ; not developed in paratype i). Metagy-
nous ; ovisacs present. Prostates very tortuous, slender, lobulated tubes in XVIII-
XX with narrow central lumen. Vasa deferentia of a side running together then
separating before joining the muscular prostatic duct near its ental limit. Penial
setae curved, moderately stout, slightly tapering ectally, the tip flattened and widened
aa
ab
be
cd
dd
dc
cb
ba
10-19
4-41
13-50
9-37
34-99
10-19
13-22
4-13
9-91
4-44
13-37
8-39
37-14
9-91
12-40
4-44
11-87
4-57
13-70
11-42
28-77
11-42
13-70
4-57
11-92
3-97
14-44
9-97
34-87
9.97
10-88
3-97
12-40
3-95
13-85
8-93
34<63
8-93
13-37
3-95
11-26
4-27
13-77
9-62
34-08
10-08
12-71
4-21
2-64
I -00
3-23
2-25
7-99
2-36
2-98
0-99
TASMANIAN EARTHWORMS 313
to form a blade with almost straight transverse or oblique somewhat crenulated end ;
ectal region, with the exception of the blade, ornamented ; ornamentation consisting
of numerous regularly spaced circlets of large acute spines, the circlets interrupted so
that the spines are in groups, of approximately 5, slightly separated from other
groups in the same circlet ; some spines bifid, most simple (holotype, paratype I, 7) ;
length of right seta = 1-6, 2-6 mm (holotype, paratype i) ; general width of shaft =
18, 25, 19 /zm (holotype, paratypes i, 7). Spermathecae 2 pairs ; diverticulum
(inseminated) digitiform but ectally widening greatly to form a bulbous swelling over
the pore ; no distinct duct developed.
Genital markings in 8 specimens of O. longus
Genital markings
Paired in
Paired (often conjoined) in
Total number
Specimen
of specimens
9/io
H
P2-7
7
I2/I3
Pi
i
13/14
Pi
i
14/15
Pi
i
15/16
Pi
i
1 6/i 7
H
Pi-3,5-7
7
19/20
H
Pi -7
8
20/21
H
Pi-3,P5-6,7(L)
7
21/22
PS
i
Median in
H = holotype ; P = paratype .
MATERIAL EXAMINED.
Cox's Bight, I46°I5'E. 43°3o'S. Map, 18/1, Tasmanian Biological Survey, Mr C. D. King,
Nov 1938 and Jun 1939 - holotype TM : K37O ; paratypes i and 5 BM(NH) : 1973.2.43-44 ;
paratypes 2-4 TM : 1^371-373 ; paratype 6 AM : ^5329 ; paratype 7 BJ : T6g.
REMARKS. O. longus is identifiable by the distinctive genital field and the
bulbous swelling of the spermatheca over its pore with an accompanying short,
digitiform diverticulum.
Oreoscolex peculiaris sp. n.
Fig. 3, 286, 32M ; Plates 81-82. Table 17
1 = 46, 60 mm, w = 5-7, 6-0 mm, s = 146, 125.
Prostomium epilobous \ with suggestions of a tanylobous condition. First dorsal
pore (5/6 ?), 6/7. Setae 8 per segment, b, c and d irregular posteriorly ; ab narrowing
and slightly nearer to the ventral midline as XVIII is approached from anteriorly or
posteriorly, but absent in XVIII. Nephropores not visible, clitellum annular
XIII-XVII. Male pores a pair in XVIII median to setal lines a, at the centre of a
large oblong-oval pad-like glandular protuberance which, basally at least, includes
the setal arcs (and setae ab) of XVII and XIX and extends laterally almost to mid
be, 4 circular, translucent pore-like markings present on the pad but discernible with
difficulty, two at the anterior and two at the posterior limit of XVIII, with centres
shortly lateral of b lines (holotype, paratype i ; similar in paratypes 2-5 ; less
19
i'3
0-4
3'7
I-O
3-o
i-i
3-4
0-4
1-6
0-6
4'3
1-6
2-5
1-8
4-2
0-6
0-8
0-6
2-1
1-2
1-9
1-3
2-1
0-6
0-7
0-6
3'3
2-2
1-8
2-1
3'4
0-6
aa
ab
be
cd
dd
dc
cb
ba
8-96
2-99
25-87
6-97
20-90
7-96
23-38
2-99
9-17
3-33
25-00
9-17
I4-58
10-42
24-58
3*75
7-48
5-44
19-72
11-56
17-68
12-24
20-40
5'44
4-87
3-90
22-40
15-10
12-18
14-12
23-38
3-90
7-62
3-92
23-25
10-70
16-34
11-19
22-94
4-02
1-94
I-OO
5-93
2'73
4-17
2-85
5-85
1-03
3i4 B. G. M. JAMIESON
developed and tending to be divided in two medianly in paratypes 6-8). Female
pores a pair anterior and slightly median to a on XIV. Spermathecal pores difficult
to observe, 2 pairs of small pores in a at 7/8 and 8/9.
TABLE 17
Intersetal distances in segment XII in Oreoscolex peculiaris
mm
aa db be cd dd dc cb ba
holotype
paratype i
paratype 5
paratype 6
standardized (u = 100)
aa ab be
holotype
paratype i
paratype 5
paratype 6
mean
interval/aft
Last hearts in XII (latero-oesophageal ; connective to dorsal vessel a delicate
filament). Supra-oesophageal in (?)-|XIV, well developed. Gizzard large and
firm, with anterior rim, in V. -Extramural calciferous glands absent ; vascularized
with high rugae in VIII-XIV, at first few ; especially numerous in XII-XIV in
which they form flattened papillae. Intestinal origin XVII ; spiral from XX
posteriorly ; typhlosole absent. Nephridia (holotype) : first nephridia large entero-
nephric (pharyngeal) tufts in V ; large tufts, decreasing in size posteriorly, in VI-
VIII, those in VI probably enteronephric, the others apparently exonephric. Small
exonephric tufts in IX-XII. In XIII and XIV the individual tubules and ducts of
the tufts have dissociated so that the meronephridia begin to spread laterally and by
XV numerous discrete micromeronephridia form bands extending far laterally, the
nephridia being astomate, avesiculate and exonephric ; approximately 15 micro-
meronephridia on each side in each segment shortly behind the prostates. Abruptly,
at the sixty-ninth segment, several dorsal nephridia which are dependent from the
anterior septum, become enlarged and acquire, each, a preseptal funnel, giving several
funnels laterally on each side in each segment ; a single preseptal funnel also is
present on the medianmost nephridium (holotype ; caudal arrangement confirmed
in paratype 3). Holandric (but incipiently metandric ?) funnels in X and XI but
iridescent in XI only ; gymnorchous ; seminal vesicles racemose, in IX (holotype)
and XII (holotype, paratype i). Metagynous ; ovisacs absent. Prostates tubulo-
racemose, running laterally in and restricted to XVIII ; vas deferens joining the
gland near its ectal end. Penial setae filiform, the tip forming a small rounded
button, ornamentation absent but longitudinal furrows present ; length of left seta
TASMANIAN EARTHWORMS 315
= 0-8 mm, general width of shaft = 12 p.m (holotype). Spermathecae 2 pairs ;
diverticulum single, clavate, uniloculate.
MATERIAL EXAMINED.
Mt Wellington, Shoobridge Bend Track, approx. 580 m, in loam and clay in eucalypt - fern
woodland, i47°i5'E. 42°55'S. Map, 14/4, Dr B. G. M. Jamieson and Mr E. A. Bradbury,
19 Aug 1971 -holotype TM : K/374 ; paratypes i and 3 BM(NH) : 1973.2.45-46 ; paratypes 2,
6 and 7 TM : 1^375-377 ; paratype 4 AM : W5330 ; paratype 5 BJ : T7O.
REMARKS. Location of the male pores median to setal lines a on a midventral,
pad-like glandular protuberance which bears pore-like genital markings, diagnoses
this species. The specific name refers to the restriction of multiple caudal nephro-
stomes to a dorsal situation distinct from the ventral funnel. Other peculiarities are
indicated in the generic definition.
Oreoscolex sexthecatus sp. n.
Fig. 3, 2gA, 32?, Q ; Plates 83-86. Table 18
1 ? (posterior amputee), w = 6-3 mm, s ? (holotype).
Prostomium epilobous f . First dorsal pore 5/6. Setae 8 per segment ; b, c and
d irregular posteriorly. Nephropores not visible. Clitellum not developed. Male
pores in b on ovoid papillae which extend from a lateral of b. Paired eye-like genital
markings with centres in b, radius ab or less, in 13/14, 14/15, 15/16, 16/17 and 19/20.
Female pores paired on small papillae anterior and slightly median of a on XIV.
Spermathecal pores 3 pairs, in 6/7-8/9, in b lines.
TABLE 18
Intersetal distances in segment XII in Oreoscolex sexthecatus
mm
aa ab be cd dd dc cb ba
holotype 1-6 0-6 1-8 1-7 6-3 1-9 1-8 0-6
standardized (u = 100)
aa ab be cd dd dc cb ba
holotype 9-73 3-54 11-06 10-62 38-94 11-50 11-06 3-54
interval/afc 2-75 i-oo 3-12 3-00 n-oo 3-25 3-12 i-oo
Last hearts in XII (latero-oesophageal, with filamentous connective to dorsal
vessel). Supra-oesophageal in ^VIII-|XIII, well developed. Gizzard large and
firm, in V ; preceded in IV by a large, thin-walled proventriculus. Extramural
calciferous glands absent. Oesophagus vascular in VII-XVI, with internal rugae,
though these are poorly developed in VII-IX and nowhere form lamellae, their
greatest development, as flattened papillae, being in XIV. Intestinal origin XVII
(? macerated) ; typhlosole ? Nephridia : astomate, avesiculate, exonephric mero-
nephridia in II to the amputation at about XXV ; very slender ducts entering the
B. G. M. JAMIESON
VII
IX
g.n
FIG. 29.
1mm
Genital fields. A, Oreoscolex sexthecatus, holotype. B, O. wellingtonensis , 14/5,
specimen i.
sp.p
body wall in front of setal follicles a, b, c and d and near middorsal : the a and b and,
anteriorly, the c and d nephridia forming spiral-looped aggregations with the appear-
ance of tufts, but presetal ducts remaining separate. A posterior fragment which
corresponds in appearance with the posterior amputee (there being only one other,
very small, worm in the sample) has 5 preseptal funnels on each side in caudal seg-
ments, the nephridia remaining exonephric. Holandric (funnels and free sperm
masses but no iridescence observed) ; gymnorchous ; very large racemose seminal
vesicles in IX and XII. Metagynous ; ovisacs in XIV. Prostates racemose,
curved linear, extending through several segments, the surface deeply lobulated ;
the central lumen moderately wide with man}'' side branches. Penial setae almost
straight, moderately slender, scarcely tapering ectally except at the tip which is
obtuse and rounded (worn ?) ; the region behind the tip furrowed ; the shaft further
entally bearing numerous scattered groups of small but attenuated spines which are
often bifid or trifid ; length = 1-5 mm, general width of shaft = 18 /mi. Sperma-
thecae 3 pairs of unusual form : duct greatly expanded with a correspondingly en-
larged internal lumen, joined by a bilobed or multiloculate diverticulum ; the sac-like
ampulla with one or two variably developed ectal outpouchings.
TASMANIAN EARTHWORMS
317
MATERIAL EXAMINED.
Lake St Clair, i46°io'E. 42°O5'S. Map, 13/5, Tasmanian Biological Survey : Jig, Mr R.
W. Kerr, Feb 1941 - holotype TM : K^8.
REMARKS. The combination of several pairs of intersegmental genital markings,
3 pairs of spermathecal pores and ectal outpouchings of the spermathecal ampulla
distinct from spermathecal diverticula, diagnoses this species.
Oreoscolex wellingtonensis (Spencer, 1895)
Fig. 3, 296, 32R ; Plates 87-90. Table 19
Cryptodrilus wellingtonensis Spencer, 1895 : 43-44, PL 3, fig. 25-27.
Notoscolex wellingtonensis ; Michaelsen, 1900 : 192-193 ; Jamieson, 19710 : 79.
1 = 140, 155 mm, w = 5-8, 5-5 mm, s = 138, 149 (specimens I and 3).
Prostomium epilobous \, closed. First dorsal pore 4/5, 3/4 ; pores very conspicu-
ous in the mid- and hind-body. Setae 8 per segment in regular longitudinal rows ;
d occasionally absent. Nephropores not visible. Clitellum annular, XIV-XVII.
Male pores on XVIII in a on a large glandular pad which extends laterally beyond b
and fills the segment longitudinally (specimen i), this pad replaced in specimens 3-5
by two large oval papillae, each bearing a male pore. A large median oblong-oval
glandular pad in 19/20 (specimens I, 2) or 20/21 (specimens 3, 4) extends laterally
beyond b, includes the setal annulus of the preceding segment and reaches that of the
succeeding segment and is traversed by a groove corresponding with the inter-
segment. A similar pad present in 12/13 in specimen 3 and a suggestion of a small
midventral pad median to a in 17/18 in specimens 1-3. Female pores paired, median
to and only slightly anterior to a, at about \aa, on XIV. Spermathecal pores 2 pairs
on small papillae, or concealed, in 7/8 and 8/9, in a lines.
TABLE 19
Intersetal distances in segment XII in Oreoscolex wellingtonensis
specimen 2
specimen 3
specimen 4
specimen 2
specimen 3
specimen 4
mean
interval/aft
aa
1-6
ab
0-6
be
2-1
1-8
cd
2-6
I'S
<fti
8-9
4'1
8-2
dc
i-i
2-6
cb
1-4
2-3
1-8
standardized (u = 100)
ba
0-6
aa
ab
be
cd
dd
dc
cb
ba
9-79
2-98
8-51
7-66
52-77
6-81
8-51
2-98
7-96
3-10
13-27
16-37
25-66
16-37
14-16
3-10
11-16
3-19
9-96
8-37
45-82
8-37
9-96
3-19
9-64
3-09
10-58
10-80
41-42
10-52
10-88
3-09
3-12
I-OO
3-42
3-50
13-40
3'4°
3-52
I-OO
Last hearts in XII (latero-oesophageal) . Supra-oesophageal moderately deve-
loped, ending at f XIII. Gizzard large, firm in V extending posteriorly to the level
3i8 B. G. M. JAMIESON
of XI, oesophagus virtually suppressed to this segment ; in XII-XIV with high
vascular internal rugae, intestinal origin XVI, deep dorsal typhlosole commencing in
XXIV (specimens I and 3). Nephridia : pairs of very large tufted nephridia in
II-V with composite ducts ; those of II apparently opening into the buccal cavity ;
those of III and IV and possibly V running forward to open ventrolaterally at the
mouth. Nephridia in VI- IX forming smaller tufts exonephric in each segment. By
XII there are dense lateral bands of meronephridia ; those of segments X and XI
being intermediate between tufts and these lateral bands. Caudally with very
numerous exonephric (?) micromeronephridia on each side, each nephridium with a
minute preseptal funnel (specimen i). Holandric (funnels iridescent in X and XI) ;
gymnorchous ; seminal vesicles racemose, in IX and XII. Prostates linear but
externally racemose, extending laterally in, and restricted to, XVIII ; narrow
central lumen present with few side branches. Penial setae moderately slender,
sinuous, the tip flattened and widened to form a blade which is terminally straight
edged with a few irregular teeth ; ectal region, with the exception ot the blade, with
a feathered appearance produced by a series of numerous composite spines, each of
which is approximately triangular but has several long and often branched lateral
pinnae ; length of left seta = i-i mm ; general width of shaft = 18 pm (specimen
i). Metagynous ; ovisacs absent. Spermathecae 2 pairs ; diverticulum (insemi-
nated) sessile, multiloculate.
TYPE-LOCALITY. Mount Wellington. Map, 14/4.
MATERIAL EXAMINED.
Domain, Hobart, i47°2o'E. 42°5o'S. Map, 14/5, Dr J. L. Hickman, 14 Aug 1954 ~ specimen i
TM : K379, 7 Nov 1955 - specimen 2 AM : W533I. Kelly's Basin, Port Davey, I45°55'E.
43°2o'S. Map, 17/1, Tasmanian Biological Survey: Mr C. D. King, Jan 1940 - specimen 3
BM(NH) : 1973.2.47. 'Tasmania', Dr J. L. Hickman - specimen 4 BJ : Tji.
REMARKS. The new material conforms sufficiently well with Spencer's account
for there to be little doubt of conspecificity. The fact that one of the new specimens
is almost topotypic and that no form more closely resembling Spencer's description
was found further supports the identification. Differences in Spencer's account are
slight, viz, length < 100 mm ; prostomium only slightly epilobous ; male and sper-
mathecal pores in ab ; clitellum extending slightly onto the dorsal surface of XVIII.
Accessory genital markings are said by Spencer to be swollen tumid ridges on XVIII,
XIX and XX, marked differences from the present specimens, but his illustration
(Plate 3, fig. 25) shows a single intersegmental marking only, in 19/20, as described in
this account.
Genus MEGASCOLEX Templeton, 1844
DIAGNOSIS. Setae, at least in the mid- and hindbody numerous (more than 8)
per segment. Prostates i pair, racemose, their pores on XVIII or, exceptionally, an
adjacent segment ; sometimes associated with penial setae. Female pores paired or
exceptionally unpaired. Spermathecal pores usually 1-5 pairs between IV and IX
(the exceptions are the few cases where the pores are fused in the midline, or where
they are numerous on each side in each segment occupied). One gizzard, in V, VI
TASMANIAN EARTHWORMS 319
or VII. Calciferous glands present or absent. Meronephric, with or without
enteronephric nephridia. Holandric, rarely metandric ; testis-sacs present or
absent. Ovaries in XIII. Spermathecae with or without free diverticula.
TYPE-SPECIES. Megascolex caeruleus Templeton, 1844 (Ceylon).
DISTRIBUTION. Ceylon and India. Australia, with Tasmania. New Caledonia.
New Zealand. Norfolk Island. Annam.
Tasmanian species
M. burniensis sp. n.
M . montisarthuri sp. n.
M. tasmanicus (Spencer, 1895).
Other species
Sixty-five Australian species, in addition to the Tasmanian complement, are
listed in Jamieson, igyic (q.v.). This is a very large genus with many species in
extra-Australian parts of its range.
REMARKS. Megascolex is a polyphyletic congeries in need of revision and con-
stitutes one of the major problems in oligochaete taxonomy (Jamieson, igyic). It is
the type-genus of the tribe Megascolecini, purely meronephric megascolecines in
which, if there is a median stomate nephridium, this opens into the intestine.
The following three Tasmanian species are provisionally retained in Megascolex,
with the definition of which they conform, pending revision of the genus. Sufficient
grounds exist to suggest that the Tasmanian species comprise one or more separate
genera but erection of the latter is deferred until the mainland species of Megascolex
are revised.
KEY TO TASMANIAN SPECIES OF MEGASCOLEX
1 Spermathecal pores 2 pairs .......... 2
- Spermathecal pores 5 pairs. Fig. 306 . M. tasmanicus (Spencer, 1895)
2 Spermathecal pores median to a, almost contiguous ventrally. Fig. 3oA
M. montisarthuri sp. n.
- Spermathecal pores in ab. Fig. 306 ...... M. burniensis sp. n.
Megascolex burniensis sp. n.
Fig. 4, 306, 32F
1 = 65 mm, w = 4-0 mm, s = 92.
Prostomium tanylobous. First dorsal pore 4/5. Setae 12 on XII ; 24 caudally
(30 counted in some segments) ; seta become irregular in the posterior 25 segments ;
zz = zy anteriorly but zz = 2-4^ posteriorly ; aa = 2-2'^ab preclitellar, aa = 3-406
postclitellar ; aa not discernible as a ventral break caudally. Nephropores not
externally visible. Clitellum annular |XIII-XVII. Male pores on small papillae
in ab, nearer a, on XVIII. Genital markings : 3 pairs of small eye-like markings
320
B. G. M. JAMIESON
sp.p
FIG. 30. Genital fields. A, Megascolex montisarthuri, holotype. B, M. burniensis,
holotype. C, M. tasmanicus, specimen i.
TASMANIAN EARTHWORMS 321
with centres approximately in a, one pair presetal in XVII, the other two pairs in
17/18 and 18/19 ; two similar presetal unpaired markings with centres in a in XIX
(right) and 19/20 (left) ; two larger median presetal markings present in XVIII and
XIX. Female pores paired, anterior and slightly median to a of XIV. Spermathe-
cal pores 2 pairs, in 7/8 and 8/9, on small papillae slightly lateral of a.
Last hearts in XIII (hearts in X-XIII latero-oesophageal with connective to
dorsal and supra-oesophageal vessel) ; supra-oesophageal in VII-XIII, well deve-
loped. Gizzard large, firm, globular with anterior rim, in V. Extramural calciferous
glands absent ; oesophagus with circumferential vascular striae in VIII-XV ;
rugae increasing posteriorly, especially well developed in XIV and XV. Intestinal
origin XVII ; typhlosole absent. Nephridia poorly preserved ; few avesiculate,
exonephric meronephridia on each side in each segment ; exonephric tufts in II-VII ;
caudally exonephric with a preseptal funnel on each side, the median nephridium
somewhat enlarged. Holandric (funnels iridescent in X and XI) ; gymnorchous ;
seminal vesicles racemose in XI and XII. Metagynous ; ovisacs present ? (a small,
flocculent mass, on the right side only, is doubtfully an ovisac). Prostates flattened
and almost square, in XVII and XVIII ; truly racemose, the muscular duct dividing
into three immediately within the gland ; vas deferens joining the duct near its
ental end. Penial setae absent. Spermathecae 2 pairs, in VIII and IX, diverti-
culum (inseminated) single, clavate, uniloculate.
MATERIAL EXAMINED.
Fern Glade, Emu River, Burnie, I45°55'E. 4i°O5'S. Map, 7/2, Dr J. L. Hickman, 24 Aug
1954 - holotype TM : K/jSo.
REMARKS. The genital field is diagnostic of this species.
Megascolex montisarthuri sp. n.
Fig. 4, soA, 326
= 95. 105 mm, s = 3-6, 5-2 mm, s = 125, 123.
Prostomium canaliculate or not, tanylobous closed. First dorsal pore 3/4, 4/5.
Setae 18 (holotype), 12 (paratype i) in XII, caudally with 24 or more in a segment ;
rows regular in the fore- and midbody, with the ventral gap = $ab and much wider
dorsal gap ; caudally with aa unchanged but other rows irregular and no appreciable
dorsal interruption of the setal circlet. Nephropores not externally visible. Clitel-
lum annular, XIII-XVII (holotype, paratype i) with weak development in XVIII
(holotype). Male pores almost contiguous midventrally in XVIII on a pair of
united papillae which lie in a depression with tumid borders which extends almost
from the setal arc of XVII to that of XIX and laterally reaches b lines ; a midventral
transverse glandular depression with tumid margins present in each of intersegments
9/10, 16/17 an<i I9/20» spanning cc and carrying a transverse series of whitish pore-
like stigmata. Female pores anteromedian to setae a at about \aa. Spermathecal
pores 2 pairs, on medianly closely apposed papillae near the anterior margins of
VIII and IX.
322
B. G. M. JAMIESON
TASMANIAN EARTHWORMS 323
Dorsal vessel single ; continuous onto the pharynx ; last hearts in XII (hearts in
X-XII latero-oesophageal, with connective to dorsal and supra-oesophageal vessel) ;
supra-oesophageal weakly developed ; gizzard very well developed, in V but extend-
ing posteriorly into IX by displacement of attenuated septa ; oesophagus widened to
form an annular calciferous gland with numerous high radial internal lamellae in XV
(holotype, paratypes, I, 8) ; similarly developed in XVI and not demarcated from
that in XV in holotype, or in XIV and less so in XIII in paratype i ; swollen and
vascular, almost reniform on each side in XVII in paratypes i and 8. Intestinal
origin XVIII reaching full width in XIX ; typhlosole strongly developed ; first
appreciable (rudimentary) in XX-XXI (holotype, paratypes i, 8). Nephridia
several exonephric, astomate, avesiculate parietal meronephridia on each side in
each segment ; a large pair of tufts in III, extending into II and IV and each sending
a sheaf of ducts (composite duct) to enter the parietes anterior to seta b of II ;
caudally the medianmost nephridium enlarged as a megameronephridium with single
preseptal funnel (exonephric) (holotype). Holandric (funnels iridescent) ; gymnor-
chous ; seminal vesicles racemose, in XI (small) and XII ; metagynous ; small ovi-
sacs in XIV ; prostates compact, racemose, restricted to XVIII (holotype, paratypes
i, 8) with ramifying internal ducts and no dominant central duct ; vas deferens
joining ectal end of duct (holotype). Penial setae absent. Spermathecae 2 pairs,
in VIII and IX (the ampullae in the segment anterior to its pores in the holotype
only) ; diverticulum (inseminated) small, clavate, uniloculate (holotype ; paratypes
1,8).
MATERIAL EXAMINED.
Mt Arthur, Weldborough Pass, 1-6 miles from eastern end, i47°55'E. 4i°io'S. Map, 9/3,
Dr J. L. Hickman, 26 Aug 1953 — holotype TM : K38i. East side of mountain i47°2o'E.
4i°i5'S. Map, 9/1, Mr A. J. Dartnall and Mr R. C. Kershaw, 15 Oct 1971 -paratypes i, 8,
16-18 BM(NH) : 1973.2.48-52 ; paratypes 2-7, 9-15 TM : 1^382-394 ; paratypes 19-25
AM : W5332-5338 ; paratypes 26-29 BJ : 1:72-75.
REMARKS. The closely apposed spermathecal and male pores and the very
distinctive accessory genital markings permit ready identification of M. montisarthuri
and jointly distinguish it from all other species.
Of 29 paratypes from Mt Arthur, all agree with the holotype in possessing the
characteristic median transverse genital markings at 16/17 and !9/20, the median
field around the two male pores being variably developed ; 25 of these have an
additional transversely punctate genital marking extending laterally into ab in 9/10.
FIG. 31. Spermathecae of Cryptodrilus. A, C. avesiculatus, holotype (right IX). B & C,
C. brunyensis : B, holotype (right VIII) ; C, paratype (right IX). D, C. dartnalli,
holotype (left IX). E-G, C. enteronephricus : E, holotype (right IX) ; F, paratype 3
(right IX) ; G, paratype 14 (right IX^. H, C. polynephricus polynephricus, specimen 12
(right IX). I, C. polynephricus ad urethrae, specimen 44 (right IX). J-M, C. p. urethrae :
J, holotype (right VIII) ; K, paratype i (right IX) ; L, paratype 8 (right IX) ; M,
paratype 20 (right IX) . N & O, C. simsoni : N, lectotype (right VI J I) ; O, specimen 2
(right IX). All except L to the scale indicated.
324 B. G. M. JAMIESON
Megascolex tasmanicus (Spencer, 1895)
Fig. 4, soC, 32H, I
Perichaeta tasmanica Spencer, 1895 : 47-48, PI. IV, fig. 37-39.
Megascolex tasmanicus ; Michaelsen, 1900 : 217 ; Jamieson, 19710 : 96.
1 = 7°> 75 mm (63-88 mm, Spencer), w = 3-5 mm (3-1 mm, Spencer), s = 89, 93
(specimens i and 2).
Prostomium epilobous § (^-|, Spencer), open ; peristomium bifid ventrally.
First dorsal pore 4/5. Setae 18 (specimen i), 25 (specimen 2) (20-22, Spencer) in
XII ; 28-30 caudally (24-28, Spencer) ; zz = 2zy anteriorly, irregular caudally,
forming a definite gap in the setal circlet in the fore- and midbody but inappreciable
at the caudal extremity ; aa = 2ab anteriorly, regular throughout, = sab caudally,
a definite gap throughout ; setae a and b absent in XVIII. Nephropores not ex-
ternally visible. Clitellum weakly developed, annular, in XIII-XVII (as Spencer).
Male pores small but conspicuous transverse slits in a or ab, on XVIII, each on the
suggestion of a small papilla, in a circular, glandular, depressed field. Accessory
genital markings incompletely developed but well-defined, presetal median trans-
verse glandular bands in IX, X, XI, and XII extending maximally to ab (specimen i)
or rudimentary in VIII-XI (specimen 2) ; a presetal, median, approximately oval,
depressed glandular patch in each of segments XVII-XXII ; similar but paired
circular presetal markings present in each of segments XX-XXIII, almost contiguous
with the median marking where one occurs in the same segment (markings median in
IX, X, XI, XX, XXI and XXII, Spencer). Female pores a pair in XIV, shortly
anterior to the setal arc and close together at about \aa ; spermathecal pores 5 pairs
of small but distinct white-rimmed slits, in 4/5-8/9, in a lines.
Internal anatomy (specimen i) : dorsal blood vessel single, continuous onto the
pharynx ; last hearts in XII (hearts in X-XII latero-oesophageal, each with con-
nective to dorsal and supra-oesophageal vessel) ; supra-oesophageal in IX-|XIII,
well developed. Gizzard muscular, but small and easily compressed, in V ; true
ovoidal calciferous glands, 3 pairs in segments X, XI and XII, each sessile, with a
narrow connection with the lateral wall of the oesophagus, the lumen of the gland
almost occluded by approximately 12 laminae which converge on the connection
with the oesophagus but do not unite ; intestinal origin XV (XVIII, Spencer) ; a
very low but definite typhlosole present, first appreciable, though very rudimentary,
FIG. 32. Spermathecae. A, Graliophilus tripapillatus, holotype (right IX). B, Pinguidri-
lus tasmanicus, lectotype (right IX). C & D, Hickmaniella opisthogaster, holotype : C,
ventral view ; D, dorsal view (right IX). E, Rhododrilus littoralis, holotype (right IX).
F, Megascolex burniensis, holotype (right IX). G, M. montisarthuri, holotype (right IX).
H & I, M. tasmanicus, specimen i : H, left IX ; I, right VIII. J, Oreoscolex bidiverti-
culatus, holotype (right IX). K, O. campestris, specimen i (right IX). L, O. irregularis,
specimen I (right VIII). M, O. peculiaris, holotype (right IX). N & O, O. longus, holo-
type : N, dorsal view ; O, ventral view (right IX). P & Q, O. sexthecatus, holotype :
P, ventral view ; Q, dorsal view (right VIII). R, 0. wellingtonensis, specimen i (right IX).
S, Pseudocryptodrilus acanthodriloides, holotype (right IX). All to the same scale
excepting E which is twice the scale indicated.
TASMANIAN EARTHWORMS
325
326 B. G. M. JAMIESON
in XVI ; caeca and muscular thickening absent. A few small exonephric tufted
nephridia on each side in IV, small nephridia apparently present in III (and II ?) ;
astomate, avesiculate, exonephric micromeronephridia numerous on each side, form-
ing a dense parietal band, in XV posteriorly ; less numerous in V-XIV. Caudally
with numerous micromeronephridia and a median nephridium which has several
spiral loops and might be considered a megameronephridium. The median nephri-
dium with a large preseptal funnel and with its duct joining a paired duct running
longitudinally on the roof of the intestine shortly lateral of the dorsal blood vessel ;
this longitudinal duct connected to and presumably opening into the intestine
posteriorly in each segment. Caudal micromeronephridia associated with preseptal
bodies which may be reduced funnels ; whether enteronephric or exonephric not
determined. Holandric (funnels iridescent in X and XI) ; gymnorchous ; seminal
vesicles racemose, very large, in IX and XII, the larger in IX. Metagynous ;
ovisacs present. Prostates racemose, bipartite, one lobe in XVIII, the other in XIX,
the muscular duct bifurcating immediately within the gland, joined entally by the
vas deferens. Penial setae absent. Spermathecae 5 pairs, diverticulum single,
longer than the ampulla, clavate, uniloculate.
TYPE-LOCALITY. Emu Bay, Tasmania, and King Island in Bass Strait. Map, 7/2.
MATERIAL EXAMINED.
Table Cape, I45°45'E. 4i°oo'S. Map, 7/1, Dr J. L. Hickman, 24 Aug 1954 - specimen *
TM : K395 ; specimen 2 BJ : Tj6.
REMARKS. M. tasmanicus is questionably distinct from the Victorian species
M. hoggi (Spencer i8g2b) and its occurrence on King Island may indicate that
it represents a Tasmanian population of the latter species. If so, it would con-
stitute the only known case of a megascolecid species shared between Tasmania
and the mainland. Spencer (1895) drew attention to the similarity of M.
tasmanicus, M. sylvaticus, M. hoggi, M. halli, M. rubra, M. frenchi and M. steeli, all
Victorian species of Spencer (i8g2b), in having a bifid peristomium, 5 pairs of sper-
mathecal pores, calciferous glands in X-XII and bilobed prostates. This combina-
tion of characters suggests a basis for recognition of a new (wholly dichogastrin ?)
genus distinct from Megascolex.
ACKNOWLEDGEMENTS
The author is deeply indebted to Dr J. L. Hickman of the University of Tasmania
for providing his incomparable collections and to Mr R. W. Sims for providing
facilities in the British Museum (Natural History) for this study. Thanks are also due
to other collectors, acknowledged in the text, and to Miss E. Pope, Dr P. Hutchings,
Mr A. J. Dartnall and Dr B. J. Smith of the Australian, Tasmanian and National
Museums for loan of material. Mr E. A. Bradbury, Mr T. Walker and Mr J. Casey are
thanked for their assistance and Miss O. Page-Hanify for her efficient typing. Mr
Walker took the photographs in cooperation with the Electron Microscope Centre,
University of Queensland, and also prepared the maps. All line drawings are by the
author. The work was made possible by Australian Research Grants Committee
grant nos. 23926o-300-R-Zool-ARGC-i2O-7i-72.
TASMANIAN EARTHWORMS 327
REFERENCES
BEDDARD, F. E. 1888. Remarks upon a species of Coccidium infesting Perichaeta. Ann. Mag.
nat. Hist. 6(2) : 434-439.
1889. On the oligochaetous fauna of New Zealand with preliminary descriptions of new
species. Proc. zool. Soc. Land. 1889 : 377-382.
BENHAM, W. B. 1907. On the Oligochaeta from the Blue Lake, Mount Kosciusko. Rec.
Aust. Mus. 6 : 251-264.
BRINKHURST, R. O. 1971. The aquatic Oligochaeta known from Australia, New Zealand,
Tasmania and the adjacent islands. Pap. Dep. Zool. Univ. Qd. 3, 8 : 99-128.
& JAMIESON, B. G. M. 1971. The Aquatic Oligochaeta of the World. Oliver and Boyd,
Edinburgh.
CLAPAREDE, E. R. 1862. R6cherches anatomiques sur les Oligochetes. Mem. Soc. Phys.
Hist. nat. Geneve, 16 : 217.
COGNETTI DE MARTIIS, L. 1910. Nuove specie deigeneriMegascolidesePheretima. Ann. Mus.
Stor. nat. Giacomo Doria, (3)4 : 327-334.
FLETCHER, J. J. i887a. Notes on Australian earthworms. Part I. Proc. Linn. Soc. N.S.W.
I (Ser. 2), 1886 : 523~574-
i887b. Notes on Australian earthworms. Part II. Proc. Linn. Soc. N.S.W. i (Ser. 2),
1886 : 943-973-
i888a. Notes on Australian earthworms. Part III. Proc. Linn. Soc. N.S.W. 2 (Ser. 2),
1887 : 375-4°2-
i888b. Notes on Australian earthworms. Part IV. Proc. Linn. Soc. N.S.W. 2 (Ser. 2),
1887: 601-620.
i88ga. Notes on Australian earthworms. Part V. Proc. Linn. Soc. N.S.W. 3 (Ser. 2),
1888 : 1521-1558.
iSSgb. Notes on Australian earthworms. Part VI. Proc. Linn. Soc. N.S.W. 4 (Ser. 2),
1889 : 987-1019.
GATES, G. E. 1959. On a taxonomic puzzle and the classification of the earthworms. Bull.
Mus. comp. Zool. Harv. 121, 6 : 229-261.
1960. On Burmese earthworms of the family Megascolecidae. Bull. Mus. comp. Zool.
Harv. 123, no. 6 : 203-282.
1962. On some Burmese earthworms of the moniligastrid genus Drawida. Bull. Mus.
comp. Zool. Harv. 127 : 297-373.
GODDARD, E. J. & MALAN, D. E. 1913. Contribution to a knowledge of South African
Oligochaeta. II. Description of a new species of Phreodrilus. Trans. R. Soc. S. Afr.
3 : 242-248.
JAMIESON, B. G. M. 1970. A revision of the Australian earthworm genus Woodwardiella with
descriptions of two new genera. /. Zool. Lond. 162 : 99-144.
197 1 a. A review of the megascolecoid earthworm genera (Oligochaeta) of Australia.
Part I - Reclassification and checklist of the megascolecoid genera of the world. Proc.
R. Soc. Qd 82, 6 : 75-86.
I97ib. A review of the megascolecoid earthworm genera (Oligochaeta) of Australia, Part
II - The subfamilies Ocnerodrilinae and Acanthodrilinae. Proc. R. Soc. Qd 82, 8 : 95-108.
19710. A review of the megascolecoid earthworm genera (Oligochaeta) of Australia.
Part III - The subfamily Megascolecinae. Mem. Qd Mus. 16, i : 69-102.
I97id. In Brinkhurst R. O. and B. G. M. Jamieson (eds.). The Aquatic Oligochaeta of
the World. Oliver and Boyd, Edinburgh.
19716. Earthworms (Megascolecidae : Oligochaeta) from Western Australia and their
zoogeography. /. Zool. Lond. 165 : 471-504.
I972a. The Australian earthworm genus Spenceriella and description of two new genera
(Megascolecidae : Oligochaeta). Mem. natn. Mus. Viet. 33 : 73-87.
I972b. A numerical study of the Australian earthworm genera Cryptodrilus Fletcher and
Trinephrus Beddar, with a new genus (Megascolecidae : Oligochaeta). Zool. J. Linn. Soc.
51, 2 : 147-175-
328 B. G. M. JAMIESON
JAMIESON, B. G. M. I973a. Earthworms (Megascolecidae : Oligochaeta) from Mount Kos-
ciusko, Australia. Rec. Aust. Mus. 28, II : 215-252.
I973b. The zoogeography and evolution of Tasmanian Oligochaeta. In W. Williams
(ed.) Biogeography and Ecology in Tasmania. Monographic Biologicae. Junk, Hague.
(In press.)
& BRADBURY, E. A. 1972. Investigation of the Australian earthworm genus Didy-
tnogaster (Megascolecidae : Oligochaeta) and discrimination of populations by analysis of
setal ratios. Pedobiologia, 12 : 111-122.
JENSZ, R. L. & SMITH, B. J. 1969. Catalogue of Baldwin Spencer earthworm types in the
National Museum of Victoria, Australia. Mem. natn. Mus. Viet. : 85-110.
LEE, K. E. 1970. Application of numerical taxonomy to the classification of Megascolecidae
(Annelida : Oligochaeta). Pedobiologia, 10 : 257-264.
MICHAELSEN, W. igoo. Das Tierreich, 10, Vermes, Oligochaeta. Friedlander, Berlin.
igo7a. Oligochaeta. In Die Fauna Siidwest-Australiens, Bd i, Lief. 2 : 117-232. Gustav
Fischer, Jena.
19075. Oligochaeten von Australien. Abh. Geb. Naturwiss. Ver. Hamburg, 19, i : 3-25.
1910. Oligochaten von verschiedenen Gebieten. Mitt, naturh. Mus. Hamburg, 27 : 47-
169.
1916. Results of Dr. E. Mjobergs Swedish Scientific Expeditions to Australia 1910-1913.
Oligochaten. K. svenska VetenskAkad. Handl., 52, no. 13 : 3-74.
1923. Oligochaten von Neuseeland und den Auckland-Campbell Inseln, nebst einigen
anderen Pacifischen Formen. Papers from Dr. Th. Mortensen's Pacific Expedition,
1914-1916, no. 17. Vidensk. Meddr. dansk. naturh. Foren. 75 : 197-240.
MULLER, O. F. 1774. Vermium terrestrium et fluviatilium. Hevniae et Lipsiae, 1773-74.
PERRIER, E. 1872. R6cherches pour servir a 1'histoire des lombriciens terrestres. Nouv.
Archs Mus. Hist, nat., Paris, 8 : 19-197.
SPENCER, W. B. i892a. Preliminary description of Victorian earthworms. Part I. The
genera Cryptodrilus and Megascolides. Proc. R. Soc. Viet. 4 : 130-156.
i8g2b. Preliminary notice of Victorian earthworms. Part II. The genus Perichaeta.
Proc. R. Soc. Viet. 5 : 1-26.
1895. Preliminary notes on Tasmanian earthworms. Proc. R. Soc. Viet. 7 : 33-54.
1900. Further descriptions of Australian earthworms, part I. Proc. R. Soc. Viet. 13 (n.s.),
pt i : 29-67.
STEPHENSON, J. 1923. Oligochaeta. In The Fauna of British India. Taylor and Francis,
London.
1924. On a new species of the genus Perionyx (Oligochaeta) from Tasmania. Ann. Mag.
nat. Hist. (9), 14, 83 : 546-547.
1925. Oligochaeta from various regions, including those collected by the Mount Everest
Expedition, 1924. Proc. zool. Soc. Lond. 1925 : 879-907.
1930. The Oligochaeta. Oxford : Clarendon Press.
1933. Oligochaeta from Australia, North Carolina, and other parts of the world. Proc.
zool. Soc. Lond. 1932 : 899-941.
TEMPLETON, R. 1844. Description of Megascolex caeruleus. Proc. zool. Soc. Lond. 12 : 89-90.
WALLACE, C. C. 1972. An examination of the classification of some Australian megascolecid
earthworms (Annelida : Oligochaeta) by numerical methods. Mem. Qd Mus. 16 (2) : 191-
209.
BARRIE G. M. JAMIESON, Ph.D.
Department of Zoology
UNIVERSITY OF QUEENSLAND
ST LUCIA
BRISBANE 4067
QUEENSLAND
AUSTRALIA
PLATES 1-12
Scanning electron micrographs of penial setae, i, Graliophilus tripapillatus, holotype.
2 & 3, Cryptodrilus avesiculatus, holotype.
4-7, C. brunyensis : 4-6, one seta ; 7, a second seta of paratype i.
8-12, C. dartnalli : 8-10, paratype 8 ; u & 12, holotype.
Bull. Br. Mus. nat. Hist. (Zool.) 26, 4
PLATES 1-12
PLATES 13-24
Scanning electron micrographs of penial setae. 13-19, Cryptodrilus enter onephricus : 13-15,
paratype i ; 16-19, paratype 14.
20-24, C. polynephricus polynephricus : 20 & 21, specimen 3 ; 22, specimen i ; 23 & 24,
specimen 5.
Bull. Br. Mus. nat. Hist. (Zool.) 26, 4
PLATES 13-24
PLATES 25-36
Scanning electron micrographs of penial setae. 25-31, Cryptodrilus polynephricus poly-
nephricus : 25, specimen 10 ; 26, specimen 8 ; 27 & 28, specimen 24 ; 29-31, specimen 34.
32-36, C. p. polynephricus ad urethrae : 32 & 33, specimen 42 ; 34-36, specimen 44.
Bull. Br. Mus. nat. Hist. (Zool.) 26, 4
PLATES 25-36
36
PLATES 37-48
Scanning electron micrographs of penial setae. 37-43, Cryptodrihis polynephricus poly-
nephricus ad urethrae : 37-39, specimen 44 ; 40-42, specimen 48 ; 43, specimen 49.
44-48, C. p. urethrae : 44-47, holotype ; 48, paratype 6.
Bull. Br. Mus. nat. Hist. (Zool.) 26, 4
PLATES 37-48
PLATES 49-60
Scanning electron micrographs of penial setae. 49-59, Cryptodrilus polynephricus urethrae :
49 & 50, second seta, paratype 6 ; 51 & 52, paratype 20 ; 53-59, 3 penial setae of paratype i,
53, lateral view of a seta ; 54, ventral view of same ; 55, dorsal view of same ; 56 & 57, lateral
and dorsal detail ; 58, lateral view of second seta ; 59, detail of third seta.
60, Rhododrilus littoralis, paratype i.
Bull. Br. Mus. nat. Hist. (Zool.) 26, 4
PLATES 49-60
SO
PLATES 61-72
Scanning electron micrographs of penial setae. 61-63, Pseudocryptodrihis acanthodriloides,
paratype 3.
64-66, Hickmaniella opisthogaster : 64, holotype ; 65 & 66, paratype i.
67, Oreoscolex bidiverticulatus, holotype.
68-72, O. campestris : 68 & 69, specimen i ; 70-72, specimen 2.
Bull. Br. Mus. nat. Hist. (Zool.) 26, 4
PLATES 61-72
PLATES 73-84
Scanning electron micrographs of penial setae. 73-80, Oreoscolex longus : 73-75, holotype
76-78, paratype i ; 79 & 80, paratype 7.
81-82, O. peculiaris, holotype.
83 & 84, O. sexthecatus, holotype.
Bull. Br. Mus. nat. Hist. (Zool.) 26, 4
PLATES
PLATES 85-90
Scanning electron micrographs of penial setae. 85 & 86, Oreoscolex sexthecatus, seta shown in
plate 83, holotype.
87-90, O. wellingtonensis , specimen i.
Bull. Br. Mus. nat. Hist. (Zool.) 26, 4
PLATES 85-90
90
PLATES 91-96
Scanning electron micrographs of penial setae. 91 & 92, Perionychella (P.) capensis, paratype i .
93, P. (Vesiculodrilus] bithecata, paratype 3. 94-96, P. (V .) dilwynnia : 94 & 95, Butler's
Gorge Road specimen ; 96, Tarraleah pipeline specimen.
Bull. Br. Mus. nat. Hist. (Zool.) 26, 4
PLATES 91-96
PLATES 97-102
Scanning electron micrographs of penial setae. 97 & 98, Perionychella (Vesiculodrilus)
hobartensis : 97, 14/6, BM(NH) : 1972.8.15 ; 98, 14/5, TM : 1^304.
99-102, P. (V .) mortoni : 99 & 100, left and right setae respectively of East Risclon specimen ;
1 01 & 1 02, specimen described from Shoobridge Bend.
Bull. Br. Mus. nat. Hist. (Zool.) 26, 4
PLATES 97-102
1O1
1O2
A LIST OF SUPPLEMENTS
TO THE ZOOLOGICAL SERIES
OF THE BULLETIN OF
THE BRITISH MUSEUM (NATURAL HISTORY)
1. KAY, E. ALISON. Marine Molluscs in the Cuming Collection British Museum
(Natural History) described by William Harper Pease. Pp. 96 ; 14 Plates.
1965. (Out of Print.) £3.75.
2. WHITEHEAD, P. J. P. The Clupeoid Fishes described by Lacepede, Cuvier and
Valenciennes. Pp. 180 ; n Plates, 15 Text-figures. 1967. £4.
3. TAYLOR, J. D., KENNEDY, W. J. & HALL, A. The Shell Structure and Mineralogy
of the Bivalvia. Introduction. Nuculacea-Trigonacea. Pp. 125 ; 29 Plates
77 Text-figures. 1969. £4.50.
4. HAYNES, J. R. Cardigan Bay Recent Foraminifera (Cruises of the R.V. Antur)
1962-1964. Pp. 245 ; 33 Plates, 47 Text-figures. 1973. £10.80.
5. WHITEHEAD, P. J. P. The Clupeoid Fishes of the Guianas. Pp. 227 ; 72
Text-figures. 1973. £9.70.
Printed in Great Britain by John Wright and Sons Ltd. at The Stonebridge Press, Bristol BS4 jNU
22 JUL1974
THE FRESHWATER FISHES OF
MUNGO AND MEME AND LAKES
KOTTO, MBOANDONG AND
SODEN, WEST CAMEROON
E. TREWAVAS
BULLETIN OF
THE BRITISH MUSEUM (NATURAL HISTORY)
ZOOLOGY Vol. 26 No. 5
LONDON: 1974
22 JULI9!
THE FRESHWATER FISHES OF RIVERS
MUNGO AND MEME AND LAKES
KOTTO, MBOANDONG AND
SODEN, WEST CAMEROON
BY
ETHELWYNN TREWAVAS
Pp 329-419; 5 Plates, 17 Text-figures
BULLETIN OF
THE BRITISH MUSEUM (NATURAL HISTORY)
ZOOLOGY Vol. 26 No. 5
LONDON: 1974
THE BULLETIN OF THE BRITISH MUSEUM
(NATURAL HISTORY), instituted in 1949, is
issued in five series corresponding to the Departments
of the Museum, and an Historical series.
Parts will appear at irregular intervals as they
become ready. Volumes will contain about three or
four hundred pages, and will not necessarily be
completed within one calendar year.
In 1965 a separate supplementary series of longer
papers was instituted, numbered serially for each
Department.
This paper is Vol. 26, No. 5, of the Zoological series.
The abbreviated titles of periodicals cited follow those
of the World List of Scientific Periodicals.
World List abbreviation :
Bull. Br. Mus. nat. Hist. (Zool.)
Trustees of the British Museum (Natural History), 1974
TRUSTEES OF
THE BRITISH MUSEUM (NATURAL HISTORY)
Issued 24 May, 1974 Price £5.60
THE FRESHWATER FISHES OF RIVERS
MUNGO AND MEME AND LAKES
KOTTO, MBOANDONG AND
SODEN, WEST CAMEROON
By ETHELWYNN TREWAVAS
CONTENTS
SYNOPSIS .
INTRODUCTION
Page
332
332
DESCRIPTION OF THE LOCALITIES ....... 333
ABBREVIATIONS AND MEASUREMENTS ...... 333
MORMYRIDAE : Mormyrus tapirus ....... 334
Brienomyrus brachyistius ...... 335
CHARACIDAE : Alestes macrolepidotus ....... 339
Alestes longipinnis ....... 339
CYPRINIDAE : Labeo camerunensis n. sp. ...... 340
Barbus batesii ........ 343
Barbus mungoensis n. sp. ...... 345
Barbus callipterus. ....... 347
Barbus camptacanthus ....... 349
Barbus thysi n. sp. ....... 350
CLARIIDAE : Clarias walkeri ........ 356
BAGRIDAE : Auchenoglanis ahli ....... 356
MOCHOKIDAE : Chiloglanis : structure of head ..... 358
Chiloglanis tnicropogon . . . . . . 361
Chiloglanis disneyi n. sp. ...... 365
MALAPTERURIDAE : Malapterurus electricus ...... 367
CYPRINODONTIDAE : Epiplatys sexfasciatus . . . . . . 367
Aphyosemion bivittatum ..... 370
Aphyosemion celiae . . . . . . 371
Aphyosemion oeseri . . . . . . 373
Procatopus : specific characters and sexual dimorphism 376
Procatopus similis ...... 384
Procatopus nototaenia ...... 385
Procatopus glaucicaudis . . . . . 386
Procatopus lacustris n. sp. ..... 386
CICHLIDAE : Hemichromis elongatus ....... 387
Chromidotilapia : definition ...... 388
Feeding mechanism : contrast with Tilapia . . . 389
Chromidotilapia finleyi n. sp. ..... 393
Chromidotilapia guntheri guntheri and C. guntheri loennbergi 398
Tilapia camerunensis ....... 402
Tilapia kotta'e ........ 405
Tilapia mariae ........ 405
Sarotherodon galilaeus ....... 408
ZOOGEOGRAPHICAL SUMMARY AND CONCLUSIONS ..... 410
ACKNOWLEDGEMENTS ......... 414
REFERENCES ........... 414
332 E. TREWAVAS
SYNOPSIS
Collections of fishes from the upper Mungo and its tributaries and from Lakes Kotto, Mboan-
dong and Soden are described and their relationships discussed. The status of Brienomyrus
longianalis is examined in the light of variation within and between populations of B. brachyis-
tius. A new species of Labeo is described. The Mungo contains two species of Barbus with
parallel-striate scales, one of which is given a new name. The three small Barbus species with
radiating striae on the scales are identical with those of Fernando Poo and one of them, here
given a new name, is so far reported only from Fernando Poo and the Mungo and Meme systems.
In the course of comparisons with this species records of the geographical distribution of B.
trispilos and B. liberiensis are critically examined and possible synonyms of B. baudoni are
discussed.
Specimens of Auchenoglanis from two localities prompt questions, but no answers, concerning
the specific characters within this genus. Two species of Chiloglanis were collected in the stony
beds of swift streams. One of them is described as new and some special features in the structure
of the head in this genus are described.
Seven species of cyprinodonts are recognized in the area, one, a Procatopus of Lake Soden, new.
The identity and synonymy of Aphyosemion oeseri Schmidt are discussed and a new description
is given of Procatopus similis Ahl.
The structure of the pharynx in Chromidotilapia is described, C. loennbergi is considered to be a
subspecies of C. guntheri, a West African species present also in the Mungo together with a new
species related to C. batesii. The new species is described and compared with samples from
Fernando Poo and South Cameroun. The Hemichromis of the area is the 'B form of H. fasciatus',
here tentatively identified with H. elongatus Guichenot.
The description of Tilapia kottae is amplified and the name T. camerunensis Lonnberg is revived
for a related species inhabiting the Mungo as well as the Meme and distinct from T. cameronensis
Holly of the Sanaga system. T. mariae of Lake Kotto is no longer considered to merit sub-
specific rank. Some notes on Sarotherodon galilaeus and its subspecies are included.
The Mungo-Meme fish fauna is seen as part of that of the forest zone from western Nigeria to
South Cameroun, having none but the most widespread species in common with the Chad basin.
Lakes Kotto and Mboandong contain a slightly modified section of this fauna and Fernando Poo
harbours twelve species common to it and the Mungo and Meme systems, adapted to life in clear
running water. This throws into greater relief the high degree of endemism in the crater lake
Barombi Mbo.
INTRODUCTION
A PREVIOUS paper (Trewavas, Green & Corbet, 1972) dealt with the taxonomy and
ecology of the fishes of Barombi Mbo, results of an expedition by three members of
Westfield College, University of London, and myself in March-April 1970. The
present paper describes the fishes collected during that expedition in the other waters
visited. The ecology of those of Kotto and Mboandong has already formed the
subject matter of a report by Corbet et al. (1973).
We were interested to get good samples from R. Mungo and its tributaries and
from the outflow of Barombi Kotto which belongs to the Meme system, as a back-
ground to the endemism in the lakes, especially Barombi Mbo ; and we wished to get
further light on the relationships of the Kotto fishes.
For the Mungo collections we are entirely indebted to Dr Henry Disney, who set
narrow-meshed traps overnight and used a hand-net to sample the river and its
tributaries at and above the bridge near Etam in the course of his field studies of
Simulium as a vector of Onchocerca. When we did not accompany him he brought
FISHES OF WEST CAMEROON 333
living fishes to our headquarters near the lake and he made further collections in
October and November, 1970.
Kotto, Mboandong and their neighbouring streams were sampled by our colleague
Dr John Griffith by means of basket traps and fyke net and by Professor Green and
Dr Corbet with hand-nets. In addition we purchased fish from cast-net fishermen
and boys with hand-lines.
Comparisons of some species with their representatives in other waters are under-
taken as a contribution to West African ichthyogeography.
DESCRIPTION OF THE LOCALITIES
The lakes have been described (Trewavas, 1962 ; Corbet et al. 1973). The
whole area is in the forest zone of the volcanic hills around Mount Cameroon.
R. Mungo at the bridge near Etam is about a quarter of a mile broad. There is a
waterfall not far above the bridge and the current is rapid. At our early visits
people could wade to the sandbanks exposed in its bed, but before we left heavy rain
had swollen it so that the sandbanks were submerged in a broad brown flood.
Between Baduma and Bolo the river was much narrower, clear and shallow, and
the same description applies to tributaries from the Blackwater upwards. The bed
of the Mangusu was dry at the beginning of our visit, but at the end the stream was in
spate and yielded a cyprinodont species.
In the Kotto area, R. Nganjoke, the outlet of Lake Kotto flowing to the Meme,
was sampled where it had become a clear shallow stream with rocky bed, in our time
crossed by stepping stones but since bridged. R. Nyoke (or Njoke), about a mile
S.E. of Lake Kotto, is not connected with the lake but is a tributary of the lower
Mungo (information from Dr Disney). The water is clear and is used by the villagers
for drinking if, as rarely happens, there has not been enough rain to be collected for
that purpose.
The Barombis living on the island in Lake Kotto obtain their drinking water from
clear streams (called Tungs) that enter the lake from the crater rim. In these we
caught cyprinodonts and Barbus callipterus.
ABBREVIATIONS AND MEASUREMENTS
BMNH = British Museum (Natural History).
SL = Standard Length, i.e. excluding the caudal fin and (except in Procatopus)
excluding any projecting part of the lower jaw.
Interorbital width in Ostariophysi includes the skinny rim of the orbit (which
covers a flexible bony flange). In Cichlidae the callipers are pressed against the
skin so that this measurement is virtually that of the bony interorbital part of the
roof of the skull.
Depth of preorbital in Cichlidae is the depth of the preorbital or lacrimal bone
measured from the middle of its orbital edge along a line continuing the radius of
the eye.
334
E. TREWAVAS
FIG. i. Sketch map of the waters of West Cameroon in which the collections were
made. Drawn by J. Green from Disney, 1971. The stream Nyoke is the unlabelled one
crossing the 'K' of Kotto.
MORMYRIDAE
Mormyrus tapirus Pappenheim, 1905
Pappenheim, 1905 : 217 ; id. 1907 : 362 pi. xiii fig. 3 (R. Lokundje = Lokoundie, southern
Cameroun) ; Boulenger, 1909: 133 fig. in ; Holly, 1927: 120 (Nachtigal Rapids and R.
Mbam, Sanaga system) ; id. 1930 : 225.
Mormyrus goheeni (Fowler ?) ; Pellegrin, 1927 : 295 (R. Nyong) ; Holly, 1930 : 225 (from
Pellegrin) .
Three specimens collected by Dr Disney in R. Wo we, October, 1970, respectively
132, 170 and 173 mm in SL.
FISHES OF WEST CAMEROON 335
These are a little smaller than most previously described specimens, so a descrip-
tion is not superfluous.
Depth of body 20-23-5% SL, length of head 27-6-28-6, length of pectoral fin
17-1-17-4, of pelvic 10-8-12-7, of dorsal base 48-0-48-5, of anal base 16-9-18-3.
Length of caudal peduncle 12-9-13-6% SL, 2-8-3-0 times its depth.
Snout narrow and curved downwards to end in the thick-lipped mouth, its length
41-48-5% length of head, 0-9-1-12 times postocular part of head. Diameter of eye
14-5-20-6% length of head.
Scales 84-98 in the lateral line, about 14-17 between lateral line and origin of
dorsal, 8-12 from lateral line to origin of anal, 13-16 around caudal peduncle.
Dorsal 62-66, the first ray minute, the next two simple. Anal 24-26, 2 or 3 of
which are simple.
Teeth 3 in upper jaw, 6 in lower (but 2 + 3 lower in one, the third of the right side
being lost or not yet formed).
The allometries revealed in comparison with earlier descriptions are the usual
negative allometry of the eye and the strong positive allometry of the snout. Holly
has measured this in terms of the eye-diameter, thus emphasizing the change, but
showing continuation of the trend observable even within the narrow size-range of our
sample.
Our specimens have fewer teeth in the upper jaw than those formerly described
(Pappenheim, Boulenger and Holly 5, Fowler 6 in M. goheeni, the last perhaps with-
out realizing the usual presence of a median tooth in Mormyridae). This may also
be related to size. Pappenheim and Boulenger found an additional tooth on each
side too in the lower jaw, making a total of 8.
Holly has already found in 7 specimens from Rivers Lokundje and Sanaga a
range of 60-73 rays in the dorsal fin. It may have been from specimens at the lower
end of this range that Pellegrin identified his Nyong specimens as M. goheeni (Fowler
gave 60, 58 and 61 for the type and two other Liberian specimens). But to my mind
the presence of low numbers in samples from Cameroun should rather be taken as
evidence that there is one species from Liberia to Cameroun than that two species
so defined are present in Cameroun.
The type of M. goheeni Fowler (1919 ; Liberia) was described as having 80 scales
in the lateral line. The anterior scales are difficult to count because of the thickened
skin overlying them and some irregularity in their alignment. The Mungo counts
not only include Pappenheim's and Boulenger's count of 90, but bring the range
nearer to Fowler's datum and make it more probable that M . goheeni is a synonym
of M . tapirus. There are no records under either name, to my knowledge, between
Liberia and Cameroun, and since Fowler's description of the types of M. goheeni no
further records from Liberia.
BRIENOMYRUS Taverne, 1971
Taverne, igyia : 101 and 108 ; id. igyib. Type species Marcusenius brachyistius Gill.
Taverne shows that this genus is distinguished from Marcusenius Gill (type species
Mormyrus cyprinoides Linnaeus) by the broader, curved mesethmoid, the absence of
336 E. TREWAVAS
lateral ethmoids and the presence of 6 circumorbital bones, the antorbital being free
from the first infraorbital.
Brienomyrus brachyistius (Gill)
Marcusenius brachyistius Gill, 1862 : 139 (Liberia ?) ; Boulenger, 1909 : 67 fig. 50 ; id. 1916 : 158 ;
Pellegrin, 1927 : 295 (Wuri and Nyong) ; Holly, 1930 : 222.
Mormyrus microcephalus Giinther, 1867 '.115 (Ogowe).
Marcusenius liberiensis Steindachner, 1894 : 67 (Liberia).
? Marcusenius longianalis Boulenger, igoia : 5 pi. iii fig. i (Niger delta) id. 1909 : 69 fig. 51 ;
id. 1916 : 159.
? Marcusenius adustus Fowler, 1936 : 247 fig. 4 (nr Kribi, Cameroun).
? Marcusenius sp. Reed, 1967 : 26 fig. 33.
Our specimens are :
SL (mm) Locality Collector
4 121-176 R. Mungo
between Baduma and Bolo R. H. L. Disney
2 79*5-133-5 R. Mungo nr Etam R. H. L. Disney
5 95'5-*37'5 R- Wowe R. H. L. Disney
i 163 R. Nganjoke J. Griffith
The holotype, of unknown locality, 'probably Liberia', had 17 dorsal and 27 anal
rays. In Cameroun the species has been recorded from R. Nyong at Dehane, R.
Wuri at Yabassi (Pellegrin, 1927), Kribi (Boulenger, 1916 : 158), Tiko near the coast
in the R. Matute N. of the Mungo mouth (Trewavas, 1962).
Our Nganjoke fish is slightly deeper-bodied than all but the smallest of the Mungo
specimens and has a deeper caudal peduncle. It resembles Boulenger's figure of B.
brachyistius.
The ii Mungo specimens are all slender fishes, only the smallest having a maximum
depth of more than 18% SL and only one having a caudal peduncle less than 3 times
as long as deep. This led me to compare them not only with B. brachyistius but also
with B. longianalis of the lower Niger.
The results, summarized in Tables i and 2, make it difficult to maintain B.
longianalis as a separate species. According to Boulenger's key (1909 : 61) the only
trenchant contrast is in the number of anal rays (31-33 v. 25-30). But one of the
syntypes of M. microcephalus Giinther (R. Ogowe), synonymized by Boulenger with
brachyistius, also has 31 anal rays and 6 other specimens from the Lower Niger,
included in longianalis, have 30-32. It is no longer justifiable therefore to draw
an arbitrary line between 30 and 31 rays. It is nevertheless true that in the lower
Niger long anals are associated with low numbers of dorsal rays (Table 3). Else-
where high anal counts are associated with rather high numbers of dorsal rays (Sierra
Leone and R. Gambia) or low dorsal counts with low numbers of anal rays (Niari-
Kouilou, Chiloango and Congo) (Table 3). In the Mungo the slender shape is as-
sociated with intermediate numbers of dorsal and anal rays, although the anal
numbers range higher than in the presumed type locality (Liberia).
FISHES OF WEST CAMEROON 337
TABLE i
Some critical characters in samples of Brienomyrus brachyistius and possible synonyms
n
D
A
A/D
sc.
depth
c.p.l/d
vert.
Liberia
10
17-18
24-27
1-5-2-0
56-63
20-6-24-8
2-2-2-9
44-45
Ghana and nr
Lagos
5
15-16
27-28
1-95-2-25
-
17-2-22-5
2-6-3-3
44-45
Nr Oban,
Calabar
3
16-17
28-29
1-9-2-2
-
18-5-20-6
2-6-3-1
45. 46
Mungo
ii
16-18
27-29
1-6-2-0
59-65
14-2-20-4
2-9-4-0
44-47
Nganjoke
i
17
26
1-9
55
17-8
2-6
45
Type of B. micro-
cephalus,
Ogowe
i
i?
3i
2-2
59
18-8
3'2
45
Types of
longianalis
2
15, J6
32,33
2-9, 3'i
60
17-5,21-7
3-i
46, 47
Others from
Niger delta
5
14-16
30-32
1-9-2-75
54-6i
18-5-22-3
2-9-3-5
45-48
Tiko, Cameroon
i
16
30
1-9
64
22-3
2-9
47
Kribi
i
17
28
2-25
65
19-8
3'9
47
Others from
Ogowe
10
15-17
27-30
2-2-2-4
59-6i
20-2-23-0
(2-9)
45-47
(in 4)
3-1-3-8
« = no. of specimens, D = dorsal fin-rays, A = anal fin-rays, A/D = length of anal base/length of dorsal
base, sc = scales in lateral line, depth = max. depth as % SL, c.p.l/d = length of caudal peduncle/its depth,
vert. = vertebrae. The Liberian samples include three syntypes of M. liberiensis.
TABLE 2
B. brachyistius from the extremes of its geographical distribution. Data as in Table i
Sierra Leone
R. Gambia (Daget)
Chiloango
Matadi
'Stanley Falls'
Niari-Kouilou (Daget)
TABLE 3
Frequencies of dorsal and anal fin-ray counts in samples of Brienomyrus
Dorsal Anal
13 14 15 16 17 18 24 25 26 27 28 29 30 31 32 33
Liberia 93 x 4 4 3
Mungo 462 1523
'tnicrocephalus' and 'longianalis' 1431 2241
Sierra Leone and Gambia 344 32141
Congo, Chiloango and Niari-Kouilou i 5 7 751
n
D
A
A/D
sc.
depth
c.p.l/d
9
16-18
27-31
1-6-2-1
(55) 59-63
18-6-23-4
2-3-3-3
i
18
30
1-9
56
21-8
2-85
6
15-16
25-27
1-8-2-1
51-56
17-6-22-0
2-6-3-0
i
13
26
1-8
53
22-2
2-4
i
15
25
1-8
50
22-9
2'55
5
15-16
25
338 E. TREWAVAS
Such evidence as we have suggests variation both within and between populations
no more than may reasonably be expected in a species occupying the lower reaches
of a series of river-systems. Although it is not a typical inhabitant of coastal
lagoons, Daget & Iltis (1965) record that it is tolerant of brackish water, so that the
possibility of occasional interchange between neighbouring populations is not ex-
cluded, perhaps in a geological rather than an historical time-scale.
The comparative material used for Tables 1-3 is that listed by Boulenger in vols
I and IV of his Catalogue under M. brachyistius and M. longianalis with the additions
listed below. Some doubt attaches to the locality of the specimen from 'Stanley
Falls' since Poll & Gosse (1963) did not find B. brachyistius in the Yangambi region,
nor did they include it in their list of species recorded from the Central Congo. The
collection by De Meuse in which it was included contained examples of several species
from Stanley Pool and of two others from Stanley Falls, Stomatorhinus microps
Boulenger and Gnathonemus petersi Gunther. Both the latter are found throughout
the Central Congo as well as in Stanley Pool and the Lower Congo, so that they do
not provide the confirmation we seek that De Meuse's collection of that date really
included fishes from Stanley Falls.
Boulenger's material included syntypes of M. liberiensis, M. microcephalus and
M . longianalis.
The specimen (BMNH 1970.10.13.1) registered as from 'Lokoja R. 1967' may be
one of the two referred to by Reed (in Reed et al. 1967 : 26 fig. 33) from a swamp near
Idah. He described them as resembling M. brachyistius but having a shorter pectoral
fin and a total length of 180 mm. The length of our specimen is now only 166 mm.
In it the right pectoral fin is reflected forwards, but both pectorals when laid back
extend just beyond the origin of the pelvics, as is usual in B. brachyistius. The short
appearance of the fin in Reed's sketch may be accounted for by the fact that the
drawing was traced from a photograph, possibly with a foreshortened fin. This fish,
with 15 dorsal and 30 anal rays and a long slender caudal peduncle, resembles the
types of M. longianalis and others from the Niger delta (Table 3).
Fowler's account of M. adustus (1936) compares it with M. batesii Boulenger, but
this species has 16 scales around the caudal peduncle and fewer anal rays. Fowler
did not mention the peduncular scales but his drawing shows fewer than 16. If we
assume that he undercounted the lateral line scales his specimens may well belong
to B. brachyistius.
B. brachyistius : material in the BMNH additional to that listed in Boulenger's
Catalogue and on p. 336 above.
BMNH register no. SL (mm) Locality Collector
1938.12.15.31 102 Akim, nr Bunsu, Ghana F. R. Irvine
1932.2.27.8 121-5 Onitsha F. R. Irvine
1970.10.13.1 150-5 'Lokoja R' W. Reed
1952.4.1.5 89-5 Tiko, Matute R. Cameroun J. Deveson
1950.9.22.1-5 49-68 S.W. Sierra Leone T. S. Jones
1958.9.18.8 107 S.W. Sierra Leone T. S. Jones
FISHES OF WEST CAMEROON 339
CHARAGIDAE
Alestes macrolepidotus (Val.)
Brycinus macrolepidotus Valenciennes, 1849, in Cuvier & Valenciennes vol. 22 : 157 pi. 639
(Senegal).
Three specimens, 42-5-58-5 mm in SL from R. Mungo at the bridge near Etam,
coll. Dr Disney, 18 March 1970 and 4 Nov. 1970.
This widespread species, recorded from the Nile, West Africa and the Congo, was
caught by us only in the part of the Mungo corresponding to its habitat in the
Yangambi region, where Gosse (1963) states that it is found near the banks and along
sand banks.
Our specimens were not full grown and were even more slender than adults (depth
26-1-29-2% SL). The gill-rakers were rather fewer than the number given by
Boulenger for the species, (7-13) + (o-i) + (15-16) on the first arch. The presence
of 10 outer teeth in the upper jaw, with the second from the symphysis inset, out of
alignment with the rest, is characteristic. The dorsal formula is ii-iii 7 or 7 + 1, the
anal iii 11-12. There are 23 or 24 scales in the lateral line, A\ rows between it and
the origin of the dorsal fin, i| from lateral line to pelvic.
Colour notes made from the smallest specimen when it had been if hours in for-
malin describe the body as pale gold and grey, the dorsal fin with pink tinge and grey
leading edge, adipose red, caudal lobes yellow with orange tinge on upper and lower
edges ; a diffuse dark blotch at end of caudal peduncle continued as a black streak
on middle caudal rays. In the preserved specimens a dark mark behind the pectoral
girdle is more evident in one specimen that was fixed in formalin than in the others
where alcohol was the fixative.
Alestes longipinnis
Brachyalestes longipinnis Giinther, 1864, Cat. Fish. vol. 5 : 315 (Sierra Leone).
One specimen from the Mungo bridge near Etam is 76 mm in SL. It is probably
male, having the longest dorsal fin-ray 140% length of head (near the values obtained
by Thys, 1967 : 31). The lateral line scales number 31 + 1, rows between lat. line
and dorsal origin 5|, between lat. line and pelvic 2|. Dorsal ii 8, anal iii 19 + 1.
Teeth in upper jaw 6 outer, evenly aligned, 8 inner; in lower jaw 8 + 2. Gill-
rakers on first arch 10 + 1 + 10, the epibranchial ones very short.
No notes were made on our living specimen, but colour notes accompanying one
collected in 1952 by Mr Deveson in R. Matute near Tiko agree essentially with those
published by Loiselle (1972) for the species in West Africa, as follows : eye brilliant
gold (iris silvery-yellow with orange crescent on top - Loiselle) ; black mark on
caudal peduncle and middle caudal rays with gold above it (metallic orange above it) ;
dorsal, adipose, caudal and anal fins with dashes of red (D, C and A rosy violet, D
orange basally). Loiselle adds that the long dorsal and anal filaments of mature
males are metallic white. From Thys's (1967) account including comparisons with
specimens from the extremes of its distribution, the colour may be more variable.
In particular, the prolongations of dorsal and anal fins may be reddish.
340 E. TREWAVAS
The species inhabits flowing water with forest shade from Guinea Republic to R.
Kribi (excluding the Volta and the Dahomey gap), as well as similar habitats in
Fernando Poo.
CYPRINIDAE
Labeo camerunensis n. sp.
PI. 3
HOLOTYPE. 162 mm SL from R. Mungo at the bridge near Etam, coll. Dr Disney.
BMNH 1973.5.14.324.
PARATYPES. Two of SL 167 and 175 mm from R. Wowe and two young of 43
and 45-5 mm from R. Mungo between Baduma and Bolo, all coll. Dr Disney.
BMNH 1973.5.14.322-3, 325-6.
These fall into a group of Labeo having the following characters : eyes supero-
lateral, snout much longer than postocular part of head, dorsal fin iii-iv 9-10, its
longest ray not or but little longer than head and its dorsal edge concave ; 16 scales
around caudal peduncle.
The branched dorsal rays are 9 in four specimens, 10 in one. The snout is tuber-
culate, without a deep furrow between the ethmoid and the jaw, projecting a short
way in front of the broad mouth. A maxillary barbel is present, hidden in the folds
of the Up in adults ; a short rostral barbel is present in one of the young, represented
by a stump or absent in the others.
There are 35 or 36 scales in the lateral line, excluding 1-3 on the caudal fin,
4|-5 rows above the lateral line in front of the dorsal fin, 3-3! between the lateral
line and the pelvic.
Proportions in the three adults are :
As % SL : depth of body 22-7-23-8 ; length of head 24-7-25-2 ; length of caudal
peduncle about 16-17, its depth 12-6-13-0 ; length of longest dorsal ray 23-0-
26-2 ; length of pectoral fin 24-2-25-0.
As % length of head : diameter of eye 20-0-20-7 ; length of snout 55-2-58-5,
1-65-2-02 times postocular part of head including skinny edge of operculum ; inter-
orbital width 41-6-46-2.
Colour fairly uniform on back and sides, each flank scale in life reddish purple with
a black base overlapped by the transparent posterior edge of the preceding scale.
Lower side of head and belly white or pale grey.
The two young already have the adult characters of large, superolateral eyes
(25-5, 26-0% length of head), a well-developed snout with some tubercles and the edge
of the dorsal fin slightly concave. There is a large round black blotch on each side
at the end of the caudal peduncle and a suggestion of a darker band along the middle
of the side.
AFFINITIES. L. annectens Boulenger, described from R. Kribi and neighbouring
rivers, has only 12 scales around the caudal peduncle (14 in one of 8 specimens
examined), but is evidently nearly related to our species. L. batesii Boulenger,
FISHES OF WEST CAMEROON 341
1911 and 1916, is known in the adult state only from the type, from R. Kribi, a fish
of SL 169 mm with a broken neck. Some young, up to 21-5 mm SL, from R. Benito
have been attributed to it by Roman (1971). The drooping head of Boulenger's
figure (1916, fig. 127), a restoration, is partly the result of damage, but the high
curved back and deep body (28-4% SL) are natural and contrast with the Mungo
fishes, whose dorsal outline is continued almost horizontally into the line of the top
of the head and snout. The caudal peduncle of L. batesii is not only deeper (14-8%
SL) and shorter (14% SL), but more compressed than in L. camerunensis. The scale
numbers are only a little higher than in the Mungo fish, those around the caudal
peduncle likewise 16, and the length of the longest dorsal fin-ray is comparable
(23-6o/0 SL).
From the Sanaga system Holly has recorded (ig27a : 131 ; ig2jb : 421 ; 1930)
L. greenii Boulenger and Pellegrin named L. chariensis var. nunensis (ig29a).
Holly gives no details but Pellegrin distinguished his 'variety' from L. chariensis by
the smaller and more numerous scales (1.1. 39, D to 1.1. 5^) and this would place it in
L. greenii (1.1. 37-38, D to 1.1. 5^-6^). Both L. chariensis and L. greenii are further
distinguished from the Mungo form by the very long last simple dorsal fin-ray
(37-43% SL in L. greenii, about the same in L. c. nunensis}. These differences out-
weigh a resemblance in the pigment pattern between the young of the Mungo form
and L. greenii, which has a conspicuous black blotch at the end of the caudal peduncle
persisting, though more vaguely delimited, at a standard length of 182 mm (see
Boulenger's original figure, 1902, pi. viii fig. 4). Pellegrin mentions no pattern
except the usual countershading, but the syntypes of L. c. nunensis were bigger
(SL 260 and 330 mm) than Boulenger's specimens of L. greenii.
L. mungoensis is a member of or closely related to the super species including L.
forskalii, and L. cylindricus, having a lower number of scales than these, but not as
low as L. annectens. A dark lateral band is found also in the young of these species.
BARBUS Cuvier & Cloquet, 1816
I. Species with parallel-striate scales and with the last simple dorsal ray thick
and rigid in its proximal part, but not serrate.
Our samples of this group comprise 75 specimens, of which 39 are young fry,
26-56 mm in SL. Good samples of adults were taken in R. Nyoke, Barombi Mbo
and R. Wowe ; R. Menge yielded one of 92 mm SL. In addition there were 14 from
Lake Soden (p. 345).
These agree in numbers of scales and fin-rays and in having two pairs of barbels,
the posterior in adults longer than the diameter of the eye (up to twice), and in the
pharyngeal dentition, but they fall into two groups on the length of the strengthened
dorsal ray and its rigid part.
The ray is longer and stronger in adults from Rivers Wowe and Menge than in
those from Barombi Mbo and R. Nyoke. This seemed to be the only character
differentiating them and as it has been found variable in some other populations I
was uncertain of its significance until I examined the fry. These have longer dorsal
342 E. TREWAVAS
30
20
10
_i
oo
-* 35
o o go oo»;W Oo oo o o
80 100 120 140 160 180 200 220 240 260
o
Q
30
25
20
0 0<b o<
O Ch O
25 30 35 40 45 50 55 60 65 70
SL mm
FIG. 2. Barbus batesii (open circles) and B. mungoensis (solid circles). Length of longest
dorsal ray as % SL in (above) adults and (below) juveniles. For adults only the rigid
part of the ray is measured, for juveniles the whole ray. With the adult B. batesii are
included the holotype and 3 other Kribi specimens ; the rest are from Barombi Mbo and
R. Nyoke. Triangles are ratios from Holly's (1927) measurements of B. brevispinis.
and caudal fins associated with' longer barbels, shorter fins with shorter barbels.
The time and/or rate of development of the barbels is a feature that Barnard
(1943 : 147) found useful in differentiating the species of South Africa.
It happens that we have a good series from fry to adults from only one locality,
R. Wowe (Dr Disney's collection), but this gives confidence in the taxonomic value
of the length of the stiffened dorsal ray at a size when the difference in barbel-length
has been outgrown.
Having separated the adults on this basis, I was able to find some taxonomic
significance in the degree of development of the lips, a notoriously variable character,
as many authors from Gunther (1868 : 84) to Daget & Iltis (1965 : 99) have em-
phasized. It seems that in the short-finned species, which I identify with B. batesii
Boulenger, the lip may become lobed, but only in large adults ; in the long-finned
species, to which I have reluctantly given a new name, quite small specimens may
develop the lobe.
The two species are contrasted as follows :
B. batesii
Last simple ray of dorsal fin 17-24% SL in adults of 116-314 mm SL, shorter
than head ; its rigid part 14-20% SL. Lower lip with a median pad, more or less
undercut, sometimes produced as a lobe in individuals over 200 mm SL. Posterior
barbel in fry of 33-53 mm SL 34'5-50-o% diameter of eye.
FISHES OF WEST CAMEROON
343
B. mungoensis n. sp.
Last simple ray of dorsal fin 25-5-30-0% SL in adults of 114-180 mm SL, as long
as or longer than head ; its rigid part 24-5-28-0% SL, 92-104% length of head.
Lower lip with an undercut or fringed median pad or with a lobe. Posterior barbel
in fry of 25-56 mm SL 71-100% diameter of eye.
Barbus batesii Boulenger, 1903
Boulenger, 1903 : 25 pi. iii fig. 2 (R. Kribi) ; id. 1911 : 43 fig. 24 ; Trewavas, 1962 : 152 (Barombi
Mbo) ; Thys van dem Audenaerde, 1967 : 34 figs, n, 12 (Fernando Poo).
In addition to the material listed by me in 1962, we now have, from our collections
of 1970 and one made by Dr C. A. Wright in 1963, the following (see also p. 345) :
18
17
2
6
4
SL (mm) Locality
69-5-267 Barombi Mbo and its feeder stream
27~53 R- Kumba above the falls
27 and 41 R. Kumba at Buea Road Bridge
28-50 R. Menge
155-202 R. Nyoke
Proportions, for three size-groups from Barombi Mbo with a separate column for
R. Nyoke, are set out in Table 4 and proportions for fry up to SL 53 mm are listed
in Table 7 and Fig. 2.
TABLE 4
Proportions in Barbus batesii in Barombi Mbo and R. Nyoke
SL (mm)
N
Depth of body (% SL)
Length of head (% SL)
Last simple D ray
(% SL)
(% length of head)
Rigid part of ray
(% SL)
(% length of head)
Length of pectoral (% SL)
Length of anal (% SL)
Length of caudal lobe (% SL)
Snout (% head)
Eye (% head)
Interorbital width (% head)
Posterior barbel/eye
Barombi Mbo
Nyoke
202-267
116-191
69-5-78-0
156-5-202
4
24-4-27-8
23-8-26-6
ii
24-8-27-4
24-5-27-5
3
24-8-29-5
26-8-28-0
4
27-8-29-4
27-7-28-9
i7«o-20'3
67-5-85-0
18-9-24-0
71-5-89-0
22-4-28-0
80-0-100-0
19-8-21-2
69-0-77-0
14-0-16-6
54-0-70-0
18-0-20-0
54-0-74-0
16-5-22-0
60-0-80-0
16-0-17-0
55-5-59-0
2I-O-22-2
20-0-23-0
22-0-23-0
22-0-24-5
19-5-20-5
24-8-30-4
33'3-39-Q
18-6-20-8
34-8-36
19-0-23-5
26-0-30-5
32-0-39-0
19-0-30-0
31-4-36-0
21-0-21-5
28-8-31-6
31-0-34-5
28-5-32-0
20-0-24-5
26-8-28-0
34-0-39-8
22-5-23-5
30-0-34-5
1-2-1-55
1-0-1-7
0-75-1-0
I-I5-I-5
344 E- TREWAVAS
Scales 24-26 : 4^ (5) : 2-2^. 5^ between origin of dorsal and lateral line on one
side of one fish. Dorsal iii-iv 8-9 in Barombi Mbo, iv 9-10 at R. Nyoke. Gill-
rakers (2-4) + (o-i) + (9-11) at Barombi Mbo, (2 -4) + ! + (11-13) at R. Nyoke.
Colour in life brassy yellow on the flanks, countershaded to dark green on the back
and top of the head and white ventrally ; anal fin colourless, or orange distally
edged with black ; base of each scale marked with a conspicuous black crescent.
For ecology see Green & Corbet in Trewavas, Green & Corbet, 1972.
Reasons for regarding B. linnellii as a synonym of B. batesii have been given already
(Trewavas, 1962) and comparison of additional material from Barombi Mbo with the
type of B. batesii and five other Kribi specimens confirms the synonymy. Three of
the Kribi fishes are bigger than our biggest ; in them the positive allometric trend of
the snout and the negative one of the eye are continued to 42% and 16% respectively
of the length of head. The interorbital width also rises to 39%. Two, of 268 and
314 mm SL, have the lower lip produced as a lobe.
The pharyngeal teeth of one Kribi specimen dissected for the purpose seem to be
in process of replacement (or final shedding ? No new teeth were found in the mem-
brane), but teeth or seats on the bones were present for the usual formula, 2.3.5-5.3.2
and the shape was as in the Barombi Mbo specimens. In both this species and B.
mungoensis the anterior tooth of the row of 5 is rather small, conical, the next stout
with a point, the others sloping backwards with a posterior point and an anterior
grinding surface. In the Kribi specimens the posterior barbel is 1-5 to twice the dia-
meter of the eye, gill-rakers (2-3) + 1 + (9-12), dorsal iv 9 in all, the last simple ray
69-88-5% length of head, its rigid part 55'5~75'5%. Scales 26-28 : 4^ : 2-2^.
The higher lateral line counts suggest a population difference in this character, but
a slight one (Table 5) . Thys's counts (1967) from 8 specimens of Fernando-Poo nearly
cover both ranges.
On the basis of Holly's descriptions (1927, 1929, 1930) neither Thys nor I can recog-
nize names given by him to Barbus of Cameroun as synonyms of B. batesii. The
nearest, as Thys remarks, is B. versluysii Holly, 1929, which has 4-^ scales from dorsal
to lateral line and fins of the right proportions ; but Holly counts 29-30 scales in the
lateral line and his figure (1930 pi. i fig. 8) shows this count to be made in the same
way as ours ; his description and figure of the lips suggest greater development than
we find in such small specimens of our populations. Probably the posterior position
of the pelvics shown in the figure (Holly, 1930 pi. i fig. 8) is a mistake since the de-
scription records them as under the first rays of the dorsal (as in B. batesii). If this is
B. batesii it extends the distribution to the Bakoko highlands.
The only possible senior synonym of B. batesii is B. compinei Sauvage, 1879, of
R. Ogowe, but a decision on this awaits further knowledge of the Barbus population
of that river.
Blache et al. (1964) record three young from the Chad basin (Guelta de Tottous)
as B. batesii, but from the data given they are more probably B. foureaui Pellegrin.
Steindachner (1914) recorded two species of Barbus from R. Ja, a secondary
tributary of R. Congo rising near the sources of the rivers of southern Cameroun.
One of them, which he identified as B. mawambiensis Steindachner (1911), seems more
like B. batesii than B. mawambiensis, the type of which came from Ituri. The fauna
FISHES OF WEST CAMEROON 345
of the Ja has much in common with that of the rivers of Cameroun, arid it is quite
possible that the distribution of B. batesii extends to it.
Barbus batesii in Lake Soden
Fourteen specimens, SL 67-5-219-5 mm, were collected by Dr J. Griffith.
We include them in B. batesii while noting the following differences.
Although the scale-counts for the lateral line are the same (24 or 25, 27 in one) the
range between the lateral line and the dorsal (3|~4|) is lower though overlapping ;
5 of the 14 specimens have 3^.
In the size-range corresponding to the middle Mbo column of Table 4 (SL 117-
219-5 mm) the length of head is 27-8-30-0% SL, longer than in Barombi Mbo but
equal to that of the Nyoke specimens.
In the same size-range the length of the last simple dorsal ray is 20-5-23-2% SL,
thus within the range for the corresponding Mbo fishes, but the rigid part is weaker
and merges so imperceptibly into the flexible part that a separate measurement
cannot be made of it. An attempt on some specimens gave 11-5-15-0% SL.
The problematic specimen recorded in 1962 from this region is clearly conspecific
with these. The minute tubercles on the snout are present in the new specimens but
also on some from Barombi Mbo.
Barbus mungoensis n. sp.
HOLOTYPE AND PARATYPES. 18 specimens, 40-5-179 mm in SL, from R. Wowe,
tributary of R. Mungo, coll. Dr Disney in October, 1970. BMNH 1973.5.14.163
(holotype, 179 mm), and BMNH 1973.5.14.164-182 (paratypes).
OTHER MATERIAL EXAMINED. 5, 30-38 mm SL, R. Blackwater, i, 92 mm SL,
R. Menge, and 2, 26 and 43 mm SL from R. Mungo between Baduma and Bolo, all
coll. Dr Disney.
Scales 24-26 (28) : 4! (5^) : 2. Dorsal iii-iv 8-9, usually iv 9. Gill-rakers on
first arch (2-5) + (o-i) + (9-12). Pharyngeal teeth 2.3.5-5.3.2.
Proportions for three size-groups are set out in Table 5. The first and second
columns correspond in size roughly to the second and third in Table 4 (B. batesii),
the third is a younger size-group. Some features of the fry are compared with young
B. batesii in Table 7 and Fig. 2. In the 26 mm fish from the Mungo the posterior
barbel is already 0-8 diameter of eye and the upper caudal lobe 31% SL, but at this
stage the last simple dorsal ray is no thicker than the following ray.
A specimen of 92-5 mm SL from R. Menge agrees better with this species than with
B. batesii. It has a very long snout (42% length of head) and flexible, lobed lips.
The flexible part of the dorsal spine is broken off and the rigid stump is only 22-4%
SL, but it is thicker than in any B. batesii of comparable size. The posterior barbel
is as long as the eye and the upper caudal lobe is 32% SL. It agrees in number of
scales (24 : 4! : 2) and gill-rakers (3 + 1 + 11) with both the Mungo B. batesii and
B. mungoensis. The few fry from the Menge are probably B. batesii and the two
species may exist together in this tributary.
346
E. TREWAVAS
TABLE 5
Barbus mungoensis from R. Wowe
SL (mm)
N
Depth of body (% SL)
Length of head (% SL)
Last simple D ray
(% SL)
(% length of head)
Rigid part of ray
(% SL)
(% length of head)
Length of pectoral (% SL)
Length of anal (% SL)
Length of caudal lobe (% SL)
Snout (% head)
Eye (% head)
Interorbital width (% head)
Posterior barbel/eye
114-5-179
3
28-29
25-4-26-8
25-5-30-0
IOO-III
24-5-28
92-104
21-0-23-8
19-22
27-3-3I-4
35-8-37-6
24-8-28-4
31-34
I-3-I-5
76-79-5
4
27-5-30-5
27-5-3I-6
30-0-34-0
106-120
26-6-30-2
84-108
22-25
23-24
31-2-33-8
29-39-5
30-32-5
28-31-4
o-86-i-i
40-5-69-5
ii
28-31
29-3I-5
30-0-34-0
91-114
26-30
80-100
22-5-23-5
3I-5-36
29-36
0-8-1-2
TABLE 6
Scales in the lateral line in Barbus batesii and B. mungoensis
24 25 26 27 {
B. batesii
R. Kribi i
R. Nyoke 3 I
Barombi-Mbo '2 9 7
R. Kumba 144
B. mungoensis
R. Wowe 365
R. Blackwater 221
TABLE 7
Contrasts in length of fins and barbels in fry of Barbus batesii and B. mungoensis
SL (mm)
D ray (% SL)
C lobe (% SL)
Pectoral (% SL)
Posterior barbel (% eye)
B. batesii B. mungoensis
(R. Kumba) (Wowe, Blackwater, Mungo)
27-53 26-56
21-26 29-35
27-32 31-36
17-21 21-23-5
25-50 71-100
Life colours are not recorded. The fry sometimes have a dark spot at the caudal
base and there may be irregular smudges on the body and caudal lobes.
RELATIONSHIPS. Although this population does not correspond exactly with
descriptions of others it appears to be one of a series beginning with B. habereri
(R. Ja) and continuing with B. foureaui into the Chad, Niger and Volta basins.
FISHES OF WEST CAMEROON 347
B. habereri Steindachner (1914 : 24 fig. 4 pi. iii fig. i) agrees with it in scale-formula
(25 : 4^ : 2|) and the well-developed lips, but has only one pair of barbels. Dr
Kahsbauer has recently kindly confirmed this for me and gives the gill-raker count
as 4 + 1 + n. According to Steindachner's figure the rigid dorsal ray is 34-6% SL.
B. brevispinis Holly, 1927 (Rivers Lokundje and Sanaga, Cameroun) in spite of its
name has a dorsal spine as long as in B. mungoensis, but has 30 scales in the lateral
line series and 5| between it and the origin of the dorsal.
Daget (1954 : 188), and following him Blache et al. (1964) and Daget & Iltis (1965),
have used the name B. occidentalis Boulenger for the Chad and western populations,
but the holotype of B. occidentalis has a short dorsal spine, its rigid part only three-
quarters length of head. Until the Ogun population is studied I think it safer to use
B. foureaui Pellegrin, igigb (syn. B. seguensis Pellegrin, 1925) for these populations.
They differ from the Mungo population in having usually 5| scales between the dorsal
and the lateral line, 12-16 gill-rakers on the lower part of the first arch and the lips
more constantly lobed.
B. lancrenonensis Blache & Miton, 1960 (see also Blache et al. 1964 : 126, fig. 60)
from R. Ngou, affluent of R. Mbere, Logone system, was described from three tiny
specimens 14-7-33-5 mm long. The authors thought that it belonged to the section
of Barbus containing small species with no enlarged dorsal ray and with radiately
striate scales, but several features point to its being the young of B. foureaui. The
radii of the scales were described as 'relativement nombreuses et faiblement conver-
gentes' ; dorsal iv 9 ; gill-rakers 10-11 on lower part of arch ; a few lines of 'pores'
on the head (as I have noted too in fry of the Mungo system) ; posterior barbel
o-6-i-o diameter of eye. It is the similarity between these and our young B.
mungoensis that drew my attention to them and leads me to suggest their identity
with B. foureaui.
II. There are three species of Barbus with radiating striae on the scales and with
the last simple dorsal ray flexible, not serrate.
Barbus (Enteromius) callipterus Boulenger, 1907
Boulenger, igoya : 486 (R. Kribi at Akok) ; id. 1911 : 167 fig. 145 ; Thys van den Audenaerde,
1967 : 48 fig. 15 (Fernando Poo).
For use of the subgenus Enteromius Cope, 1869, see Greenwood (1970), who shows
that the name should replace Beirabarbus.
This species was caught in R. Mungo between Baduma and Bolo, in the Mungo
tributaries Wowe and Menge and was the only small Barbus caught by us in the
Kotto area, being taken in the lake itself, in Tung Nsuia, a stream entering it from
the crater rim, and in the outflowing stream Nganjoke.
All Kotto specimens had the characteristic black patch on the dorsal fin leaving
the tip of the fin pigment-free. In life the proximal part of this fin varies from pale
yellow or salmon pink to orange and the caudal fin from pale yellow to orange-red.
In a well-coloured fish the top of the iris is red. In all, the head is brassy yellow.
348 E. TREWAVAS
Already at 14-18 mm SL these colours are developed though still at the pale end
of the range.
In samples from the Mungo and its tributaries, where it was abundant, the black
pigment on the dorsal fin is much less intense, often faint, less sharply demarcated
than in the Kotto samples or the types, and may extend nearly to the end of the
first three branched rays.
No significant differences in numbers of scales, fin rays and gill rakers could be
found between the two populations and the pharyngeal teeth are alike.
Comparison in proportions between Kotto and Mungo samples is blurred because
in the preserved material all Kotto specimens (except the two juveniles) are bigger
than all Mungo specimens. The four syntypes of B. callipterus are also smaller
than the Kotto specimens, so that when, as with length of head and diameter of eye,
the Mungo specimens are more like the types the difference between both and the
Kotto fishes can be ascribed to allometry. Only in the height of the dorsal fin is the
Mungo sample possibly peculiar. This ratio is :
in 8 specimens of 64-7-73-5 mm SL from Kotto 24-0-29-6% SL
in 4 syntypes of 50-5-63-5 mm SL from Kribi 27-2-29-9% SL
in 8 specimens of 48-7-59-5 mm SL from R. Blackwater 28-5-32-3% SL
When allowance is made for individual variation and negative allometry even this
may not be significant.
Thus the question is left open whether the different pigmentation of the dorsal
fin in the Mungo and Kotto populations is environmentally controlled or indicates
a genetic relationship of the respective populations with others showing the same
characteristic.
I have discussed this with Mr A. Indrasenan, who has the same problem in Northern
Nigeria. We have both examined the syntypes of B. deserti Pellegrin, 1909, in Paris
and London. As the Hopsons noted (1965 : 113) these all have the dorsal fin
damaged at the tip. In the least damaged Mr Indrasenan and I found some
indication of dorsal fin pigmentation more like that of the Mungo samples than the
types of B. callipterus and if there are two species involved it is possible that the
Mungo population should be linked with B. deserti. The type locality of the latter
is in the Algerian Sahara, but populations showing one or the other type of pigmenta-
tion are found in several places between this and Cameroun.
ECOLOGY. See Corbet et al. 1973. In Lake Kotto the food consists mainly of
bottom-living invertebrates, with chironomid larvae predominating.
MATERIAL preserved in 1970
BMNH register no. SL (mm) Locality Collector
1973.5.14.42-44 14-18 R. Nganjoke J. Green and
S. A. Corbet
1973.5.14.73-81 64-72 Barombi Kotto J. Green and
S. A. Corbet
1973.5.14.114 44-5 R. Wowe R. H. L. Disney
1973.5.14.45-49 37-44 R- Mungo between
Baduma and Bolo R. H. L. Disney
FISHES OF WEST CAMEROON
349
BMNH register no. SL (mm)
1973.5.14.50-72, 82-85 26-56
1973.5.14.86-110 4°-59
1973.5.14.111-113 40-43
Locality
R. Menge
R. Blackwater
R. Mungo at bridge
near Etam
Collector
R. H. L. Disney
R. H. L. Disney
R. H. L. Disney
Barbus camptacanthus Boulenger, 1911
Fig- 3
Puntius (Barbodes) camptacanthus Bleeker, 1863 : in pi. xxiii fig. 2 (Fernando Poo).
Barbus camptacanthus ; Boulenger, 1911 : 166 fig. 144 ; Thys van den Audenaerde, 1967 : 40
fig- 13-
This fish was abundant in the Mungo tributaries and in R. Bille (Meme system),
but we did not find it in the Kotto area.
The pigment pattern in adults is as shown in the figures of Boulenger and Thys,
with a black line along the middle of the flank swollen into an elongate patch on the
caudal peduncle and another from below the origin of the dorsal fin forwards, not
reaching the operculum. As Thys notes, there is variation in the shape of the
anterior patch ; in young this is not elongate but rounded and from it a fainter
pigment streak extends vertically upwards to the base of the first dorsal rays
(Fig. 3). The fins are all coloured from yellow to red according to the density of
the chromatophores. There is no golden sheen on head or flanks.
n I U UrK
10 mm
FIG. 3. Barbus camptacanthus. Outline drawing of young to show the melanin pattern.
In some older specimens from Fernando Poo and the mainland the anterior blotch
ceases to become a conspicuous part of the pattern and shrinks or is obscured by the
general darker colour. Markings at the bases of the anterior scales of the lateral
line may be more prominent than the mid-lateral streak. In Bleeker's figure of an
350
E. TREWAVAS
adult of about 100-130 mm the lateral streak is intensified at anterior and posterior
ends and faint horizontal lines mark the scale-rows above it.
Boulenger (1911 : 166) and Thys (1967) noted the presence of 'nuptial tubercles.'
These or their crater-like seats are present in most adults of both sexes on the pre-
orbital (lacrimal) and first infraorbital bones, sometimes also on the second, and have
sharp points directed slightly backwards. In our material they are already present
at SL 66-5 mm, rudimentary at 61-5 mm and absent in smaller fishes, but in a sample
from Lomee district they are present at 54 mm.
Specimens in the BMNH range to SL 130 mm, but ours were all young. Possibly
the swift waters in which our collections were made are less favoured by adults
except for breeding.
Thys (loc. cit.) has discussed the distribution of B. camptacanthus. It is abundant
throughout the forested region of Cameroun and Gabon and is found southwards to
the Congo and westwards into the Niger delta.
MATERIAL collected in 1970
BMNH register no.
I973.5.I4-I28
I973-5-I4-H5-II9
1973.5.14.120-126, 130-159
1973.5.14.160-162
SL (mm) Locality
1973.5.14.127
1973.5.14.129
44
Collector
R. H. L. Disney
R. H. L. Disney
R. H. L. Disney
R. H. L. Disney
R. -H. L. Disney
R. Kobe at Etam R. H. L. Disney
66-5
R. Bille
25-52
R. Wowe
23-61-5
R. Menge
16-5, I9-5
and 41-5
R. Moliba
60
R. Kumba at Buea
Road Bridge
Barbus thysi n. sp.
Barbus trispilos (nee Bleeker) ; Thys van den Audenaerde, 1967 : 44 fig. 14 (Fernando Poo).
HOLOTYPE. 53 mm SL from R. Kake, at Kake village, Mungo system, BMNH
1973.5.14.284.
PARATYPES. 10, 31-63-5 mm SL from R. Blackwater, Mungo system, BMNH
1973.5.14.204-213.
Named for Dr D. Thys van den Audenaerde, who collected this species in Fernando
Poo and recorded the difference in colour pattern between it and B. trispilos.
DESCRIPTION. Based on 28 specimens from the Mungo system, including the
holotype and paratypes, and one from R. Bille, Meme system. For some details
further specimens from the Mungo samples are used.
Depth of body 26-33% SL, length of head 24-5-30 (33-3 at 22 mm SL). Distance
between snout and origin of dorsal fin 46-5-51-5%. Diameter of eye 27-36% length
of head.
Anterior barbel at standard lengths below 30 mm just reaching anterior edge of
eye, but at SL over 30 mm reaching beyond middle of eye, usually to posterior edge
or beyond, in two specimens to posterior end of operculum ; its length 1-05-3-1
FISHES OF WEST CAMEROON 351
times diameter of eye, 30-93% length of head, 9-7-23-5% SL. Posterior barbel in
adults usually extending to preoperculum or beyond, in a specimen of 63-5 mm well
beyond posterior edge of operculum ; 1-3-3-6 times diameter of eye, 43-103% length
of head, 9-6-26-0% SL (see also Table 8).
Gill-rakers on outer side of first arch (0-2) + (o-i) + (4 or 5). Pharyngeal teeth
2.3-5-5-3-2.
Scales 25-28 in lateral line, excluding i or 2 on base of caudal fin (26-30 including
these) ; 4! from lateral line to dorsum in front of dorsal fin, 2, 2 f or 3 between lateral
line and pelvic fin.
Dorsal iii 8 (iii 9 in one). Anal iii 5.
Colour. Countershaded from grey-brown on the back to white on the belly with
metallic brassy colour on operculum and flanks contrasting handsomely with the
black pigment. Usually three conspicuous black spots on each side at the level of
the horizontal myoseptum, the middle largest and most intense, the caudal well
marked and the anterior variable, smaller and often fainter than the others, occasion-
ally absent. Some part of the middle spot below last dorsal ray. Distance between
middle of first spot and middle of second contained 1-36-2-0 times in distance be-
tween middle of second and middle of caudal spot, rarely (in only 2 of 35 measured)
less than 1-4. Lateral line scales often each with a vertical black mark at its base ;
similar markings on the anterior scales of the row below and on the scales of a variable
number of rows above until they merge in the general dark colour of the back.
Often a small dark spot at origin of dorsal and occasionally, especially in young, a
pigment patch above origin of anterior anal rays. Fins usually colourless, but caudal
sometimes yellow or pinkish at base. In formalin-fixed specimens a subcutaneous
grey band extends between caudal and middle spots.
The intensity of the markings varies with the locality, the sample from R. Kumba
below the pollution of the town being the most dingy.
Variation in the barbels. The wide ranges in barbel lengths are partly due to
individual variation and partly to a positive allometric trend, even between standard
lengths of 30 and 65 mm (much steeper at smaller sizes). As well as a broad allo-
metric band there are exceptional individuals and subpopulations with higher or
lower averages. Examples of the former are one of 31 mm SL from the Black water
with barbels of 19-6 and 19-8% SL, ratios not matched in any other of less than
57 mm except another individual, the holotype, of 53 mm (R. Kake) in which the
anterior is 23%, the posterior 19-8% SL. But the Black water ratios include other
high ones, four fishes of 58-65 mm having anterior barbels measuring more than
20% SL, whereas our only other Kake individual is average for its size. In Mpanga
Water and R. Kumba below the town, on the other hand, barbel lengths are below
average.
The six specimens from 'British Cameroons', probably from the coastal area,
38-65 mm in SL, have rather short barbels, the anterior 7-5-10-5% SL, the posterior
9-6-13-0%, and the barbels measured by Thys (1967) in the Fernando Poo specimens
are in the lower part of our ranges (maximum 52-6% length of head).
In short, although this is a long-barbelled species several factors combine to pro-
duce a wide range of variation in this character and little taxonomic value can be
352 E. TREWAVAS
attached to local differences in the average lengths. Nevertheless, the longest
barbels were found in specimens of 50-70 mm SL from swift, clear streams and this
may prove to have some ecological significance.
The species in Fernando Poo. Thys's description (1967) includes some with slightly
deeper body and slightly shorter head and at least one with 5^ rows of scales above the
lateral line, but clearly refers to the same species as that of the mainland. Its
removal from B. trispilos removes also the difficulty he found in explaining its
geographical distribution.
Young. Four young of 9-5-11-5 mm SL were taken in R. Blackwater. The
squamation is not complete but they have a strong spot at the end of the caudal
peduncle. This excludes B. callipterus and since B. camptacanthus was not caught
by us in the Blackwater they probably belong to B. thysi. There is no sign of barbels
at 9-5 mm and at 11-5 they are mere stumps. Of the lateral spots only the caudal
is present, but it is strong ; a thin black line, composed of both superficial and deep
melanophores, extends from this forwards to below origin of dorsal fin. There is
some pigment on the head and back, a strip mid-ventrally in front of the pelvics and
a pair of ventrolateral strips behind the pelvics.
In a 22 mm specimen from R. Kumba at Buea Road Bridge the anterior barbel
barely reaches the eye and the posterior is still shorter ; the caudal spot is strong,
the middle spot very small and the first has not yet appeared. In one of 29 mm from
the same batch both pairs of barbels are present, but short, each about 2 mm long.
There is still only one black spot, the caudal, but there is a little pigment in the
position of the future middle spot. The subcutaneous streak is well marked. In
this locality the pigment spots are weak in adults so their poor development in the
29 mm specimen may not be typical of the species.
DISTRIBUTION. B. thysi is known with certainty only in the Mungo, Meme, the
short rivers between their mouths and Fernando Poo, but a record by Radda (i97ia :
84) of 'Barbus trispilus' from R. Lobe, South Cameroun, probably indicates that it
lives in other Camerounian rivers. Roman (1971) does not record it from Rio Muni.
AFFINITIES. Comparison with B. trispilos and B. sublineatus.
B. thysi resembles B. trispilos Bleeker (1863) in numbers of scales and fin-rays, in the
well-developed barbels and the presence of three distinct black spots on each side. It
resembles also B. sublineatus Daget (1954) in these features, but the latter nearly always
has additional spots between the second and caudal spots, and all the marks are more
elongate than in B. trispilos and B. thysi. The second spot of B. sublineatus is, however,
in the same position as that of B. thysi, whereas that of B. trispilos is more posteriorly
placed, usually wholly behind the last dorsal ray, a difference noted by Thys in 1967.
If the distance between the centres of the first two spots be denoted by a, that
between the centres of the second and caudal spots by b, then a comparison of the
three species gives :
B. trispilos B. sublineatus B. thysi
N = 31 N = 15 N = 35
b/a 1-02-1-37 (1-4) 1-4-2-25 1-36-2-0
Mean 1-18 1-73 1-62
FISHES OF WEST CAMEROON 353
Daget & Iltis describe (1965) variation in the position of the spots in B. trispilos
from Ivory Coast, but in the variants from the typical condition the middle spot is
more posterior and thus the difference from B. thysi is exaggerated.
A further difference between B. trispilos and B. thysi is in the colour pattern of the
young. In a formalin-fixed specimen of 18-5 mm SL from R. Wei Wei (Ghana) all
three pigment spots are present on each side, there is a strong spot at the origin of
both dorsal and anal fins and no subcutaneous lateral streak ; the barbels are each
about 2 mm long. At 26-3 mm the pigment pattern is the same and the barbels are
respectively 3-5 and 3-8 mm long. Thus from our admittedly inadequate numbers
of young it seems that both adult pigment pattern and barbels develop earlier than
in B. thysi.
In two young B. sublineatus of 26 and 28 mm SL four lateral spots are present and
there is a strong anal spot ; the barbels at SL 28 mm are shorter than in B. trispilos
of 26-3 mm.
In comparing lengths of barbels it must be taken into account that the Hopsons'
(1965) measurements are shorter than my own (see Table 8). In their paper of 1965
they tabulate measurements for 24 specimens of each species. For comparison I
give in Table 8 my own measurements for smaller samples of B. trispilos and B.
sublineatus, in the case of the former from the Hopsons' material (though not neces-
sarily the specimens used for measuring), in the latter from the Cross River sample
mentioned below. The lower minimum ratio in the Hopsons' figures may be the
result of including smaller specimens in which the definitive length of the barbels
had not been attained. Ranges and means in B. thysi are raised by the inclusion of
two specimens respectively from the Kake and Blackwater in which both barbels
are 20% SL or over, but in n of the 26 B. thysi measured the anterior barbel is over
15% and in ii the posterior is 17% or over.
Finally, although the length of barbels is too variable to be used as a diagnostic
character between B. thysi and B. trispilos the evidence suggests that they may grow
longer in B. thysi and they are definitely longer in both than in B. sublineatus and
B. perince Riippell. The last named invites comparison because, although in the
Nile its scale count is higher than in B. thysi in the Chad basin and at Panyam the
lateral line count may be the same (but usually there are 5^ D-l.l.).
TABLE 8
Length of barbels as % SL in three species of Barbus as measured by the Hopsons
(1965 : 127 and 129) and by the author
B. trispilos B. sublineatus B. thysi
E. T. Hopson E. T. Hopson
SL
It
Anterior
Mean
Posterior
Mean
38-5-61
27-65
39-5-57
28-40
31-68-5
8
24
9
24
26
10-25-15-5
9-8-I2-7
8-7-10-6
5-3-9-1
10-3-23-6
12-2
II-I
IO-2
7'3
14-8
I3-8-I6-9
II-7-I5-4
10-1-14-5
7-8-11-2
10-9-26-0
I5'1
13-8
12-3
9-2
16-1
354 E. TREWAVAS
Distribution of B. trispilos
The type locality, 'Dabo Krom, Guinee', is in southern Ghana ; most of Bleeker's
West African material was sent by Pel, then Resident at Cape Coast Castle. Mr S. A.
Whyte tells me that 'krom' is a suffix meaning 'town' and that Dabo Krom is a small
town not far west of Accra. The species is not reliably recorded east of Ghana and
the Hopsons did not even find it in the Volta. They accepted Norman's (1932)
record from Sierra Leone, but I find that the specimens on which this was based are
B. eburneensis Poll, 1941, exhibiting the four lateral spots, broad infraorbital bones
and short barbels of that species. The Hopsons ignored the Nigerian records in
Boulenger's list of 1916 (p. 267) repeated by Trewavas & Irvine in Irvine et al.
(1947). I have re-examined the two samples on which these records were based ;
neither is B. trispilos.
One of them is listed by Boulenger (loc. cit.) as specimens 16-27 °f ^. trispilos,
collected at Omalu, headwaters of R. Aboina, Cross River system, by Major G. E.
Bruce and registered as BM (NH) 1911.3.30.33-42. So from the catalogue list and
the register we should expect to find 10 specimens in the jar. In fact I found 12 ;
2 of them were B. trispilos and 10 were B. sublineatus Daget, 1954. I suspect that
the two B. trispilos had been put in this jar by mistake at some time when the ten
B. sublineatus were being compared with specimens from Ghana. This uncertainty
disqualifies them as evidence that B. trispilos occurs east of Ghana.
The distribution of B. sublineatus is typically in the savannah region of Ivory Coast
and Ghana and in the upper and middle Niger ; the above record extends it to the
headwaters of the Cross system. Daget & Stauch (1963 : 92) do not record it from
the tributaries of the Benue, but they suggest that B. chlorotaenia Boulenger, which
they do record, may be identical with B. sublineatus and its name a senior synonym.
I am not convinced that this is so. The specimens of 'B. chlorotaenia' from Ejura,
Ghana, so recorded by Trewavas & Irvine in Irvine et al. (1947) are B. sublineatus,
but I would now hesitate to regard them as conspecific with the types of B. chlorotaenia
until more is known of the population of the type locality of the latter, R. Omi, east
of Lagos. The lateral band in the types is more uniform than any described for
B. sublineatus and life colours as described are rather different.
The other Nigerian sample listed as B. trispilos comprises four specimens of Barbus
liberiensis Steindachner, 1894, measuring 27-0-38-5 mm SL. They were presented
by J. P. Arnold in 1913 and said to be from Sapelle, Niger delta. This species, first
described as a 'variety' of B. camptacanthus , was synonymized by Boulenger with
B. trispilos, but Greenwood (1962) redescribed the types and showed that they,
together with some specimens from Sierra Leone, belong to a distinct species. I
think it possible that some of the specimens from Ivory Coast assigned by Daget &
Iltis (1965) to B. trispilos may also be B. liberiensis. The young specimens from
Sapelle agree with this species in numbers of scales - 22 + 1 or 2 in the lateral line,
3 1 from lateral line to origin of dorsal fin - in the characteristic decurved profile and
the pattern of three large round spots on the flanks and a black leading edge to the
dorsal fin. There is no black spot at the origin of the anal fin, also in agreement
with the western samples. The barbels (anterior 6-5-9-2% SL, posterior 10-3-11-4%)
are relatively a little shorter than in Greenwood's specimens, which were all much
FISHES OF WEST CAMEROON
355
bigger. The lateral spots in B. liberiensis are nearly equidistant. Using the same
formula as on p. 352, in the five Sierra Leone specimens b/a = 1-05-1-2, in those from
'Sapelle' bja = i-o or the middle spot is marginally nearer to the caudal than to the
anterior.
Greenwood, like Steindachner, noticed the resemblances between this species and
B. camptacanthus. I think it is even closer to B. eburneensis, having also a rather
broad infraorbital bone in the adult.
A record from the Niger delta is even more surprising for B. liberiensis than it
would be for B. trispilos and this one should be treated with the caution necessary
for aquarists' material Although Arnold himself was very careful some of his sup-
pliers may have been less meticulous.
Comparison with other species
I have considered as possible relatives of B. thy si also B. lepidus Pfaff of the upper
and middle Niger and the Chad basin, a species that Daget (1954) distinguishes from
B. sublineatus by the absence at all ages of an anal spot. Its pattern of lateral spots
is different too from both B. sublineatus and B. thy si but, as in the latter, in the very
young there is a caudal spot and a narrow streak running forwards from it. It has
shorter barbels than B. thysi. Barbus tetrastigma Boulenger (1913) of the northern
tributaries of the Congo has an additional lateral spot and the anterior spot is well
above the lateral line instead of impinging upon it as in B. thysi.
Although in the Mungo and Meme systems and Fernando Poo B. thysi is not accom-
panied by either of these related species and can therefore be said to represent them
geographically, there does not seem to be enough evidence to consider it a subspecies
of one of them rather than another and for this reason I propose specific rank for it.
MATERIAL EXAMINED
BMNH register no.
1973.5.14.284 holotype
1973.5.14.204-213 paratypes
1973.5.14.258-283
1973.5.14.202
1973.5.14.214-240
I973-5-I4-305-32I
1973-5-14-285-290
I973-5.I4-29I-304
Locality
R. Kake, Mungo
system
R. Blackwater
R. Blackwater
R. Kake
R. Mungo between
Baduma and Bolo
R. Kumba at Buea
Road Bridge
R. Kumba above
waterfall
Mpanga Water,
tributary of R.
Kumba
Collector
R. H. L. Disney
R. H. L. Disney
R. H. L. Disney
R. H. L. Disney
R. H. L. Disney
R. H. L. Disney
R. H. L. Disney
R. H. L. Disney
356 E. TREWAVAS
BMNH register no. SL (mm) Locality Collector
1973.5.14.192-201 37-63-5 R. Mungo at bridge
near Etam R. H. L. Disney
I973-5-I4-24I-257 47'5~68 R. Wowe R. H. L. Disney
I973-5-I4-203 48 R. Bille, Meme system R. H. L. Disney
1952.8.30.2-5 38-54-5 British Cameroons Deveson
1937.11.24.2-3 57-5, 65 British Cameroons E. Roloff
Synonyms of Barbus baudoni Boulenger, 1918
In the course of these determinations I had occasion to look again at the syntypes
of B. baudoni. As Blache et al. (1964 : 126) have pointed out, Boulenger was
mistaken not only in stating that the types lacked barbels (they are very short),
but also in giving the number of scales around the caudal peduncle as 8 instead of 12.
This removes the only distinction between B. baudoni and B. svenssoni Johnels,
1954, and also between both and B. voltae Hopson, 1965. B. baudoni therefore, or
B. baudoni with its close relatives, has a wide distribution in the savannah zone from
R. Shari, through the Lake Chad area, the upper tributaries of the Benue (Daget &
Stauch, 1963 : 92), in the upper Volta and the Gambia.
CLARIIDAE
Clarias walkeri Giinther was the only species caught in the rivers and in Barombi
Kotto and was present in the basin of Barombi Mbo together with C. maclareni (see
Trewavas et al. 1972).
The Clarias of Lake Soden, recorded from one specimen by me in 1962, as from the
neighbouring village of Lisoni, will be described from further specimens by Dr J.
Griffith.
BAGRIDAE
AUCHENOGLANIS Gunther, 1865
Auchenoglanis ahli Holly, 1930
Holly, 1930 : 201 pi. i fig. 9 ; Trewavas, 1962 : 155 fig. i.
Three specimens 93.5, 125 and 132 mm SL from R. Nganjoke (BMNH 1973.5.14.
327-329) and three of about the same size range from R. Nyoke (two had their
caudal ends bitten off by an unknown predator) (BMNH 1973.5.14.330-1 and 333) ;
one of SL 117-5 mm from R. Mungo at the bridge near Etam (BMNH 1973.5.14.332).
These belong to the section of the genus in which the interneural plate in front of
the dorsal fin is narrow, hidden beneath the skin and not meeting the occipital pro-
cess. The pectoral spine is coarsely serrate on its posterior edge.
In the following description figures for the Mungo specimen are given separately
in brackets and Table 9 shows the length of barbels in the same two groups in com-
parison with Holly's data for A . ahli and A . pietschmanni.
FISHES OF WEST CAMEROON 357
Depth of body 18-20-5 (20)% SL, length of head 32-5-34-5 (32)%, 1-40-1-45 (1-5)
times its width. Diameter of eye 11-13 (14-5)% length of head, interocular width
25-4-28-8 (28-3). Premaxillary tooth band 2-5-4 times as long (from left to right)
as wide.
Dorsal spine, including its flexible tip, 14-4-16-6 (15-3)% SL, 43-50 (48)% length
of head. Pectoral spine, excluding its flexible tip, 17-19 (17)% SL, 49-5-56-0
(53'5)% length of head. Soft dorsal rays 7 (f.5 and Mungo specimen) or 8 (f.i).
Anal 9-10 (n), the first 2 or 3 simple. Caudal fin rounded.
Gill-rakers (2-3) + 1 + 7 (2 + 1 + 7) on the first arch.
Specimens from the streams near Kotto, like the types of A . pietschmanni and A .
ahli, have on the body numerous dark spots about the size of the eye with cloudy
edges and (like A. ahli) similarly vague-edged vertical bars on the flanks ; the fins
are spotted. The Mungo specimen has smaller and fewer spots and the vertical
bars are narrower.
I have placed these specimens in A. ahli rather than in A. pietschmanni Holly
(1926 : 158 and ic)2jb : 208 fig. 3) because they agree with the former in the number
of anal rays (13-14 in A. pietschmanni). Holly distinguished A. ahli especially by
the longer barbels, but the types of this species were much smaller than those of A .
pietschmanni and the difference in relative barbel length is in accordance with a usual
allometry in catfishes. Our Mungo specimen fits quite well into a series including
both (Table 9), those from the neighbourhood of Kotto not so well. The types of
A . pietschmanni were from R. Mbam, those of A . ahli from the Bakoko Highlands.
Auchenoglanis.
Mungo Nyoke and Nganjoke A. ahli A. pietschmanni
(1970) (1962) from Holly
N 16165
SL (mm) 117-5 93'5-I32 77'5 67-105 249-317
Barbels
Maxillary 65-0 43-8-62-0 76-5 87-100 48-7-66-3
Outer mandibular 112-5 72-0-94-0 H5'O 122-137 98-5-103-0
Inner mandibular 57.3 42-0-52-7 55-3 39~43 41-0-51-0
Anal rays n 9-10 10 10-11 I3~I4
* For the types of A . ahli Holly's ratios have been converted to % and for A . pietschmanni his measure-
ments in mm have been similarly converted.
Whether A. pietschmanni and A. ahli are distinct, whether we have one or two
species in the Mungo and Meme systems and the relationship of these to Holly's
species can be decided only after study of the variation, individual or allometric,
within the populations in the characters indicated in the table and in colour pattern.
Together the specimens concerned represent closely related populations.
In the same area (Bakoko Highlands) Holly has recorded A. ballayi Sauvage.
This species of the northern tributaries of the Congo and R. Nyong is distinguished
358 E. TREWAVAS
from the group to which our samples belong by the smaller, kidney-shaped patch of
premaxillary teeth and the weakly serrate pectoral spine (see Trewavas, 1962).
A. monkei Keilhack (1910), from the delta of R. Wouri, differs in having the occipital
process meeting the interneural plate, and its premaxillary tooth-band was described
as small and kidney-shaped, twice (adult) or three times (young) as long as wide.
MOCHOKIDAE
CHILOGLANIS Peters, 1868
The species of Chiloglanis are small catfishes that live under stones in clear
running water. The ventral mouth is surrounded by a sucking disc that forms a large
part of the ventral surface of the head and incorporates the bases of the barbels ; its
lower surface is patterned with bosses of thickened epithelium set with taste buds.
The eyes and two pairs of nostrils are dorsal.
The mandibular teeth are S-shaped like those of Synodontis (see Poll, 1971),
grouped together in one or two rows in the middle of the jaw. The second row when
present is probably a replacement row. The teeth of the upper jaw, also S-shaped,
are borne on the premaxillae and, some, in the membrane immediately behind the
premaxillary symphysis.
Structure of the mouth and disc
I have studied this by means of a dissection of a C. batesii Boulenger from South
Cameroun, an alizarin preparation of C. micropogon, longitudinal and transverse
sections of two young specimens probably of C. micropogon and sections of part of the
disc of an older C. micropogon.
The bones of the jaws and ethmoid region are similar to those of Atopochilus
guentheri Schilthuis as figured by Starks (1926, fig. 15), the maxilla being rather well
developed, lying dorsally to the premaxilla and articulating with the short palatine.
The A i section of the adductor mandibulae has a musculous insertion on it and to its
distal end is attached the core of the maxillary barbel. The maxilla and its muscle
probably constitute the main agent controlling the suction of the disc. The pre-
maxilla is shaped like a rigid lobster claw or a wishbone, with the stem of the claw
towards the midline, where its dorsal surface bears an apophysis for articulating
with the ethmoid. At the symphysis right and left bones are separated by a
fibrous pad. The anterior limb of the claw is curved with its concave surface behind
and grooved. Attached to the groove and in the tissue behind are two to five uneven
rows of relatively large S-shaped teeth. Only the small end of the S projects
through the membrane of the disc with a brown, pointed or chisel-shaped tip.
Behind these teeth is the posterior, straight limb of the premaxilla and this bears
on its posterior mediad surface some much smaller teeth, shaped as a more gently
curved S with their tips pointed backwards into the mouth. The lateral of these
are attached to the bone, but others are quite free from it in the membrane between
right and left bones. In C. batesii the attached and free teeth compose a uniform
group, but in C. micropogon the median teeth are still smaller and sections show that
FISHES OF WEST CAMEROON
3S9
pmx
pmx
mdt
2 mm
FIG. 4. Chiloglanis batesii, specimen from R. Ja. A. Dorsal view of bones of anterior part
of head, with premaxillary teeth indicated on left. B. Isolated premaxilla. C. Outlines
of tooth patches from below, ap = apophysis for ethmoid, d = edge of disc, md.t. =
mandibular teeth, meth = mesethmoid, mx = maxilla, pal = palatine, pmx — premaxilla,
t = isolated premaxillary tooth and bases of some in situ, pmx.t. i = main patch of
premaxillary teeth, pmx.t.2. = patch of smaller premaxillary teeth.
the membrane in which they lie forms a minute pouch opening backwards, the teeth
at the edge of the pouch forming a transverse series across the midline. Examined
with a dissecting microscope these latter appear as a row of minute ciliiform teeth
on a small membranous curtain. This pouch is not the oral valve, which is a more
posterior, wider, thin-walled pouch.
Boulenger (1907^ and Whitehead (1958) refer to these smaller premaxillary
teeth as vomerine, but, as we have seen, the lateral ones are attached to the pre-
maxilla and although the anterior end of the vomer is above the median teeth it is
separated from them by connective tissue ; their position in front of the oral valve
also shows them to belong to the premaxillary series.
Fig. 4 shows these structures in a dissected C. batesii and Fig. 5 in an alizarin
preparation of a smaller C. micropogon. The teeth have been omitted from Fig. 5,
but supporting structures of the disc, not obvious in dissection, have taken the stain
3 6o
E. TREWAVAS
pmx
mx
I mm
FIG. 5. Chiloglanis micropogon. A. Cartilaginous supports of posterior part of disc. B.
Right half of skeletal tissues of anterior part of head. Coarse stipple hyaline cartilage,
fine stipple fibrocartilage. d = edge of disc, fo = frontal fontanelle, fr = frontal, l.eth =
lateral ethmoid, meth = mesethmoid, mx = maxilla, pal = palatine, pmx = premaxilla,
/ = space between arms of premaxilla where the main premaxillary teeth are situated.
with a purplish colour and are inserted. Sections show them to be composed of
fibrocartilage. One extends from above the anterior limb of the premaxilla on each
side forwards, and near the anterior rim of the disc bends towards the midline, where
it is continuous with its counterpart of the other side, the whole recalling the front
bumper of a car. Over it lies a median pad of less collaginous tissue. Another is
attached to the posterior limb of the premaxilla and is directed obliquely outwards
behind the maxilla and parallel to it. In the part of the disc representing the lower
lip are two pairs of bars composed of fibrocartilage incorporating a narrow rod of
hyaline cartilage. Each is divided at its distal end in a palmate manner into three or
FISHES OF WEST CAMEROON 361
four twigs (Fig. 5 A). They are probably the cores of the incorporated mandibular
barbels and they support the posterior part of the disc like umbrella ribs. Between
the supporting structures the connective tissue of the disc is loose and includes wide
spaces.
The sensory buds on the ventral surface have been mentioned. On the dorsal
surface of head, trunk and fins, in the deeper epidermal layer, are large club cells
staining pink with Masson's triple stain (like the huge cells of the pectoral gland) ;
and in the external layer of the epidermis over the whole surface and inside the mouth
are much smaller mucus cells, staining pale blue. I use the nomenclature of Bhatti
(1938), who followed Rauther, 1937 and others.
The Mungo specimens
The Mungo specimens of Chiloglanis belong to two species, one of which I identify
with C. micropogon Poll, 1952 ; the other I describe as new, naming it after the
collector C. disneyi.
Both species have the mandibular barbels almost entirely incorporated in the disc,
a row of minute ciliiform teeth in the membrane behind the premaxillary symphysis
and very small mandibular teeth in comparison with several other species.
They are distinguished from each other (see Table 10) by the size of the eyes in
adults and their distance from the posterior nostril ; by the interorbital width ; by
the position of the pelvics, more posterior in C. disneyi ; the shape of the caudal fin,
more deeply forked in C. micropogon ; the number of soft dorsal rays, 6, rarely 5,
in C. micropogon, 4 or 5 in C. disneyi ; the number of mandibular teeth, 8-12 in
C. micropogon, 15-20 in one row in adults in C. disneyi ; and in adults by details of
the premaxillary dentition. In addition our biggest specimens differ in the pattern
of bosses on the ventral surface of the oral disc.
The proportions are difficult to measure consistently because of different positions
in which the disc has been fixed, but those used for specific distinction are sufficiently
contrasted in adults to be reliable.
Chiloglanis micropogon Poll, 1952
Poll, 1952 : 228 figs. 3, 4 (Eastern sources of the Congo, affluent streams of Rivers Ulindi (Kabare
Territory) and Luhoho (West Kivu)); Daget, 1954 : 3°7 &S- II6 (Rivers Tinkisso and Niandan,
upper tributaries of R. Niger).
Several specimens of this species were caught from under stones by Dr Disney in
the upper Mungo at Baduma, its tributaries Blackwater, Wo we and Dilolo and in R.
Bille, Meme system. It has not previously been recorded from Cameroun and this
locality is distant from both the type localities and the upper Niger system whence
Daget recorded it. Some hitherto unnamed Chiloglanis in the BMNH, collected
by Welman in 1934 in R. Kaduna, Nigeria, comprise ten of this species and two C.
niloticus waterloti Daget ; and a single specimen from R. Lobi, South Cameroun,
included by Boulenger in C. batesii, also proves to be C. micropogon.
362
E. TREWAVAS
The description set out in Table 10 and the following paragraphs is based on n
specimens from the Mungo system and R. Bille, 21-5-35-5 mm in SL.
Head tuberculate in adult. Maxillary barbel 0-6-1-45 times diameter of eye ;
outer mandibular barbel absent or scarcely longer than the other fringing scallops of
the lower lip, inner no longer.
TABLE 10
Critical characters in specimens of Chiloglanis of the Mungo system and R. Bille assigned
to C. micropogon and C. disneyi
SL (mm)
Length of head (% SL)
Distance P-V (% SL)
Base of adipose (% SL)
Distance adipose to rayed
D/base adipose
Length of snout (% length of
head)
Diameter of eye (% length of
head)
(% post-cephalic length)
(times in distance from
nostril)
(times in interorbital width)
D soft rays
Anal branched rays
Teeth lower jaw (one row, the
longer)
C. micropogon
(n specimens)
21-5-35-5
36-0-42-8
19-5-22-5
11-6-16-9
1-15-2-0
53-65
11-75-18-8
7-7-11-3
0-65-1-0
1-05-2-0
6 (or 5)
5-7
C. disneyi
Holotype
Wowe Blackwater
35
27-0
2-25
70
12-4
6-5
1-9
2-5
4
5
20
25
36-0
28-0
2-36
68
n-i
6-25
1-6
2-2
5
6
15
C. disneyi ?
Dilolo
24-5
24-5
38-0
36-0
26-0
26-0
18-4
16-0
1-2
i'5
56
62
15-0
16-7
9-6
9'4
0-85
0-8
1-6
1-4
4
5
5
5
II
Teeth of the ventral face of the premaxilla in four series at the widest part, with
bluntly pointed or chisel-shaped tips. A pair of smaller groups of very small teeth
on the posterior face of the premaxillae, nearly meeting at the symphysis and behind
the symphysis minute, extremely slender teeth, the posterior of which are arranged
in a transverse row, with their tips appearing as a row of ciliiform teeth on a short
membranous curtain (in which their bases are embedded).* The mandibular teeth
are S-shaped, smaller than in many species, forming one row of 8-n teeth, sometimes
with a second row of replacement teeth. If the earlier set of teeth persists behind the
replacements the latter may be a little more numerous (nearer the definitive number) .
Among the specimens used for this description two rows (8 + 10 and 10 + 10) are
present in two. The Lobi specimen has n + n and the holotype (Poll, 1952)
12 + 12. In smaller specimens two rows are naturally more frequent.
Pectoral spine short, 12-4-17-8% SL. Dorsal spine from a little shorter than the
pectoral to a little longer. Usually 6 branched rays in the dorsal, but 5 in one speci-
men. Anal with 7-9 rays, 2 of which are simple. Caudal forked, the middle rays
Dr Poll has kindly re-examined the types at my request and finds a similar row of fine teeth in them.
FISHES OF WEST CAMEROON
363
FIG. 6. Dorsal view of heads of, left, Chiloglanis disneyi and, right, C. micropogon.
being 10-17% SL, usually 10-15%, the longest rays 1-5-2-2 times as long, forming
bluntly pointed or rounded lobes.
Ground colour yellow in life. Top of head irregularly clouded with dark grey.
Behind this three black bands on the back which vary in width from narrow in some
of the Dilolo fishes to continuous in the one from R. Bille ; these extend on to the
sides where they are usually less intense and spread, anastomosing irregularly. A
dark vertical streak at base of caudal ; a broad black band crossing each caudal lobe.
Dorsal fin spine dark ; a dark mark on leading pectoral edge and a streak from it
crossing the fin, but only on the rays, absent in paler fishes. Pelvic and anal fins
may be lightly marked. Ventral surface without melanin.
A female of 28 mm has ovarian eggs of 1-7 mm diameter, one of 35-5 mm of 1-9 mm.
The genital opening is behind the anus between a pair of low ridges and behind it is
a tapering urinary papilla. A similar papilla present in the male probably transmits
the urinogenital duct.
The gut of one of the sectioned Mungo specimens contains filamentous blue-green
algae and a few diatoms. Another contains a small arthropod, probably an ostracod,
that had been feeding on the same algal mixture. The specimen stained with alizarin
has a segmented invertebrate in its stomach. These observations point to a habit
of feeding on the algal film of stones and its microfauna.
The small C. micropogon (up to 23 mm SL) from R. Kaduna resemble these in most
respects, but the length of the adipose fin is variable and may even exceed its distance
from the rayed dorsal, which has 6 soft rays in all. The dark colour is paler, much
less contrasted with the ground colour than in most of the Mungo specimens or than
in the C. niloticus waterloti found with them. A note from the collector (J. B. Wei-
man) states that they were clinging to algae-covered stones.
E. TREWAVAS
FIG. 7. Chiloglanis micropogon. Ventral view of disc showing sensory tubercles
and teeth.
The syntypes of C. batesii, 27-5-33-5 mm in SL, have in common with C. micro-
pogon the short barbels and the proportions of the head and eyes. They differ from
it in lacking the differentiated ciliiform teeth, in the somewhat larger and fewer
mandibular teeth (in the syntypes respectively 6 + 1, 6 + 4, 7, 7, 8), in the longer and
stronger pectoral spine (17-7-24-8% SL) and in the produced caudal lobes with a
dark band continuous around the fork (see Boulenger, 1911, fig. 363). I find 6 soft
dorsal rays in all the syntypes, not 5 as reported by Boulenger ; the last two are
quite separate at the base.
C. micropogon appears to be related to C. brevibarbis Boulenger, a species of the
upper reaches of the rivers of Kenya, both eastward flowing and those entering Lake
Victoria from the North and East (Whitehead, 1958). In this species the barbels
are less reduced, there are only 5 soft rays in the dorsal fin and there are usually no
ciliiform teeth at the back of the posterior premaxillary group. Only in one speci-
men of several examined have I found such teeth, but they were of the same size as
the other posterior premaxillary teeth. This species is represented in the BMNH by
specimens mostly bigger than any taken in the Mungo.
MATERIAL EXAMINED
BMNH register no.
1909.4.29.97
I935-5.29.25-34
I973.5.I4-346
1973.5.14.347-354
Locality
R. Lobi, S. Cameroun
R. Kaduna, Nigeria
R. Bille, Meme system
R. Blackwater, Mungo system
FISHES OF WEST CAMEROON 365
BMNH register no. Locality
1973.5.14.355-359 R. Dilolo, Mungo system
I973-5-I4-36o-363, 371 R. Mungo, nr Baduma
1973.5.14.364-370 R. Wowe, Mungo system
Chiloglanis disneyi n. sp.
HOLOTYPE. $ 35 mm in SL from R. Wowe, collected by Dr Disney in October,
1972, BMNH 1973.5.14.342.
The most important proportions and meristic characters of the holotype and three
other specimens probably of this species are set out in Table 10. The following
additional details are taken from the holotype.
Head smooth, without tubercles. Maxillary barbel 1-3 times diameter of eye ;
barbels of lower lip scarcely projecting beyond edge of disc. Pattern of bosses on
ventral surface of rostral half of disc forming a series of frond-like units.
Teeth of ventral face of premaxillary with chisel-shaped tips, in 3-4 rows, the
right and left groups well separated from each other. Between and behind them the
minute teeth of the posterior face of the bone apparently on the 'stem of the claw'
(see description on p. 358 above) ; the posterior of these forming a median transverse
row of ciliiform teeth in the membrane, which is not raised as a 'curtain'. Mandi-
bular teeth very small, close-set, in one row of 20.
Pectoral spine 18-5% SL ; dorsal spinous ray 17% SL, as long as its soft rays,
which number 4. Anal with 7 rays, of which the first two are simple. Caudal emar-
ginate, with bluntly pointed lower lobe (the upper is incomplete), its middle rays
16% SL, rays of lower lobe less than 1-3 times as long.
Markings on body running together irregularly, giving a generally dark colour
above the flanks ; ventral surface white in preserved fish. A dark band at base of
caudal fin and another crossing all the rays.
A second specimen, 25 mm .in SL, from R. Blackwater (BMNH 1973.5.14.345), as
can be seen from Table 10, agrees with this and contrasts with C. micropogon in the
posteriorly placed pelvics and adipose, the longer snout, smaller eye and higher
number of mandibular teeth. Its caudal fin is less deeply forked than in C. micro-
pogon, but the premaxillary teeth and the pattern of bosses on the disc show no
contrast with C. micropogon, perhaps because it is too young to have developed the
specific pattern.
Two Dilolo specimens (BMNH 1973.5.14.343, 344) are placed tentatively with
C. disneyi because of the posterior position of the pelvics, but their eyes are no smaller
than in C. micropogon of the same size and the numbers of mandibular teeth are
within the range for this species, although C, disneyi probably passes through a stage
in which they are as few.
AFFINITIES. C. disneyi shows resemblances to four other species of Chiloglanis,
each described and known only from a single specimen. These are : C. congicus
Boulenger, 1920, from Stanley Falls (SL 75 mm), C. marlieri Poll, 1952, from a
remote tributary of the Congo west of Lake Kivu (SL 45-6 mm), C. polyodon Norman,
1932, from the headwaters of R. Bagbwe, Sierra Leone (SL 44-6 mm) and C. lamottei
366
E. TREWAVAS
FIG. 8. Chiloglanis disneyi. Ventral view of head showing sensory tubercles
and teeth.
Daget, 1948, from Mont Nimba (SL ca. 43 mm). These all have high numbers of
mandibular teeth, posteriorly placed pelvics and small eyes well separated from each
other and from the posterior nostril and all except C. congicus have fewer than six
branched rays in the dorsal fin. - At least in C. polyodon and C. marlieri there are no
tubercles on the top of the head.
I have seen the types of C. polyodon and C. marlieri and rely on the descriptions of
Boulenger and Poll (1952 : 231) for C. congicus and of Daget for C. lamottei. Com-
parison is not easy because of difference in size and preservation - the first three were
fixed in alcohol and the edges of the disc are reflected over the ventral bosses, which
are themselves flabby in contrast to our formalin-fixed specimens. In C. polyodon
and C. marlieri the bosses have not the frond-like structure seen in the holotype of
C. disneyi, but even in the Mungo we cannot be sure that this is specific. The 18
mandibular teeth of C. congicus are described as bunched together ('en bouquet',
Poll) and this, together with the six branched dorsal rays, may be held to distinguish
if from C. disneyi. The high number (44 Poll, 41 my count) of lower teeth in C.
marlieri is made up of one row of 26, a widely interrupted inner row of 6 + 7 and 2
incompletely erupted outer teeth at the midline. Whether this should be inter-
preted as an inner, partly shed row, a functional row and two of a future replacement
series, or whether it foreshadows the multiseriate dentition of Atopochilus I do not
know. The line of teeth occupies a greater percentage of the width of the broad
mouth-opening than that of C. disneyi does of the narrower mouth, and its unique
appearance is against regarding it as conspecific with any of the other species under
consideration, at least until more is known of the populations represented by the
types. The posterior premaxillary teeth are of one size without an inner row of
FISHES OF WEST CAMEROON 367
ciliiform teeth. In C. polyodon the 15 inner and 14 outer mandibular teeth are
bunched together ; the posterior teeth of the upper jaw are almost confined to the
row of ciliiforms, with only 2 or 3 others, while the main premaxillary tooth-patches
are large, on relatively massive bones (as in C. marlieri). The mandibular barbels
are very short but the maxillary barbel is 1-6 times the diameter of the eye and the
caudal fin is more deeply forked than in C. disneyi. C. lamottei has 14 mandibular
teeth in one row, which according to the figure is more extended than in C. disneyi.
Although the distribution of C. micropogon accepted here suggests that species or
species-groups of Chiloglanis may have a wide and discontinuous distribution in
swift, clear waters, there does not at present seem to be enough evidence to justify
uniting the Mungo form here named C. disneyi specifically with any other.
Young Chiloglanis
Smaller specimens have not yet developed the features here relied on to distinguish
the two Mungo species. The diameter of the eyes is more than their distance from
the nostril, the teeth are not fully developed and the bosses on the disc are discrete.
The caudal has a lunate posterior margin at SL about 12 mm and the adipose varies
in extent, in some being longer than its distance from the rayed dorsal (cf. the small
C. micropogon from Kaduna). The dorsal formula is I 5 in two, I 6 in the rest.
They are all tentatively placed with C. micropogon.
MALAPTERURIDAE
Malapterurus electricus
Silurus electricus Gmelin, 1789, I : 1351 (Nile).
The electric catfish was abundant at both stations of R. Mungo and in the Wo we
and is the classical species. Thys (1967 : 62) records it also from a rivulet near
Barombi Kotto and discusses its occurrence in Fernando Poo. It is evidently one
of the most successful of African freshwater fishes and is common throughout its
wide distribution in west and central Africa excluding Lakes Victoria and Malawi
and the upper Zambezi, although it is present in the middle and lower Zambezi and
in R. Pungwe (Jubb, 1961).
Two specimens from R. Wowe were preserved (BMNH 1973.5.14.334, 335).
GYPRINODONTIDAE
Epiplatys sexfasciatus Gill, 1862
(Figs 9 and 10)
Gill, 1862 : 136 (Gaboon River ?) ; Trewavas, 1962 : 159 (Barombi Kotto and Barombi Mbo) ;
Scheel, ig68a : 383, figs. ; Trewavas et al., 1972 : 47, 80.
Haplochilus infrafasciatus Giinther, 1866 : 313 (Old Calabar) ; Lonnberg, 1903 : 39 (coastal
river at Sanye, near Barombi Kotto).
Epiplatys sexfasciatus rathkei Radda, i97ob : 152 fig. 3 (R. Kake, nr Kumba).
368
E. TREWAVAS
FIG. g. Epiplatys sexfasciatus $ from Barombi Mbo, a six-barred specimen.
Barombi name at Barombi Mbo : longo katta.
This proved to be one of the most abundant species in the Kumba area and is one
of the three non-endemic fishes in Barombi Mbo, where it is common around the edge
of the lake and in both inflowing and outflowing streams. We did not catch it in
Lake Kotto itself nor in Mboandong, our only specimens from this area being two
females, one caught by hand-net in Tung Nsuria, a stream entering Kotto from within
the crater, the other from R. Nyoke.
The male is richly coloured. The flanks are green, countershaded to dark green
on the back and yellowish below. Each scale of the upper half of the body carries
a red-purple spot. The dorsal fin has a black band within a narrow white margin
and there is some black on the posterior rays ; the centre of the fin is yellow with red
spots. The bluntly pointed or rounded caudal has a black marginal band above and
below and within this a yellow band that reaches the edge on the middle rays and
bears red spots and streaks in its upper section ; the middle and proximal part of the
fin is greenish with darker rays. Both anal and pelvic have black marginal and
yellow submarginal bands ; the proximal part of the anal is like the middle of the
caudal. The head is dark with a gold spot on the occiput, the throat pale with a dark
streak parallel to the lower jaw behind its edge. There are six to nine near- vertical
dark stripes on the lower half of the body, the first behind the pectoral base, the last
extending right across the base of the caudal. Stripes additional to the basic six
are frequently shorter or narrower than the others.
The female is paler to watery buff ; dorsal and caudal fins are watery green with
faint crimson spots, the other fins colourless. In some females observed the vertical
bars were shorter than in males and might disappear altogether. In one in an
aquarium they were seen to disappear on the side towards the light while they re-
mained on the other. A vague horizontal band appeared when the fish was chased
and disappeared when it was narcotized.
FISHES OF WEST CAMEROON 369
The fins of males are longer, the dorsal and anal in mature fishes extending beyond
the base of the caudal, the pectoral beyond the origin of the pelvic and the latter
beyond the origin of the anal. Males collected reached a length of 65 + 22 mm,
females 55-5 + 16-5 mm.
One female contained transparent spherical eggs i mm in diameter with one or
more transparent globules within them. According to Scheel, ripe eggs are bigger -
1-4- 1-6 mm - and Loiselle (1971) gives 2-0 mm.
The type locality of E. sexfasciatus is a queried 'Gaboon River', and its
distribution extends in coastal areas from the Congo to the Togo Hills. Scheel,
who (i968a) compared samples from several populations morphologically, cyto-
logically and by breeding experiments, found reduced fertility in hybrids
between individuals from Barombi Mbo and western Nigeria, but concluded
that the species comprises a chain of populations within which it is not
profitable to name subspecies. But Radda (igyob) has named the population
near Kumba E. s. rathkei and Loiselle (1971 : 37) describes the most western
population as E. s. togolensis.
Radda distinguished E. s. rathkei by its rich colouring and by the higher number of
vertical black bars (usually 8-10), but he did not compare with the population of
Gabon, the presumed type locality. I have seen only two preserved specimens from
Gabon and they have only six bars, but so have many specimens from Barombi
Mbo and the Mungo. Scheel (1968 : 392) states that the pelvics in populations of
South Cameroun and Gabon are 'short' in contrast to those of western Nigeria,
without quantifying his statement. Loiselle's definition of 'short' is 'not reaching
the anal fin', but this in inadequate. Western samples available to me do not
contain enough mature males for a satisfactory comparison. Fig. 10 shows that the
pelvic length in 28 males of 38-65 mm SL from Barombi Mbo and the Mungo system
is 18-7-25-8% SL. In 5 males of 40-7-49-6 mm from near Lagos I find a ratio of
22-9-33-2%. Both samples include males of varying degrees of maturity and the
figures show that Scheel's observation for the species in South Cameroun and Gabon
can be extended to include the Mungo population in this character. I do not
therefore consider that on present knowledge there is any justification for the use of
a subspecific name to distinguish the Mungo members of E. sexfasciatus from those
of the type locality. We found no difference between those from the lake and the
streams.
E. sexfasciatus is reported from Fernando Poo by Boulenger (1915) and Thys
(1967). Thys found it the most abundant freshwater fish on the island ; he dis-
cussed its distribution and synonymy. In Fernando Poo the vertical bars on each
side number 6-9 and the pelvics in males up to about 70 mm SL are not greatly
prolonged. Their length is given in terms of the length of head and so is not
directly comparable with Fig. 10. In Thys's figured male the pelvic is about 20%
SL and extends to about the 4th anal ray. The island populations evidently agree
with those of the opposite mainland in essential features.
Recently (1971) Scheel has merged Epiplatys with Aplocheilus McClelland, but
until the evidence is given I use the name under which this species is well
known.
E. TREWAVAS
O
32
30
28
26
24
22
20
18
16
O
V
SL
O
O
O
O
O
34 38 42 46 50 54 58 62 66
SL mm
FIG. 10. Epiplatys sexfasciatus. Length of pelvic fins as % SL. O 6*6* from southern
Nigeria ; • 6*6* and © $? from Barombi Mbo and tributaries of R. Mungo.
Aphyosemion bivittatum (Lonnberg, 1895)
Fundulus bivittatus Lonnberg, 1895 : 190 (a rivulet near the waterfall of R. Ndian, between
Rivers Meme and Cross).
Aphyosemion bivittatum ; Scheel, ig66a (characteristics and interrelationships of populations) ;
id. i968a : 116-127, illustr. ; Thys van den Audenaerde, 1967 : 63 fig. 19 (Fernando Poo) ;
Radda, i97ib : fig. 8 ; id. i97ic : 128.
A. bivittatum was caught in the following localities : R. Kake at Kake village
(2 males), BMNH 1970.12.14.10-11 ; Tung Nsuia, flowing into Barombi Kotto
(several), 1970.12.14.31-33 ; Edge of Mboandong (abundant, with A. oeseri) ;
1970.12.14.16-30 ; the stream Nyoke, south of Kotto, tributary of lower Mungo,
1970.12.14.12-15.
Males are easily distinguished from the other West Cameroonian species of
Aphyosemion by the long streamers of the dorsal fin and upper and lower corners of
the caudal. Both sexes and juveniles are further distinguished by the presence of
two dark transverse stripes across the lower jaw ('chin-straps') separated by a white
or yellow area.
The two longitudinal stripes that give the species its name were present in the
living fish only in juveniles and some females. This is in agreement with the
observation of Scheel (i966a) that the stripes appear only in females in fear or making
FISHES OF WEST CAMEROON 371
submissive movements. In a male the two lead-grey stripes appeared on narcotiza-
tion and on death.
This species also has fewer and larger scales than the other two - 25-26 in a
longitudinal series, 12 around the caudal peduncle. Dorsal rays number 10-12,
anal 12-15. These ranges are understandably narrower than those given by Scheel
(i968a) in a much larger number of specimens from several populations.
Breeding colours were most marked in a male of 31 mm SL from R. Kake, from
which the following notes were made.
Abdominal wall green, chest silvery ; scales of other parts of body each with a
crimson spot, those of three upper rows continuous with three crimson streaks on
the gill-cover ; a fourth crimson streak below eye ; two black chin-straps with
yellow between them ; inter- and suboperculum yellow. Dorsal fin with crimson
spots near its base and with yellow tip prolonged to extend well beyond base of
caudal. Upper and lower tips of caudal also bright yellow and prolonged ; base of
anal fin greenish with dark rays ; edge of fin black. Pectorals greenish yellow ;
pel vies yellow.
In less mature males the anal has no black edge, but the edge may be pinkish. In
another male soon after death the anal had crimson spots basally and a crimson
margin ; the pelvics were yellow with a crimson edge, base and inner ray. Markings
that appear black in life often become crimson immediately on death, presumably by
contraction of the melanophores that are mixed with red pigment.
The biggest specimen in our collection is a $ of SL 41 mm, whose ovary contained
clear, spherical eggs of 1-6 mm diameter, each with one or more minute clear globules
inside it.
Scheel (i966a) has made a study, illustrated by fine colour-photographs, of
different populations assigned to this species, which he regards as a superspecies.
Some of the names in his tentative synonymy (ig66a : 262), the types of which were
from a wide area from R. Kribi to Lagos, may be open to doubt, but at least the
distribution he gives, from Togo (?) and Dahomey to Rio Muni, is that of the group of
populations to which our species belongs. Scheel has brought much evidence,
including chromosome studies and crossing experiments, to establish the degree of
relationship among them.
Thys (1967) recorded the species from Fernando Poo.
Among Scheel's colour-photographs of I966a, his fig. 7 portrays a <$ from R.
Mboumboula and fig. 10 one from near Buea. Neither of them exactly corresponds
to the description above, but I think the differences are no more than could be
accounted for by modifications due to lighting and printing and by variation and
phase of maturity (cf. Foersch's experience with aquarium-bred generations of
A. oeseri, see p. 375, below).
Aphyosemion celiae Scheel, 1971
Scheel, 1971 : 52, fig. on p. 55 (Mambanda, nr Kumba) ; id. igy2a : 193 (chromosomes) ; Radda,
1971 : 128 fig. 4, lower (Mayefe stream, nr Mambanda).
372 E. TREWAVAS
Our material consists of : 5 <£<£, 25-29 mm in SL, 9 $$ of 25-28 mm and 17 im-
mature of 15-18-5 mm from Mangusu, most of them collected by Dr Disney a few
days after the stream had been seen to be dry in that place ; a <$ of 25-5 mm from the
neighbouring R. Moliba, also coll. Dr Disney ; 2 <$<$ and 9 $$, i7'5~34 mm in SL from
Lake Soden, coll. Dr J. Griffith. BMNH 1970.12.14.74-111.
Meristic characters, checked on 15 specimens from the Mungo tributaries and n
from Lake Soden, are : D 10-12 ; A 14-15 ; scales in a longitudinal series 29-31,
around caudal peduncle 13-16. Caudal rounded.
A colour sketch from one of our Mangusu specimens, a living, narcotized male,
shows the characteristic dark arc, here violet, on the caudal as well as a band of the
same colour on dorsal and anal, parallel to the margin but separated from it by a
narrow white streak and a broader orange-yellow marginal band. The outer zone
of the caudal was orange, but was interrupted on the middle rays by the dark blue-
green ground colour of the fish as Scheel (1964 : 326) describes in one variant of A.
cinnamomeum. The proximal part of the dorsal fin was streaked with red. The
pelvics had a proximal blue-green zone streaked with red, followed by violet and red
zones. The pectoral was yellow and no extension of this colour on to the body was
noted. The body colour of this and other river males was green or blue-green with
scattered purple spots. Red streaks were present on the head.
In females there are vague dark spots on the dorsal and upper parts of the caudal
fins.
Notes on the Lake Soden population agree with this, although when they were
observed (in aquarium) the body colour was pale light green to straw-coloured and
the elements of the caudal pattern of males changed in intensity as the fishes swam
about. They agreed with A. celiae in contrast to the related A. cinnamomeum in
having the scales of the flank and caudal peduncle each with a carmine spot. In a
well-coloured male dorsal, anal and caudal fins each had a broad orange marginal
zone and a neutral proximal zone, between which a dark band appeared and dis-
appeared, that of the caudal fin being looped as in A. celiae and A. cinnamomeum
and in no other species known to me. The pectoral fin was yellow with a whitish
lower edge.
Both the Mungo and Lake Soden specimens agree with A. celiae and differ from
A, cinnamomeum in the presence of red or purple spots on the flanks, in the lower
numbers of fin-rays and the small pores of the preopercular canal of the lateral line
system. A. cinnamomeum, according to Clausen (1963) and Scheel (igGSa), has
D 13-14, A 16-17, and the preopercular pores are enlarged. Gyldenholm & Scheel
(1971) and Scheel (i972a) also report a difference in karyotype between the two
species, A. celiae in the type locality having 10 haploid chromosomes with 20
chromosome arms, A. cinnamomeum 20 chromosomes and 31 (Scheel, 1971) or 30
(i972a) arms.
The type locality of A . cinnamomeum is a stream about 70 km north of Kumba,
but whether in the Mungo or the Cross system is not clear. In any case the localities
where both species have been taken are in a relatively small area where affluent
streams of Mungo, Cross and Meme rise. They share the peculiar caudal colour
pattern and Scheel (1971 : 57) reports that both show signs of a short diapause
FISHES OF WEST CAMEROON 373
before the egg begins to develop. In this connection it is interesting that this was
the first species found when R. Mangusu flowed in its bed again, although of course
the development of these specimens had already proceeded, probably in some up-
stream pool that had had at most a shorter period of desiccation than the river at
the locality of our collection.
Aphyosemion oeseri (Schmidt, 1928)
Panchax oeseri Schmidt, 1928 : 165 fig. on p. 166 (Fernando Poo).
? Panchax vexillifer Meinken, 1929 : 255 fig. ('Westkiiste Afrikas').
Aphyosemion cameronensis (nee Boulenger, 1904) ; Boulenger, 1915 : 48 (Fernando Poo specimens
only) ; Thys van den Audenaerde, 1965 : 316 ; id. 1967 : 72, but not fig. 21 (Fernando Poo).
Aphyosemion santa-isabellae Scheel, 19685 : 332, fig. i (Fernando Poo).
? Aphyosemion scheeli Radda, 19700 : 178, figs i and 2.
Aphyosemion santaisabellae ; Foersch, 1971 : 20, 4 figs ; Berkenkamp, 1972 : 795, fig. on p. 796.
Aphyosemion calliurum s.l. ; Radda, 19710 : 128 fig. 4.
Aphyosemion oeseri ; Trewavas et al., 1972 : 47 (inflowing stream of Barombi Mbo) ; Roloff,
1972 : 381 (Fernando Poo).
Our preserved collections comprise :
Length (mm) Locality BMNH register no.
1 33 + 9'7 R. Moliba, upper Mungo 1970.12.14.63
6 24-5 + 7 to 36-0 + 9 Inlet stream of Barombi Mbo 1970.12.14.64-69
i 28-1 + 8-6 Barombi Kotto, near outlet 1970.12.14.70
10 18-0 + 6-5 to
29-0 + 8*2 Tung Nsuia in the Kotto basin 1970.12.14.34-43
21 14-5 + 5 to 24-5 + 7 Mboandong 1970.12.14.44-62,71-72
DESCRIPTION OF THESE SPECIMENS. Scales 29-31 in a longitudinal series (ex-
cluding one or two on the caudal fin), 14-16, usually 16 around the caudal peduncle.
D 9-11 ; A (12 ?) 13-16 (mean 14-1). Tip of dorsal not extending to vertical of
base of caudal fin ; origin of dorsal above 4th to 6th ray of anal. Caudal fin pointed
in young, becoming rounded in female, cricket-bat shaped in male.
Often a grey or black spot on chin (no 'chin-straps').
Female watery green or brown, with a few red spots on flanks ; dull reddish spots
on dorsal and upper part of caudal fins ; pelvics and anal greenish yellow with grey
margins.
Male with bluish or greenish flanks with red spots on many of the scales, either
irregularly scattered or partly confluent to form three main longitudinal lines from
operculum to above anal fin ; no vertical markings on caudal peduncle ; two or
three wavy red streaks on head below and behind eye. Dorsal fin greenish with dull
red spots and sometimes a very narrow yellow edge ; similar spots on proximal part
of caudal, which has a broad yellow band along its lower margin and a narrower one
along the upper, both of them submargined with a dark grey or dark red band.
Pelvics and anal yellow or salmon pink proximally, outside this a dark red band or
row of contiguous spots that may be delimited with grey or light blue, and distally
374 E. TREWAVAS
a broad bright orange band. Pectorals with a yellow to orange lower margin sub-
margined with grey.
Many males lack the yellow on the tip of the dorsal and on the upper caudal mar-
gin, i.e. these probably develop later than the lower yellow marginal bands.
In the living fish the occiput has a metallic golden or silver spot.
Size : Our biggest and most colourful male had a total length of 45 mm (36 + 9)
and the biggest female 42-7 mm (33 + 9-7).
ECOLOGY. See Green & Corbet in Trewavas et al. 1972, and Corbet et al. 1973.
DISCUSSION OF THE SYNONYMS. The fin-ray numbers given by Meinken for A.
vexillifer are D 9, A 14. Scheel gives for A. santaisabellae D 10-12, A 14-16 (mean
14-6) ; his biggest specimen was 41 mm in SL.
The history of these names has been summarized by Scheel (i968a : 377 ; 1969 :
74), who dismissed A. oeseri as an invalid name. With this I cannot agree. The
name was published by Schmidt (1928), attributed to Ahl and accompanied by a
photograph and full colour description. In the same year Oeser (1928 : 374) had
described his capture of cyprinodonts in the neighbourhood of Santa Isabel, Fernando
Poo, and wrote that his collections, mainly herpetological, would be handled by
Dr Ahl. Ahl never published on the cyprinodonts and although Schmidt no doubt
relied on Ahl's authority for considering it new to science he satisfied the requirements
of Art. 50 of the International Code as the author to whom the name should be
attributed. The photograph in Schmidt's article very well matches that of Scheel's
A . santaisabellae, and when to the one is added Schmidt's description and to the other
Foersch's account of the aquarium offspring of specimens collected by Scheel as
near as possible to Oeser's collecting ground there can be no doubt that they are the
same species. Schmidt's article is the one piece of evidence (and the most important)
that Scheel failed to see.*
Herr Roloff (1972) has also given reasons for adopting this name for the Fernando
Poo population.
Meinken's P. vexillifer was based on specimens of a consignment of about a hundred
of this species brought from 'Westkiiste Afrikas' by a friendly seaman to the Tlaty-
Tischrund', Hamburg, where it was shown in October, 1928, the month of Schmidt's
article on 'Neuimportierte Fische'. It does not seem far-fetched to suggest that this
seaman was a member of the crew of the ship on which Dr Oeser travelled, though
Dr Oeser (in litt.} does not think so. Oeser started his journey in a ship of a Hamburg
line about the beginning of May, 1928, reaching Teneriffe on 5 May. The ship
called at Liberia, Gold Coast, Fernando Poo and several other places (but not
Nigeria or Cameroons) southwards to Wai vis Bay, stopping a few days at each port,
then turned homewards, a journey that might have occupied up to three months.
Oeser collected very few fishes and none of the others mentioned by him could be
confused with A. oeseri or A. vexillifer. Meinken himself realized (1929 : 257) the
possible identity of P. vexillifer with P. oeseri, which he too considered not validly
* I have the following statement from Dr Oeser: "... the fishes named after me, Panchax oeseri by
Herbert Schmidt in 1928, were collected by me in Fernando Po as described in my article of 1928.
Freiburg d. 20.10.1973 Dr. med. Richard Oeser."
FISHES OF WEST CAMEROON 375
described. His drawing of A . vexillifer was reproduced by Arnold & Ahl (1936 : 303)
and copied by Sterba (1959) and Scheel (igGSa), becoming progressively less like the
original. Scheel (igGSa) considered it a synonym of A. calliurum, but the absence
of lyre-like corners of the caudal in the original drawing does not favour that species,
though no doubt it, with A . oeseri, belongs to the 'calliurum-group' of the genus.
Schmidt stated that Dr Oeser brought only two specimens, both males ; he does
not say whether these types were preserved, but there is no record of them in the
Berlin Museum (Dr Karrer, in litt.}. The types of P. vexillifer were deposited in the
Magdeburg Museum as no. 23/1929.
The next question concerns the specific identity of our mainland populations with
that of Fernando Poo. Schmidt (1928) described the anal fin as having an outer
orange band, then a row of red spots, then a yellow zone streaked with bluish. The
margin of the dorsal fin was porcelain-white (cf . yellow in ours and Scheel's) and both
upper and lower margins of the caudal were orange-yellow. In Scheel's types there
were in the male no colour-zones in anal and pelvic fins, which were uniformly
yellow. In Berkenkamp's figure the anal is yellow with a few red spots. But Dr
Foersch (1971) got, in his first generation bred from Scheel's collections, a male with
a dark red submarginal band in the anal fin, although this later disappeared. In the
second generation a male developed the same colour-zones in pelvic and anal that
we found ; they were permanent and appeared also in offspring sired by him
(Foersch, loc. cit. fig. on p. 21). The difference, at least at first, was in the apparent
timing, the upper yellow margins appearing before the lower in the Fernando Poo
strain, the reverse in the mainland. Scheel also states that the Fernando Poo fishes
had no grey or black spot on the chin, as the mainland ones often had. But in other
species, e.g. A. obscurum (Ahl), this is said to appear in some individuals, so it is
evidently facultative.
Arnold & Ahl (1936 : 293) described a species that Arnold received from near
Douala and the lower Calabar River. They identified it as A. cameronense, but
noted that it diverged somewhat from the types. Their description fits our specimens
rather well. Their figure and description of 'A . oeseri' were taken from a subsequent
importation and do not seem to be this species.
The species at present known as A. ahli Myers 1933 (see Scheel, 19683. ; Foersch,
1968 : 367) differs from A. oeseri in having a lyre-shaped caudal fin in mature males.
The aquarium strain described by Scheel and Foersch is from R. Mboumboula near
Douala and is similar to ours except for the caudal and some vertical red markings
on the caudal peduncle. The original description of A. ahli by Meinken (1932)
under the name of 'Panchax (Aphyosemion) calliurus Boul. var. caeruleus' is accom-
panied by a very pale colour-plate, but the description corresponds fairly well to
the photographs of the Mboumboula strain by Scheel and Foersch, and to a trans-
parency from the same strain kindly sent to me by Mr Roloff. This population
seems to be related to A. oeseri, but at least subspecifically distinct. The specimens
referred to 'P. (A.) cameronensis' by Meinken in the same publication (1932) may well
be A . oeseri itself.
The name A. cameronense (Boulenger, 1904) has been widely applied to cyprino-
donts from Cameroun and eastern Nigeria, but Scheel (igGSa) has rightly restricted
376 E. TREWAVAS
it to the species whose males have two red longitudinal bands on the body, one of
which runs close to the ventral outline. (He still (ig72a : 193) mentions specimens
from the Mungo system, but we did not find any.) Among the specimens listed in
this species by Boulenger (1915) are two collected by Seimund in Fernando Poo and
some from the lower Calabar River collected by Dr Leiper. These are now all faded,
but none has any trace of two red bands, which have survived preservation in alcohol
in the types of A . cameronense. The Seimund specimens, on meristic numbers and
the shape of the caudal fin, are probably A. oeseri. Two males of Boulenger's
Calabar sample have traces of the caudal and anal colour-zones, but their fin-ray
numbers are rather low D 9 (f.5), n (f.i), A n (f.i), 12 (1.3), 13 (f.i), 14 (f.i). They
are probably another species of the 'calliurum group'.
We have thus no firm evidence that the distribution of A. oeseri extends to the
Calabar River, and it has not been reported south of Mount Cameroon.
Scheel and Radda have both collected recently in the Kumba area. The species
that Radda (19710) calls 'A. calliurum s.l.' is probably our A. oeseri. One hesitates
to suggest that Scheel, who knows A. cameronense from the type locality, would use
that name for our species and fail to recognize its resemblance to the Fernando Poo
population that he described as A . santaisabellae , but after hesitating I do suggest it.
The chromosome number given in his Table 3 (i972a) for the Mungo samples is
n = 16, chromosome arms 23, that of A. 'santaisabellae' n = 20 with 20 arms. The
correspondence between various populations of a cyprinodont phenotype and their
karyotypes is so irregular that Scheel has almost ceased to use the word 'species'. It
would be interesting, however, to know the results of breeding experiments between
the A . oeseri of Fernando Poo and the Mungo fish that we are calling by the same
name.
The name A. scheeli was given by Radda to an established aquarium strain of
unknown origin (illustrated also by Scheel, I968a, p. 140, as the ' Burundi- Aphyo-
semion'). It is very much like our species and if the latter should prove to be
different from the island form the name scheeli may be applicable, which would be
unfortunate in view of the long aquarium pedigree of the types.
Procatopus Boulenger, 1904
Boulenger, 1904 : 20. Type species P. nototaenia Boulenger.
The Aplocheilichthyinae are distinguished by having the pectoral fins attached at
about the middle of the height of the body and the pelvics below or not far behind
them. In the genus Procatopus the pelvics are in the most anterior position and the
branchiostegal membrane is produced in adult males beyond the edge of the oper-
culum as a bright yellow appendage.
Clausen (1959) divided Procatopus into two subgenera, the second, Andreasenius
Clausen, differing from the nominate subgenus in the more posterior position of the
pelvics, usually lower numbers of dorsal and anal fin-rays, a more slender body and
in mature males by the elongation of upper and sometimes lower caudal rays.
FISHES OF WEST CAMEROON 377
Besides the type species, P. aberrans Ahl, Clausen named five other species of this
subgenus in West Cameroon and Nigeria.
The nominal species of the subgenus Procatopus are :
P. nototaenia Boulenger, 1904 ; R. Lobi (Lobe), South Cameroun.
P. similis Ahl, 1927 ; Logobaba in the delta of R. Wouri.
P. abbreviatus Pellegrin, i929b ; Yabassi, R. Wouri.
P. glaucicaudis Clausen, 1959 ; an upper tributary of the Cross system at the road
from Kumba to Mamfe, 93 km north of Kumba.
After examining types of the named species, Clausen recognized P. similis and
assigned specimens from R. Kumba to it (one presented to BMNH). He noted that
numbers of scales and fin-rays are alike in all species and relied for specific distinctions
on the position of the pelvics and details of coloration.
Scheel (1970) was inclined to consider each subgenus to represent one species and
except in this way did not use the subgeneric concept. But in an appendix to the
same article he modified this view, at least as regards the number of species, as a
result of further collecting.
Populations sampled by us include two species of P. (Procatopus) and one agreeing
better with P. (Andreasenius) , but to some extent intermediate.
Our samples from R. Bille and the tributaries of the Mungo were collected by Dr
Disney mainly in narrow-mesh basket traps, but from R. Blackwater by hand-net.
Those from the inlet stream of Barombi Mbo were taken by Barombi women in
closely woven baskets. The Lake Soden population was sampled by means of a
butterfly net.
Method of measurement. Instead of the standard length Clausen used the 'post-
premaxillary length'. I use instead a standard length including (but for this genus
only) the projecting lower jaw.
In counting scales in a longitudinal series I begin with two above the pectoral fin-
base. This gives two more than Boulenger's (and Pellegrin's ?) counts, but agrees
with Clausen and Thys (1967).
Evaluation of specific characters
The position of the pelvics may be expressed in three ways :
(a) Clausen's 'pelvics index', i.e. the horizontal distance between pectoral and
pelvic bases as % distance from pelvic base to origin of anal fin. I have found these
measurements to have a low repeatability. The distance P-V is about 0-5-2-0 mm
measured as a projection and gives results that vary even with one operator on one
fish. Moreover, my measurement on the paratype of P. glaucicaudis in the BMNH
gives an index far outside the range recorded by Clausen for this species, and I am
evidently not measuring in the same way.
(b) The distance from the tip of the lower jaw to the pelvic ('pre-pelvic distance')
as % SL (Fig. 12). This was used by Poll & Lambert (1965) for Hypsopanchax and
is more reliable in my hands than (a) .
(c) The position of the pelvic origin relative to the lower ends of the ribs (Fig. n
and Table n). Because of the relative translucence of the tissues, even in alcohol,
378
E. TREWAVAS
dls
FISHES OF WEST CAMEROON
379
it is possible to see the ribs through the skin. The first rib is an exception, being
usually hidden by muscles ; in an alizarin preparation it can be seen emerging from
behind the cleithrum accompanied by the postcleithrum, a smaller bone. The
second slopes downwards and backwards immediately behind the pectoral base and
serves as a landmark when the first is obscure. It does not quite reach the ventral
surface and I have described the attachment of the pelvic as 'at 2nd rib' when the
anterior edge of the base of its anterior ray is at a point continuing the line of the rib.
TABLE n
Position of pelvics in populations of Procatopus
P. similis
R. Bille
Rivers Kumba, Mpanga and Kobe
and inflowing stream of
Barombi Mbo
P. nototaenia (syntypes)
P. glaucicaudis (paratype)
P. glaucicaudis ?
Kotto inlet stream
R. Black water
R. Menge
R. Wowe
P. lacustris (Lake Soden)
SL (mm)
Attachment of pelvics in
relation to ribs
24-5-30-2 Before ist
25-5-45-5 At or before ist
25'5~36 Between ist and 2nd to just behind 2nd
25'5 Just before 2nd
23, 28-5 Between 2nd and 3rd
I9~35'5 ist-2nd
34-39-5 At or just behind 2nd
29-34'5 Between ist and 2nd to between 2nd and
3rd
22~33*5 2nd to 4th, usually between 2nd and 3rd
The anterior position of the pelvics is unique among African cyprinodonts and it
is not surprising to find it farther back in the young. From R. Blackwater we have
some very small fry in which, at 10-5-12 mm SL, it is at a point 33-43% of the dis-
tance from pectoral to anal, i.e. Clausen's index is 50-75. The definitive position is
evidently reached quite early. In P. similis the pelvics are attached in front of the
first rib already at 18-5 mm SL.
Scheel (1970) did not use the position of the pelvics except as a distinction between
the two subgenera or, as he conceived it, between P. nototaenia and P. aberrans.
Perhaps he too found Clausen's index impractical.
Produced caudal rays in mature males characterize the populations of subgenus
Andreasenius, but this is a feature that may appear in only a few males of a collec-
tion even when they are intact (cf. Scheel, 1970) and the tips of the caudal rays often
become damaged in preserved specimens (cf. Clausen on P. plumosus, 1959 : 278,
our own collection from Lake Soden and the types of P. nototaenia) .
Scheel (1970 : 13) found in a tributary of R. Lobe, very near the type locality of
P. nototaenia, a population of Procatopus having the upper and sometimes lower caudal
rays elongated as streamers. Since the number of dorsal fin-rays in the sample
was also low he naturally expressed doubt as to whether P. nototaenia, or some of its
syntypes, might be more closely related to P. aberrans than to the populations of
380 E. TREWAVAS
similis
nototaenia
(acustris
28 30 32 34 36 38 40 42 44
FIG. 12. Procatopus. Ranges, means, standard deviations and standard errors of means
of prepelvic distance as % SL in P. similis (n = 26), P. nototaenia (n syntypes), P.
glaucicaudis (holotype and 2 paratypes), 'P. glaucicaudis ?' of Mungo system (n = 10),
P. lacustris (n = 14), 8 specimens from three populations in the neighbourhood of Mamfe
(P. nigromarginatus, P. roseipinnis, P. plumosus) and P. gracilis of western Nigeria
(n = 4).
Fernando Poo and southern Nigeria usually assigned to P. nototaenia. I have
examined the caudal fins of the 16 syntypes of P. nototaenia. Of the 14 males only
6 have the upper caudal rays intact ; in 4 of these the fin is truncate, in the other 2
the third principal ray from the top is very slightly longer than the next, but there
is nothing that could be called a 'streamer'. All these have the deep body and ad-
vanced pelvics characteristic of P. (Procatopus) as defined by Clausen (see Fig. 12)
and melanophores are absent from the dorsal streaks described as yellow by Boulenger.
Although the mean numbers of dorsal and anal rays are lower than in our populations
of P. similis (Table 12) they are higher than those found by Scheel in his Lobe
population (D 8-9, mean 8-3). Scheel does not mention the position of the pelvics,
but it seems possible that both subgenera (or superspecies) are represented in the
Lobe system.
Coloration. The general appearance of Procatopus is of a bluish translucence with
the vertical fins variously tinted yellow or orange with a delicate edging of deeper
orange (of which only the melanophore components remain in the preserved fish) ;
the dorsum is grey with paired orange streaks or rows of spots variously mixed with
or replaced by melanin ; sometimes also a median dorsal streak.
Clausen gave specific value to colour differences between populations of P.
(Andreasenius) , but Scheel, collecting later in the type localities, found that the
colours no longer corresponded to Clausen's descriptions and concluded that they
might change from generation to generation.
In contrast to Clausen's data notes on colour by ourselves and others are not
sufficiently systematic to give a measure of the variation within populations, still
less to give reliable contrasts between populations.
FISHES OF WEST CAMEROON 381
Iris. Types of P. glaucicaudis (from Clausen) - anterodorsally black, a gold spot
behind this, else silver. 'P. glaucicaudis ?' in R. Menge-red dorsally, else silver.
P. similis in R. Kumba - yellow dorsally.
Paired dorsal streaks or series of spots. P. nototaenia types - paired yellow and
median orange stripes (Boulenger) . P. glaucicaudis types - brownish or yellowish
grey streaks (Clausen). 'P. glaucicaudis ?' in R. Menge - orange colour masked by
melanin, seen only when looked for. 'P. glaucicaudis ?' in Kotto basin - more dark
pigment than in P. similis. P. similis in stream entering Barombi Mbo - conspicu-
ous orange spots ; in R. Kumba - orange spots mixed with melanophores ; in R.
Bille - orange spots present and between them a mid-dorsal orange streak. The
population in Lake Soden - grey streaks, median streak black.
Fins. Dorsal and anal fins are generally described as pale yellow or flesh
colour proximally, yellow or orange distally (our notes on R. Kumba and R.
Menge specimens, Clausen's on R. Kumba and on types of P. glaucicaudis).
Examined more closely, the pale proximal colour is produced by sparse
melanophores and yellow pigment cells mixed, the distal by denser yellow cells.
In a probably more intensely coloured P. similis from R. Kumba Clausen
described these fins as bright yellow proximally, bright orange distally. At the
other extreme we described the dorsal of a male from R. Menge ('P. glaucicaudis ?')
as pale greenish with orange margin, its anal as greenish yellow with red spots.
The anal and caudal often bear red or brownish spots. Scheel (1970) gave a wide
range of fin colour in samples from tributaries of R. Wouri, from almost colourless
through lemon, golden and orange to blood red. Scheel's journey extended from
the type locality of P. similis in the delta to that of P. abbreviatus at Yabassi and
evidently the Procatopus populations of that river system require further
analysis.
These examples show that it is risky to use colour for diagnosis until more
systematic studies have been made of it. No reliance can yet be placed on fin
colour, but on the evidence P. nototaenia and P. similis (as identified here) have
more yellow on the dorsum than P. aberrans and the species grouped with it as
Andreasenius and in this feature P. glaucicaudis and the Lake Soden species are
intermediate.
Numbers of fin-rays. The lowest ranges are those given by Ahl for P. aberrans -
D 6-8, A 13-15 - but Ahl was apt to under-count (see Holly, 1930) and I consider
here only Clausen's counts and my own. The main contrast is between the species
of P. (Andreasenius} on the one hand, in which they are D 7-9, mode 8, and A 14-17,
mode 15 (nearly approached by 16), and P. (Procatopus} on the other, with D 8-12,
mode 10 or n in different species and A 15-20, mode 17 or 18 in different species.
See Table 12. Between them the Lake Soden sample is intermediate with strong
modes of 10 for the dorsal and 16 for the anal.
Depth of body. On the whole, species of P. (Andreasenius} are more slender than
P. (Procatopus}, but this is a character showing positive allometry with the standard
length.
Chromosomes. Scheel (1970) states that in species from both sections of the genus
the karyotype appears to be identical, with n = 24.
382 E. TREWAVAS
TABLE 12
Frequencies of dorsal and anal fin-ray numbers in samples of populations of Procatopus
Dorsal Anal
8 9 10 ii 12 Mean SD 15 16 17 18 19 20 Mean SD
P. similis
R. Bille 2 I 10-33 °'58 2 ! I6'33 °'58
Rivers Kumba, Mpanga
and Kobe 288 11-33 0-69 37621 17-53 1-07
Barombi Mbo inlet stream 15 3 11-17 °*38 3 n 4 18-05 °'^4
P. nototaenia syntypes I 3 10 2 9-81 0-75 2562 16-53 °'84
P. glaucicaudis ?
Fernando Poo (Thys) i 14 7 10-27 °'55 81851 16-97 °'74
Kotto inlet stream 2 2
Rivers Menge, Wowe and
Blackwater 10 5 3 10-61 0-78 584 16-94 °'75
P. glaucicaudis types
(from Clausen) 651 10-58 0-67 1812 i7'33 0-89
P. lacustris 4 23 i 9-89 0-42 i 22 4 16-11 0-42
Sexual dimorphism
Our catches proved to contain very few females and the explanation is almost
certainly selectivity of gear acting on a size difference between the sexes. In Lake
Soden and the inlet stream of Barombi Mbo the mesh was finer and we got more
females. Even in P. similis, where males reached 45-5 mm SL our biggest female
was only 33 mm. This size difference was also noted by Clausen (1959).
I confirm the presence in mature males of the prolonged branchiostegal rays that
form a bright yellow ornament at the lower edge of the gill-cover. In P. similis this
was beginning to grow out at SL 25-5 mm. In our biggest male from Lake Soden
(34 mm SL) the appendage is still not as long as in even smaller P. glaucicaudis and
P. similis, and in smaller specimens the branchiostegal membrane projects only
slightly and almost evenly beyond the opercular bones, with a crenulate edge.
In mature males the posterior rays of dorsal and anal fins may be elongated to
form a short point, but the most striking sexual difference in these fins is the elonga-
tion of the base of the anal in males (Table 13). This brings the origin of the dorsal
over the middle of the anal in males and over the posterior rays in females (although
in none of ours is it over the last ray as in Thys's fig. 23 of 1967). The same dimor-
phism was described by Poll & Lambert (1965) in Hypsopanchax jobaerti and Hylo-
panchax silvestris, the latter having also a slight, but even prolongation of the
branchiostegal membrane, but only to the degree manifested by immature males
of Procatopus.
In mature males of all populations examined the pelvic fin is longer than in females
and, as Clausen described for P. glaucicaudis, it is the third ray that is longest or the
third with the outer fork of the fourth. In one of the females of the syntypes of
P. nototaenia the first ray is produced a little beyond all the others, which are sub-
equal. In females of P. similis this elongation of the first ray is exaggerated (Fig. 13).
FISHES OF WEST CAMEROON
383
FIG. 13. Procatopus. Pelvic fins of, A, $ syntype of P. nototaenia, SL 26-5 mm ; B,
$ (30 mm) and 6* (42 mm) of P. similis ; C, $ (31 mm) of P. lacustris.
TABLE 13
Sexual dimorphism in length of anal fin base in species of Procatopus
Species and
locality
P. lacustris
Lake Soden
P. glaucicaudis ?
R. Wo we
R. Menge
R. Blackwater
Kotto
P. nototaenia
R. Lobi (types)
P. similis
R. Kumba
R. Kobe
Feeder stream of
Barombi Mbo
R. Bille
fo.
of SL (mm)
Anal base
No. of
SL (mm)
Anal base
tt
(% SL)
?$
(% SL)
5
27-33-5
20-5-24-5
4
26-5-31
18-22
3
29-34-5
25-27-5
5
34-39-5
23*5-27
2
25-5, 32-5
23-5, 20
6
24-36
25-3-29
i
28-6
25-2
I
23
20
14
29-5-36
26-4-30-5
2
25-5, 32
I9-6, I9'4
13
38-5-45-5
27-5-3I-5
2
29-30-5
I9-I9-3
2
29, 35*5
27-5-30-2
I
25-5
25-5
21
25-6-42
(24-2)
6
22-5-33
19.3-24-9
25-3-29-7
2
24-5-30
27-30-4
I
26-5
22-6
Summary
P.
lacustris <$
20-5-24-5
$ 18-0-22
•o
P.
glaucicaudis <$
23-5-29-0
$ 20-0-23
•5
P.
nototaenia $
26-4-30-5
$ 19-4. J9
•6
P.
similis $
(24-2) 25-3-31-'
; $ 19-0-25
•5
384 E. TREWAVAS
Contrast with this is probably what Clausen had in mind when he described the pelvic
rays of P. glaucicaudis as decreasing in length gradually. In this the two females
that we have from R. Menge and the Lake Soden females agree with P. glaucicaudis.
This suggests that the shape of the pelvics is not only sexually dimorphic but also
characterizes females of P. similis and P. nototaenia vis-a-vis the other species, but
its value needs confirmation from more females of P. nototaenia and the populations
here included in P. glaucicaudis.
Conclusions on Procatopus of Cameroons
The types of P. similis, two of which I have examined, have the pelvics in the
extreme anterior position, before the first rib. The fins originate before or at the
first rib also in our specimens from Rivers Kumba, Mpanga, Kobe, the inlet stream
of Barombi Mbo and R. Bille, which are therefore identified as P. similis. They
provide the further distinguishing characters of conspicuous yellow spots or streaks
on the dorsum, high mean numbers of dorsal and anal fin-rays and an elongate first
pelvic ray in females, all of which have yet to be checked in the type locality.
The population of Lake Soden has the pelvics significantly farther back (Table n
and Fig. 12), no conspicuous yellow markings on the dorsum and lower numbers of
dorsal and anal rays. It is very similar to P. nigromarginatus Clausen and the other
populations of the Mamfe area, which Scheel suggests should be included in P.
aberrans. In comparison with these, its higher mean numbers of rays and its peculiar
ecology, together with its isolated position, justify a new name, although it may be
only a subspecies of P. aberrans.
The types of P. nototaenia and specimens identified here as P. glaucicaudis from
the upper Mungo tributaries are intermediate in the position of the pelvics and
numbers of fin-rays. They seem to differ from each other in the amount of yellow
on the dorsum and in the shape of the pelvics in females, but more data are required
on these features.
The population of Fernando Poo according to Thys's description seems to agree
with the Mungo P. glaucicaudis (position of pelvics, shape of pelvics in females,
absence of yellow on dorsum) as do a few specimens taken in the Kotto area.
Queries still hang over the population of the Wouri system from which Pellegrin
named P. abbreviatus and those of R. Lobe in which, as well as P. nototaenia, there is
according to Scheel (1970) a species resembling P. aberrans.
Procatopus similis Ahl, 1927
(Figs. 11-13, Tables 11-13)
Ahl, 1927 : 79 (Logobaba, eastern side of Wouri delta).
LECTOTYPE. SL 37 mm, Zool. Mus. Berlin 17772.
LECTOPARATYPE. SL 36 mm, Z.M.B. 22712. Both selected in 1965 by Clausen,
but selection not published. Examined by me in 1972.
FISHES OF WEST CAMEROON 385
Our collections comprise 56 specimens, 18-5-45-5 mm in SL, from the inlet stream
of Barombi Mbo, from R. Kumba and its tributary Mpanga Water, R. Kobe (all
Mungo system) and R. Bille (Meme system). BMNH 1973.5.14.508-548.
The details of the following description are taken from a varying number of these
as the variation seemed to demand. The types are included in the ranges but not in
the means.
Pelvics attached below pectoral base, at or (usually) in front of first rib ; pre-pelvic
distance 27-5-36% SL (Table n, Figs, n, 12). Depth of body 29-40% SL, of
caudal peduncle 17-2 1. Dorsal fin-rays 10-12, mode n,mean 11-15 or 11-3 in samples
from different populations. Anal 16-19 (20), mode 18, mean 17-5 or 18-1. Scales
29 or 30 in a longitudinal series, 12 around caudal peduncle. Outer ray of pelvic in
female produced to vent or origin of anal, other rays remaining short (Fig. 13).
Paired row of orange spots on the back conspicuous, in some a median orange
streak between them.
The specimen with 20 anal rays has the 20th ray double, consisting of an anterior
branch weakly forked and a posterior simple branch, the two fused or closely applied
at the base. The well-defined modes in fin-ray numbers in the sample from the inlet
stream of Barombi Mbo (Table 12) are the mark of a restricted and isolated popula-
tion.
ECOLOGY. Inhabits running water and feeds on insects, taken mainly from the
surface but also from the bottom (Green & Corbet in Trewavas et al., 1972).
There is no true contrast with P. nototaenia in Ahl's description, but the examined
types have the pelvics originating in front of the first rib. The fin-ray numbers are
equivocal (D 10, n ; A 17, 18).
Procatopus nototaenia Boulenger, 1904
(Fig. 12, Tables 11-13)
Boulenger, 1904 : 20 (R. Lobi, South Cameroun) ; id. 1915 : 79, fig. 66 (Lobi, Kribi and Cross
Rivers) .
DESCRIPTION OF THE SYNTYPES. 14 &£ and 2 £$, 25-5-36-0 mm in SL.
Pelvics attached just behind pectoral base, from behind first rib to behind second ;
prepelvic distance 32-0-37-5% SL (Table n). Depth of body 31-34% SL, of caudal
peduncle 18-4-20-8. Dorsal fin-rays 8-u, mode 10, mean 9-8 ; anal 15-18, mean
16-5. Scales 28 or 29 in a longitudinal series, 12 around caudal peduncle. Outer
ray of pelvic in (one) female produced a little beyond the others (Fig. 13). Paired
yellow and median orange stripes on the back conspicuous (Boulenger).
Boulenger does not state whether he relied on the collector's field notes for colour
or on the condition of the preserved fishes. In the position corresponding to the
orange spots of P. similis there is a pair of melanin-free strips. Of the two female
syntypes only the smaller has undamaged pelvics.
The four Kribi specimens recorded by Boulenger are all males, 27-34 mm in SL.
The pelvics are attached between the first and second ribs or just behind ; the pre-
pelvic distance is 32-0-33-5% SL. The few specimens in the BMNH from the Cross
386 E. TREWAVAS
River system and lower Niger are also males, so that without information on colour
or shape of female pelvics there is no reason to remove them from P. nototaenia.
Clausen (1959) included Nigerian specimens in this species, and Scheel (1970) pub-
lished a photograph of one under this name.
Procatopus glaucicaudis
Stenholt-Clausen, 1959 : 268, figs, i, 2 and Table 2 (stream of Cross River system, 93 km north
of Kumba) ; Radda, 1970 : 239 (stream near Baduma, upper Mungo system).
DESCRIPTION OF 16 SPECIMENS (14 <&£, 2 $$). 24-0-39-5 mm in SL from Rivers
Menge, Wowe and Blackwater, tributaries of the Mungo. BMNH 1973.5.14.429-444.
Pelvics attached behind pectoral base, behind first rib to between second and
third ; pre-pelvic distance 32-38% SL (Table n). Depth of body 26-4-38-0% SL,
of caudal peduncle 16-3-20-3. Dorsal fin-rays 10-12, mode 10, mean 10-6 ; anal
16-19, mode equally 16 or 17, mean 16-94. Scales in a longitudinal series 28-30,
around caudal peduncle 12. Lower jaw with a slight mental process. Females
with pelvic rays decreasing gradually in length from outer to inner.
Yellow tinge in the dorsal streaks inconspicuous, masked by melanin.
Males reaching 39-5 mm, females 32-6 mm in SL.
Eleven young of 7-2-19-0 mm were collected with the adults in R. Blackwater.
At lengths up to 12 mm the pelvics are situated more posteriorly, but at 19 mm
they are already below the second rib. Two of 16 and 19 mm have been included
for fin-rays in Table 12.
Radda (1970) identified specimens from a stream near Baduma (probably from the
Wowe or Moliba) as P. glaucicaudis. The type locality is about 48 km farther
North over the watershed separating the Mungo from the Cross system.
The reasons for including our specimens in P. glaucicaudis and for suggesting that
the two from the Kotto basin and those of Fernando Poo recorded tentatively by
Thys (1967) as P. nototaenia may also belong to this species are given on p. 384.
Conversion of Clausen's ratios to % SL as used by me give lower but overlapping
ranges for depth of body and of caudal peduncle (respectively 24-0-32-5 and 13-7-
16-8), but Poul Winther's illustrations (Clausen, figs, i, 2) could serve also for our
samples.
The assignment here of the various samples to P. nototaenia or P. glaucicaudis
gives an impression of the interruption of the distribution of P. nototaenia and the
singularity of the fauna of Fernando Poo -Mungo that is hardly justified in view of
the uncertainty expressed above, but the specific identity of the Fernando Poo and
Mungo populations (excluding P. similis) seems reasonably certain and is in line with
decisions about other species.
Procatopus lacustris n. sp.
(Figs. 11-13, Tables 11-13)
HOLOTYPE. # of 33-5 mm SL ; allotype $ of 30 mm ; BMNH 1973.5.14.456-7,
from Lake Soden.
FISHES OF WEST CAMEROON 387
PARATYPES. 16 $$, 15 $$ and 19 sex indet., 17-5-33-0 mm in SL, also from Lake
Soden, 1973.5.14.458-507, collected by Drs Green, Corbet and Griffith in April,
1970.
Pelvics attached behind pectoral base, from second rib to behind third or even (in
specimens of 17-21-5 mm SL) at the fourth ; pre-pelvic distance 36-0-41-5% SL.
Depth of body 26-2-29-0% SL, of caudal peduncle 15-9-18-8. Dorsal fin-rays 9-11,
mode 10, mean 9-9 ; anal 15-17, mode 16, mean 16-1. Scales 29-31 in a longitudinal
series, 12 around caudal peduncle. Pelvic rays in female gradually decreasing in
length from outer to inner. Process of branchiostegal membrane of male not
reaching full development below SL 34 mm, and then less prolonged than in other
species.
No conspicuous orange or yellow spots in the dorsal streaks, the median of which
is grey.
The mouth is set at a slightly steeper angle than in other species and there is a
slight mental process.
Males reach 34 mm SL, females 31 mm.
The food consists of planktonic Crustacea (Green, 1972 : 298 ; Green & Corbet,
in preparation).
Known only from Lake Soden.
As can be seen from Fig. 12, the position of the pelvics in this species is more like
that in the subgenus Andreasenius than in other species of P. (Procatopus). Un-
fortunately the caudal fins of all our specimens except two were damaged. In the
two it is truncate, but we cannot be sure that in mature males the corners may not
grow out as streamers. If that were so it would conform to the definition of sub-
genus Andreasenius, although the number of dorsal fin-rays is higher than in other
species.
The existence of a Procatopus in Lake Soden was already known to Dr Thys van
den Audenaerde before our visit.
CICHLIDAE
Hemichromis elongatus (Guichenot, 1859)
Chromichthys elongatus Guichenot in Dumeril, 1859 (Gabon).
Hemichromis fasciatus form B, Burchard & Wickler, 1965 ; Burchard, in Reed et al., 1967 : 131 ;
Corbet et al., 1973 : 314.
This species was not caught by us in the upper Mungo or its tributaries, the only
specimen recorded in the Kumba area being one from the Bille, an upper tributary
of the Meme.
In Barombi Kotto and Mboandong it was one of the commonest species around
the edges of the lakes and it was caught not only in Nganjoke, the outlet stream of
Kotto (Meme system) but also in R. Nyoke, the stream near Kotto that is a tributary
of the Mungo. It was included in Mr Deveson's collection from near Tiko, 1951-2.
The characteristic five black blotches along the flanks were faintly prolonged to
the dorsum. An intense black opercular spot had an iridescent red mark above and
below it. The lower parts of the head and belly were flushed red in most individuals
388 -E. TREWAVAS
and the scales of the flank each shone with a green iridescent spot. The dorsal
lappets, upper edge of soft dorsal and posterodorsal corner of the caudal were pale
with a red edge. The iris was black or grey, in some with a red circum-pupillar ring.
The metallic green spots on the scale rows are characteristic of the 'B form' of
H. fasciatus (Burchard & Wickler, 1965). Mr Loiselle informs me (in litt.} that this
is sufficiently evident in preserved specimens to allow him to use for it the name
elongatus, until now included in the synonymy of H . fasciatus Peters, on the ground
that known specimens from the broad type locality of C. elongatus ('Gabon') belong
to the B form, although the type specimen, preserved for over a century, no longer
shows anything of the diagnostic colour-features. Mr Loiselle, who has a paper in
preparation detailing his evidence, also noted a second colour-pattern criterion,
applicable only to specimens of 100 mm SL or more. I have applied these tests to
numerous samples in the BMNH, including some known to belong to the 'B form',
and I find considerable support for Loiselle's claim.
On the same evidence the Kotto population also appears to be H. elongatus,
although the essential test of breeding coloration and behaviour has yet to be applied.
Both Dr P. H. Greenwood and I have spent considerable time trying to find
structural differences between the two species, without success.
The smallest ripe female recorded from Kotto was 60 mm in SL, one from Mboan-
dong 52 mm. The genital papilla of the male is surrounded by a small scalloped
fringe.
Young of about 16 mm SL are distinguishable from young Tilapia and Sarothero-
don by the bigger mouth, thicker lips, minute, sharply pointed unicuspid teeth, the
presence of two series of vague melanophore concentrations on each side, one along
the dorsum, the other along the middle of the flank (not continuous stripes like those
figured by Gosse (1963) for H. fasciatus and also present in some young from Sierra
Leone), and the absence of a tilapia-mark.
The ecology is described by Corbet et al. (1973).
MATERIAL PRESERVED
BMNH register no. SL (mm) Locality
1969.1.13.12-15 28-29-5 Kotto, coll. C. A. Wright
1973.7.18.106-109 52-79 Mboandong, 1970
1973.7.18.110 105 R. Bille, 1970
1973.7.18.111-113 76-5, 61, 15 Kotto, 1972
CHROMIDOTILAPIA Boulenger, 1898
Boulenger, 1898 : 151. Type species C. kingsleyae Boulenger.
This genus, formerly merged in Pelmatochromis Steindachner by Pellegrin (1904),
Boulenger (1915) and Regan (1922), has been revalidated as a subgenus by Thys
(ig68a) for the species with a hanging pad in the pharynx and no microbranchiospines
(but excluding some small species with a large ocellar spot on the caudal and a sexual
differentiation in the length of the pelvic fins). It is in this sense, but with full
FISHES OF WEST CAMEROON 389
generic rank as implied later by Thys (ig68b), that I now use it for the species in
the region under review.
Feeding mechanism : contrast with Tilapia
The characteristic hanging (or boot-shaped) pad (Fig. 14) is a hyperdevelopment of
the glandular and sensory pad present in most (all ?) cichlid fishes on the roof of the
pharynx anterolaterally to the upper pharyngeal tooth-plates. The ankle of the
boot on its anterolateral side is delimited by a groove with unspecialized epithelium,
above and behind which lie the gill-rakers of the first epibranchial, which in their
simple slender shape are sharply differentiated from the swollen and tuberculate
ceratobranchial rakers (Fig. 16). The pad itself is filled with rather loose, fatty
connective tissue, rich in nerves and capillaries, in the midst of which (as also in
cichlids with a sessile pad) is a flange of bone and cartilage that is an extension of the
second epibranchial (see Trewavas, 1973). Sections show (PI. 3 Figs, a and b) that
the hillocks of the pad are richly supplied with taste buds, while scattered at the
surface of the stratified epithelium are small spherical mucus cells ('beaker cells',
but they have no stalk) staining sky blue with Masson's trichrome. Sections of the
ceratobranchial of the first gill-arch reveal that its rakers and tubercles are also
thickly set with taste buds distally while the epithelium of their proximal parts
includes some small mucus-cells. There are tubercles on the roof of the buccopharynx
and the inner side of the suspensorium as well as on the rakers, probably also bearing
taste buds.
The sessile pads of Tilapia mariae and Sarotherodon melanotheron (PI. 3 Fig. c and
d) contrast histologically with the hanging pad. In them there are far fewer taste
buds, the small beaker cells are more plentiful, and the main constituents of the
epithelium are tubular secretory cells that extend through the depth of the stratified
epithelium (T. mariae) or nearly (S. melanotheron). Their secretion occupies nearly
the whole cell and takes very little stain, appearing very pale blue with Masson ;
their nuclei are basal.
Such cells are completely absent from the hanging pad of C. guntheri, but similar
long and abundant gland cells were described by Imhof (1935) in blennies that graze
on the algal film of underwater surfaces in contrast to carnivorous members of the
same or related genera. Rauther (in Rauther & Leiner, 1937), who quoted Imhof,
himself found the same correlation between the presence of such cells and a diet of
unicellular or filamentous algae, instancing Plecostomus and Garra and the pharyngeal
sac of Curimatus.
Fishelson (1966, figs 9-13) found both sensory and secretory cells in the mouth,
pharynx and oesophagus of Tilapia tholloni and T. macrocephala ( = S. melanotheron)
at the larval stage ; at this stage the epithelium is not so many-layered, but the size
and abundance of the gland cells show that they are the precursors of the deep cells
of the adult.
Stolk (1957) described columnar secretory cells in the pharyngeal pads of Pseudo-
crenilabrus multicolor [= Haplochromis multicolor (Schoeller)] . In this case he was
looking for evidence of a secretion relevant to the survival of the mouth-brooded
young, but their similarity to the cells in species of Tilapia that do not practise oral
390
E. TREWAVAS
FIG. 14. Chromidotilapia guntheri, side view of head with gill-cover cut away to show the
boot-shaped 'hanging pad' and the differently shaped gill-rakers on the epi- and cerato-
branchial of the first arch. From Irvine & Trewavas, 1947.
muscle--
upper
pharyngeal
teeth
gill-cover--
b"
FIG. 15. Roof of mouth and pharynx in, A, Chromidotilapia guntheri and, B, Sarotherodon
melanotheron, to show the upper pharyngeal teeth and different development of the pads
and grooves on the roof of the pharynx.
FISHES OF WEST CAMEROON 391
incubation suggests that their function may be connected with feeding too. This
species has the dental equipment of an omnivore. What substances are secreted is
still unknown.
Similar secretory cells have been described in the pharynx of microphagous
anuran tadpoles by Kenny (1969), who refers to concentrations of such cells as 'food-
traps'. Greenwood (1953) recognized the same function for mucous secretions in the
pharynx of Sarotherodon esculentus (Tilapia esculenta) .
The series of epibranchial rakers in Chromidotilapia is long in comparison with
that of most cichlids and in C. guntheri they may exceed in number those of the
ceratobranchial. The frenum at the hinge between suspensorium and first gill-arch
is at the upper end of the epibranchial, leaving a free passage between the groove
and the upper part of the parabranchial chamber. The groove pinching off the
hanging pad is continued around the front of the pad transversely to the midline
and is sharply demarcated in front (Fig. I5A.
Elements of such an epibranchial arrangement are found in some species of
Pelmatochromis in which the pad is sessile (see Trewavas, 1973), but in these the
histology of the pad is unknown. The complete mechanism, including the hanging
pad, is found in several cichlid genera not very closely related to Chromidotilapia.
Examples are Tylochromis (West Africa, Congo basin and Lake Tanganyika),
Callochromis (Lake Tanganyika) and Geophagus (South America). In Geophagus
jurupari figured by Pellegrin (1904 : 104 fig. 27) the gill-rakers are on the edge of the
pad instead of close to the epibranchial and are of a specialized structure,* but in
G. Pellegrini Regan they closely resemble those of Chromidotilapia both in structure
and position and there are intermediate conditions in other species. The species
in which this mechanism is developed are all bottom feeders and typically have the
mouth set low at the end of a long snout with a deep preorbital bone. The feeding
method of Chromidotilapia guntheri is described by Burchard (in Reed et al. 1967)
as a vacuum-cleaner-like action and that of Geophagus by Sterba (1962 : 712) as a
continual grubbing with its snout in the sand. The findings of Poll (1956 : 226ff.)
on the food of Callochromis and those of Corbet et al. (1973) on Chromidotilapia
g. loennbergi during our expedition confirm that the food is taken from the
bottom.
It is plain that the arrangement of pad, grooves and epibranchial must affect the
currents at the back of the pharynx, causing them to be different from those in
fishes with a sessile pad, no transverse groove on the roof, a short series of blunt
epibranchial rakers and the frenum near the epi-ceratobranchial joint.
Several species of the genus Lethrinops of Lake Malawi feed, as C. g. loennbergi
does, on chironomid larvae. They too have a low-set mouth, long snout and deep
preorbital, giving the same physiognomy as in Chromidotilapia, but their epibranchial
arrangements are unmodified from the pattern usual in Haplochromis, to which they
are related. A diagram of their feeding movements published by Fryer (1959) and
Fryer & lies (1972) shows the snout burrowing into the sand and finally the sand
being shot out at the branchial opening.
* Macroscopically they look like batteries of taste buds.
392
E. TREWAVAS
FIG. 16. Chromidotilapia. Gill-rakers of first arch in, A, C. guntheri of SL 69-5 mm from
R. Menge and, B, C. finleyi of 76-5 mm from R. Blackwater. With each an oral view
of part of the ceratobranchial. The broken line is the outline of the gill.
It may be that the hanging pad mechanism is a more efficient means of attaining
this end, and the unusually plentiful supply of taste buds points to the selection at
this place of items of food value from the general intake. The shape of the groove
around the stem of the pad aird the nature of the epibranchial gill-rakers (capable of
a degree of filtering) suggest the possibility of re-entry of part of the respiratory and
food-bearing current from the parabranchial chamber into the pharynx, but there
are difficulties in reconciling this with the known bucco-pharyngeal movements in
more normal perciform fishes (cf. Osse, 1969). Clarification must await investigation
by experimental techniques.
The lips of C. g. loennbergi are thick and broad and the teeth are frequently em-
bedded in the gums with the crowns worn off. Sectioned, the lower lip is found to
have a stratified epithelium containing here and there a sensory bud, and with the
cells of three outer layers flattened in a plane parallel to the surface. The derrm's
has a thick outer layer of collagen fibres and a tough fibrous connective tissue con-
taining a bar of hyaline cartilage sheathed in fibrocartilage. Such thick lips occur
in other populations of this species and must be an effective bull-dozing tool.
Species in West Cameroon
In the Mungo tributaries and the outlet stream of Barombi Kotto two species of
Chromidotilapia were found, C. guntheri (Sauvage) and a species related to C.
batesii (Boulenger). I suggest that C. loennbergi (Trewavas) of Lake Kotto should
be regarded as a subspecies of C. guntheri and I describe the other species as new.
FISHES OF WEST CAMEROON
393
C. guntheri shows the most advanced development of the feeding mechanism out-
lined above and the other species belongs to the least specialized part of the genus,
having a shorter snout and preorbital, more teeth in the jaws, fewer epibranchial gill-
rakers and fewer tubercles on the ceratobranchial rakers (Fig. 16). It also has a
lower modal number of vertebrae (25 or 26, mode 26 ; cf. 26 or 27, mode 27, in C.
guntheri) and some of the scales have a gothic arrangement of the circuli (roman in
C. guntheri), both of which I interpret as relatively primitive characters.
Chromidotilapia finleyi n. sp.
(PI. 5 and Fig. 16)
? Pelmatochromis nigrofasciatus (nee Pellegrin, part) ; Boulenger, 1915 : 393, nos. 1-4, 8, 9, 17.
? Pelmatochromis batesii (nee Boulenger, 1901, part) ; Thys van den Audenaerde, 1967 : 91 fig. 28
(Fernando Poo) ; Roman, 1971 : 225 fig. 98 and pi. on p. 285.
HOLOTYPE. $ 73 mm SL, R. Nyoke, tributary of Mungo, SE of Lake Kotto.
BMNH I973-5-M-553-
PARATYPES. All collected during the expedition of 1970 or later in 1970 by Dr
Disney.
SL (mm)
74
29-44
63
BMNH register no. Sex
1973.5.14.554 ?
1973.5.14.568-71 juv.
I973-5-I4-555 ?
Locality
R. Nganjoke, outlet of Kotto
Stream W of Kotto
R. Mungo, between Baduma and
1973.5.14.573-576 indet. 32-5-71
1973.5.14.556-562 2 juv.,
i ?, 4 3$ 50-75
1973.5.14.563-567 I&I?,
3 indet. 43-5-62-5
1973.5.14.572 juv. 31-5
Bolo
R. Menge, tributary of Mungo
R. Blackwater, tributary of Mungo
R. Wowe, tributary of Mungo
R. Kobe, tributary of Mungo at
Ikiliwindi
Named in honour of Mr Lee Finley, who generously provided the photograph of
PL 5-
DESCRIPTION. Proportions as % SL (holotype in brackets). Depth of body
36-40 (40), length of head 35-6-39-8 (37-8), length of pectoral fin 24-28 (24), length
of caudal peduncle 10-5-16-5 (12-0), o-6-i-o times its depth (0-75).
Proportions as % length of head. Length of snout 34-4-41-0 (37-5), diameter of
eye 25-0-34-0 (25-8), depth of preorbital 18-1-21-8 (20-6), interorbital width 24-2-
27-3 (26-6), length of lower jaw 36-6-39-0 (37-4), of upper jaw 31-4-34-6 (33-7).
Cleft of mouth at low angle with the horizontal, 10-20 degrees, exceptionally
30 degrees. Maxilla reaching vertical from front edge of eye. Gill-rakers on first
arch (3-6) + (o-i) + (8-io), (5 + 1 + 10 in holotype). Lower pharyngeal bone
arrowhead shaped, the blade about half as long as the median length of the toothed
area, the teeth pointed, firm, not crowded.
394 E. TREWAVAS
Scales some with gothic arrangement of circuli, some roman ; 26-28 in the lateral
line series, 3-3^ (3!) between origin of dorsal and lateral line, 16 around caudal
peduncle ; 3 or 4 series on cheek.
Vertebrae 25 (f.2) or 26 (£.4).
Dorsal XV 9-10 or XVI 9, usually (including holotype) XV 10, total 24 in 3,
25 in 21.
Anal III 7-8, mode III 7.
Pelvic with first two (or i|) soft rays produced as a filament or streamer, in adults
of both sexes beyond origin of anal.
Caudal bluntly rounded or truncate with rounded corners.
Colour in life of holotype, a $ with ovarian eggs of 2-5 mm long diameter : lower
flanks pink, lower parts of head with green and blue iridescence, lower lip white,
opercular spot black ; spinous dorsal iridescent silver with yellow lappets and no
black spots ; caudal without spots, its upper half yellow with some silver iridescence,
lower half dusky ; pelvic with first i| soft rays grey and some blue iridescence on the
inner rays.
Three immature from a stream near Kotto, 39-46 mm in SL, were dull green-grey
in colour, with the dorsal lappets and upper edge of soft dorsal and caudal edged with
red and with two or three series of green spots on the caudal. Preserved males
from R. Wowe have no caudal spots, but those from R. Blackwater have them, dark
grey as preserved.
Some, but not all, young of 31-39 mm SL have a dark round mark at the base of
the anterior part of the soft dorsal, simulating a tilapia-mark.
Preserved specimens have nearly uniform countershaded ground colour, some with
very faint and irregular indications of vertical bars below the dark back. Some that
were pale straw-coloured when alive are uniformly whitish as preserved. In one or
two there are a few oval clear spots on the soft dorsal and caudal.
The colours shown in Mr Finley's photographs (PI. 5) agree very well with this
description. Another photograph, not used here, illustrates a pale colour with two
black horizontal bands said to be characteristic of a fish experiencing fear. Mr
Finley, who succeeded in breeding the species in an aquarium, tells me (in lift.} that
the tilapia-mark may be present in both sexes.
ECOLOGY. C. finleyi appears to be confined to clear running water ; we did not
catch it in the lakes nor at the Mungo bridge where the water is brown, silt-laden and
set with sandbanks. This is in contrast to C. guntheri which is found in both types
of habitat.
AFFINITIES. Table 14 shows a comparison between C. finleyi, C. batesii as now
restricted and certain samples from R. Kribi and Fernando Poo now considered to
be nearer to C. finleyi than to C. batesii.
(i) Under C. batesii I consider only the three syntypes, another specimen from R.
Benito and six of the Kribi specimens included by Boulenger in P. nigrofasciatus
under the impression that P. batesii was its junior synonym. Two of the syntypes
and two of the Kribi specimens are mature males and these have small dark spots
on the caudal fin and on the posterior rays of the dorsal and anal (see Boulenger,
FISHES OF WEST CAMEROON
395
TABLE 14
Chromidotilapia batesii, C. finleyi and related populations
N
SL (mm)
Depth (% SL)
Head (% SL)
Pectoral (% SL)
Caudal (% SL)
Preorbital (% head)
Interorbital (% head)
l.j. (% head)
u.j. (% head)
Vertebrae 25
26
27
Fin-rays
D modal formula :
Totals 24
25
26
Anal (soft)
7
8
Scales 1.1.
Gill-rakers
Upper
Lower
C. batesii
Benito
4
56-93
36-39
34-5-37
23-5-26-5
30-3I-5
21-5-22-5
26-28
32-5-35-5
3
i
XV 10
3
i
4
26-28
3-5
10
C. batesii ?
Kribi
6
61-98
37-40
34-5-37-5
25-26-5
32-5-35 (<J)
3I-3I-5 (??)
20-5-25
27-5-3I-5
37-40 (43)
3I-5-36-5
i
2
2
XV 10
3
3
3
3
26-28
4-5
9-n
C. finleyi
C. finleyi ?
Mungo and
Fernando
Kribi
Meme
Poo
18 ( + 3)
5
9
43-74-5
42-86
68-5-90
36-40
35-5-39
37-5-41-5
35-5-40
36-39
36-5-40
24-26 (28)
22-25
22-5-29
28-5-31-5
28-32
(27-5) 30-33
18-23
20-22
19-21-5
24-27-5
22-24-5
24-27
36-5-42-0
38-39
38-5-43-5
31-5-36-0
33-34-5
35-5-40
2
-
i
4
2
6
—
3
-
XV 10
XV 10
XV 9
3
-
6
21
5
3
5
26-28
3-6
8-10
5
27, 28
3-5
9-n
8
i
26 or 27
5-7
8-n
For each sample are given the Standard Length (SL) ; number of specimens used (N) (for C. finleyi
three small specimens are used for meristic characters only) ; proportions as % SL ; depth of preorbital,
interorbital width and lengths of lower (l.j.) and upper (u.j.) jaws as % length of head; and certain
meristic characters counted in the usual way. For the gill-rakers of the first arch one at the epi-cerato-
branchial joint is omitted.
1915 fig. 265). The preserved marking of the females, all considerably bleached,
consists of a vague dark blotch at the base of the anterior part of the soft dorsal.
The characteristic grouping of the spots in males is not found in C. finleyi (nor in the
Kribi specimens now omitted from C. batesii) and resembles the pattern in C.
kingsleyae and C. caudifasciatus. C. batesii is close to the latter, which differs from
it mainly in the more elongate body (depth of body = length of head = 33-3% SL ;
cf. in C. batesii depth 36-40, head 34-5-37-5% SL) and lower number of gill-rakers
(7-8 on lower part of anterior arch).
In addition to the difference in colour-pattern, the numbers of vertebrae and dorsal
and anal fin-rays have lower means in C. finleyi than in C. batesii (Table 15), although
samples are too small for confidence in the significance of the differences. Likewise
a different but overlapping range of relative length of head must be considered in
396
E. TREWAVAS
FIG. 17. Chromidotilapia sp. from R. Kribi, BMNH 1908.5.30.182 and 187. Above (J,
below $ to show contrast in preserved colour pattern. A damaged caudal fin of another
cj is shown for variation in pattern.
relation to the fact that most of our specimens of C. finleyi are smaller than most of
those available of C. batesii.
(ii) Other Kribi and Efulen specimens include mature males with a colour-pattern
different from both C. batesii and the Mungo-Meme C. finleyi, characterized by the
presence as preserved of white oval spots narrowly outlined in black on dorsal and
caudal fins (Fig. 17). Vertical dark bars on the body are conspicuous in both sexes.
Whether this pattern and the life colours can be included in the variation of C.
finleyi must await further study of both populations. Meristic characters are nearer
to those of C. finleyi than C. batesii, but among the proportions the size of the mouth,
FISHES OF WEST CAMEROON 397
as expressed in the length of both jaws, is greater than in either and the cleft is
usually at a higher angle with the horizontal (20-40 degrees, cf. 10-30 degrees in
C. finleyi, 20-25 degrees in C. batesii).
(iii) The Fernando Poo specimens listed by Boulenger (1915) are not well enough
preserved to determine the sex either from viscera or papilla. They have irregular
black markings on the posterior part of the spinous dorsal fin and on the soft dorsal.
Those on the male figured by Thys (1967) are a more regular variant of this pattern.
In proportions they are similar to C. finleyi and have not the large jaws of mature
Kribi specimens. Unlike these or the Mungo fishes sampled they include some with
27 vertebrae. I assign them provisionally to C. finleyi.
(iv) C. schoutedeni (Poll & Thys, 1967), from clear running water in the tributaries
of the Congo at Yangambi and of the Lualaba, seems to belong to this group within
the genus, and elements of the colour-pattern are reminiscent of the Fernando Poo
samples. It differs from the other populations considered in having 29 scales in the
lateral line series and the lower part of the cheek naked below 3 series of small
scales.
Finally, I conclude that there are populations in the Mungo and Meme, in Fernando
Poo and R. Kribi related to but specifically distinct from C. batesii. Whether they
are conspecific with each other or must be subdivided can only be decided after
further work on living specimens.
If they are conspecific it is evident that in the Kribi system the areas of C. finleyi
and C. batesii overlap and this may be true also of R. Ntem. The photograph of
'C. batesii' reproduced by Roman (1971) in colour (p. 285) and monochrome (fig. 98,
p. 225) does not show the spot pattern characteristic of the vertical fins of male
C. batesii (see Boulenger, 1915, fig. 265). The sex is not given ; it may be $ although
the silver band on the spinous dorsal is less striking than is typical in Chromidotilapia.
As in the holotype of C. finleyi the upper half of the caudal fin is yellow. Vertical
bands on the body are as well marked as in the Kribi C. finleyi (Fig. 17). I suggest,
therefore, that Roman's figure represents C. finleyi and not C. batesii. Alternatively,
if it is C. batesii then the females of these two species are more alike than the males.
Males with spotted vertical fins like those of the types of C. batesii were not en-
countered in the Mungo.
Evidence from Scheel's chromosome studies
After most of the above was written I read Scheel's (i972b) report that he found
two species in the neighbourhood of Kribi corresponding to the definition of C. batesii
and very difficult to distinguish on preserved material. One of them, the 'northern
form', has a chromosome number (zri) of 42, the other 44, although there are 80
chromosome arms in both. He states that they are different in reproductive be-
haviour and colour in life but gives no details. The photograph of 'C. batesii' that
he reproduces is the black-and-white one of Roman (1971).
Scheel's observation confirms my distinction of two Kribi forms, but leaves un-
resolved the questions of the identity of one of them with the Mungo species and of the
full nature of the distinction from C. batesii.
398 E. TREWAVAS
Chromidotilapia guntheri
Hemichromis guntheri Sauvage, 1882 : 317 pi. v fig. i (Assini, Ivory Coast).
For synonyms and comments on the synonymy see Trewavas, 1962 : 162.
DISTINGUISHING CHARACTERS. Snout long, 36-7-46-0% length of head (in speci-
mens of SL 42-5-91-5 mm), depth of preorbital 21-8-27-7%. Teeth in adult in 2
series, occasionally a third, the 6-8 anterior of lower jaw stronger than those of the
rest of the outer series and usually set forwards at a greater distance from the inner
row. Gill-rakers on the ceratobranchial tuberculate, those of outer series of first
arch with 3-6 tubercles. Vertebrae 26-27, mode 27. Scales 28, rarely 29 in lateral
line series, 3^-4 from origin of dorsal to lateral line. Female with a broad silver-
bronze band on the spinous dorsal fin, extending on to anterior soft rays and bearing
small dark shapeless spots in 1-3 irregular rows or irregularly arranged ; and with a
few very faint spots or none on the caudal. Leading edge of pelvic light iridescent
blue in $, white in <£.
The snout and preorbital measurements are still higher in bigger specimens.
Comparison of populations (Tables 15 and 16)
Both the Mungo population and C. g. loennbergi agree with this diagnosis, but they
differ from each other and from samples from Ghana and Nigeria as follows.
Ghana and Nigeria
Interorbital width (at SL 68-93 mm) 25-28% length of head. Gill-rakers on
first epibranchial 7-10 (8-13 according to Loiselle, in the press). Total rays in
dorsal fin 25-28, mode 27, moclal formula XVI u.
Upper Mungo and tributaries
Interorbital width (at SL 64-5-91-5 mm) 24-5-28% length of head. Gill-rakers
on first epibranchial 7-11, usually 9-10. Total rays in dorsal fin 24-27, mode 25,
modal formula XV 10.
L. Kotto and its outlet R. Nganjoke (C. g. loennbergi)
Interorbital width (at SL 73-90 mm) 21-2-26-0% length of head. Gill-rakers on
first epibranchial 5-8. Total dorsal rays 25-27, mode 26, modal formula XVI 10.
The numbers of specimens on which the meristic comparisons are based can be
seen in Tables 15 and 16.
This comparison shows that the Mungo samples differ from those from western
localities in having fewer dorsal rays and that C. g. loennbergi differs from both in the
lower number of epibranchial gill-rakers as well as a lower but overlapping range for
the interorbital width.
C. g. guntheri in the Mungo system
Our collection comprises 44 specimens caught in traps at five localities in the upper
Mungo system, but 20 of these were taken at one time in R. Menge and 5 others came
from the same locality at a later date ; 18 of the 23 specimens giving the modal
formula of XV 10 were from these samples. This may therefore give a local and
FISHES OF WEST CAMEROON 399
TABLE 15
Upper (epibranchial) gill-rakers in samples of Chromidotilapia guntheri*
C.g. loennbergi C.g. guntheri
Upper rakers
Kotto
Nganjoke
5
6
i
3
2
7
8
5
2
if
9
10
ii
Mungo Ghana
(2)
3 2
6 9
ii 6
9 2
6
* One at the epi-ceratobranchial joint is not counted. £ signifies right or left side only.
even temporary picture of the fin formula in the Mungo and does not justify a sub-
specific name.
A female of 64 mm SL with ovarian eggs of 2-5 mm long diameter had a general
dark olive colour with green iridescence on two longitudinal rows of scales above the
level of the pectoral fin. The opercular spot was black with blue iridescence and
this blue extended downwards on the opercular edge. The dorsal fin had the
characteristic broad silver-bronze band and the spots on it were in one series. The
lower lip was pale yellow and the leading edge of the pelvic iridescent light blue.
The caudal was without spots. A male had a spotted caudal and the dorsal fin was
pinkish without any iridescence.
C. g. loennbergi (Trewavas 1962 : 160 fig. 4)
The narrower interorbital on which I relied for the distinction of the Kotto popula-
tion in 1962 distinguishes the majority of our larger sample, but would hardly be
enough to maintain even its subspecific rank without the supporting distinction of
the lower number of epibranchial gill-rakers.
The general colour is paler than in the Mungo population, giving the impression of
a yellow fish. The flanks vary from watery green to yellow, sometimes soiled with
melanin.
There is a black opercular spot and the lower parts of the head are pale iridescent
blue, the branchiostegal membrane white or pink. The chest is white (?) or pink ($)
and the belly and lower parts of flanks are often tinged pink. The dorsal and caudal
fins have yellow ground colour ; in mature fishes the upper edge of the soft dorsal
and upper corner of the caudal are black with white submargin. The spinous dorsal
of the female bears numerous small black spots in 1-3 irregular series or quite
irregularly arranged, extending on to the soft dorsal ; in the sexually mature female
these spots overlie a broad silver-bronze band extending along the whole of the
spinous dorsal and the lappets are grey tipped with black. The male has neither the
metallic band nor the black spots on the dorsal, but the soft dorsal may have faint
spots that range from clear (on a yellow ground) to grey. The caudal is usually
immaculate in females but spotted in males. The anal is grey, sometimes clouded
E. TREWAVAS
vO
H
w
u
M
1
3
X
1 >
^o X
O
-o 2
X
-*. M
•2 «
*S
£ rS
8
PH
O
X
o X
>• 't' 00 O
X M
O U-)
•5 O
^^
^^ VH
S 3
^s
a
j
o
'a1
I
bo
T3
d
O
FISHES OF WEST CAMEROON
401
or washed pink. The pectoral is clear yellow. The pelvic is colourless in the $
with iridescent blue leading edge, in the <$ pinkish grey with leading edge and a
short filament white.
On narcotization or death a vague longitudinal dark band or series of blotches
appears along each flank. There are some differences between this colour pattern
and that described by Loiselle (i972a, b) for a Ghana population. He records a
lateral longitudinal dark band in the living fish more regular and conspicuous than
in either West Cameroonian population, a magenta ventral blotch in the female
and carmine instead of black edges to soft dorsal and caudal fins. More surprisingly,
an iridescent band along the dorsal fin appears in photographs labelled both male and
female, whereas I (and others) have found it only in females. Also on the branchio-
stegal membrane and 'gular region' (= chest according to Loiselle in litt.) where I
record a pink colour Loiselle records orange red, possibly a local or behavioural
difference. Loiselle observed mature fishes in aquaria and so was able to see colour
displays in life, and we did not.
The ecology of C. g. loennbergi is described by Corbet et al. (1973). The main
food consists of chironomid larvae. The structures of mouth and pharynx concerned
with feeding are described on pp. 389 - 391 of the present paper.
Two specimens were taken in R. Nyoke, which, although it flows past the crater
of Barombi Kotto not more than half a mile away, has no connection with the lake
or its outlet Nganjoke. Their upper gill-rakers and interorbital width are both in
the overlap zone between C. g. guntheri and C. g. loennbergi and I cannot place them
definitely in either. The types of C. boulengeri (Lonnberg) which in 1962 I synony-
mized with C. guntheri are respectively bigger (SL no mm) and smaller (SL 62 and
36-5 mm) than our Kotto samples. They have rather broad interorbitals, respectively
28-0, 25-8 and 25-0% length of head, and I counted the upper gill-rakers as 6, 8 and
9. Their status must await further knowledge as to how far in R. Meme the charac-
teristics of the Kotto population extend. The part of the Meme that yielded these
types is not recorded. If they belong to the Kotto population their name must
replace C. g. loennbergi, but I have not at present enough evidence to make such a
change.
Distribution of C. guntheri
Samples in the BMNH range from Sierra Leone to R. Benito. Although some
doubt has been thrown on the extreme records (LoiseUe, loc. cit.) we have no reason
to doubt them. Those from 'Murray Town, Sierra Leone' were collected by Mr P.
Carmichael Lowe and registered in 1920. This expedition had as its main object the
collection of birds and Sierra Leone is the only mainland territory from which Mr
Lowe's birds and fishes of that date are registered. The species has not been sub-
sequently collected in Sierra Leone.
Of the two specimens recorded from R. Benito (Boulenger, 1915 : 389 nos 28-29)
one is C. batesii. The other is a male of 128 mm SL, bigger than all our specimens of
402 E. TREWAVAS
C. kingsleyae, the only other long-snouted species with which it might be confused.
It differs from 7 specimens of C. kingsleyae from the Ogowe system (61-96 mm in
SL) in the broader interorbital (27-5% length of head, cf. 19-6-23-8% in C. kingsleyae}
and higher number of epibranchial gill-rakers (10, cf. 6-7 in C. kingsleyae) and seems
to be a genuine C. guntheri.
Of two Kribi fishes listed as C. kingsleyae (reg. no. 1908.5.30.185-6) one is either
C. kingsleyae or C. guntheri. It has the narrow interorbital of C. kingsleyae but a
number of upper gill-rakers characteristic of C. guntheri (9 left, 10 right). Neither
this nor the R. Benito specimen has any spots on the fins.
The evidence suggests that C. guntheri extends to R. Benito and C. kingsleyae
takes its place in the Ogowe.
The upper Ogowe evidently contains other species of Chromidotilapia represented
by the types of Pelmatochromis regani Pellegrin, 1907, and P. haugi Pellegrin, igiga,
which I have examined. Both are bigger than any C. kingsleyae in the BMNH and
they have more lower gill-rakers, respectively 14 and 15 (cf. 9-11 in C. kingsleyae,
but there are sometimes additional tubercles below them). P. haugi may be a
synonym of C. kingsleyae, which it resembles in the long snout (49% length of head)
and deep preorbital bone (28-6%). The bigger P. regani has a preorbital only about
25% of the head length and some of its pharyngeal teeth are enlarged.
Tilapia camerunensis Lonnberg, 1903
Tilapia lata var. camerunensis Lonnberg, 1903 : 41.
Tilapia lata (part, nee Gunther) ; Trewavas, 1962 : 166.
In 1962 I considered Lonnberg's type to be conspecific with the small specimen
(SL 88-2 mm) that is the type of T. lata (Gunther), but I now think that T. lata is a
possible synonym of T. guineensis (Bleeker in Gunther, 1862). The two names were
proposed on the same page of Giinther's Catalogue, T. lata from 'West Africa', T.
guineensis from 'Ashantee'. Bleeker's specimens of T. guineensis, the type in
Leyden and a specimen in the BMNH, are recognizable members of the lagoon
populations of Ghana. The type of T. lata, less typical in shape, nevertheless has no
feature that would exclude it from the same species as now understood, and since
its type-locality is so vague there is every reason for adopting T. guineensis as the
valid name, as has been done by Thys (i968b, 1971).
It seems possible, however, that the type of T. camerunensis, from 'high up in
R. Meme', is conspecific with the Mungo population, and this is not T. guineensis.
Our specimens were all young, collected by Dr Disney from R. Mungo : i of SL 35 mm,
between Baduma and Bolo ; 6 of SL 42-5-75 mm from the bridge near Etam
(BMNH 1973.5.14.1089-1095). We also have one of 135-5 mm collected by Mr
David Blair in R. Blackwater (1973.5.14.1088). Lonnberg's type is 102 mm in SL.
DESCRIPTION of these specimens, with the figures for the type in brackets.
Depth of body 42-5-46-0 (46-5)% SL, length of head 32-8-35-0 (32), length of
pectoral fin 31-5-33-0 (32-4), middle caudal rays 26-27 (25 ?).
FISHES OF WEST CAMEROON 403
Snout with straight or evenly decurved upper profile, 36-5-44-0 (44)% length of
head, diameter of eye 24-5-37-5 (27-8), negatively allometric, depth of preorbital
bone 20-5-27-0 (24-5), positively allometric, interorbital width 30-0-33-6 (33-8),
length of lower jaw 34-39 (31).
Lower jaw falling a little short of upper in front. Teeth in 3 or 4 rows in upper
jaw, a few of a 5th row in the largest, 3 rows in lower jaw, with a short 4th in largest ;
24-46 (38) in outer series of upper jaw ; outer stout, bicuspid with truncate major
cusp, inner tricuspid.
Gill-rakers (2-3) + (o-i) + (8-9) on the first arch.
Lower pharyngeal bone stout, its width 37-2-38-7% length of head in two specimens
of 75 and 64-5 mm SL respectively, with short blade and tricuspid posterior teeth
(not examined in type).
Scales in 3 (3-4) rows on cheek, 29 (28 left, 29 right) in lateral line series, rather
small on chest.
Dorsal XVI n (f.6) or XVI 12 (f.2 and type). Anal III 8 (£.3 and type) or III 9
(f-5).
Vertebrae 28, the modal number in T. (Coptodori). Mesethmoid meeting vomer
in a brief suture on each side.
The Mungo specimens differ from T. guineensis of the lagoons of Ghana and Nigeria
in the shorter pectoral fin and broader pharyngeal bone. The dorsal formula XVI
12 is fairly common in T. guineensis, but I have not found XVI n among 66 speci-
mens from Lagos to Chiloango and Thys does not record it for this species in southern
Cameroun. Neither is it recorded for T. dageti Thys, 1971, the species known to
Daget (1954), Daget & Iltis (1965) and Blache et al. (1964) as T. melanopleura.
The dorsal formula distinguishes it too from T. cameronensis Holly of R. Sanaga.
Among the 26 specimens described by Thys (1966) none has XVI n and only four
XVI 12, and III 8 is a rare anal formula. This species also has a very broad lower
jaw and the gill-raker numbers range higher. I have examined, as well as the holo-
type, four T. cameronensis from the Stanford University collection, all from R.
Sanaga, and three of them have the modal dorsal formula given by Thys, XV 14,
the fourth has XIV 14.
T. kottae differs from the Mungo samples in the fewer scales on the cheek and
smaller pharyngeal bone as well as in colour, but XVI n is not an uncommon for-
mula. T. zillii differs in the shape of head and body and usually has both horizontal
and vertical stripes. T. nyongana Thys, 1971, though having meristic characters
close to the Mungo samples, has a steep, straight profile from occiput to tip of snout
and usually XV dorsal spines. Our living Mungo specimens were perhaps too young
to have developed the silvery spots on the head characteristic of T. nyongana and
there is no trace of them in the preserved adult. The geographical relation of the
two populations, with T. cameronensis of R. Sanaga intervening, makes their identity
unlikely.
The Mungo population of T. (Coptodon) has in fact its own characteristics and
although it seems unwarranted to regard it as a species it is hard to say of which
species it might be a subspecies. It is only tentatively bracketed with the Meme
population as T. camerunensis.
404 E. TREWAVAS
Tilapia kottae Lonnberg, 1904
Lonnberg, 1904 : 135 ; Boulenger, 1915 : 200 fig. 127 ; Trewavas, 1962 : 164.
Barombi names : fikunle (young), pindu (large specimens).
We collected the species only in the lakes, Barombi Kotto and Mboandong, where
it was very abundant. Traps set in the streams near Kotto yielded T. mariae and
Chromidotilapia but no T. kottae. Details of the ecology are given by Corbet et al.
(1973)-
The description given earlier (Trewavas, 1962) covers also the additional specimens,
with slight extensions in the ranges of allometric characters corresponding to the
further ranges of size. The length of the lower pharyngeal bone in 5 specimens of
SL 95-113 mm is 24-0-25-5% length of head, its width 30-0-32-3, and the blade is
0-5-0-73 of the length of the toothed part. In a specimen of SL 68 mm the relative
size of the bone is a little more, length 26-4%, width 36-2. Most specimens have
only 2 rows of scales on the cheek, or at most an incomplete third row, in some on
one side only.
Meristic characters, as might be expected in a restricted population, are very
uniform, with narrow ranges and strong modes. The scales in the lateral line series
number 29 in 17 of 21 specimens, 28 in 2, 30 in 2. There are 9 gill-rakers on the lower
part of the arch in 10 of 14 specimens, and 8 anal soft rays in 13 of 14. The dorsal
rays were counted in 29 specimens and gave a total of 27 in 22 specimens (XV 12 or
XVI n), 26 in 7 and 28 in 2.
Of the specimens analysed, 8 of SL 54-5-100-5 mm were females, 13 of 56-7-
113-0 mm males. The gonads of the smallest $ were far advanced and a <$ of 65 mm
had enlarged testes. No differences were found between the sexes unless perhaps a
filament terminating the pelvic fin is confined to larger males. Ripe eggs are pale
olive-green (not yellow as found in preserved specimens in 1962), the biggest ovarian
eggs seen having a long diameter of 1-8-2-0 mm.
Small individuals are yellow or olive-green, often with grey vertically elongate
blotches along the middle of the flanks ; these may alternate with a more dorsal
series of blotches or two of the latter may join with one of the former to give a
dorsally forked bar (as in T. ogowensis and often in the young of other related species) .
Never was such a mid-lateral longitudinal band seen as is usual in T. zillii, nor a
band punctuated with blotches as in T. tholloni. This vertical pattern is already
present in fry of 15 mm and the breadth of the bars distinguishes them at sight from
other species in Lake Kotto. The tilapia-mark in the young is round, intensely
black, with a clear ring, bigger than in T. mariae or S. galilaeus. It persists until it is
masked by the spread of melanin. From SL 24 mm upwards the throat, interoper-
culum and suboperculum and the sides of the chest become black, and blue iridescence
sometimes seen on the gill-cover of the young soon becomes masked by melanin.
Chest and belly may remain white or white unevenly soiled with melanin. The lower
part of the flank may be pink.
Melanin extends all over the body and fins in bigger fishes, which have either a
soiled appearance or are almost totally black, usually with a white lower lip and
sometimes with the lappets of the dorsal fin yellow or orange. Upper and lower
FISHES OF WEST CAMEROON 405
halves of the caudal may contrast in colour, but the colour varies, the upper half
being pale yellow or with a pink reticulum enclosing watery green spots, the lower
half yellow, soiled yellow or black ; or the whole fin may be covered more or less
densely with melanin.
RELATIONSHIPS. Behaviour studies on species of Tilapia inspire caution in the
use of the predominance of vertical or horizontal dark bands on the body as evidence
of relationship (see especially fig. i in Voss, 1969). Yet the fact that in neither living
nor preserved specimens did we see longitudinal bands, either continuous or repre-
sented by blotches on the vertical bars, differentiates it from T. zillii and T. tholloni,
in which these are present. The rather big scales on the cheek, usually in only two
rows distinguish it from T. ogowensis and T. ' camerunensis' and from the latter it
also differs in colour, the more abruptly straight profile from snout to occiput and
the usually smaller pharyngeal bone and thicker lips.
A species recently described by Thys (1972), T. bemini, from a crater lake draining
to an upper tributary of the Cross River system, is, as Thys suggests, very similar in
shape, meristic characters, thick lips, only two rows of scales on the cheek and yellow-
ish colour, but in this the dentition of jaws and pharynx is more refined and only a
few of the posterior pharyngeal teeth are tricuspid. These two species probably
had a common ancestor at no very distant period, and this cannot have been very
different from T. camerunensis.
ECOLOGY. Although T. kottae belongs to a section of the genus Tilapia whose
relatively coarse dentition fits it to include vascular plants in the diet, Corbet et al.
(1973) report that in Lake Kotto the main food consists of the phytoplankton in
which this lake is very rich. This may be reflected in the slightly smaller pharyngeal
bone of this species in comparison with T. camerunensis.
In Mboandong, where the phytoplankton is less dense, organic debris accounted
for a bigger percentage of the stomach contents. Here, as in the case of many species
of Tilapia and Sar other odon, the young of SL 9-22 mm were more carnivorous than
the adults (Corbet et al., loc. cit.).
MATERIAL PRESERVED
BMNH register no. SL (mm) Locality
1973.5.14.1117-1140 55-H4 Kotto, 1970
1973.7.18.219-258 n-39 Kotto, 1972
Tilapia mariae Boulenger, 1899
Boulenger, 1899 : 122 pi. xi fig. i (Niger Delta) ; Thys, 1966 : 36, pi. v and text-fig. 4 (Ivory
Coast to South Cameroun).
Tilapia microcephala (nee Bleeker) ; Lonnberg, 1903 : 41 (R. Meme).
Tilapia dubia Lonnberg, 1904 : 137 (Barombi Kotto) ; Boulenger, 1915 : 189 (Kotto).
Tilapia meeki Pellegrin, 1911 : 185 ; id. 1914 : 63 pi. ii fig. 2 (Ivory Coast) ; id. 1928 : 9 (R.
Sanaga) .
Tilapia heudeloti (part, nee Dumeril) ; Boulenger, 1915 : 175 (no. 34 only, R. Meme).
Tilapia mariae dubia (excl. syn. T. haugi) ; Trewavas, 1962 : 166 (Barombi Kotto).
406 E. TREWAVAS
Barombi name : finjongo.
I formerly (Trewavas, 1962) expressed doubt as to whether this population had
more affinity with T. haugi ( = T. cabrae) than with T. mariae. The discovery that
sexually mature individuals of the Kotto population develop the characteristic
lateral blotches of T. mariae (see R. A. Whitehead, 1962) confirms the action of Thys
(1966) in putting the geographical boundary between the two species farther South.
I also agree with him in abandoning trinominal nomenclature within this species,
although the two features on which the subspecies dubia was maintained receive some
confirmation from the new material.
Inter orbital width. Table 17 shows different but overlapping ranges for this ratio
in the Kotto and other populations, but the small numbers examined together with
the allometry cast doubt on the significance of the difference. The values given by
Thys (1966 : 37) for other localities in the Cameroons match better with my Kotto
ranges than with those of western samples. Possibly there is a cline in this character.
Pectoral fin. Table 18 shows that although the pectoral lengths cover nearly the
same range in and outside the Kotto population, 8 of the 9 specimens in which it is
more than 38% SL are Kotto fishes. The range given by Thys for other specimens
of Cameroun does not exceed 37-3%. Since the pectoral is probably used more in
still than in flowing waters for keeping station and locomotion this is not an un-
expected difference and may not be genetically based.
Colour. The colour-pattern, as R. A. Whitehead (1962) showed, undergoes a
change from a series of dark vertical bands in the young to a series of up to six dark
blotches along the middle of each side in the adult. According to Whitehead's
TABLE 17
Tilapia mariae : Interorbital width as % length of head in
Lake Kotto and other localities
At SL 62-100 mm At SL 101-150 mm
Barombi Kotto 34-0-39-0 (« = 13) 36-3-43-6 (« = n)
Mean SL in mm
Other localities 40-0-42-5 (n = 9) 41-4-48-8 (n = 9)
Mean SL 128 mm
TABLE 18
Tilapia mariae : Length of pectoral fin as % SL in Lake Kotto and
other localities : frequencies
% Kotto Other localities
32-0 i i
32-1-34-0 o 2
34-1-36-0 3 4
36-1-38-0 6 3
38-1-40-0 6 i
40-1-41-5 2 O
FISHES OF WEST CAMEROON
407
observations, the change is related to sexual maturity - he found both to occur at a
smaller size in one of the localities studied by him than in the other.
Thys (1966) found in South Cameroun two types of adult coloration, one with dark
blotches on a yellow ground, the other black. In Barombi Kotto both changes
occur in the same individual, but not necessarily at the same time. The blotches,
when they occur, alternate with the positions of the bands as can be seen in transi-
tional cases where bands and blotches occur together. An uneven spread of melanin
may leave bands still visible for instance on the lower half of the body, or may allow
some blotches to be seen.
The following examples illustrate these changes, 'stg' signifies gonads just
starting to swell, 'rpg' a later stage, in females with eggs not yet to definitive size,
'rp' with eggs at full size.
SL 91 mm $ rpg : 5 blotches between the fading bands.
SL 91-2 mm $ rp (eggs 2-2^ mm) : 4 incipient blotches between 2nd to 6th bands.
SL 102-8 mm $ rpg (eggs 1-5 mm, green) : 5 marked blotches on a green ground.
SL 105-5 mm ? rP : general dark colour, suggestion of three dark blotches on left
side.
SL 116-0 mm $ rp (eggs partly discharged) : general very dark green colour, no
bands or blotches visible.
SL 118-5 mm ? rPg : 4 bands on lower half of body.
SL 105-6 mm $ stg : vague grey bands on lower half of left side ; body green with
bases of scales black.
SL 135-0 mm (J, gonads small : six dark vertical bands.
SL 135-0 mm (£, gonads small : vague traces of bands obscured by black to dark
green.
SL 145-0 mm £ stg : bands on a green ground ; fins dark (after 2 days in formalin
body all-black).
SL 150 mm <£ : body and fins dark green to black.
Any more precise appraisal of the correlations must be based on more living ma-
terial. Thys reports observations by Wickler that suggest a dependence of the blotch
pattern on the assumption of territorial behaviour, and the association of the black
colour with a non-sexual phase. The latter correlation is not supported by our
observations on the Kotto samples.
However dark the fish may be, the upper edge of the soft dorsal and upper corner
of the caudal are red or pink, sometimes briefly underlined with white or yellow.
The dorsal lappets may have narrow pink or yellow edges.
The food of T. mariae in Lake Kotto consists mainly of phytoplankton (see
Corbet et al., 1973) and this is not what would be expected from its dentition, which
consists of several rows of spoon-shaped teeth, the outer with one notch, the inner
with two. This is the typical equipment of an aufwuchs-grazer, which probably is
T. mariae's more usual role (cf. Fagade, 1971, on the population of Lagos Lagoon).
As in the case of T. kottae, the unusual richness of Lake Kotto in phytoplankton has
determined its special feeding habit.
We caught T. mariae in Lake Kotto and its outlet stream Nganjoke ; it is recorded
from R. Meme. Its habitat throughout its range is in the fresh waters of the coastal
408 E. TREWAVAS
plains and it is not surprising that it was not found in the upper Mungo, separated
as this is from the coastal reaches by falls and rapids. Holly (19270) reported it from
R. Mbam, Sanaga system, much farther inland, but Thys (1966 : 42) has seen Holly's
material and includes in T. mariae only the specimens from the lower Sanaga.
We have no information on breeding from this population (although it is now well
documented from elsewhere). A single young of 16-5 mm SL caught in 1972 with
young T. kottae at a beach at the perimeter of Lake Kotto was distinguished from
other species by a gill-raker count of 12, very small notched teeth and the presence
of seven faint vertical bars on the flanks some of which extended on to the dorsal
fin (see also photograph by Marcuse, 1971).
MATERIAL PRESERVED
BMNH register no. SL (mm) Locality
1973.5.14.1097-1115 91-150 Kotto
1973.5.14.1096 115 R. Nganjoke
Sarotherodon galilaeus (Linnaeus, 1758)
Sparus galilaeus Linnaeus, 1758 : 282 (Tiberias).
Tilapia macrocephala (nee Bleeker) ; Lonnberg, 1904 : 135 (Kotto).
Tilapia galilaea ; Trewavas, 1962 : 177 (Kotto).
Sarotherodon galilaeus Trewavas, 1973 : 23.
Measurements on 5 additional specimens from Barombi Kotto and 4 from Mboan-
dong confirm the observation x)f 1962 that the depth of body in this population is less
than the average - 35-0-46-5% SL (in other populations sampled, except one,
43'5~56'0%). The meristic characters associated with the metameres are in agree-
ment with those of other West African populations (excluding S. g. multifasciatus) ,
namely, vertebrae 29 in all (mode in West Africa and Lake Albert 29, in the Jordan
Valley and other rivers of Israel 30) ; scales in lateral line series 30 (f.6) or 31 (f>4)
(mode in West Africa and Lake Albert 30, in Asia and the Nile 31) ; total dorsal
rays 27 (f.2) or 28 (f.7) (mode in West Africa and Lake Albert 28, in Asia and the
Nile 29).
The colour in adults is pale brassy yellow, white ventrally. The discontinuous
vertical bands that are present in individuals of all populations, and, as students of
behaviour have shown, possibly in all individuals in certain emotional states, are
absent from most of our specimens, but traces of them are present in three. A
juvenile of SL 69-5 mm had after preservation the fainter, more continuous grey
vertical bands common in juvenile Sarotherodon and a faint tilapia-mark.
The fins were almost colourless, but in a favourable light the soft dorsal was seen
to have pale yellow spots on a pale grey ground. The pelvics were faintly tinged
with orange and their leading edge and short filament were white.
The juvenile of 69-5 mm was silvery instead of yellow.
Other proportions and structures are as in other populations of 5. galilaeus,
including the pharyngeal bones, which are no smaller than elsewhere (cf. 'slightly
FISHES OF WEST CAMEROON 409
smaller', Trewavas, 1962). I do not find any justification for considering this popu-
lation a distinct subspecies. It differs from most others in the more slender body,
but the S. galilaeus of Nungua Reservoir, near Accra, are also relatively slender
(depth 38-49% SL) and the difference is no more than might be attributed to 'condi-
tion' - although I do not suggest that either population is underfed.
Neither do I think it useful to use a subspecific name to distinguish the West
African populations from those of the Nile and the Jordan Valley. Schultz (1942)
and Thys (1969, 1971) have used pleuromelas Dumeril in this sense at specific or
subspecific level without giving any distinguishing characters. Although the modal
numbers of vertebrae, scales and dorsal rays are one less in West Africa than in
Asia and the Nile, the lower modes are also those of the small Lake Albert sample
available (15 specimens). Biparental mouth-brooding is now known to be practised
by the species in West Africa as well as in the Jordan Valley (lies & Holden, 1969).
The present geographical gap between the Nile and Chad populations is evidently a
recent one - 5. galilaeus is present in Jebel Marra between Chad and the White Nile
and in R. Draa, north of the Sahara.
It is natural to compare the Kotto 5. galilaeus with those of R. Sanaga, Cameroun,
named T. sanagaensis by Thys in 1966. This was distinguished by the blue-grey
to purplish colour of the dorsum and the series of clear round spots on a background
of the same colour on the dorsal and caudal fins. From the description I have given
of the Kotto colouring it is plain that the basis of this colour-pattern is there too,
but lacks the contrast given in the Sanaga by the darker ground colour.
The figure of the type given by Thys (1966, pi. i) shows a very galilaeus-like facies.
But the head in this fish, and also in one from R. Lorn (tributary of Sanaga) in the
BMNH, is relatively small - 32% SL, the lowest ratio in any S. galilaeus seen by
me. Holly (ic)2jb : 220) obtained the same ratio in specimens from R. Mbam. The
head length in Thys's seven specimens, however, ranged to 35'5%, in Sanaga fishes
seen by me to 34-5%. The range in Kotto and Mboandong is 35-37%. Elsewhere,
even among topotypes of S. galilaeus, the lower limit of the ranges is 32-5%, so a
small head seems a character of doubtful taxonomic significance.
I do not doubt that the Sanaga population has its own characteristics, and I note
Thys's report that it retains its distinctive colour vis-a-vis others imported to the
same ponds from Ubangi-Chari, but I doubt if any useful purpose can be served by
naming it. However that may be, the Kotto population does not share these
peculiarities and I would retain the latter in S. g. galilaeus.
This species was caught by fishermen with cast-nets in the open water and not
(except young) in traps set at the shore. We found no brooding individuals. Ripe
ovarian eggs were olive-green. The food, mainly phytoplankton, is described by
Corbet et al. (1973).
Professor Green and Dr Corbet caught numerous young of 21-37 mm SL at the
landing beach of Kotto Island in 1972. Like many young Sar other odon, these are as
deep-bodied as the adults (depth 39-47% SL). The teeth, smaller than in T. kottae
of the same size, are triangular in shape with the apex of the triangle at the attached
end and the free edge divided by a notch. Dorsal fin-rays, counted in eight, number
27 in seven (XV 12 or XVI u), 26 (XIV 12) in one. A small but distinct
4io E. TREWAVAS
tilapia-mark is present and the edges of the vertical fins are narrowly black
(preserved specimens). Faint grey vertical bars are present on the flanks.
MATERIAL PRESERVED
BMNH register no. SL (mm) Locality
I973-5-I4-779~7&2 122-145 Mboandong
1973.5.14.783 69-5 Kotto, 1970
1973.7.18.173-196 21-38 Kotto, 1972
ZOOGEOGRAPHICAL SUMMARY AND CONCLUSIONS
In the neighbourhood of the three lakes of West Cameroon visited by our expedi-
tion the streams of the Mungo and Meme systems are not far apart and even inter-
digitate in this very uneven territory, whose topography is the result of vulcanism
from the middle Tertiary to the present day. River-capture in the upper streams is
readily understandable, and Barombi Mbo, whose present outlet is to the Mungo, is
believed formerly to have drained to the Meme (Dusen, quoted by Trewavas, 1962 :
150). The stream Nyoke, from which villagers living close to the Kotto crater get
their drinking water, is a tributary of the Mungo, although Lake Kotto drains to
the Meme. The southern sources of the Cross River system are not far from the
sources of the Mungo and there is evidence of faunistic relationship with them, but
our knowledge of the Cross fauna is too incomplete to contribute much to the present
conclusions.* Topographically therefore we have good reason to treat the Mungo
and Meme systems together and although our sampling of the Meme system was
confined to Lake Kotto, its outlet Nganjoke and an upper tributary, R. Bille,
there is some justification in treating them as a faunistic unit too.
Consideration of the list of Mungo -Meme fishes prompts certain comparisons :
(1) Comparison of our upland collection with a small collection of fishes made
near the mouth of the Mungo by Mr Deveson in 1951 and 1952.
(2) The relationship of the Mungo-Meme fauna to that of Fernando Poo.
(3) The relationship of the fauna of the rivers to those of the three lakes.
(4) The relationship of the Mungo-Meme-Fernando Poo fauna to that of the rest
of West Africa.
(i) Mr Deveson's collections were made in the neighbourhood of Tiko and his
biotopes included both still waters and streams with a considerable current. A list
of his species follows with the relevant parts of his field notes.
a. Species also present in our collections.
Brienomyrus brachyistius (or longianalis ?) : in rapid water under shelter of rocks
and weeds at side of river ; in loose shoals of about 20 fishes.
Alestes longipinnis : in open water with a current.
Barbus thy si : in water flowing at about 6 knots.
* In the mainstream of the Cross River marine and brackish water species extend inland as far as the
neighbourhood of Bambui. Messrs J. Grimshaw and L. Torrans, American Peace Corps workers in the
region, showed me specimens of Pomadasys jubelini and Trachynotus goreensis collected there.
FISHES OF WEST CAMEROON 411
Epiplatys sexfasciatus : common in rivers and agricultural drains ; seemed to
prefer more acid waters than other species collected.
Chromidotilapia guniheri : in long grass on flooded land.
H emichromis fasciatus (or elongatus ?) : in both still and running water, acid or
alkaline.
b. Species not in our collections and not recorded from Fernando Poo.
Nannaethiops unitaeniatus : in a forest stream, tributary of R. Matute, spaced
along stream with no sign of shoals.
H emichromis bimaculatus Gill : ecology as H. fasciatus.
Ctenopoma kingsleyae (Giinther) : in still, sheltered parts of small rivers ; feeds on
surface insects and comes to surface for air.
Channa obscurus (Giinther) : found in flooded grasslands.
The species common to both collections were found in biotopes similar to those
in which we caught them, except that C. guniheri, which we found to be the more
adaptable of the two species in the Mungo, was found in a biotope not encountered
by us. Ctenopoma and Channa are adapted to lowland swamps liable to oxygen
deficiencies. In the lowland forest streams near Douala Thys (1967 : 139) also
reports these two as well as Calamoichthys calabaricus. The basis of selection of
biotopes by Nannaethiops and H emichromis bimaculatus is unknown to me.
(2) In his analysis of the freshwater fish fauna of Fernando Poo, Thys (1967)
places twelve species in his group A, true freshwater fishes with little or no salt
tolerance in nature. All of these are present in our Mungo-Meme collections. Al-
though three of Thys's identifications have since been corrected, the new names are
here held to apply also to the respective Mungo populations (Barbus thysi, Aphyo-
semion oeseri and Chromidotilapia finleyi). The question has been raised (p. 397)
of the possible subspecific distinctness of the island Chromidotilapia ; and the
specific identity of the Mungo A . oeseri with the type population of the island may
with advantage be further tested ; but the relationships are at any rate very close.
Emphasizing the identity of the island and mainland faunas is the apparent restric-
tion of two species, A . oeseri and B. thysi (but see p. 352) to Fernando Poo and the
Mungo and Meme basins.
(3) Kotto and Mboandong. Eleven species were found within the crater of Kotto.
Four of these, the cyprinodonts E. sexfasciatus, A. bivittatum, A. oeseri and Procatopus
sp. were caught by us only in the inflowing streams. The others were Barbus
callipterus, Clarias walkeri and the five Cichlidae, including the endemic subspecies
C. guniheri loennbergi and T. kottae, endemic to this and the small lake Mboandong.
From what has been written on p. 405, it will be seen that I consider it possible that
T. kottae, T. camerunensis and T. bemini form a group of taxa (species ?, subspecies ?)
peculiar to the Mungo, Meme and Cross systems. T. mariae is a freshwater Tilapia
abundant in the lower reaches of river systems from South Cameroun to Western
Nigeria and again in Ivory Coast, in the forest zone. It was caught in both Kotto
and its outlet. Sar other odon g. galilaeus is not recorded elsewhere in Cameroun, the
population of R. Sanaga being sufficiently distinct from it to have been regarded as
another species. Adults were caught in the open water of both lakes and young
around the margins. The widespread H emichromis elongatus has flourishing colonies
412 E. TREWAVAS
in both Kotto and Mboandong. With T. kottae, it lives near the edges of Mboandong,
and here too are abundant A . bivittatum and A . oeseri. The rim of this shallow lake
with its carpet of fallen leaves and its constant rain of terrestrial insects resembles a
forest stream and it occurred to us that S. galilaeus, so untypical of such an environ-
ment, might have been stocked there from Kotto. It was being caught by cast-nets.
To summarize, the fish fauna of Lake Kotto shows no endemism that on structural
grounds can be considered to extend beyond the subspecific level (except possibly
T. kottae). It is a part of the Mungo-Meme fauna whose members are well fitted
to take advantage of the eutrophic ecology of Lake Kotto.
Mbo (see Trewavas et al., 1972). Thys estimates (1937 : 135) that the Fernando
Poo fresh waters have been effectively isolated from the mainland for 8000-10 ooo
years. Unfortunately I know of no evidence for the date at which the crater of
Barombi Mbo was formed or when volcanic disturbance within it last exterminated
life in its waters. But the contrast between the extreme endemism of Barombi
Mbo and the non-endemism of the Fernando Poo fish fauna can hardly be explained
by differences in the degree or time of isolation, both of which may well be less for
the lake than the island.
The contrast is associated with two facts :
(1) The ecology of the fish biotopes in the island is identical with that of the
Mungo-Meme tributaries - hilly forest streams - that of the lake is contrasted.
(2) The lake was colonized by members of the family Cichlidae, a family represented
by only one species on the island, and that preadapted for the island biotopes. To
say that the Cichlidae are notoriously able to adjust their feeding and reproductive
habits to a lake environment is of course only to push the question farther back.
Our comparison of the Mbo, fauna has been mainly with that of the neighbouring
Mungo streams, but the possibility that Mbo formerly drained to the Meme reminds
us that the Kotto section of the Meme fauna includes two possible ancestors of
Mbo cichlids, T. mariae and 5. galilaeus.
Soden. Of the four species present in this lake, Barbus batesii and the Clarias
are members of the Mungo-Meme fauna with population characteristics peculiar to
the lake. Aphyosemion celiae relates the lake fauna to that of the Upper Mungo,
but further exploration of the Meme may discover it there too. The special ecology
of Procatopus lacustris may be made possible by the absence of competitors, though
probable predators (Barbus and Clarias) are present.
(4) Table 19 provides a conspectus of the distribution of the species of the Mungo-
Meme fauna as far as present knowledge will take us. Leaving out of account the
ten widespread species that head the list, it is clear that the fishes of West Cameroon
belong to the fauna of the coastal forest belt from South Cameroun to Lagos. The
emphasis is on South Cameroun - note especially that the section of the genus
Auchenoglanis to which our species belongs is not reported West of the Cameroons.
Since we expect our knowledge of Cameroun fishes soon to be extended by the work
of Dr Thys van den Audenaerde it would be premature to go into further details.
The small number of species common to our area and the Chad basin supports the
increasingly recognized division between forest and savannah fish faunas in West
Africa.
FISHES OF WEST CAMEROON
413
o
t/5 \-r
to f~\
^ I — ^
*o> ®
ifl
ho g
.s o
o^ a
TJ
w a
J 3
PQ o
B
X!
now
O
C/)
Xi
O
o
H
ri
Q
Cameroun (SC),
ast (Iv), Liberia
(Cd), Nile (Nil).
ed together; H.
eans the West
tapirus includes
he taxonomy of
eneath the skin
South
ory Co
basin
(R
Gh)
, C
atus
ns
Mormyru
state of
n hidden
,
, Iv
ad
are
y
fi
Gabon (Ga), Rio Muni
(Da), Togo (To), Ghana(
le and Upper Niger (UN)
s fasciatus and H. elong
its subspecies; B. mung
including B. mungoensis
ause of the unsatisfacto
te in front of the dorsal
he areas considered.
basin (Co), Chiloango (Ch),
rom Lagos (La), Dahomey
(Gm), Senegal (Se), Midd
, as follows: Hemichromi
on; 5. galilaeus includes
or longer than the head,
nius, is included bec
the interneural pla
occur also outside t
o
nd fro
bia
d,
E.«
-. CO T-l
e j= o
a«|u^
3 bo S ^3
g
a
am
e u
trib
lon
s A
d t
ead
Con
, inl
), G
are
dist
as l
nus
ded
re
bbreviated as fol
o (FP), Niger del
Guinea and Guin
xa bigger than
ith broadly the
st simple, stiffen
topus, including
Auchenoglanis a
ess. Some of t
are a
ndo Po
ues
o
S S rt «
?<58<sa
(5»ft
ug
s
e
h
cie
h
!l£.i!«5ie!
•-Pfih-S" ^ g|S,
li
er
Po
ocal
C F
p
o
s
g
^•|I|J|8
CC.« <« 3^ ^ >,
•S H-i.-cCuJTiOai'-irr
oo oSo^Htn^iCS
s sPI-flr.?.!
llflljilM
^ 03 (J> »H ^ »^ ^"
C
n
h
e
e
S
s
cu
a
o
g
o
o hi
Th
West
(Li),
In
bimac
Afric
M. g
that
and
414 E. TREWAVAS
ACKNOWLEDGEMENTS
The visit to Cameroon in 1970 with Drs S. A. Corbet, J. Green and J. Griffith was
financed by grants from the Royal Society, the Central Research Fund of the
University of London and the Godman Exploration Fund and I should like to renew
my thanks to these and to the members of the Helminthiasis Research Unit at
Kumba for all the help they gave us.
I have had the benefit of all the resources available at the British Museum (Natural
History) for statistical advice (Dr M. Hills), radiography (Mr G. Howes), microscope
sectioning (Mr D. Cooper), alizarin preparations (Mr J. Chambers), photography
(Mr P. A. Richens and Mr J. V. Brown) as well as of its incomparable library and
collections.
As already stated in the Introduction, most of the fishes of the Mungo tributaries
were collected by Dr R. H. L. Disney, who also provided the model for the sketch
map (Fig. i). On the enigmatic species of Labeo I have had the benefit of discussions
with Mr Gordon Reid, a newcomer to taxonomy ; and I thank Dr Mary Whitear for
an interesting discussion on histology.
The drawings for Figs 6, 7, 8, 9, n and 17 were made by Sharon Chambers. Mr E.
Finley kindly provided the colour-photograph of PL 5.
REFERENCES
AHL, E. 1927. Neue afrikanische Fische der Familien Anabantidae und Cyprinodontidae.
Sber. Ges. naturf. Freunde Berl. 1927 : 76-81.
ARNOLD, J. P. & AHL, E. 1936. Fremdldndische Susswasserfische. Wenzel & Sohn, Braunsch-
weig.
BARNARD, K. H. 1943. Revision of the indigenous fishes of the S.W. Cape region. Ann. S.
Afr. Mus. 36 : 101-202.
BERKENKAMP, H. O. 1972. Zwei eierlegende Zahnkarpfen von Fernando Poo. Das
Aquarium, Wuppertal Elberfeld, Jg. 6 : 795-797, 2 figs.
BHATTI, H. K. 1938. The integument and dermal skeleton of Siluroidea. Trans, zool. Soc.
Lond. 24 : 1-79, pis i-xi.
BLACHE, J. & MITON, F. 1960. Poissons nouveaux du bassin du Tchad et du bassin adjacent
du Mayo Kebbi II. Cyprinoidei. Bull. Mus. Hist. nat. Paris (2) 32 : 143-153.
BLACHE, J., MITON, F., ILTIS, A. & LOUBENS, G. 1964. Les poissons du bassin du Tchad et du
bassin adjacent du Mayo Kebbi. Mem. O.R.S.T.O.M. Paris, 483 pp., 147 figs.
BLEEKER, P. 1863. Memoire sur les poissons de la cote de Guinee. Natuurk. Verh. Holland.
Maatsch. Wet. Haarlem 18, no. 2. 136 pp., pis.
BOULENGER, G. A. 1898. A revision of the African and Syrian fishes of the family Cichlidae.
Part i. Proc. zool. Soc. Lond. 1898 : 132-152, pi. xix.
- igoia. On the fishes collected by Dr W. I. Ansorge in the Niger Delta. Proc. zool. Soc.
Lond. 1901 : 4-10, pis ii-iv.
Les Poissons du Bassin du Congo. Bruxelles. pp. i-lxii, 1-532, pis i-xxv.
Description of a new fish of the family Cichlidae from the French Congo. Ann.
Mag. nat. Hist. (7) 8 : 114-116.
1902. Additions a la faune ichthyologique du bassin du Congo. Ann. Mus. Congo 2 : 19-
57, pis. vii -xvi.
1903. Descriptions of new freshwater fishes from Southern Cameroon. Ann. Mag. nat.
Hist. (7) 12 : 435-441.
1904. Descriptions of new West African Fishes. Ann. Mag. nat. Hist. (7) 14 : 16-20.
igo7a. Descriptions of three new freshwater fishes discovered by Mr G. L. Bates in South
Cameroon. Ann. Mag. nat. Hist. (7) 20 : 485-487.
FISHES OF WEST CAMEROON 415
BOULENGER, G. A. iQC^b. Fishes of the Nile, in Zoology of Egypt. London.
1909-1916. Catalogue of the Fresh-water Fishes of Africa. Vol. i (1909), 2 (1911), 3 (1915)
and 4 (1916). London : British Museum (Natural History).
1913. Sur une petite collection de poissons recueillis dans 1'Uele par la mission dirigee par
M. Hutereau. Rev. Zool. afr. 2 : 155-161.
1918. On some fishes from the Shari River, with descriptions of two new species. Ann.
Mag. nat. Hist. (9) 2 : 426-427.
1920. Poissons recueillis au Congo Beige par 1'expedition du Dr Christy. Ann. Mus.
Congo. Zool. (i) 2 fasc. 4. 38 pp., pis xxiii-xxv.
BURCHARD, J. & WICKLER, W. 1965. Eine neue Form des Cichliden Hemichromis fasciatus
Peters (Pisces : Percif ormes) . Z. Zool. Syst. Evolutionsforsch. 3 : 277-283, 2 text figs.
CLAUSEN, H. S. - see Stenholt-Clausen.
COPE, E. D. 1869. On some new species of American and African fishes. Trans. Amer.
Philos. Soc. 13 : 400-407.
CORBET, S. A., GREEN, J., GRIFFITH, J. & BETNEY, E. 1973. Ecological studies on crater
lakes in West Cameroon. Lakes Kotto and Mboandong. /. Zool. Lond. 170 : 309-324.
CUVIER, G. F. & CLOQUET, H. 1816. Barbus (Ichtyol.), in Dictionnaire des Sciences naturelles
IV Suppl. : 6.
DAGET, J. 1954. Les poissons du Niger superieur. M6m. Inst. fr. Afr. Noire No. 36 : 5-391.
1960. La faune ichthyologique du bassin de la Gambie. Bull. Inst.fr. Afr. Noire 22 : 610-
619.
1961. Le pare national du Niokolo-Koba. xxxv. Poissons. M6m. Inst. fr. Afr. Noire
No. 62 : 325-362.
1961. Poissons du Niari-Kouilou recoltes par Mm. Ch. Roux, J. Ducroz et J. P. Troadec
(Afrique noire - region Gabon-Congo). Bull. Mus. nail Hist. nat. (2) 33 : 577-586.
DAGET, J. & ILTIS, A. 1965. Poissons de Cote d'lvoire (eaux douces et saumatres). M6m
Inst. fr. Afr. Noire No. 74. X1 + 3&5 pp.
DAGET, J. & STAUCH, A. 1963. Poissons de la partie Camerounaise du bassin de la Benoue.
Mem. Inst. fr. Afr. Noire No. 68 : 85-107.
DISNEY, R. H. L. 1971. Association between blackflies (Simuliidae) and prawns (Atyidae),
with a discussion of the phoretic habit in Simuliids. /. Anim. Ecol. 40 : 83-92.
DUMERIL, A. H. A. 1859. Reptiles et poissons de 1'Afrique occidentale. Arch. Mus. nat.
Hist. Paris, 10 : 137-268, pis. xiii-xxiii.
FAGADE, S. O. 1971. The food and feeding habits of Tilapia species in the Lagos Lagoon.
/. Fish Biol. 3 : 151-156, i text fig.
FOERSCH, W. 1968. Neue Aphyosemionarten aus Kamerun. Aquar.- u. Terrar.-Z. 21 : 365-
369, u figs.
1971. Aphyosemion santaisabellae Scheel 1968. Aquar.- u. Terrar.-Z. Jg. 24 : 20-23, 4 n§s-
FOWLER, H- W' 1919. The fishes of the United States Eclipse Expedition to West Africa.
Proc. U.S. nail. Mus. 56 : 195-292.
1936. Zoological results of the George Vanderbilt African Expedition of 1934. Part III.
The freshwater fishes. Proc. Acad. nat. Sci. Philad. 88 : 243-335.
FRANZ, J. ig66b. Aphyosemion cinnamomeum. Trop. Fish Hobbyist 14 (9) : 11-16, 3 figs.
FRYER, F. 1959. The trophic interrelationships and ecology of some littoral communities of
Lake Nyasa with especial reference to the fishes, and a discussion of the evolution of a group
of rock-frequenting Cichlidae. Proc. zool. Soc. Lond. 132 : 153-281.
GILL, T. 1862. On the West African genus Hemichromis and descriptions of new species in
the Museums of the Academy and Smithsonian Institution. Proc. Acad. nat. Sci. Philad.
1862 : 134-139.
GMELIN, J. F. 1789. Linnaeus. Sy sterna Naturae, Edn XIII, vol. i.
GOSSE, L. 1963. Le milieu aquatique et 1'ecologie des poissons dans la region de Yangambi.
Ann. Mus. r. Afr. Centr. 8° Zool. no. 116 : 113-249, pis. i-x.
GREENWOOD, P. H. 1953. Feeding mechanism of the cichlid fish Tilapia esculenta Graham.
Nature. Lond. 172 : 207, i fig.
416 E. TREWAVAS
GREENWOOD, P. H. 1962. The identity and status of Barbus camptacanthus var. liberiensis
Steindachner, 1894 (Pisces, Cyprinidae). Bull, aquat, Biol. 3 : 121-125, * text-fig.
- 1970- A revision of the cyprinid species Barbus (Enteromius] radiatus Peters, 1853, with a
note on the synonymy of the subgenera Beirabarbus and Enteromius. Rev. Zool. Bot. afr.
82:1-13.
GUNTHER, A. 1865. Pisces, in Zool. Rec. for 1864.
- 1866. Catalogue of Fishes, vol. IV. xv + 388 pp.
- 1867. New fishes from the Gaboon and Gold Coast. A nn. Mag. nat. Hist. (3) 20 : 110-117.
- 1868. Catalogue of Fishes, vol. VII.
GYLDENHOLM, A. O. & SCHEEL, J. J. 1971. Chromosome numbers of fishes, i. /. Fish
Biol. 3 : 479-486.
HOLLY, M. 1926. Einige neue Fischformen aus Kamerun. Am. Akad. Wiss. Wien, 1926 :
155-159.
- i927a. Mormyriden, Characiniden und Cypriniden aus Kamerun. Sber. Akad. Wiss. Wien
(i) 136 : 115-150.
Siluriden, Cyprinodontiden, Acanthopterygier und Mastacembeliden aus Kamerun.
Sber. Akad. Wiss. Wien (i) 136 : 195-230, 7 text-figs.
- I927C. Weitere Fische aus Kamerun. Sber. Akad. Wiss. Wien (i) 136 : 422-427.
- 1928. Fische aus dem Scharifluss, gesammelt von Herrn A. Weidholz, darunter drei noch
nicht beschriebene Formen. Zool. Anz. 79 : 183-190, figs. 1-3.
- 1930. Synopsis der Siisswasserfische Kameruns. Sber. Akad. Wiss. Wien 139 : 195-281.
HOPSON, A. J. 1965. Four new species of Barbus (Pisces, Cyprinidae) from Ghana. Rev.
Zool. Bot. afr. 71 : 245-257.
HOPSON, A. J. & HOPSON, J. 1965. Barbus (Pisces, Cyprinidae) of the Volta region. Bull.
Br. Mus. nat. Hist. Zool. 13 : 101-149.
ILES, T. D. & HOLDEN M. J. 1969. Biparental mouth brooding in Tilapia galilaea. J. Zool.
158 : 327-333-
IMHOF, K. 1935. Morphologic und Histologie der Mund- und Kiemenhohle der Blenniidae.
Z. wiss. Zool. 147 : 1-58, 12 text-figs.
JOHNELS, A. G. 1954. Notes on fishes from the Gambia River. Ark. Zool. (2) 6 : 327-411.
KEILHACK, L. 1910. Uber einige von Herrn Dr H. Monke in Duala (Kamerun) gesammelte
Fische. Mitt. zool. Mus. Berl. 5 : 119-124.
KENNY, J. S. 1969. Feeding mechanisms in anuran larvae. /. zool. Lond. 157 : 225-246,
2 pis, 1 6 text-figs.
LONNBERG, E. 1895. Notes on fishes collected in the Cameroons by Mr Y. Sjostedt. Ofvers
K. Vetensk-Akad., Fork. Stockholm No. 3 : 179-195.
- 1903. On a collection of fishes from the Cameroon containing new species. Ann. Mag.
nat. Hist. (7) 12 : 37-46.
- 1904. On some fishes from the lakes of the Cameroon Mountain. Ann. Mag. nat. Hist.
(7) 13: I35-I39.
LOISELLE, P. V. 1971. The biology of Epiplatys sexfasciatus Gill, 1862 (Teleostomi : Cyprino-
dontidae : Rivulinae) in Southeastern Dahomey and the Togo Hills. /. Amer. Killifish
Ass. 7 : 29-53, 6 figs.
- 1972. Alestes longipinnis, the African longfin tetra. African Aquarist 5 (n) : 13-15,
3 figs and (12) : 13-14.
MARCUSE, G. 1971. Tilapia mariae. Tropical Fish World 2 (4) : cover and p. 8.
MEINKEN, H. 1929. Uber eine neue Panchax-Art (Panchax vexillifer n. sp.). Bl. Aquar.
Terrarienk. 40 : 385-389, 2 figs.
- 1932. Die calliurus Gruppe der Gattung Panchax. Aquarium, Berl. 1932 : 3-5, pi. xviii.
MYERS, G. S. 1924. New genera of African poeciliid fishes. Copeia, no. 129 : 41-48.
- X933- The genera of Indo-Malayan and African cyprinodont fishes related to Panchax and
Nothobranchius. Copeia, 1933 : 180-185.
- 1938. Studies on the genera of Cyprinodont fishes. XIV. Aplocheilichthys and its
relatives in Africa. Copeia, 1938 : 136-143.
FISHES OF WEST CAMEROON 417
NORMAN, J. R. 1932. A collection of fishes from Sierra Leone. Ann. Mag. nat. Hist. (10)
10 : 180-185.
OESER, R. 1928. Westafrikareise II. Bl. Aquar. Terrarienk. 39 : 373-376.
OSSE, J. W. M. 1969. Functional morphology of the head of the perch (Perca fluviatilis L.) :
an electromyographic study. Neth. J. Zool. 19 : 289-392, 15 pis, 24 text-figs.
PAPPENHEIM, P. 1905. Ein neuer 'pseudo'-elektrischer Fisch aus Kamerun. Sber. Ges.
naturf. Fr. 1905 : 217-219.
- 1907. Zur Systematik und Variationsstatistik der Mormyriden, hauptsachlich aus den
deutsch-afrikanischen Schutzgebieten. Mitth. zool. Mus. Berlin 3 : 339-367, pis 11-13.
PELLEGRIN, J. 1900. Poissons nouveaux ou rares du Congo fran9ais. Bull. Mus. Hist. nat.
6 : 348-354-
- 1904. Contribution a 1'etude anatomique, biologique et taxonomique des poissons de la
famille des cichlides. Mem. Soc. zool. Fr. 16 : 41-401, pis. iv-vii.
- 1907. Sur une collection de poissons recueillie par M. E. Haug a Ngomo (Ogooue). Bull.
Soc. philomath. Paris (N.S.), 9 : 17-42, i pi.
- 1909. Description d'un Barbus nouveau du Sahara. Bull. Mus. nat. Hist. Paris
1909 : 239-240.
Sur un cichlide nouveau de 1'Ogooue appartenant au genre Pelmatochromis. Bull.
Soc. zool. Fr. 44 : 102-105.
- igigb. Poissons du Gribingui recueillis par M. Baudon. Description de sept especes
nouvelles. Bull. Soc. zool. Fr. 44 : 201-214.
- 1925. Description d'un barbeau nouveau du Niger. Bull. Mus. Hist. nat. Paris, 1925 :
281-282.
- 1927. Mormyrides du Cameroun recueillis par M. Th. Monod. Description d'un genre,
de quatre especes et d'une variete. Bull. Soc. zool. Fr. 52 : 294-300.
- i929a. Description d'une variete nouvelle du Labeo chariensis Pellegrin, recueillie au
Cameroun par M. Th. Monod. Bull. Soc. zool. Fr. 54 : 288-289.
- i92gb. Silurides, Cyprinodontides, Acanthopterygiens du Cameroun recueillis par M.
Th. Monod. Description de cinq especes et deux varietes nouvelles. Bull. Soc. zool. Fr.
54 : 358-369-
- 1936. Contribution a 1'ichtyologie de 1'Angola. Arq. Mus. Bocage, 7 : 45-62.
PETERS, W. C. H. 1857. Neue Chromiden Gattung. Mber. K. preuss. Akad. Wiss. 1857 :
4°3-
- 1868. Ueber eine von dem Baron Carl von der Decken entdeckte neue Gattung von Welsen
(Chiloglanis deckeni) und einige andere Siisswasserfische aus Ostafrika. Mber. k. preuss.
Akad. Wiss. 1868 : 598-602.
POLL, M. 1952. Poissons de rivieres de la region des lacs Tanganika et Kivu recueillis par G.
Marlier. Rev. Zool. Bot. afr. 46 : 221-236, figs 1-6 and map.
- 1954- Poissons de foret des environs de Yangami (Stanleyville) recueillis par A. Hulot.
Ann. Mus. Congo in 8° Zool. 1 : 56-68.
- 1971. Revision des Synodontis africains (famille Mochocidae). Ann. Mus. r. Afr. centr.
8° Zool. no. 191. 497 pp., xiii pis.
POLL, M. & GOSSE, J. P. 1963. Contribution a 1'etude systematique de la faune ichthyologique
du Congo Central. Ann. Mus. r. Afr. centr. 8° Zool. no. 116 : 43-101, 4 pis.
POLL, M. & LAMBERT, J. G. 1965. Contribution a 1'etude systematique et zoogeographique
des Procatopodinae de 1'Afrique centrale (Pisces, Cyprinodontidae) . Bull. A cad. r. Sci.
Outre-Mer, 1965 : 615-631.
POLL, M. & THYS VAN DEN AUDENAERDE, D. 1967. Description de Pelmatochromis schoutedeni
sp. n. du Congo oriental (Pisces, Cichlidae). Rev. Zool. Bot. afr. 75 : 383-391, fig. i.
RADDA, A. C. i97oa. Auf Fischfang in Kamerun. Aquar. Terrar. Z. Jg. 23 : 237-241,
illustr.
- i97ob. Die Epiplatysarten Ostguineas, mit vorlaufigen Beschreibung von Epiplatys
sexfasciatus rathkei nov. subspec. Aquaria, 4 : 145-152, 8 text-figs.
- i97oc. Vorlaufige Beschreibung von Aphyosemion scheeli nov. spec. Aquaria, 4 : 177-180.
4i8 E. TREWAVAS
RADDA, A. C. 19713. Cyprinodontidenstudien im siidlichen Kamerun. i. Das Gebiet um
Kribi. Aquaria, 5 : 77-88, 9 figs.
— I97ib. Cyprinodontidenstudien im siidlichen Kamerun. 2. Das Tiefland der Kiiste.
Aquaria, 5 : 109-121, 10 figs.
19710. Cyprinodontidenstudien im siidostlichen Kamerun. 3. Das Gebiet um Kumba.
Aquaria, 5 : 125-134, 8 figs.
RAUTHER, M. & LEINER, M. 1937. Echte Fische. 5. Lief., in Bronn's Klassen u. Ordnungen
des Tierreichs 6, Abt. i, Buch 2 : 629-910.
REED, W., BURCHARD, J., HOPSON, A. J., JENNESS, J. & YARO, I. 1967. Fish and fisheries
of Nigeria. Min. Agric. Northern Nigeria. 226 pp.
REGAN, C. T. 1922. The classification of the fishes of the family Cichlidae. II. On African
and Syrian genera not restricted to the Great Lakes. Ann. Mag. nat. Hist. (9) 10 : 247-264.
ROLOFF, E. 1972. Aphyosemion oeseri H. Schmidt 1928. Aquar. Terrar. Z. Jg. 25 : 381-383,
4 figs.
ROMAN, B. 1971. Peces de Rio Muni, Guinea Ecuatorial (Aguas dulces y salobres). Barcelona.
295 pp, 118 text-figs (6 col.), 5 col. pis.
SAUVAGE, E. 1882. Notice sur les poissons du territoire d'Assinie (Cote d'Or). (Mission
scientifique de M. Chaper.) Bull. Soc. zool. Fr. 7 : 313-325, pi. v.
SCHEEL, J. J. 1964. Aphyosemion cinnamomeum. Aquar. J. 35 (7) : 324-328, 6 figs.
ig66a. Notes on phenotypy, distribution, and systematics of Aphyosemion bivittatum
Loennberg), with remarks on the chromosome number in the Rivulinae. Aquar. J . 38 :
261-278, figs 1-15.
ig66b. Taxonomic studies of African and Asian tooth-carps (Rivulinae) based on chromo-
some numbers, haemoglobin patterns, some morphological traits and crossing experiments.
Vidensk. Meddr. dansk. naturh. Foren. 129 : 123-148, pis vi-xviii.
I968a. Rivulins of the Old World. Jersey City : T.F.H. Publ. 473 pp., indices and many
figs.
I968b. A new species of Aphyosemion (Cyprinodontidae, Pisces) from Fernando Poo.
Revue Zool. Bot. afr. 78 : 332-342, i fig.
1969. Zur Identifizierung voaOesers Aphyosemion. Aquar. Terrar. Z. Jg. 22 : 74-76, i fig.
1971. Aphyosemion franzwerneri and Aphyosemion celiae, two new rivulins from Cameroon.
Trop. Fish Hobbyist 19 (Jan.) : 48-49, 52-57, 60-66.
i972a. Rivuline karyotypes and their evolution (Rivulinae, Cyprinodontidae, Pisces).
Z. zool. Syst. Evolut.-forsch. 10 : 180-209, mcl- P^8- i~v> text-figs.
I972b. Cichlid chromosomes : the nominal genus Pelmatochromis. Advanced Aquarists'
Mag. May, 1972 : 6-7, 2 figs.
SCHMIDT, H. 1928. Neuimportierte Fische. Aquarium, Berlin, 1928 : 165-166, figs.
SCHOELLER, C. H. 1903. Ein neuer Chromis. Bl. Aquar. Terrarienk. 14 : 185-187, 203-206,
fig. on p. 185.
STARKS, E. C. 1926. Bones of the ethmoid region of the fish skull. Stanford Univ. Publ.
Biol. 4 : 139-338, 58 text-figs.
STEINDACHNER, F. 1894. Die Fische Liberia's. Notes Leyden Mus. 16 : 1-96, pis. i-iv.
1914. Zur Fischfauna des Dscha, eines sekundaren Nebenflusses des Kongo, im Bezirke
Molundu, Kamerun. Denkschr. Akad. Wiss, Wien, 89 : 1-64 pis. i-ix.
STENHOLT-CLAUSEN, H. 1959. Description of two subgenera and six new species of Procatopus
Boul., a little-known West African genus of cyprinodont fishes. Vidensk. Meddr. dansk.
naturh. Foren. 121 : 261-291, 10 figs.
1963. Description of three new species of Aphyosemion Myers from Nigeria and Cameroun.
Vidensk. Meddr. dansk. naturh. Foren. 125 : 195-206, i pi, i fig.
STERBA, G. 1959. Susswasserfische aus oiler Welt. Uranie-Verlag, Leipzig/Jena.
1962. Freshwater Fishes of the World. English version by D. W. Tucker. London : Vista
Books.
STOLK, A. 1957. Pharyngeal glands of the cichlid Haplochromis multicolor (Hilgendorf) .
Proc. K. ned. Akad. Wet. 60 C : 567-577, 15 figs.
FISHES OF WEST CAMEROON 419
TAVERNE, L. igyia. Note sur la systematique des poissons Morymriformes. Le probleme
des genres Gnathonemus Gill, Marcusenius Gill, Hippopotamyrus Pappenheim, Cyphomyrus
Myers et les nouveaux genres Pollimyrus et Brienomyrus. Rev. Zool. Bot. afr. 84 : 99-110.
- 19710. Osteologie des genres Marcusenius Gill, Hippopotamyrus Pappenheim, Cyphomy-
rus Myers, Pollimyrus Taverne et Brienomyrus Taverne. (Pisces, Mormyriformes.) Ann.
Mus. r. Afr. centr. 8° Zool. no. 188. vii + 144 pp., 3 pis.
THOMINOT, A. 1886. Sur quelques poissons nouveaux appartenant a la collection du Museum
d'Histoire Naturelle. Bull. Soc. philomath. Paris (7) 10 : 161-171.
THYS VAN DEN AUDENAERDE, D. F. E. 1965. List of the freshwater fishes presently known
from the island of Fernando Poo. Bonn. zool. Beitr. Jg. 16 : 316-317.
- 1967. The freshwater fishes of Fernando Poo. Verh. k. vlaamse Acad. Wet. Lett. Sch.
Kunst. Belgie (Wet.), Jg. 29 : no. 100. 167 pp. 47 figs.
- ig68a. A preliminary contribution to a systematic revision of the genus Pelmatochromis
sensu lato (Pisces, Cichlidae). Rev. Zool. Bot. afr. 72 : 349-391, 17 figs, 2 tables.
ig68b. An annoted bibliography of Tilapia (Pisces, Cichlidae). Documentn. zool. Mus.
r. Afr. centr. No. 14, xl + 4o6 pp.
1971. Some new data concerning the Tilapia species of the subgenus Coptodon (Pisces,
Cichlidae). Rev. Zool. Bot. afr. 84 : 160-168, i fig.
1972. Description of a small new Tilapia (Pisces, Cichlidae) from West Cameroon. Rev.
Zool. Bot. afr. 85 : 93-98, figs i & 2.
TREWAVAS, E. 1962. Fishes of the crater lakes of the northwestern Cameroons. Bonn.
Zool. Beitr. 1962 : 146-192.
1973- On the cichlid fishes of the genus Pelmatochromis with proposal of a new genus
for P. congicus ; on the relationship between Pelmatochromis and Tilapia and the recognition
of Sarotherodon as a distinct genus. Bull. Br. Mus. nat. Hist. (Zool.) 25 : 1-26, 14 text-figs.
TREWAVAS, E., GREEN, J. & CORBET, S. A. 1972. Ecological studies on crater lakes in West
Cameroon. Fishes of Barombi Mbo. /. Zool. Lond. 167 : 41-95, pi. i.
Voss, J. 1969. Contribution a l'6thologie des poissons cichlides : Tilapia zillii (Gervais,
1848). Rev. Zool. Bot. afr. 79 : 99-109.
WHITEHEAD, P. J. P. 1958. A new species of Chiloglanis (Pisces, Mochocidae) in Kenya. Ann.
Mag. nat. Hist. (13) 1 : 197-208.
WHITEHEAD, R. A. 1962. The life history and breeding habits of the African cichlid fish
Tilapia mariae and the status of T. meeki. Proc. zool. Soc. Lond. 139 : 535-543, 2 pis.
E. TREWAVAS
Department of Zoology
BRITISH MUSEUM (NATURAL HISTORY)
CROMWELL ROAD
LONDON SW7 5BD
PLATE i
Mormyrus tapirus, two specimens respectively 132 and 173 mm in SL from R. Wowe.
Bull. Br. Mus. nat. Hist. (Zool.) 26, 5
PLATE i
-7
PLATE 2
Brienomyrus brachyistius. A, B and C from R. Mungo, near Etam ; D from R. Nganjoke.
The standard lengths are, A 78-5 mm, B 119 mm, C 174 mm and D 156 mm. Although the
length of the Nganjoke fish, D, is between those of B and C, its shape is nearer that of the much
smaller specimen, A. In A the epidermis of the top of the head has become detached.
Bull. Br. Mm. nat. Hist. (Zool.) 26, 5
PLATE 2
PLATE 3
Labeo camerunensis , paratypes (adult and young).
Bull. Br. Mus. nat. Hist. (Zool.) 26, 5
PLATE 3
PLATE
Sections through the pharyngeal pad of, A and B, Chromidotilapia g. loennbergi and, C and D,
Tilapia mariae. In A and B a hillock bearing taste buds is seen at different magnifications. In
A some small mucus-cells can be seen near the surface of the epithelium. C and D are different
parts of the same pad ; in C two taste buds have been grazed, in D one, but the most noticeable
elements are the deep secretory cells that occupy the greater part of the epithelium.
Bull. Br. Mus. nat. Hist. (Zool.) 26, 5
PLATE 4
PLATE 5
Chromidotilapia finleyi, female. Photograph by Mr Lee Finley of Bridgeport, Conn., U.S.A.
The specimen was taken within a 24 km radius of Kumba and belonged to Mr Ray Heimbuch
of Westhaven, Conn.
Bull. Br. Mus. nat. Hist. (Zool.) 26, 5
PLATE 5
A LIST OF SUPPLEMENTS
TO THE ZOOLOGICAL SERIES
OF THE BULLETIN OF
THE BRITISH MUSEUM (NATURAL HISTORY)
1. KAY, E. ALISON. Marine Molluscs in the Cuming Collection British Museum
(Natural History) described by William Harper Pease. Pp. 96 ; 14 Plates.
1965. (Out of Print.) £3.75. "
2. WHITEHEAD, P. J. P. The Clupeoid Fishes described by Lacepede, Cuvier and
Valenciennes. Pp. 180 ; n Plates, 15 Text-figures. 1967. £4.
3. TAYLOR, J. D., KENNEDY, W. J. & HALL, A. The Shell Structure and Mineralogy
of the Bivalvia. Introduction. Nuculacea-Trigonacea. Pp. 125 ; 29 Plates
77 Text-figures. 1969. £4.50.
4. HAYNES, J. R. Cardigan Bay Recent Foraminifera (Cruises of the R.V. Antur)
1962-1964. Pp. 245 ; 33 Plates, 47 Text-figures. 1973. £10.80.
5. WHITEHEAD, P. J. P. The Clupeoid Fishes of the Guianas. Pp. 227 ; 72
Text-figures. 1973. £9.70.
Printed in Great Britain by John Wright and Sons Ltd. at The Stonebridge Press, Bristol BS4 jNU
22 JUL1974
NOTES ON SOME ECHINODERMS
FROM SOUTHERN AFRICA
A. M. CLARK
BULLETIN OF
THE BRITISH MUSEUM (NATURAL HISTORY)
ZOOLOGY Vol. 26 No. 6
LONDON: 1974
22 JULJ97.
NOTES ON SOME ECHINODERMS
FROM SOUTHERN AFRICA
BY
AILSA M. CLARK
Pp. 421-487 ; 3 Plates ; 16 Text-figures ; i Map ; 3 Tables
BULLETIN OF
THE BRITISH MUSEUM (NATURAL HISTORY)
ZOOLOGY Vol. 26 No. 6
LONDON: 1974
THE BULLETIN OF THE BRITISH MUSEUM
(NATURAL HISTORY), instituted in 1949, is
issued in five series corresponding to the Departments
of the Museum, and an Historical series.
Parts will appear at irregular intervals as they
become ready. Volumes will contain about three or
four hundred pages, and will not necessarily be
completed within one calendar year.
In 1965 a separate supplementary series of longer
papers was instituted, numbered serially for each
Department.
This paper is Vol. 26, No. 6, of the Zoological
series. The abbreviated titles of periodicals cited
follow those of the World List of Scientific Periodicals.
World List abbreviation :
Bull. Br. Mus. not. Hist. (Zool.)
© Trustees of the British Museum (Natural History), 1974
TRUSTEES OF
THE BRITISH MUSEUM (NATURAL HISTORY)
Issued 7 June, 1974 Price £3.40
NOTES ON SOME ECHINODERMS
FROM SOUTHERN AFRICA
By AILSA M. CLARK
SYNOPSIS
The main part of this paper deals with noteworthy species of echinoderms (excepting holo-
thurians) from around the mainland of southern Africa south of the Tropic of Capricorn, with
appendices on a few species from off-lying localities including the Vema Seamount, Walter's
Shoal, a peak of the South- West Indian Ocean Ridge and Madagascar. Most of the specimens
came from the Ecological Survey of the University of Cape Town but some were collected in
conjunction with the International Indian Ocean Year, notably by the 'Anton Bruun' ; a few
from the British Museum collections are also dealt with. Three new species of ophiuroids are
described and the ranges of four other ophiuroids and two unstalked crinoids are extended to
southern Africa. A lectotype is selected for Astropecten granulatus natalensis John and taxono-
mic changes are made to several nominal species of other Asterozoa, including transfers to
different genera, alterations in rank and synonymies (see p. 483).
INTRODUCTION
THIS paper brings up-to-date details of noteworthy echinoderms (excepting holo-
thurians) collected since I last reported on South African echinoderms in 1952 by
Professor J. H. Day and his co-workers at the University of Cape Town during their
long-term Ecological Survey of South African and adjacent waters. Following the
exhaustive studies of H. L. Clark (1923) and Mortensen (1933) it is not surprising
that only three undescribed species - all ophiuroids - were found. Tantalizingly,
no further material came to light of some of the little-known species such as the
crinoids Decametra durbanensis A. M. Clark and Monachocrinus perrieri (Koehler
and Vaney), the asteroids Tethy aster pacei (Mortensen) (originally Anthosticte pacei) ,
Ceramaster trispinosus H. L. Clark, Anseropoda novemradiata (Bell) and Mithrodia
gigas Mortensen, or the ophiuroids Amphiura acutisquama A. M. Clark, A. linearis
Mortensen, Ophiogymna capensis (Liitken) and Ophioderma wahlbergi Miiller and
Troschel.
The main part of the text deals with specimens from around the mainland of
southern Africa south of the Tropic of Capricorn (c. 23 1° S), which just extends to
include Inhambane in Mozambique on the east coast but not quite to Walvis Bay in
South-West Africa on the west. There are six species new to the area, the crinoid
Oligometra serripinna (P. H. Carpenter) and the ophiuroids Anamphiura valida H. L.
Clark and Amphiophiura trifolium Hertz extending to Natal from Mauritius, the
Amirante Islands, the Zanzibar area and Mombasa respectively, while the crinoid
Tropiometra magnified A. H. Clark with the ophiuroids Ophiocirce inutilis Koehler
and Ophiopallas paradoxa are extended to southern Mozambique from localities
further north in East Africa.
Since this paper went to press I have received from M. Jangoux a copy of his important paper 'Les
Asteries de 1'Ile d'Inhaca (Mozambique) (Echinodermata, Asteroidea). i. Les especes recoltees et leur
repartition geographique.' Ann. Mus. R. Afr. Cent., ser. 8°, Zool. No. 208: 1-50, 13 figs., 7 pis.
424
AILSA M. CLARK
In addition, a few specimens from the off-lying Vema Seamount, Walter's Shoal,
Madagascar and the South-west Indian Ocean Ridge are dealt with in appendices.
Luderitz Bay
30 ^ Richard's Bay
Orange R.
\ x Ridge
JvPort Nolloth Walter's
Peak
Hondeklip Bay
Lambert's Bay
Saldanha Bay
33 > Tugela R.
The Haven
Dassen I.
Table Bay
Cape Point
False Bay
Cape Hangklip
Hermanus
Algoa Bay
Cape St Francis
Cape Agulhas
Mossel Bay
21 23 \ x Still Bay
Breede R.
100 metres
500 metres
MAP Southern Africa with inset of larger area including Madagascar, showing the three
offshore localities dealt with' in Appendices I, II and IV. The even numbers represent
latitude S and the odd ones longitude E.
Class CRINOIDEA
Family GOLOBOMETRIDAE
Oligometra serripinna occidentalis A. H. Clark
Oligometra serripinna var. occidentalis A. H. Clark, 1911 : 33-34 ; 1947 : 239-240, pi. 26, fig. 136,
pi. 28, figs. 150-151, pi. 30, fig. 163.
MATERIAL. AFR 1028 J, 28°28'S : 32°25'E (off N. Natal), 27 metres ; 3 specimens.
NAD 2W, 30°47'i'S : 3O°29-i'E (S of Durban), 44 metres ; i specimen.
DESCRIPTION. The largest specimen from sample AFR 1028 has the centrodorsal
discoidal with the perfectly flat dorsal pole 2-3 mm in diameter. The cirri are
arranged in a single crowded and in some parts staggered ring around the edge.
The longest are c. 10 mm long with 22 segments, all shorter than broad.
The division series are narrow, rounded laterally, and the adjacent ones are well
separated from one another. The proximal arm syzygies are regularly placed at
brachials 3 + 4, 9 + 10 and 14+15. The breadth at the first syzygy is i-i mm and
the length from the proximal edge of the IB^ to 9 + 10 is 5-5-6-0 mm.
The proximal pinnules are markedly prismatic but not the distal ones. The
latter have about 21 segments, most of which are half again to twice as long as broad
ECHINODERMS FROM SOUTHERN AFRICA 425
and they measure 5-5-6-0 mm in length. P2 is much stouter than either Px or P3
and only tapers beyond the proximal half. The last three or four segments are small.
The arms are patterned with dark stripes, especially at some of the joints. These
are maroon in spirit but in life the colour was banded yellow and brown.
Some numerical details of this and two other specimens from Natal are given in
Table I together with those for two specimens from Cargados Carajos, north of
Mauritius, seen earlier in the British Museum collections by A. H. Clark. The large
TABLE i
Numerical data from three specimens of Oligometra serripinna occidentalis from
Natal and two from Cargados Carajos, in that order
Arms
Length Breadth Cirri PI P2 P3
at 3 + 4 No. Segs. Segs. L. Segs. L. Segs. L.
c. 50 i'i XX 16-22 14 4-0 16, 17 5-5 14 4-0
50 i-o XIX 17-21 15 4-2 16, 17 6-0 13 3-5
33 o*9 XVII 17-19 12 3-5 14 5'° 12 3-5
c. 80 1-4 XIX 22-29 J4 5'° J6 7-0 14 5-0
c. 55 i-i XVII 19-21 14 4'o 17 6-0 14 4-0
number of cirrus segments in the fourth specimen in the table is noteworthy but
smaller specimens from that locality agree in the counts with those from Natal. The
relative proportions of the proximal pinnules appear to bfe fairly consistent but their
shape is rather variable. In the smaller specimen from Cargados Carajos included
in the table the segments are shorter and much more ornate than usual in this
subspecies with flaring distal processes. However, another from the same locality
is indistinguishable from those from Natal in having the early pinnules very flat
and the last seven or eight segments slightly flared at the sharp outer edge distally,
each following segment being narrower so that the profile of the tapering part of the
pinnule is serrated without being markedly spinose.
The largest specimen from Cargados Carajos has the dorsal ridges of the middle
cirrus segments rising at each end to a slight peak and a similar modification is
shown in the smallest Natal specimen tabulated. In the other material examined
the ridges appear fairly level, in side view resembling a single small acute tubercle
arising at about the middle of the length of the segment or slightly towards the
distal end.
REMARKS. It is possible that Decametra durbanensis, which I described in 1951
on the basis of a single specimen also from Natal (29°30-8'S : 3i°23'E) in 68 metres
may prove to be synonymous with Oligometra serripinna occidentalis. The holotype
is relatively large with an arm length of 90 mm and breadth at 3 + 4 2-0 mm. The
cirri are very stout and the dorsal surfaces of the proximal segments are studded
with small tubercles, so that they appear very different to the smooth segments of
the present material. The centrodorsal is similarly discoidal with marginal cirri,
numbering as many as XXVII and with up to 32 segments. However, the segments
of the proximal pinnules are no more numerous than those of 0. serripinna occi-
dentalis, though the proportions of the first three pinnules are similar.
426 AILSA M. CLARK
In his key to the genera of Colobometridae (1947 : 9) A. H. Clark distinguished
Decametra from Oligometra in the following terms :
'&1 The proximal cirrus segments bear dorsally a transverse ridge which distally
becomes a pair of tubercles or small spines, or rarely a single median spine ; Pa is
absent ; P2 is rounded or rounded prismatic with the distal ends of the segments
uniformly spinous or smooth . ... . . . Decametra
k* Cirrus segments all with a transverse ridge dorsally except for the penultimate
which bears the opposing spine ; Pa present (or rarely absent on some arms) ;
P2 more or less strongly prismatic with the distal portion of the prismatic ridges
on each segment produced into more or less broad finlike processes, or with the
distal ends of the prismatic ridges bearing tufts or spines . . Oligometra'
In the holotype of Decametra durbanensis Pa is absent on all but two of the arms.
In all four specimens of Oligometra serripinna occidentalis from Natal Pa is present
throughout and the same is true (with the exception of single arms of two specimens)
in eight out of the nine from Cargados Carajos. However, in the ninth and largest
with the arm breadth 1-4 mm Pa is absent on all the arms and Pb is also missing on
two arms one of which even lacks Pc as well. A. H. Clark also noted that the two
small specimens from John Murray Expedition st. 45 (South Arabian coast) lack Pa ;
one of them has the arms only 25 mm long. Oligometra serripinna macrobrachius
from the East Indies regularly lacks Pa. In the Red Sea Decametra chadwicki may
have more arms with Pa than without, as I noted in 1967. Clearly this character
alone is unreliable as a generic distinction.
With regard to the dorsal armament of the cirrus segments, A. H. Clark included
in Decametra several species, notably the type species D. modica, in which the trans-
verse dorsal ridge of the middle segments never resolves itself into a pair of tubercles
on the distal segments. Unfortunately no examples of these species larger than
50 mm arm length have yet been described and it is possible that this character only
manifests itself with increasing size.
As for the development of processes or spines on the segments of P2, judging from
the descriptions of various specimens of Oligometra serripinna in A. H. Clark's
monograph, it is very variable and in the subspecies occidentalis it is usually very-
slight.
Oligometra serripinna is the type species of Oligometra. If further studies of these
three characters substantiate their unreliability as being of generic weight, then
Decametra A. H. Clark, 1911 will be better referred to the synonymy of Oligometra,
established three years earlier.
RANGE. These records extend the range of Oligometra serripinna south-west
from Mauritius into South African waters.
Family TROPIOMETRIDAE
Tropiometra magnified A. H. Clark
Tropiometra magnified A. H. Clark, 1937 : 90-91, pi- i, fig- r > J947 '• 266-268, pi. 33, figs. 170-
173 ; A. M. Clark, 1972 : 131-132.
ECHINODERMS FROM SOUTHERN AFRICA 427
MATERIAL. FED 5\V, 24°46'S : 35°i8'E (S of Inhambane, Mozambique), no
metres ; 2 broken specimens.
An error in my paper of 1972 concerning this species (p. 131, bottom line) needs
to be corrected. The cirrus length should be 'up to 80 mm' (not 4-0-4-5 mm).
RANGE. This record provides an extension of range southwards from Kenya and
the Gulf of Aden.
Family THALASSOMETRIDAE
Crotalometra magnicirra (Bell)
Antedon magnicirra Bell, 19050 : 141, pi. 4.
Crotalometra magnicirra : Gislen, 1938 : 17-18 ; A. H. Clark, 1950 : 97-100.
MATERIAL. This species is only represented in the present collections by a sample
of 5 specimens from the South-West Indian Ocean Ridge (see Appendix, p. 482); as it
has never been properly described, details are given here of a syntype, B.M. registered
number 1904.6.28.16-19 (part), Tieter Faure' no. 12885-6, East London, Buffalo
River bearing N 15 miles distant, 567 metres.
DESCRIPTION. The arm length is c. no mm ; the breadth at 3 + 4 is 1-7 mm and
the length from the proximal edge of the IBrx to the syzygy at 3 + 4, including a
IIBr series of four ossicles, is n mm.
The centrodorsal is rounded hemispherical ; 6-5 mm in basal diameter and 4-0 mm
across the convex dorsal pole. The cirrus sockets are in ten very definite columns,
two in each interradius, with a narrow bare wedge-shaped concave radial area
separating the pairs of columns from each other. There are three sockets in each
column, though the apicalmost one may be obsolete and the peripheral one may bear
only an immature cirrus. The number of mature cirri is therefore best estimated
at c. XXV. Of the few remaining mature peripheral cirri, most have c. 63 segments
and measure c. 55 mm in length, which is approximately half the arm length. The
ninth or tenth is the transition segment and is the longest, measuring c. 2-0 mm by
i-o mm median breadth. The proximal segments are hardly at all constricted in
the middle or flared distally. Beyond the tenth, the segments become relatively
shorter so that the seventeenth is as long as broad and the more distal ones are
broader than long. The shorter segments are all evenly flared towards their distal
ends on the dorsal side and end abruptly, so that the profile of the cirrus is very
serrated.
The division series have lateral flanges with straight edges so that adjacent ones
are in close apposition. The synarthrial tubercles are very low and rounded, hardly
evident at all. There were probably 20 arms (two IIBr series are broken). All the
IIBr series are 4 (3 + 4). The first brachial syzygy is at 3 + 4, the second at 17+18
or beyond. The distal intersyzygial interval is from four to eight muscular joints.
PD on IIBr2 has 22-27 segments and is c. 12 mm long. Px has c. 19 segments
and is c. 9 mm long. P2 with c. 12 segments is 5 mm long and P3 is similar to P2.
The distal pinnules have c. 14 segments and measure c. 6-5 mm. The proximal
428 AILSA M. CLARK
pinnules are markedly prismatic at the base, with a high dorsal crest which becomes
displaced after the first two or three segments towards the inner angle. The seg-
ments are short but flared distally and distinctly spinose. The gonads are not at
all enlarged. The distal pinnules are markedly prismatic, the segments not more
than twice as long as wide and with the side and covering plates distinct.
VARIATIONS. The 15 other syntypes of Crotalometra magnicirra are variable with
regard to certain characters. The centrodorsal is usually flattened hemispherical
but may resemble a truncated high cone with the apex more or less convex, or a
complete low cone about two-thirds as high as broad basally. The number of cirri
varies according to the size of the centrodorsal. There may be some columns with
four sockets but then the apicalmost is usually obsolete or the peripheral cirrus is
immature. The maximum number of cirri found is XXXIII but commonly there
are c. XXV.
One specimen has only ten arms, the longest remaining being 75 mm long, having
lost c. 15 mm. The breadth at 3 + 4 is 1-9 mm and the length from IBrt to 3 + 4
is only 5-0-5-5 mm, no IIBr series intervening. Most of the other specimens have
20 arms or at least 15. Only one has a IIBr series of only two ossicles, all the rest
are of four. The arm breadth at 3 + 4 is usually 1-8-2-0 mm and the length to this
syzygy including a IIBr4 series is 11-12 mm.
The longest cirrus found measures c. 60 mm and the greatest number of segments is
64. The few specimens with both a complete arm and a complete mature cirrus
show a ratio of just over 2:1.
In the ten-armed specimen P2 (equivalent to Px in the other specimens since no
PD occurs) has 12 segments and measures c. 5 mm ; P3 with n or 12 segments is
4-5 mm and P4 with n is 4-0 mm.
Two specimens from the same sample as the one described have a syzygy at
brachials i + 2 on some or all arms following a IIBr series, as noted by Gislen (1938).
One of them has only eight arms based on IIBr series remaining attached and six
of these have a syzygy at I + 2. In the second specimen all 14 arms of this kind
remaining also show such an initial syzygy. Both specimens are rather slender and
it may be significant that none of the stouter individuals share this distinction.
Several of the syntypes have an abrupt change to a paler colour at the first
brachial syzygy or sometimes at IIBr3 + 4, as if they have regenerated from that
point, indicating damage so drastic that survival and recovery is surprising when
feeding powers must have been badly impaired for a prolonged period.
Although the majority have the division series and brachials quite smooth, in
one specimen there are distinct thorns along the edges of the joints.
REMARKS. This species is sympatric with Glyptometra sclateri (Bell), with which
it has a number of characters in common. Glyptometra belongs to the family
Charitometridae and Crotalometra to the Thalassometridae but both of these fall
within the superfamily Tropiometrida. The two species both have conspicuous side
and covering plates along the ambulacra, prismatic distal as well as proximal
pinnules, usually more than 10 arms with the IIBr series of four ossicles, well-plated
discs and laterally flared division series and proximal brachials. Apart from the
ECHINODERMS FROM SOUTHERN AFRICA 429
very different cirri, the long ones of Crotalometra having a distinct transition segment,
the present material suggests that Crotalometra magnicirra can be distinguished from
G. sclateri by the normal occurrence of the first brachial syzygy at brachials 3 + 4
(though 1 + 2 may occur) and by the smaller pinnules, the genital ones of Crotalo-
metra having only a vestige of the lateral expansion so distinct in Glyptometra. Also
the centrodorsal seems to be relatively higher in most specimens of C. magnicirra,
the height usually about two-thirds the basal diameter, while the shape may be
quite conical in some specimens.
Family CHARITOMETRIDAE
Glyptometra sclateri (Bell)
Antedon sclateri Bell, 19050 : 140, pi. 3.
Pachylometra sclateri (pt.) : H. L. Clark, 1923 : 234 [armless adult only ; non P. sclateri (?) :
Gisl6n, 1938 : 18-20, which represents Gislenometra perplexa A. H. Clark, 1947, like H. L.
Clark's small specimens.]
Glyptometra sclateri : A. H. Clark, 1950 : 268-270.
MATERIAL. Again this species is only represented in the present collections by
material from the South-West Indian Ocean Ridge (see Appendix, p. 482) but since
it has never been fully described details are given here of a syntype, British Museum
registered number 1904.7.3.6-10 (part), 'Pieter Faure' no. 12711, East London
bearing NW ^ N distant 18 miles, 457-549 metres.
DESCRIPTION. The arms are all broken but were probably at least 80 mm long ;
the breadth at 3 + 4 is 2-0 mm and the length from the proximal edge of the IBrx to
the syzygy at 3 + 4, including a IIBr series of four ossicles, is 9-0 mm.
The centrodorsal is flattened hemispherical, the rugose dorsal pole 3-5 mm in
diameter, the peripheral diameter 6-0 mm and the height 2-3 mm. Some of the cirrus
sockets seem to be arranged in columns but others are crowded out of alignment.
The cirri number c. XXXV and have up to 18 segments, their length being at the
most 22 mm. The longest segments are the sixth to the eighth, the sixth on one
cirrus measuring 1-8 mm in median length and i-i mm median breadth. The follow-
ing segments are relatively shorter, except for the penultimate, which is abruptly
narrower than the antepenultimate ; its opposing spine is terminal and directed
distally, as is usual in the family Charitometridae. The antepenultimate and pre-
ceding short segments are all slightly flared at their distal ends, the more proximal
ones especially on the ventral side, the distal ones on the dorsal side.
The division series have lateral flanges with straight sides so that the adjacent
ones are closely apposed. The synarthrial tubercles are low and barely appreciable.
There are 21 arms. All the IIBr series are 4 (3 + 4) and the only IIIBr series is
2 (1 + 2). The proximal syzygies are rather irregular in position ; on 12 arms the
first is at 1 + 2 with the following joint a muscular one, on eight arms there are
syzygies at both 1 + 2 and 3 + 4, while the last arm has a syzygy at 2 + 3. The
brachials of the outer half of the arm are triangular, approximately as long as wide.
The intersyzygial interval is usually five, sometimes six, muscular articulations.
430 AILSA M. CLARK
PD on IIBr2 has 28-30 segments and is c. 10 mm long. It is curled inwards over
the disc and has short, squarish segments. Pl has c. 27 segments and is 9-5-10-0 mm
long, rather similar to PD. P2 has 19 or 20 segments and length c. 8 mm. It is
deeper than PD and P1 at the base but tapers more. The segments after the first
six or seven are longer than broad but none are as much as twice as long as broad.
P3 is the first genital pinnule ; it has c. 15 segments and is 5-5-6-0 mm long. The
first three segments taper like those of P2 but the fourth expands markedly and the
fifth and sixth are conspicuously wide, the seventh tapers abruptly and the following
segments are much longer, the distal ones twice as long as broad. P4 and the other
genital pinnules have a similar expansion of a few of the proximal segments after
an initial tapering.
The disc is densely studded with small granuliform plates.
VARIATIONS. Of the 13 other specimens in the British Museum collection - 12
syntypes and one from among the syntypes of Crotalometm magnicirra - the arm
number is as follows : 10, n, n, 13, 14, 15, 17, 17, 17, 18, 19, 20 and 20. In the
ten-armed specimen, the arms are 90 mm long. The smallest specimen is an eleven-
armed one with an arm length of c. 80 mm and breadth at 3 + 4 of 1-8 mm. Most of
the specimens have all the IIBr series of four ossicles but one has two out of seven
IIBr series of two ossicles and another has one out of ten similarly abbreviated. The
synarthrial tubercles may be much more prominent than in the specimen described.
The occurrence of proximal arm syzygies is very variable ; a few specimens have
the first syzygy at 3 + 4 on some arms arising from IIBr series, rather than the usual
1 + 2, as well as on the arms arising from IBr series. One specimen with 19 arms is
unusual in having six with 2 + 3, though nine others have 1 + 2, three have both 1 + 2
and 3 + 4 and one (the only arm arising from a IBr series) has 3 + 4 alone. Another
specimen with 20 arms has none with 2 + 3, but 16 with 1 + 2, three with 1 + 2 and
3 + 4 and one with 3 + 4 alone.
The centrodorsal varies in shape from thick discoidal to flattened hemispherical
but is never as conical as in some specimens of Crotalometra magnicirra ; the dorsal
pole ranges from as little as 2-0 mm in diameter to as much as 5-0 mm. There are
usually ten cirri around the dorsal pole and in some specimens, particularly the
smaller ones, the other cirri may be arranged in quite regular columns in line with the
adapical ones. However, in larger individuals there are usually more than ten
peripheral cirri and the sockets become crowded irregularly in the interradii, though
sometimes they leave bare a space in each radius, which may be wedge-shaped and
concave as in C. magnicirra. One specimen with a particularly regular columnar
arrangement of the cirrus sockets and well-marked radial spaces has all its cirri
broken off short and can only be distinguished from C. magnicirra by the irregularity
of the proximal arm syzygies and the larger pinnules with some expanded segments
on the genital ones.
The proximal pinnules vary to some extent. One has PD with 23 segments, Px
with 19, P 2 with 1 6 and P3 with 13 ; while another has P2 on an arm arising from a
IBr series with 22 segments and length 9-5 mm. The ten-armed specimen has P!
with 22 segments but only 6-5 mm long.
ECHINODERMS FROM SOUTHERN AFRICA 431
Class STELLEROIDEA
Subclass ASTEROIDEA
Family LUIDIIDAE
Luidia sp. cf. L. avicularia Fisher
See : Fisher, 1919 : 172-175, pi. 43, fig. i, pi. 44, fig. 2, pi. 46, fig. 2.
MATERIAL. SCD 74!*, 32°33'S : 28°38'E (off Bashee River mouth, N of East
London), 55 metres ; i specimen.
Unfortunately this nine-armed Luidia has had all its arms broken (or bitten) off
short and regeneration is only partial. The large bivalved pedicellaria on the oral
face of each jaw plate and the similar but sometimes smaller one on the proximal
abradial ends of most adambulacral plates, distinguish it from the similarly nine-
armed L. maculata Muller & Troschel, which has been recorded from Natal by
Mortensen, 1933, the latter having three-bladed pedicellariae on the multiple rather
than single actinal plates. The present specimen also differs in having an enlarged
spinelet or spine on most of the lateral abactinal paxillae, as in L. avicularia, known
from the Philippines and Banda Sea, which also agrees in the form of the pedicel-
lariae. Further specimens are needed to settle the specific identification.
Family ASTROPEGTINIDAE
Astropecten granulatus natalensis John
PI. i, fig. i
Astropecten granulatus : H. L. Clark, 1923 : 250-251.
Astropecten granulatus natalensis John, 1948 : 5, pi. i, figs, i, 2.
MATERIAL. SCD 74(3, 32°33'S : 28°38'E (off Bashee River), 55 metres ; i specimen.
NAD 8R, 29°53'6'S : 3i°o4-6'E (off Durban), 38 metres ; 3 specimens.
NAD I2D, 29° 46'S : 3i°i7'E, 110-130 metres ; i specimen.
NAD 26B, 29°53'5'S : 3i°o6-i'E, 71 metres ; 2 specimens.
NAD 52A, 29°29'S : 3i°45'E, 86 metres ; 5 specimens ['Anton Bruun' st. 39iC].
DESCRIPTION. The specimen from East London is much larger than the others
so far recorded from South Africa, in which R was less than 40 mm ; it has R
72-74 mm ; r is 16 mm and br (at the fourth superomarginals) also 16 mm ; the
paxillar breadth in the same position is n mm.
There are 34 superomarginals in a complete series, all lacking spines. The
inferomarginal plates each bear two large almost equal spines at the upper end,
aligned slightly obliquely, with a third fairly large spine proximal to them usually
present as well as a smaller fourth spine below but distal ; on the ventral face are
several more spaced spines along the distal side, each longer than the plate, giving a
fairly shaggy appearance.
432 AILSA M. CLARK
The paxillae are arranged in transverse rows each side of the mid-radius, though a
few are slightly irregular ; 23 or 24 paxillae correspond to the first ten supero-
marginals. The larger proximal paxillae have c. 10 central spinelets and 15-20
peripheral ones but some are larger still.
TABLE 2
Numerical data from six specimens of Astropecten granulatus natalensis from
near Durban
R r Br Paxillar No. of
at SM4 Br superomarginals
43 ii'o 9'5 5'5 21
38 ii-o 9-5 6-5 19
35 9'5 8-5 5-5 18
30 7'5 7'° 4'0 18
27 8-0 6-0 4-0 16 (17)
20 5-5 5-0 2-5 16
There are only two actinal plates in each series. The first ten inferomarginals
correspond to 16 adambulacral plates.
AFFINITIES. The six specimens from the vicinity of Durban included in Table 2
are the least badly damaged ones. They also have only two actinal plates in each
series, like John's material In this they differ from the holotype of Astropecten
anacanthus H. L. Clark, 1926, also taken in the vicinity of Durban, which has three
actinal plates. H. L. Clark ptherwise distinguished A. anacanthus as having rela-
tively narrower paxillar areas (and consequently more conspicuous superomarginals)
and 'totally different' ventral spinulation. The holotype of A. anacanthus has
R/r 44/11 mm, br is given as 12 mm with paxillar br at the same level 8 mm. Judging
from H. L. Clark's rather poor photograph (1926, pi. i, fig. i), his br measurement
must have been taken proximal to the first superomarginal, i.e. between two adjacent
interradial edges. Since the arms flare out at the base, I think that a better estimate
of their breadth in Astropecten species is given by measuring at the level of the fourth
superomarginals, or the fifth as Doderlein does. In the specimen with R 43 mm in
Table 2, br between two interradii is 12-5 mm and the paxillar breadth here is 8 mm.
The superomarginals appear more conspicuous than those of the holotype of A.
anacanthus, not less. Nor can I detect any significant difference in the armament
of the inferomarginals, which John pointed out is very variable. All the specimens
of A. granulatus natalensis have at least some of the proximal paxillae with the
peripheral spinelets distinctly higher than the central ones, which H. L. Clark did
not observe in A. anacanthus. Possibly this, together with the difference in the
number of actinal plates, may provide a valid distinction ; also the relative number
of marginal plates in each series may be significant. There are 27 in the holotype of
A. anacanthus at R 44 mm, compared with only 21 in the first specimen of natalensis
in Table 2.
The relative number of superomarginals was also used by John to distinguish
between Astropecten granulatus from Australia and natalensis from South Africa ;
ECHINODERMS FROM SOUTHERN AFRICA 433
the large specimen from East London reinforces this, having 34 superomarginals at
Re. 74 mm, compared with 40 in an Australian specimen with R 77 mm. The
smaller specimens also support this difference. The East London specimen is also
notable for having two almost equal large inferomarginal spines on the proximal
plates. John has already noted that smaller specimens from Natal have a second
inferomarginal spine distinctly larger than in Australian specimens of A. granulatus.
LECTOTYPE DESIGNATION. John did not designate a holotype for the subspecies
natalensis and it is clearly undesirable that the aberrant specimen (no. 29) with a
large spine on each of the first superomarginals, of which he illustrated the upper side
(1948, pi. i, fig. i) be treated as lectotype, while the other specimen he depicted
(his pi. i, fig. 2) is much smaller (R 22 mm) and shown in ventral view. I therefore
designate the specimen with R/r 37/11 mm, reg. no. 1904.4.20.120-122 (part),
ref. no. 10723 as lectotype. Some numerical data from it are included in John's
tables and a photograph of it is now given (PI. i, fig. i).
Astropecten polyacanthus phragmorus Fisher
PL i, fig. 2 ; PI. 2, fig. i
Astropecten acanthifer phragmorus Fisher, 1913 : 604.
Astropecten phragmorus : Doderlein, 1917 : 178 ; Fisher, 1919 : 65-67, pi. n, fig. 5, pi. 14, fig. i ;
Mortensen, 1940 : 61 ; Clark & Rowe, 1971 : 44.
Astropecten polyacanthus : H. L. Clark, 1923 : 249.
MATERIAL. PED I2A, 24°46'S : 34°5o'E (Mozambique, between Delagoa Bay
and Inhambane), 22 metres ; 3 specimens ['Anton Bruun' st. 3726].
TAXONOMIC RANK. As noted in 1971, Mortensen (1940) has queried the specific
validity of Astropecten phragmorus, having found many specimens intermediate
between it and A . polyacanthus in the Persian Gulf, an experience which I have shared
in dealing with material from that area. Apart from the three specimens from south-
ern Mozambique, there is an older one in the British Museum collections, ref. no. 49
(details unknown) sent by the Cape of Good Hope Government. (This is not to be
confused with the specimen of the same number illustrated as Astropecten granulatus
natalensis by John.) All four have the complete series of superomarginal spines
characteristic of A . phragmorus, most of the spines less conspicuous in size than the
comparable ones of A. polyacanthus, in which the second plate (sometimes also the
third) is reduced and spineless, exaggerating the prominence of the pair of spines
on the first superomarginals of each series in each interradius (PI. 2, fig. 2). H. L.
Clark (1923) also noticed that the specimen from Natal which he referred to A.
Polyacanthus had smaller superomarginal spines than usual and, although he did
not say that the spine series was complete, he did suggest that South African speci-
mens might be subspecifically distinct from A. polyacanthus. I agree with this
ranking and accordingly now reduce A. phragmorus to the status of a subspecies of
A. polyacanthus.
It may be noted here that in some specimens of Astropecten polyacanthus phrag-
morus the superomarginal plates are less high than is usual in A. polyacanthus
434 AILSA M. CLARK
poly acanthus, as shown by a comparison of figs, i and 2 in Plate 2. However, the
apparent discrepancy in the number of plates is an illusion caused by the different
magnifications ; in fact, the two specimens photographed have a comparable number
of marginal plates relative to R.
RANGE. The range of Astropecten poly acanthus phragmorus is therefore extended
to south-east Africa ; it is otherwise known from the Philippines (the type-locality)
and the Persian Gulf.
Dipsacaster sladeni capensis A. M. Clark
Leptopty 'chaster kerguelensis : Bell, igosa : 242-243. [Non L. kerguelensis Smith, 1879.]
Dipsacaster sladeni : H. L. Clark, 1923 : 246-247 ; Mortensen, I933a : 237.
Dipsacaster sladeni capensis A. M. Clark, 1952 : 204, pi. 17.
MATERIAL. SCD 2030, 34°5i'S : 23°4i'E (SE of Knysna), 184 metres ; i speci-
men.
R/r is 87/33 mm = 2-6/1 and there are 35 superomarginals in each series, agreeing
with the relative number found in previous South African specimens and significantly
higher than in D. sladeni Alcock from the Bay of Bengal.
Psilaster acuminatus Sladen
Psilaster acuminatus Sladen, 1889 : 225-228, pi. 40, figs, i, 2, pi. 42, figs. 7, 8 ; H. L. Clark,
1923 : 248-249 ; Mortensen, I933a : 236-237.
MATERIAL. AFR 7776, 29°i8'S : I5°35'E (SW from Port Nolloth), 384 metres ;
i specimen.
TRA 73F, 32°o6'S : i6°37'E (W from Lambert's Bay), c. 310 metres ; i specimen.
TRA I2gj, 32°26'S : i6°38'E, 480 metres ; i specimen.
AFR 831 A, 35°i5-3'S : i8°39-3'E (S from False Bay), 547 metres ; i specimen.
REMARKS. The northernmost specimen (AFR 777) shows some minor differences
from the others. Its madreporite is larger, diameter i-i mm, its centre nearer the
interradial edge, 3-0 mm, R/r being 30/8-5 mm in comparison with specimen TRA 73,
in which the same measurements are 0-75 mm, 3-4 mm and 29/8-5 mm. Its arma-
ment of furrow and marginal spines and granules is consistently more attenuated
and it even has a slightly elongated granule at the distal edge of some superomarginals
about one-third the height from the upper end of the plate, forming an incipient
superomarginal spine. Its adambulacral plates have more pronouncedly angular
furrow margins. As preserved, it has no anal cone, unlike the other specimens.
In comparison with Bathybiaster vexillifer, this species has relatively fewer supero-
marginals and these do have a small horizontal face rather than being restricted to
the lateral face of the arms. . B. vexillifer (at least the holotype of Phoxaster pumilus
(Sladen), which is believed to be synonymous) has 40 marginals at R c. 40 mm, in
comparison with 30 at R 30 mm in one of the present specimens of Psilaster acu-
minatus.
ECHINODERMS FROM SOUTHERN AFRICA 435
Family GONIASTERIDAE
Ceramaster patagonicus euryplax H. L. Clark
Ceramaster patagonicus var. euryplax H. L. Clark, 1923 : 262-264, pi. 14, figs, i, 2 ; 1926 : u ;
A. M. Clark, 1952 : 204-205.
Ceramaster chondriscus H. L. Clark, 1923 : 258-260, pi. 14, figs. 5, 6 ; Mortensen, I933a : 242-243.
MATERIAL. AFR72QC, 3i°22-8'S : i6°20-2'E (W from Hondeklip Bay), 365
metres ; 3 specimens.
TRA 7C, 48 miles W by S of Cape Town, c. 34°S : I7°E, 402 metres ; i specimen.
REMARKS. R/r is 35/21 mm = 1-67/1, 50/30 mm = 1-67/1, 51/29 mm = 1-76/1
and 60/32 mm = i-88/i. The number of superomarginals in each series is respec-
tively 12, 15 (or 16), 14 and 19. The smallest specimen has most of the upper surface
of all the superomarginals bare but the larger ones have only small bare patches on
the distal plates.
The superomarginals are squarish in shape, the paxillar spinelets are angular and
the distal subambulacral spines are hardly at all enlarged, so these specimens run
down to Ceramaster chondriscus in H. L. Clark's key (1923). Mortensen and I have
already pointed out the lack of distinction between C. chondriscus and C. patagonicus
euryplax and I believe that they are synonymous. In fact, the former has page
priority in H. L. Clark, 1923 but, since I regard the difference between South African
and Patagonian specimens as less than a specific one, it seems better to retain the
existing trinomial than to introduce the combination C. patagonicus chondriscus.
Since there is a geographical distinction, the rank is that of a subspecies rather than
a variety.
Family OREASTERIDAE
Genus ASTERODISCIDES nom. nov.
Asterodiscus Gray, 1847 : 75 [Non Asterodiscus Ehrenberg, 1839, Protozoa.] Type-species :
Asterodiscus elegans Gray, 1847.
I am obliged to Dr D. L. Pawson and Miss M. Downey for notification of the
homonymy of this generic name.
Asterodiscides elegans Gray
Asterodiscus elegans Gray, 1847 : 75 ; Macnae & Kalk, 1969 : 129 ; Clark & Rowe, 1971 : 34, 40.
MATERIAL. NAD 2oM, 29°58'S : 3i°O2'E (off Durban), 49 metres ; i specimen.
RANGE. This record provides a small extension of range from Inhaca.
Family OPHIDIASTERIDAE
Hacelia capensis Mortensen
Hacelia superba var. capensis Mortensen, 1925 : 152.
436 AILSA M. CLARK
MATERIAL. NAD giF, 29°n'S : 32°02'E (SE of Richard's Bay, Natal), 70 metres ;
i specimen ['Anton Bruun' st. 3576].
DESCRIPTION. Unfortunately this specimen (only the second to be taken) is
small with R/r only 15/3 mm = 5/1 ; br basally is 3 mm and at half R also 3 mm
since the arms only taper in their distal half.
There are nine complete longitudinal series of plates on each arm (not counting
the adambulacrals), one dorso-lateral series each side of the carinal row, two marginals
and one actinal each side, with eight series of papulae between them. The granula-
tion is slightly coarser on the middles of the plates. There is also a partial second
series of actinal plates proximally, above the main series, consisting of two rounded
plates on the disc and very base of the arm and about six other very narrow plates
corresponding in position and very close to the inferomarginals. There are 16
(or 17) superomarginals in each series, the last seven to ten with a large bare central
area, as also on most of the carinal plates except for the first two to four. The ter-
minal plates are very large and convex. The papulae are relatively few, only one or
two in each area, those below the inferomarginals single. The subambulacral spines
are up to twice as long as broad ; the two furrow spines of each plate are slightly
spaced from those of adjacent plates ; there are no granules on the furrow faces of
the plates. Pedicellariae are numerous, tong-shaped, their valves very compressed
beyond the base and not at all expanded sideways terminally. Their sockets have
prominent flanged edges. There is usually a pedicellaria in each papular area above
about five consecutive superomarginals at about one-third the arm length and
several others in other papular areas on the upper side but not below.
NOMENCLATURE. Hacelia superba H. L. Clark, 1921 from off Barbados in the
West Indies was referred by A. H. Clark (1948) to the synonymy of H. floridae
(Perrier, 1881), which species Downey (1971) has referred back to Tamaria without
reference to H. superba in her synonymy. The holotype of H . superba was the largest
specimen, R 75-78 mm, and had actinal papular areas below the inferomarginal
plates with up to eight papulae in each ; accordingly it could not be referred to
Tamaria in which the absence of proper actinal papular areas is diagnostic. In
H . floridae (and H . superba if distinct) the arms taper evenly from the base in con-
trast to this South African specimen and also differ in having pedicellariae on the under
side, not above, and small terminal plates. Also only the holotype of H. superba
had any distal marginal plates medially bare ; judging from the descriptions of Verrill
(1915, as Ophidiaster alexandri] and Downey all small specimens are fully granulated.
In spite of its small size, I am sure that this South African specimen is not con-
specific with the West Indian species, whether or not that is a Hacelia. Also it
looks more like an Ophidiaster superficially, with which the eight series of papulae
agree but of course the fact that there are not ten series, as in Hacelia, could well be
due to the small size. Mortensen did not even give the size of his specimen - the
holotype - and only distinguishes it from H. superba by the more numerous centrally
bare plates including dorso-lateral as well as carinal plates. Pending discovery of
larger South African specimens, it seems best to retain the combination of capensis
with Hacelia.
ECHINODERMS FROM SOUTHERN AFRICA 437
Family ASTERINIDAE
Asterina gracilispina H. L. Clark
Asterina gracilispina H. L. Clark, 1923 : 286-287, pi. 16, figs. 3, 4 ; Mortensen, iQ33a : 255-256.
MATERIAL. FAL 7200, 34°07'S : i8°45'E (False Bay), 20-26 metres ; i specimen.
CP 7ioA, no data ; i specimen.
AG 2E, Cape Agulhas (nearly 35°S : 20°E) ; i specimen.
MB I3C, 34°04'i7"S : 22°i3'53"E (Mossel Bay) ; i specimen.
SCD I55U, 34°Q3'S : 25°59'E (Algoa Bay), 85 metres ; i specimen.
SCD I77K, 34°2o'S : 23°3i'E (SE of Knysna), 56 metres ; i specimen.
Unfortunately most of these specimens are in poor condition ; the two from Mossel
Bay and Cape Agulhas are the best. They have R/r respectively 6/4-5 mm = 1-3/1
and 12/9 mm = 1'3/1- The holotype was dried and this probably accounts for the
choice of specific name which conflicts with Mortensen's comment that his wet
specimen has the abactinal spinelets blunt, not 'short, sharp' as in the type. The
present wet specimens agree in general with Mortensen's but their abactinal spinelets
extend all over the plates, not leaving a bare central patch and average about
12 per plate in the smaller and c. 15 in the larger. Other spine counts include : actinal
spines 3-5 (in a straight row) and 2 or 3 ; subambulacral spines 3 and 3 to 2 ;
furrow spines 4 to 3 and 4 or 5 ; suboral spines 2 and 3 sometimes 4 ; oral furrow
spines 6 and 9. The madreporite is only visible in the larger one. This is also
true of the specimen from south of Algoa Bay, where R/r is 10/7-5 mm = 1-3/1 and
the actinal spines number 3-5 on each plate.
CP 7ioA is rather flattened, the arms distinctly petaloid ; R/r is 18/12 mm = 1-5.
It retains some purple colour on the upper side except peripherally and small patches
near the centre which are buff-yellowish ; there are 2-4 actinal spines, usually 3 ;
4 furrow spines ; 2 or sometimes 3 subambulacral and 3 suboral.
Asterina burtoni Gray
Asterina burtonii Gray, 1840 : 289 ; H. L. Clark, 1923 : 283 ; Smith, 1927 : 641-645.
Asterina burtoni : A. M. Clark & Rowe, 1971 : 68-70, pi. 9, figs. 4, 5.
MATERIAL. PEA 2P, Maxixe, Mozambique (c. 25°S : 32°E), LW ; 7 specimens.
JAN 26G, Jangamo reef, Mozambique (c. 24°S : 35°E) ; i specimen.
MOR 6F, Morrumbene, Mozambique (c. 23°S : 35°E) ; i specimen.
One of the first sample has six arms and three madreporites ; the rest have five
and one, as usual in East African specimens.
The specimen from Jangamo reef has unusually long arms, R/r 25/9 mm = 2-8/1.
From above it closely resembles the larger of the two specimens from Zanzibar which
I called Paranepanthia sp. in 1971 ; that has R/r 30/11 mm = 2-7/1. Both have a
distinct boundary between the mid-radial and lateral 'fields' of abactinal plates,
though in the Jangamo specimen the innermost plates of the lateral series are not
noticeably larger than the second row of plates, unlike the Zanzibar specimen.
However, from the under-side there is a marked difference from Paranepanthia in the
30
438 AILSA M. CLARK
armament of the actinal plates, which consists in JAN 26 of fans of three to five
spines, similar to the subambulacral spines, as usual in Asterina burtoni, but of clusters
of 7-12 on most plates of the Zanzibar specimen.
Patiriella dyscrita (H. L. Clark) new comb.
Asterina dyscrita H. L. Clark, 1923 : 284-285, pi. 16, figs. 5, 6.
Asterina (Patiriella) exigua (part) : Mortensen, iQ33a : 252-255.
MATERIAL. BRE 78F, c. 33°S : ig°E (Breede River estuary) ; 3 specimens.
E 36, c. 33°S : 25 °E (Port Elizabeth), below LWS ; i specimen.
NOMENCLATURE. Dartnall (1971) has reviewed Patiriella exigua and recognizes
two species from South African waters, one (the true P. exigua) with downwardly
directed gonopores and the other with upwardly directed ones. As quoted in
Dartnall's paper, I think that the latter is conspecific with H. L. Clark's Asterina
dyscrita. Since Dartnall has split off a tropical Indo-West Pacific species that also
has dorsal gonopores as Patiriella pseudoexigua, clearly he does not consider the
different alignment of these pores warrants a generic distinction. As A . dyscrita is
otherwise so similar morphologically to P. exigua, I propose to refer it to Patiriella.
Preserved colourless specimens where the position of the genital openings is often
difficult to detect are all likely to be referred to P. exigua, though in life the greenish-
blue of its lower side will distinguish that species from P. dyscrita.
Family SOLASTERIDAE
Lophaster quadrispinus H. L. Clark
Lophaster quadrispinus H. L. Clark, 1923 : 295-297, pi. 18, figs. 1,2; 1926 : 21 ; Mortensen,
1933 : 272.
MATERIAL. AFR 73oA, 3i°29'S : i6°O3'E (W of Lambert's Bay), 459 metres ;
i specimen.
TRA 8A, 42 miles W by N of Cape Town (c. 33°S : i7°E) ; i specimen.
WCD 2I4A, 34°29'S : i8°i6'E (W of Cape Point), 400 metres ; i specimen.
Whereas H. L. Clark gives the number of furrow spines on each plate in the proxi-
mal half of the arm of the holotype (R 70 mm) as four and Mortensen (size not given)
as five to six, no. AFR 730A with R 60 mm has only three furrow spines on most
plates, only a few proximal ones with four and at least one with only two. The
same is true of the dry specimen, TRA 8 A, R again c. 60 mm, in which only about two
proximal plates of each series have four spines, the rest three.
Family EGHINASTERIDAE
Henricia reticulata (H. L. Clark) new comb.
Echinaster reticulatus H. L. Clark, 1923 : 290-292, pi. 15, figs, i, 2 ; Mortensen, i933a : 264-265.
ECHINODERMS FROM SOUTHERN AFRICA 439
NOMENCLATURE. Fisher (1940 : 272) has pointed out Mortensen's inconsistency
in referring Henricia ornata (Perrier) back to Echinaster on the grounds that it has
no actinal papulae but leaving E. reticulatus, with such papulae, still in Echinaster.
Unfortunately, however, Fisher's own attempt to distinguish between Echinaster
and Henricia on the supposed longitudinal webbing of the innermost subambulacral
spines in Echinaster is not reliable. Although some specimens of the type-species,
E. sepositus (Lamarck) [perhaps properly called E. sentus (Retzius)], are preserved
with the skin-covering appearing markedly thickened and giving some effect of
longitudinal webbing, the usual condition is for the spine sheaths to be independent,
with a distinct crease between adjacent plates.
In the armament of the abactinal plates, H. ornata does span both genera, speci-
mens with coarser, spaced spinelets resembling Echinaster, whereas others in which
there are linear or even doubled series of spinelets along the reticulations of the skele-
ton agree with Henricia. This is a notoriously difficult family to classify.
Family ASTERIIDAE
Marthasterias glacialis forma africana (Miiller & Troschel)
Asteracanthion africanus Miiller & Troschel, 1842 : 15.
Asterias africana : Doderlein, 1910 : 252.
Marthasterias africana : H. L. Clark, 1923 : 306.
Marthasterias glacialis var. africana : Mortensen, i933a : 273-274, pi. 16, fig. i.
MATERIAL. FBY I33J, False Bay, no details ; i specimen.
R/r is c. no/c. 20 mm = 5-5/1, maximum br c. 30 mm but possibly less in life
since the arms have flattened in preservation somewhat. Even allowing for this,
the arms are unusually short and fat. It is also unusual in having about ten proxi-
mal actinal plates each with a spine, simulating a third inferomarginal spine as they
are in series. The pedicellaria wreaths on the outer inferomarginal spines wrap well
round and appear to be more or less fused to some of the inner spines as well. How-
ever, some northern specimens of M. glacialis also show this. There are not very
many dorso-lateral spines but the carinal plates have up to four spines on a single
plate. The more lateral parts of the dorso-lateral areas are fairly bare but not so
obviously so as in the forma rarispina. The largest straight pedicellariae have
broad rounded tips, sometimes with digits, unlike Fisher's description for northern
M. glacialis (1928) as 'slender lanceolate to ovate'.
Perissasterias polyacantha H. L. Clark
Perissasterias polyacantha H. L. Clark, 1923 : 307-309, pi. 18, fig. 3 ; 1926 : 29-30, pi. 6 ; Mor-
tensen, i933a : 278.
MATERIAL. AFR 736N, 30°42-4'S : i5°59'2'E (SW of Hondeklip Bay), 201 metres ;
i specimen.
This specimen is dried ; it has R only c. 100 mm, whereas H. L. Clark's material
had R 260 to over 300 mm. This probably accounts for it having only three adam-
bulacral spines, when he described five to seven. The carinal spines are tapering
44o AILSA M. CLARK
and bluntly pointed, unlike Perissasterias oUusispina H. L. Clark, 1926, the only
specimen of which (R c. 160 mm) has low capitate spines. The latter has four
adambulacral spines on most plates and I think will prove to be conspecific with
P. polyacantha when more material is available ; the shapes of abactinal spines are
often very variable in this family.
Subclass OPHIUROIDEA
Family GORGONOCEPHALIDAE
Astrocladus euryale (Retzius)
PI. 3, figs i and 2
Asterias euryale Retzius, 1783 : 243.
Euryale verrucosum Lamarck, 1816 : 537.
Gorgonocephalus verrucosus : Lyman, 1882 : 262-263 '• Bell, 19055 : 260.
Astrocladus verrucosus : Doderlein, 1910 : 256.
Astrocladus euryale : Doderlein, 1911:28, 40, 106 ; H. L. Clark, 1923:319; Mortensen,
I933a : 293-296, figs. 21, 22, pi. 18, fig. 7.
MATERIAL. AFR 8426, 34°34'S : i9°i8'E (W of Cape Agulhas), 31 metres ; i
specimen.
SCD 32U, 33°38-6'S : 26°547'E (W of Port Elizabeth), 55 metres ; i large and 2
small specimens.
SCD 42K, 32°i5-2'S : 28°577'E (off Bashee River), 47 metres ; i small specimen.
SCD 57 W, 33°37'S : 26°56-6'E (W of Port Elizabeth), 46 metres ; 5 specimens.
SCD giC, 33°03'S : 27°55'E (S of East London), 27 metres ; 4 specimens.
SCD I55P, 34°03'S : 25°59'E (S of Port Elizabeth), 85 metres ; 2 specimens.
SCD 355A, 34°03'3'S : 25°43'5'E, 38 metres ; i specimen.
VARIATIONS. In his key to the family Gorgonocephalidae, Doderlein (1911 : 28)
used the presence or absence of tentacle scales (or arm spines) before the first fork
in the arms as one of the main dichotomies, Astrocladus being included among the
genera in which these are absent (at least in adult specimens). However, of the
specimens now referred to Astrocladus euryale, only the one from west of Cape
Agulhas (AFR 842C) does not have spines before the first fork. It has disc diameter
(d.d.) 60 mm. All the others come from east of Cape Agulhas and have spines
developed before the first fork to some degree. Apart from this, I cannot discern
any significant difference between them to warrant a specific, let alone a generic
distinction. Of the 'Challenger' specimens from False Bay, the largest (d.d. 75 mm)
has two very short spines adjacent to several pores before the first fork, including
those of the third arm segment, just projecting from the granulation, though con-
secutive segments all have spines only after the first fork. A smaller specimen from
the same station (d.d. 40-44 mm) has short spines, at first two then three, starting
at the second segment but another with d.d. 47 mm shows no spines at all before the
first fork. The largest of sample SCD giC (d.d. 45 mm), as well as the smaller ones,
has three spines at the second tentacle pores increasing to four, though the largest
from SCD 57\V of similar size (d.d. 40-45 mm) only has two very short spines on
ECHINODERMS FROM SOUTHERN AFRICA 441
these early segments. Clearly there is much more variation in the degree of reduction
of these spines with growth than has hitherto been allowed for. Mortensen (1933)
thought them suppressed from about d.d. 20 mm, while H. L. Clark (1923) gave the
critical size as 15 mm. The holotype of Mortensen's Astroconus capensis has d.d.
30 mm and superficially resembles Astrocladus euryale in the tubercles on the radial
ribs. It has two arm spines by the second and following pores before the first arm
fork, which presumably prompted its inclusion in Astroconus rather than Astrocladus.
It also lacks the belts of hook-bearing platelets which Mortensen discovered on the
distal arm segments of Astrocladus euryale. These platelets are found at least in the
largest specimen from SCD giC (d.d. 45 mm) (and from the third fork onwards,
not just distally) but, as noted above, this specimen also has arm spines before the
first fork.
It seems to me that a reappraisal of the taxonomic characters in use for the family
Gorgonocephalidae in the light of ontogenetic changes and variation is badly needed.
It is possible that some use can be made of the average number of segments between
successive forks on the arms but this distinction may also prove to be untenable.
It may be noted that most of the specimens now referred to Astrocladus euryale have
only six to eight, sometimes nine, segments between most of the forks on the main
branches, whereas the holotype of Astroconus capensis has 12-13 on the main
branches and seven or eight only on the side branches. The type-locality of Mor-
tensen's species was off Natal (c. 3O°S : 3i|°E). Most of the specimens he referred
to Astrocladus euryale were from False Bay and Walker Bay (west of Cape Agulhas)
and were large with d.d. 50-60 mm but he also referred specimens from East London
(size not given) to the same species.
Family OPHIAGANTHIDAE
Ophiomitrella corynephora H. L. Clark
Ophiomitrella corynephora H. L. Clark, 1923 : 322-324, fig. 2, pi. 19, figs. 5, 6 ; Mortensen,
I933a : 33!-333, figs. 48, 49.
MATERIAL. AFR 743F, 3O°S : i5°E (W of Hondeklip Bay), 102 metres ; 4
specimens.
WCD 2511, 33°o6'5'S : I7°55'5'E (near Saldanha Bay), 79 metres ; 17 specimens.
WCD 2igC, 34°42-8'S : i8°o8-8'E (SW of Cape Point), 360 metres ; 6 specimens.
SCD 4A, 34°3o'S : 24°4o'E (SW of Cape St Francis), 102 metres ; 2 specimens.
SCD I77C, 34°2o'S : 23°3i'E (SE of Knysna), 56 metres ; 7 specimens.
SCD 249E, 34°48'S : 23°39'E (SE of Knysna), 146 metres ; 4 specimens.
SCD 32oF, 34°i5'S : 25°5o-5'E (S of Port Elizabeth), 108 metres ; i specimen.
Contrary to H. L. Clark's figure, not all specimens of Ophiomitrella corynephora
have the radial shields completely separated. In several smaller ones the radial
shields of some pairs at least are just contiguous, for instance in one from SCD 4 A
with d.d. 3-5 mm, which also has relatively large radial shields, 0-75 mm long or
nearly half the disc radius.
One of the seven from SCD J-77C has unusually short disc stumps and arm spines.
Unfortunately the ventral side is concealed since it is wrapped around a gorgonian.
442 AILSA M. CLARK
Ophiacantha sp. juv., aff. O. striolata Mortensen
MATERIAL. WCD I40K, 34°i6'S : i8°i5'E (SW of Table Bay), 158 metres ; i
specimen.
AFFINITIES. The d.d. is only i-i mm. The mouth plates are immature with the
first ventral arm plate and second oral tentacle superficial. The distal oral papilla
is present but the second one is not yet developed, leaving a diastema between the
distal papilla and the first one, which is close alongside the apical papilla (or outer-
most tooth). The disc is armed with relatively long trifid stumps. The radial
shields are not visible. The oral papillae are markedly rugose and many of the arm
spines have lateral thorns for about the basal one-third or half of their length. The
lateral plates are very moniliform (constricted proximally and flared distally) and
their texture is striated like that of the small ventral arm plates. It is this last
character particularly which suggests affinity with Ophiacantha striolata. Neither
Koehler (1923) nor Cherbonnier (1962) mentioned striations on the arm plates of
Ophiacantha angolensis, while a specimen of Ophiacantha baccata of similar small size
already has oral papillae like those of the adult, even to the third one being distinctly
coarser than the second. However, it does have relatively large trifid stumps on the
disc and the proximal arm spines are similarly thorny in their basal halves. In
0. baccata the arm plates are not striated.
If this specimen can be referred to Ophiacantha striolata, then it extends the range
of the species westwards, other records being from the Durban area (Mortensen)
and a new one from off East London (SCD 29yF, 33°09'S : 28°02'E, 84 metres).
Amphilimna cribriformis sp. nov.
Fig. i
MATERIAL. NAD 336, 29°38'S : 3i°36'E (NE of Durban), 200 metres ; i speci-
men ['Anton Bruun' st. 3900].
NAD 35 W, 29°35'S : 3i°38'E, 150 metres ; i specimen [St. 39oL].
NAD 4oV, 29°34'S : 3i°39'E, 118 metres, 17 specimens including the holotype.
[St. 390?.]
NAD 430, 29°34'S : 3i°39'E, 115 metres ; 2 specimens [St. 39oN].
NAD 52E and 556, 29°29'S : 3i°45'E, 86 metres ; 22 specimens [St. 3916 & B].
FED 19 J, 25°07'S : 34°34'E (between Delagoa Bay and Inhambane, Mozambique),
112 metres, 4 specimens [St. 372L],
DESCRIPTION. The holotype has the disc somewhat distorted in preservation,
having d.d. 5-5-6-5 mm. It is similar to that of Amphilimna olivacea, having fine
indistinct scaling with scattered slender tapering spinelets, in this specimen up to
c. 0-5 mm long. The radial shields are narrow, length : breadth c. 0-8 : 0-15 mm,
contiguous for most of their length but separated by a (possibly unnatural) notch
distally. The upper end of the genital plate just distal to each shield is armed with a
pair of tapering spines stouter than the disc spines.
ECHINODERMS FROM SOUTHERN AFRICA
b
443
FIG. i. Amphilimna cribriformis sp. nov. Holotype. NAD 4oV. a. Dorsal view of part of
disc (apparently incompletely regenerated) and two arm segments which would normally
be overlain by the disc, showing transitionally modified upper arm spines and rudimentary,
non-contiguous dorsal arm plates, b. Two jaws and the first few arm segments, the distal
oral papillae omitted from the series on the extreme left, the indistinct sutures between
the ventral and lateral arm plates proximal to the tentacle pores not drawn in. c and d.
Dorsal and ventral views of the twelfth free arm segment, the perforations in the dorsal
arm plate only drawn on half the plate ; the lines of dashes indicate the underlying ver-
tebrae seen by transparency. The scale measures i mm.
The oral shields are rounded triangular, broadest distally and about as long as
broad, the adorals widely separated from each other by the width of the rounded
proximal angles. The oral plates are long. On each jaw there are two to four
asymmetrically placed apical papillae superficial to the outermost broad, rounded
tooth and flanked on each side by an almost superficial, elongated oral tentacle
scale filling in the diastema between the apical group of papillae and the three spini-
form distal papillae each side, the two outer of which arise from the edge of the adoral
shield.
The disc appears to have shrunk back from the bases of the arms slightly since the
dorsal arm plates of the first few free segments are rudimentary and their upper
arm spines are very flat and abbreviated, though not fused together like those of
the basal segments adjoining the genital slits. The succeeding dorsal arm plates
become broader than long, approximately ovate, but still thin and semi-transparent,
having a sieve-like appearance (hence the specific name) owing to the single layer of
perforations. The underlying parts of the lateral arm plates and vertebrae can
accordingly be discerned, at least when wet.
The first ventral arm plate is fairly large, its superficial part appearing triangular
and swollen ; the distal edge is convex. The second plate has the distal edge
almost straight and is relatively broad between the tentacle pores but the following
444 AILSA M. CLARK
plates soon become narrow in the proximal part, flare out abruptly distal to the pores
and rejoin the lateral plate each side, while the distal edge sweeps back each side for
more than a third of its extent.
On the first seven segments all the spines but the lowest of each series are modified
into a wing-like flange, as also in Amphilimna olivacea. Beyond the disc the arm
spines are still flattened but separate, tapering to sharp tips ; they number up to
six and none exceed the segment in length. Owing to the flattening, their internal
cavities are small.
The first eight to ten segments have two tentacle scales to each pore, a short
proximal one on the lateral arm plate, at first almost in series with the lowest spine
and itself resembling a short stumpy spine, while the second scale is based on the
ventral arm plate about half-way along the side of the pore and is at first long and
spiniform but soon becomes reduced and lost altogether by about the tenth (rarely
the twelfth) segment ; simultaneously the proximal scale becomes flattened and
shifts to the edge of the pore, adopting the more usual lid-like form.
AFFINITIES. In comparison with Amphilimna olivacea from both sides of the
tropical Atlantic and adjacent coasts, recently redescribed and figured by Cher-
bonnier (1962), Thomas (1967), Madsen (1971) and Thomas & Schoener (1972), this
new species from the Indian Ocean side of southern Africa differs in the following
ways : the presence of usually two well-developed spines at the upper end of each
genital plate rather than one ; the triangular shape of the oral shields with a flattened
distal side rather than the rhombic shape so constantly found in A . olivacea, with the
adorals widely separated rather than contiguous (or nearly so) interradially ; the
more asymmetrical apical oral papillae, usually numbering three or four, sometimes
two, rather than usually only two ; the greater breadth of the dorsal arm plates
and their delicacy ; and the lesser elongation of the arm spines, not exceeding the
segment length.
I agree with Thomas (1967) that the genus Amphilimna does not belong in the
family Amphiuridae. He has ranged it instead in the Ophiacanthidae, which is
rather a heterogeneous assemblage and badly in need of revision. Some of its mem-
bers show considerable affinity with the Chilophiurida, including the Ophiocomidae
and Ophionereidae.
Family AMPHIURIDAE
Amphiura albella Mortensen
Amphiura albella Mortensen, I933a : 359-361, fig. 67.
MATERIAL. ABD I2H, 30°O9'S : 3i°37'E (E of Durban), 930 metres ; i specimen
['Anton Bruun' st. ^8gE].
This specimen has d.d. 6-0 mm, compared with 5 mm in the holotype. It differs
in having the radial shields relatively larger, their length almost equal to half the
disc radius, length : breadth c. 1-4 : 0-5 mm = 2-8:1. One radius is abnormal
with a single very broad radial shield. There are five arm spines proximally, com-
pared with four in the type. There are two tentacle scales only on the proximal
parts of the arms, the remaining pores having only the one on the lateral arm plate.
ECHINODERMS FROM SOUTHERN AFRICA 445
Amphiura Candida Ljungman
Fig. 2
Amphiura Candida Ljungman, 1867:318-319; Mortensen, iQ33a : 361. [Non A. Candida:
Marktanner-Turneretscher, 1887, nee Koehler, igc^a.]
Amphiura kalki Balinsky, 1957 : 3~5> ng- x> pi- J> ngs- I> 2 • A. M. Clark m Clark & Rowe,
1971 : 80, 97.
MATERIAL. MOR 506, c. 23^8 : 35|°E (Morrumbene estuary, Mozambique),
3-5 metres ; 9 specimens.
MOR i3iY, same locality, 2 metres ; 2 specimens.
FIG. 2. Amphiura Candida Ljungman. MOR 5oC. D.d. 5*0-5*5 mm. a. Two jaws and
the first two arm segments ; most jaws have more or less offset apical papillae, not fully
infradental, as in the one on the left ; the distal oral papillae are all foreshortened,
b. Dorsal view of twentieth free arm segment showing the truncated, slightly hooked,
middle arm spines. The scale measures i mm.
NOMENCLATURE. Mortensen's unilateral declaration of the name Amphiura
Candida as a 'species delenda' has no validity without supporting action by the Inter-
national Commission on Zoological Nomenclature, under the code of which the
designation of a Japanese specimen as neotype by Koehler (1904) is also invalid.
Since the type locality is Mozambique - for long a neglected collecting area - and
not South Africa, the absence of the species among the South African collections
studied by Mortensen is not surprising. It is unfortunate that Balinsky accepted at
face value Mortensen's rejection of the name A. Candida and so introduced the new
name Amphiura kalki, which is surely synonymous. He himself 'had no criterion
for distinguishing it from A. Candida' .
The Morrumbene specimens agree with Ljungman's description as far as that goes,
except that the distal oral papilla is not really squamiform, being quite thick,
though short and slightly broadened with a rounded end. It arises from the distal
end of the oral plate just adjacent to the edge of the adoral shield. Unfortunately
Ljungman omitted to mention the arm spines, which have the distinctively large
number of up to eight.
Amphiura capensis Ljungman
Amphiura capensis Ljungman, 1867:320; Lyman, 1882: 129, pi. 18, figs. 14-16; Koehler,
1908 : 634 : Doderlein, 1910 : 253-254, pi. 5, fig. 2 ; H. L. Clark, 1923 : 327 ; Mortensen,
446 AILSA M. CLARK
? Amphiura angularis : H. L. Clark, 1923:327-328; Mortensen, iQ33a : 354. [Non A.
angularis Lyman, 1879.]
Amphiura adjecta Mortensen, i933a : 355-357, fig. 62.
Amphiura compressa Mortensen, I933a : 357-358, figs. 63, 64.
MATERIAL. LU 53 Y, Luderitz Bay (c. 26|°S : I5°E), shore ; 2 specimens.
SWD 86G, 27°3o'S : I5°25'E (S of Luderitz Bay), 35 metres ; 7 specimens.
HB 5D, Hondekl'p Bay (c. 30°S : i7°E) ; 3 specimens.
PP 48, Paternoster (c. 32°S : I7°E), shore ; i specimen.
SB 1760, I78X, i87A, i88F, 205?, Saldanha Bay (c. 33°S : i8°E), 13-15 metres ;
1 6 specimens.
LB 116, 524Q, 5556, Langebaan Lagoon (c. 33°S : i8°E) ; 7 specimens.
TRA 76N, 32°4i'S : i8°O3'E (N of Saldanha Bay), 27 metres ; i specimen.
TRA 132!, 34°i9'S : i8°3o'E (False Bay) ; 3 specimens.
FAL igD, 453H, 4570, 467!^, 493M, 5050, 578T, False Bay, 29-80 metres ; 19
specimens.
TRA 1028, near Mossel Bay (34°S : 22°E) ; i specimen.
KKN 42Z, Buffalo Bay, near Knysna (34°S : 23°E), intertidal ; i specimen.
SCD 1550, 34°O3'S : 25°59'E (Algoa Bay), 85 metres ; 3 specimens.
SCD 294!!, 33°04'S : 27°57'E (near East London), 84 metres ; 6 specimens.
SCD 3236, 34°i5'S : 25°5o-5'E (S of Algoa Bay), 108 metres ; i specimen.
SCD 365E, 33°5O'7'S : 25°47'5'E (Algoa Bay), 36 metres ; i specimen.
VARIATION. Like H. L. Clark I cannot distinguish specifically between specimens
from deeper water down to c. 180 metres (as at 'Challenger' st. 141) and those taken
by shore collecting. The only possibly significant difference found in a detailed
examination of 24 specimens was in the length of the radial shields in relation to the
disc radius, the ratio being i : 3-5 and i : 3-2 in two out of the three 'Challenger'
specimens compared with a range of i : 2-2-2-8 in 21 of the Ecological Survey speci-
mens from depths down to only 84 metres. However, the third 'Challenger' specimen
(d.d. 5-8 mm) has the ratio 1:2-7.
I counted up to only seven arm spines in the 'Challenger' specimens and I suspect
that the presence of eight in Lyman's figure was an artist's mistake. The maximum
number of spines basally tends to increase with size from five at d.d. <4'5 mm to
seven at >8 mm, though there is some individual variation. The upper spines at
least are distinctly flattened, usually broadest at about the middle of their length but
slightly tapering in smaller specimens, as in the holotype of Amphiura adjecta with
d.d. 4 mm, or conversely spatulate with the tip more or less truncated or even ex-
panded like an axe-head, as in the large holotype of A. compressa (d.d. n mm).
The length of the longest spines is 1-0-1-4 (normally i-i or 1-2) times the segment
length, compared with 1-5 times as long in the spines of the holotype and paratype
of A . angularis Lyman from south of the Antarctic Convergence, which also number
up to only five, even though d.d. is as much as 9-2 and 7-5 mm, and are tapering,
though the middle ones are blunted at the tip.
Although H. L. Clark and Mortensen have both referred some South African
specimens to Amphiura angularis, the latter realized in 1936 that these were dis-
tinguished from the type material by having the disc scaled ventrally rather than
ECHINODERMS FROM SOUTHERN AFRICA 447
abruptly naked below with only a few scattered scales. In fact, I find that nearly all
the South African specimens which I am attributing to A . capensis have the scaling
deficient in a small area immediately distal to the oral shields, as Mortensen shows in
his figure of A . adjecta. Only three specimens show perfectly continuous scaling ;
conversely six specimens have larger bare areas ventrally, one of them (from south
of Luderitz Bay - the northernmost locality) has an abrupt edge to the scaling below
the ambitus, though another from the same station has more extensive ventral
scaling.
Another variable feature is the occurrence of a distinct primary rosette on the disc,
which occurs in only about half the specimens ; anyway the rosette decreases in
relative size with growth. It is possible that the variable occurrence of the rosette
as well as the differences in the ventral scaling can be attributed to the viviparous
habit, which may render the disc more than usually liable to deformity, rupture or
loss and regeneration.
The shape of the distal oral papilla is always somewhat elongated but varies from
blunted conical to flattened ; only occasionally is it more acute.
Growth changes include modifications in the shape of the ventral arm plates.
These usually have a straight distal edge but it is convex in the smallest specimens,
as in the type of Amphiura adjecta, while three specimens (two of them relatively
large) have the edge slightly concave, as in the type of A . compressa.
The adoral shields are contiguous interradially in the smallest specimen studied
(d.d. 2-5 mm), again as in the holotype of Amphiura adjecta, and almost so in several
of the other smaller specimens, but distinctly separate in the majority.
The least number of scales on a line between the radial shields interradially
increases from four to seven at d.d. <4 mm to c. 20 at 10-8 mm, with usually 9-13
at d.d. 5-7 mm. Again there is some individual variation.
There is a tendency for reduction in the relative length of the radial shields with
growth, as usual in amphiurids, but only to the extent of changing the mean from
i : 2-6 to i : 2-5 when the larger Survey specimens (d.d. 5 mm or more) are treated
separately from the smaller ones. The length : breadth ratio of the radial shields
changes from a mean of 2-35 : i in 12 specimens with d.d. <5 mm to 2-60 : i in a
similar number of larger specimens.
Considering all these variable characters, I find myself quite unable to distinguish
the four species recognized by Mortensen and consequently am referring both
Amphiura adjecta and A. compressa to the synonymy of A. capensis, simultaneously
restricting A. angularis to antarctic specimens.
Amphiura grandisquama natalensis Mortensen
Fig. 3
Amphiura grandisquama var. natalensis Mortensen, iQ33a : 353-354, fig- 60.
MATERIAL. ABD iG, 29°2i'S : 3i°58'E (NE of Durban), 370 metres ; 5 speci-
mens. ['Anton Bruun' st. 358C.]
448 AILSA M. CLARK
These specimens are all very small, d.d. <3«5 mm, and are notable for the extreme
length of the lowest arm spine, sometimes equalling the length of three segments.
The single, relatively large, tentacle scale appears rounded in contrast to the rather
pointed scales illustrated by Mortensen (i933b) in a specimen from St Helena. The
oral plates are distinctly flattened so that the proximal end of the first oral tentacle
scale is on the same level as the infradental papilla and the second tentacle is almost
superficial, approaching the structure of Amphilepis except that the teeth are broad.
FIG. 3. Amphiura grandisquamanatalensisMoTtensen. ABD iG. D.d. 3-25 mm. a. Two
jaws. b. Side view of seventh free arm segment, the middle spines foreshortened. The
scale measures i mm.
Amphiura simonsi A. M. Clark
Fig. 4
Amphiura simonsi A. M. Clark, 1952 : 215-217, fig. 2.
MATERIAL. FB 10560, FAL 64!, g6Z, i88W, 2o6E, 2ioQ, 234!!, 374*!, 3950,
4196, 4226, 4470, 45oA, 7036, 8o4N, False Bay, 27-75 metres ; 36 specimens but
only 16 of them with discs.
SCD 94R, 34°2i'S : 25°4i'E (SW of Port Elizabeth), no metres ; i discless speci-
men.
SCD H4V, 34°29'S : 2i°49'5'E (SW of Mossel Bay), 73 metres ; i discless specimen.
SCD ngC, 34°33'S : 2i°52'E, 76 metres : i discless specimen.
SCD 2090, 33°58-8'S : 25°42-2'E (Algoa Bay), 27 metres ; 2 specimens with discs
detached, but present.
NAD 850, 29°io'S : 3i°5i'E (S of Richards Bay), 42 metres ; i discless specimen.
['Anton Bruun' st. 3566.]
DESCRIPTION. The holotype of this species was also from sample FB 1056 but
lacked the disc. A complete specimen from this sample was only sent to me later.
Its disc is partially detached and rather deformed but probably measured between
6 and 7 mm in diameter. The oral frame (from distal edge of an oral shield to the
distal edge of the first ventral arm plate opposite it) measures 2-0 mm. The radial
shields are relatively small, c. 0-9 mm long and 0-3-0-4 mm broad ; they are con-
tiguous only at the distal end and diverge proximally. The rest of the upper side
of the disc is covered with small, even-sized scales and the primary plates cannot be
ECHINODERMS FROM SOUTHERN AFRICA
449
FIG. 4. Amphiura simonsi A. M. Clark. 'Topotype.' FB 10560. D.d. 6 mm (oral frame
diameter 2-0 mm), a. Dorsal view of part of semi-detached disc, the scaling very trans-
parent and the abradial limits of the radial shields seen by transparency shown by dashes,
b. One jaw showing the ill-defined limits of the outer of the two distal oral papillae each
side. c. Fully modified middle arm spine with barb in this specimen. The scale measures
i mm for a and b and 0-5 mm for c.
distinguished. The ventral side is abruptly more flexible and appears to be com-
pletely lacking in scales.
The outer of the two distal oral papillae is less well developed than the inner one
and may be more of an extension of the rim distal to the second oral tentacle than an
articulated papilla. The oral tentacle scale is small and rounded, not projecting
much from the side of the oral plate, and well within the slit.
There are four arm spines proximally, the lowest one somewhat flattened and
tapering to a blunt or rounded tip, half as long again as the segment ; the second
spine gradually develops the conspicuous glassy hook beyond the base of the arm, its
complete modification being reached by about the fortieth segment, where it shows
an abrupt constriction after the opaque basal part ; the transparent hook is armed
with a barb, not found in the holotype. The two upper spines are extremely flat and
paddle-shaped or elliptical, widest in the middle or distally, with the tip broadly
rounded. As many as twelve of the basal segments may lack the single tentacle
scale, compared with only two in the holotype.
Amphiura sp. indet. A.
MATERIAL. SCD 3236, 34°i5'S : 25°5o|'E (S of Algoa Bay), 108 metres ; i small,
damaged specimen.
The disc is very distorted. The radial shields seem relatively long and were
probably only contiguous distally. The disc was probably bare below. There are
two spiniform distal oral papillae each side of the jaw, of which the outermost and
largest arises from the adoral shield, while the other is on the distal end of the oral
plate, separated from the infradental papilla by a diastema revealing the oral tentacle
scale in the slit. The oral formula is m,om,m + t. The oral shields are rhombic
and the adorals, with the exception of one contiguous pair, are just separate inwardly.
The dorsal arm plates are almost circular, though the distal edge is slightly flattened
medially. There are four short, stout, blunt arm spines and one poorly calcified
tentacle scale.
450 AILSA M. CLARK
? Amphiura sp. indet. B
MATERIAL. NAD 4oZ, 29°34'S : 3i°39'E (NE of Durban), 118 metres ; i discless
specimen ['Anton Bruun' st. 390?].
This specimen again has two distal oral papillae but rounded, not spiniform and
the inner one of them arises partly from the adoral shield so that the formula is
m,on^nm + t. It has some resemblance to Amphiura koreae Duncan. The dorsal
arm plates are broad, oval. There are three acute arm spines and one tentacle scale,
unlike A. koreae, which has two.
? Amphiura sp. indet. C
MATERIAL. FAL 4160, 34°i2'5'S : i8°37'E (False Bay), 48 metres ; i oral frame
and arm bases and a detached disc, ? of the same specimen.
The disc has very long narrow radial shields equal to about half the radius and was
probably bare below. The oral shields are very short and broad, the adorals relatively
large and broadly contiguous. There is one short rounded distal oral papilla. The
four arm spines are extremely flat. The proximal pores are scaleless but then a
narrow scale develops along the edge of the ventral arm plate.
Amphipholis similis Mortensen
Fig. 5a
Amphipholis similis Mortensen, I933a : 363-364, fig. 69.
MATERIAL. TRA 5QF, off Bull Point (c. 34°S : 2i°E), 70 metres ; i specimen.
SCD I44N, 34°46'S : 22°O5'E (S of Mossel Bay), 93 metres ; i specimen.
NAD 460, 29°35'S : 3i°42'E (NE of Durban), 138 metres; i specimen ['Anton
Bruun' st. 3908].
This species is liable to confusion with small specimens of Amphioplus (Lymanella)
integer with the fourth oral papilla undeveloped or more or less completely concealed
behind the enlarged third one. However, the Amphipholis can be distinguished by
the third papilla being much more than twice as broad as the second, markedly
broader than in A. integer. Also the tentacle scales are not so much enlarged in
Amphipholis similis and the arm spines are all shorter than the segment and taper
more.
All three specimens are small with d.d. not more than 3 mm, the size of the holo-
type, but the one from Natal has four arm spines proximally, whereas the holotype
had only three as also in Amphioplus (Lymanella} integer at all sizes.
Amphipholis strata Mortensen
Fig. 5b-d
Amphipholis strata Mortensen, iQ33a : 361-363, fig. 68, pi. 19, fig. 20.
MATERIAL. FAL 2i6W, 467!,, 493M, False Bay (c. 34°S : i8|°E), 29-42 metres ;
5 specimens.
ECHINODERMS FROM SOUTHERN AFRICA
b
451
FIG. 5. a. Amphipholis similis Mortensen. NAD 4&C. D.d. 3 mm. One jaw. b-d.
Amphipholis strata Mortensen, b and c. FAL 467!^. D.d. 1-9 mm. b. Dorsal view of
whole disc and one arm base. c. One jaw and the adjacent interradius. d. FAL 493M.
D.d. 7-0 mm. Ventral view of disc edge between two radii from which the entire arms
have been lost. The scale measures i mm.
TRA 1320, 34°i9'6'S : i8°30-5'E (False Bay), depth? ; 4 specimens.
MB 74 J, 770, 34°og'S : 22°O7'E (Mossel Bay), 12-24 metres ; 2 specimens.
SCD 1550, 34°03'S : 25°59'E (Algoa Bay), 85 metres ; i specimen.
DESCRIPTION. All but one of the False Bay specimens are fairly large with
d.d. >6 mm. Only interstitial scales occur between the plates of the rosette. At
d.d. c. 6 mm there are usually only four scales in a line between the radial shields
across each interradius, but six in the largest specimen (d.d. 8-5 mm).
A modification not noticed by Mortensen is the presence of a row of enlarged
squarish disc scales just below the ambitus across each interradius, numbering eight,
sometimes nine, in each area between the genital plates at d.d. c. 7 mm (fig. 5d).
These are less noticeable in the largest specimens.
The oral frame is rather like that of Amphioplus (Lymanella) integer except for the
absence of a fourth oral papilla and a more squared-off sculptured appearance of the
three papillae with the third one particularly broad, as usual in Amphipholis.
The relatively long blunt-tipped arm spines are also reminiscent of Amphioplus
integer.
The small specimen (fig. 5b, c) has d.d. only 1-9 mm ; its two intact arms measure
5-0 and 6-1 mm. It has a huge rosette on the disc without even interstitial scales
between the plates and with only one row of very small scales between each primary
radial and the relatively short broad radial shields. Interradially there are only two
scales spanning the space between the radial shields distal to the primary interradial.
452 AILSA M. CLARK
The row of square scales below the ambitus is even larger than in the adult specimens,
occupying half the length of each ventral interradial area. The main difference from
the larger specimens is in the extremely short arm spines, only about half as long as
the segment.
At the same station in False Bay several young specimens of Amphioplus (Lyman-
ella) integer of similar size were also collected. These differ conspicuously in having
more numerous disc scales, the plates of the rosette already almost completely
separated, while the uppermost scales of the ventral side of the disc are unmodified
and the arm spines just exceed the segment in length. At d.d. 2 mm the fourth oral
papilla is not yet developed in the young A . integer.
Amphioplus (Lymanella) furcatus Mortensen
? Ophiophragmus gibbosus Ljungman, 1867 : 316.
Amphiura incana : Bell, 19055 : 258. [Non A. incana Lyman, 1879.]
? Amphioplus gibbosus : Koehler, 1927 : 21-23, pi- 4. fig8- 5. 6.
Amphioplus furcatus Mortensen, I933a : 370-372, fig. 75.
MATERIAL. NAD 83?, 29°io'S : 3i°4o'E (E of Tugela River mouth, N Natal),
33 metres ; i complete and 2 discless specimens ['Anton Bruun' st. 3566].
[Also 4 syntypes, Tugela River mouth, B.M. no. 1904.6.8.1-3.]
NOMENCLATURE. After studying numerous specimens from the Persian Gulf,
Mortensen (1940, Danish Scientific Investigations in Iran, part 2 : 95-97) concluded
that his Amphioplus furcatus is synonymous with A. hastatus from Mozambique and
other tropical Indo-West ^Pacific localities. Although unfortunately most of the
few specimens of A . hastatus in the British Museum collections are relatively small,
d.d. <4 mm, I am not convinced that Mortensen was correct in this case. The four
syntypes of A . furcatus and one Survey specimen which are complete with the disc
all have relatively finer scaling and the radial shields relatively smaller, narrower and
more tapering proximally than in the specimens of A . hastatus studied and the syn-
type figured by Koehler (1927, pi. 3, fig. 2) . Also, contrary to Mortensen's description,
two of his four syntypes do have the rosette just distinct though the plates are
relatively small and widely separated, even in the smaller of the two with d.d.
c. 3-3 mm, whereas in specimens of A. hastatus with the original disc the rosette is
very conspicuous, its plates separated by only one line of scales. The two smaller
specimens of A . furcatus have about 1 1 scales across each interradius along the shortest
line between two neighbouring radial shields but those with d.d. 4 mm or more have
about 13 such scales. These figures compare with minima of five scales in an Arabian
specimen of A. hastatus with d.d. 3-25 mm (Clark & Rowe, 1971, fig. 24a) and seven
scales in the figured syntype with d.d. 4 mm. Such discrepancies seem to me too
large to overlook.
Yet another question of synonymy concerns the identity of the holotype and only
known specimen of Ophiophragmus gibbosus Ljungman, 1867, type locality Port
Natal, d.d. only 2-33 mm (2-5 mm according to Koehler). In the contiguous radial
shields, four oral papillae, three arm spines and two tentacle scales, Ljungman's
description of this species agrees both with Amphioplus furcatus and with A. integer
ECHINODERMS FROM SOUTHERN AFRICA 453
Ljungman, type locality also Port Natal. The presence of a rosette may be matched
by both species at this small size, though it may be less conspicuous in A . furcatus.
However, the description of the margin of the disc in 0. gibbosus as having a row of
rounded or tubercular papillae (as characteristic of the genus Ophiophragmus) fits
with neither. It seems though that Ljungman made a mistake in so describing the
margin of the disc since Koehler (1927) re-examined the holotype and could not find
any marginal papillae or tubercles, the only modification being that the marginal
scales were slightly elevated and projecting. This observation prompted him to
remove the species to Amphioplus. The projection of the marginal scales could
resemble the condition found in A . furcatus, where all five specimens with discs have
the free edges of the uppermost row of ventral scales extended into spinose, some-
times bifid, projections. Thanks to the kindness of Dr P. A. Andersson I have been
able to borrow the holotype of A. gibbosus from the Riksmuseum, Stockholm, but
unfortunately it is now in an even worse condition than when Koehler saw it, badly
decalcified with many of the superficial plates completely eaten away. At the same
time I borrowed the holotype of A . integer, also collected at Port Natal by Wahlberg,
which is in good condition. This fact in itself implies affinity of A. gibbosus with A.
furcatus, even the recently collected Survey specimens of which are not strongly
calcified, their arm plates being semitransparent, showing the underlying vertebrae
beneath, in contrast to A. integer. Conversely, however, the arm spines of A.
gibbosus were described as disproportionately stout by Koehler, agreeing with A.
integer, not with the relatively slender tapering spines of A . furcatus. In view of this
contradiction, coupled with the hopeless condition of the holotype, I think that it is
best to reject the name Ophiophragmus gibbosus.
TABLE 3
Numerical data from the four syntypes (i, 2, 4 and 5) of Amphioplus furcatus with discs
and one specimen from NAD 83 (3)
Disc diameter (d.d.) 5-2 5-0 4-0 3-7 3-3
Radial shield length i-o 0-9 0-8 0-7 0-75
Disc radius : radial shield length 2-5 : i 2-75 2-4 2-7 2-2
Radial shield length : breadth 2-4:1 2-6 2-5 2-3 2-3
Minimum no. interradial scales 13 13 13 n n
Arm segment length : spine length i-o : i i-i i-o i-o 1-3
The number of interradial scales is from a linear series across the interradius between neighbouring
radial shields. The spine lengths were taken from the middle arm spine of about the third free segment.
[The small discrepancies in the ratios of disc radius to radial shield length compared with the figures
above them are due to using the radial disc radius, which may be slightly less than half the diameter
and to approximations in the conversion from micrometer scale units to the nearest tenth of a millimetre.]
Amphioplus (Lymanella) integer (Ljungman)
Fig. 6
Amphipholis integra Ljungman, 1867 : 313.
? Ophiophrcgmus gibbosus Ljungman, 1867 : 316.
454 AILSA M. CLARK
Amphiura Integra : Koehler, igc^a : 65-66, figs. 16, 17.
Amphioplus integer : H. L. Clark, 1923 : 330-331 ; Mortensen, i933a : 368-370, figs. 73, 74.
? Amphioplus gibbosus : Koehler, 1927 : 21-23, pi- 4> figs- 5. 6.
Amphioplus (Lymanelld) integer : A. M. Clark, 1970 : 52.
MATERIAL. LAM 29!!, 32°05'S : i8°i7'E (Lambert's Bay), 20 metres ; i speci-
men.
SBS 33K, Lambert's Bay, 10 metres ; i specimen.
FAL io8E, 4&7L, 486E, 745?, False Bay (c. 34°S : i8|°E), 7-62 metres ; 21
specimens.
F 169, False Bay ; i specimen.
HM 8E, Hermanns (34°S : I9°E) ; i specimen.
CI 4D, Cape Infanta (34°S : 2O°E) ; i specimen.
KKN 4iQ, Buffalo Bay, near Knysna (34°S : 23°E), intertidal ; i specimen.
SCD 329A, 34°04'S : 23°23'E (E of Knysna), 22 metres ; 5 specimens.
SCD 2606, 33°48'S : 25°47'E (Algoa Bay), 26 metres ; 3 specimens.
LLL 8Y, East London (33°S : 27°E) ; i specimen.
MOR 396, 4iX, 44A, 4gM, 506, 5iV, yoV, 77M, 10711, 1158, ii7F, I27E, 1288,
i64N, i68Q, i7iE, Morrumbene estuary, Mozambique (23°S : 35°E) ; c. 55 specimens
including some discless and others juvenile.
VARIATION AND SYNONYMY. There is considerable variation in the form of the
disc, at least partly correlated with regeneration following its loss, the most obvious
consequence of which is absence of the rosette. When regeneration is incomplete
the radial shields are abnormally short. The scales may make a fairly smooth cover-
ing or they may be somewhat thickened and if the disc is flattened in preservation
then the marginal scales at the junction between the dorsal and ventral ones tend to
project, though they do not have the spinose prolongations found in Amphioplus
furcatus. Koehler (1904) described the marginal scales of the small holotype of
Ophiophragmus gibbosus Ljungman (from Port Natal like that of A. integer) as erect.
As detailed under the heading of A. furcatus, it is debatable whether 0. gibbosus is
conspecific with that species or with A . integer. The holotype is now unrecognizable.
In 1971 (Clark & Rowe : 102-103), on the basis of some of the specimens from
Morrumbene, I treated Amphioplus integer as a species with relatively narrow radial
shields, thinking that Koehler's diagrammatic figure of the holotype (1904, fig. 16)
showed them as shorter than reality, especially as Mortensen's specimen (also from
Durban) has relatively longer shields, length : breadth c. 2-4:1. However, since
the Cape Province specimens collected by the Ecological Survey now examined show
some divergence in comparison with the Morrumbene specimens and both lots show
considerable variation in shield shape and arm spine length, I borrowed the holotype
of A. integer from the Riksmuseum, Stockholm, thanks to Dr P. A. Andersson. Its
locality falls between the two groups of specimens geographically. The length :
breadth ratio of its shields is only c. 1-9 : i, Koehler's figure being actually fairly
accurate in this respect, and since the rosette is present there is clearly no regenera-
tion in process. This ratio falls within the range of 1-6-2-5 : I (mean 2-2 : i) found
in 18 specimens from around Cape Province, compared with a range of 2-3-3-7 : J
(mean 2-8) for 15 specimens from Morrumbene.
ECHINODERMS FROM SOUTHERN AFRICA
455
0
FIG. 6. Amphioplus (Lymanella) integer (Ljungman). Holotype. Port Natal. Stock-
holm Museum no. 771. D.d. 5 mm. a. Dorsal view of part of disc and one arm base. b.
Uppermost arm spine viewed perpendicularly, c. One jaw. The scale measures i mm.
As for relative arm spine length, the maximum length of the middle spine compared
with the length of the corresponding segment is 1-5:1 in the holotype, 1-4-1 -8
(mean 1-6) : i in the Cape specimens and 1-0-1-25 (mean i-i) : i in the Morrumbene
specimens. The specimens from the False Bay area seem to have relatively the
longest spines.
Possibly the strictly South African specimens with their broader radial shields
and longer spines merit a subspecific distinction from those taken at Morrumbene.
Unfortunately Balinsky (1957) does not comment on either of these characters in his
specimens from Inhaca, midway between Durban and Morrumbene.
It should be noted that Koehler's 1904 figure of the holotype of Amphioplus integer
exaggerates the shape and length of the uppermost arm spines, which in fact barely
exceed the segment in length (Fig. 6b) and are flattened and only slightly expanded
at the tip, appearing somewhat truncated when foreshortened but rounded when
viewed from above. None of the other specimens I have studied have the uppermost
spine modified to this extent.
Amphioplus (Unioplus) falcatus Mortensen
Fig. 7
Amphioplus falcatus Mortensen, iQ33a : 365-367, figs. 70, 71, pi. 19, figs. 18, 19.
Unioplus falcatus : Fell, 1962 : 16.
Amphioplus (Unioplus) falcatus : A. M. Clark, 1970 : 44, 49.
MATERIAL. ABD I5E, 29°42'S : 3i°38'E (NE of Durban), 350 metres ; i discless
specimen ['Anton Bruun' st. 39oE].
NAD 3iV, 29°37'5S : 3i°33'E, 175-200 metres ; i specimen lacking arms [St.
390H].
31*
456
AILSA M. CLARK
NAD 35V (part), 29°35'S : 3i°38'E, 150 metres ; 9 discless specimens and 4 discs
[St. 39oL].
NAD 466, 29°35'S : 3i°42'E, 138 metres ; 2 discless specimens [St. 3908].
Judging from this series of specimens, the normal oral formula in this species is
clearly m,om,m + t, the tentacle scale rarely being quite in series with the oral
papillae. Only on very occasional jaws is a fourth oral papilla present.
The great length of the radial shields, more than half the disc radius, and their
sickle-shaped (falcate) form, with a distinct kink at about a quarter of the way from
the distal end, is very conspicuous and in addition many of the scales have an orange
patch, giving a peculiar spotty appearance. On one disc particularly the pigmented
areas are also slightly elevated and the profile is correspondingly uneven. The arm
plates and spines are also partially orange coloured.
The largest specimen has d.d. 10 mm.
This species is sympatric with Amphioplus pectinatus but the single elongated
tentacle scale and the fewer oral papillae immediately distinguish it even if the disc
with its unusual radial shields is lacking.
FIG. 7. Amphioplus (Unioplus) falcatus Mortensen. NAD 35V. a. Dorsal view of part
of detached disc, d.d. 10 mm. b. Two jaws and the first arm segment, c. Dorsal view
of twentieth free arm segment. The scale measures i mm.
Amphioplus (Amphioplus) pectinatus Mortensen
Fig. 8
Amphioplus pectinatus Mortensen, I933a, : 367-368, fig. 72.
MATERIAL. NAD 3iW, 29°37'5'S : 3i°33'E (NE of Durban), 175-200 metres ;
5 discless specimens and i disc, form A ['Anton Bruun' st. 390!!].
NAD 35V (part), 29°35'S : 3i°38'E, 150 metres ; 6 discless specimens, form A
[St. 3QOL].
ECHINODERMS FROM SOUTHERN AFRICA
457
FIG. 8. Amphioplus (Amphioplus) pectinatus Mortensen. a and b. Holotype. Copen-
hagen Museum ; from sketches by Dr Madsen. a. Oblique view of disc radially showing
'combs' distal to the radial shields, b. Eighth and ninth dorsal arm plates, c-e.
NAD 3iV. Form A. c. Dorsal view of part of detached disc, d.d. c. 8 mm. d. Two
jaws and first two arm segments, e. Dorsal view of fifteenth free arm segment, f-i.
NAD 6iR. Form B. f. Oblique radial views of two different detached discs, g. Two
jaws and first two arm segments with h. Enlargement of oral slit between two jaws,
i. Dorsal view of fifteenth free arm segment. The scale measures i mm for c-g and i
and 0-5 mm for h.
NAD 460, 2g°35'S : 3i°42'E, 138 metres ; i discless specimen, form A [St. 3908].
NAD 520, 29°29'S : 3i°45'E, 86 metres ; 3 discless specimens and i disc, form B
[St. 39iC].
NAD 550, 29°29'S : 3i°45'E, 86 metres ; 2 discless specimens and 3 discs, form B
[St. 39iB].
NAD 6iR, 29°26'S : 3i°46'E, 77 metres : 3 discless specimens and 4 discs, form B
[St. 39iF].
VARIATION. As indicated above, two forms, A and B, are now attributed to
Amphioplus pectinatus from the Survey collections, the first from samples NAD 3iW,
35V and 460 and the second from 526, 550 and 6iR. I am by no means sure that
458 AILSA M. CLARK
these are conspecific but, as none of the discs found are attached to oral frames,
there is an element of doubt that they do correspond. In addition, the intact
holotype of A. pectinatus with d.d. only 5 mm appears to have tentacle scales like
form A but dorsal arm plates and discs as in form B. I am indebted to Dr Madsen
of the Copenhagen Museum for comments on and sketches of Mortensen's two type
specimens. He notes that Mortensen's fig. 72 is reasonably accurate except with
regard to the projections distal to the radial shields in fig. 72b, which appear rather
as in the detail in fig. 72c. Also Madsen's drawing of the eighth and ninth dorsal arm
plates of the holotype shows them as more convex distally, being more nearly ellip-
tical than triangular in shape, though in fact in the paratype (d.d. 4 mm) their shape
is more triangular, agreeing with fig. 72b. The paratype also has the disc scaling
fairly coarse and its rosette of primary plates is distinguishable. Both types have
the second from outermost oral papilla distinctly enlarged and the smaller specimen
has the distalmost papilla reduced in size. The oral tentacle scale (or second papilla)
is superficial in the paratype and in series with the other papillae but in the holotype
it is slightly inset. Incidentally, the longitude of the type locality is the subject
of a printer's error and should have been 3i°i9'E, not 49', from which station several
other amphiurids including Amphiolus falcatus were also taken.
The differences between the two present forms are as follows :
First oral tentacle scale single (Fig. 8d)
Dorsal arm plates with a more or less
distinct median distal angle (Fig. 8e)
Arm spines five on a few proximal
segments (six in one specimen, d.d.
7 mm)
Some tentacle pores with only one scale
rather than two ; the scale on the ventral
arm plate short (Fig. 8d)
Usually two oral tentacle scales adjacent
on the side of the oral plate, the proximal
one sometimes almost superficial, the
distal one tending to be hidden behind
the second oral papilla (Fig. 8g)
Dorsal arm plates almost elliptical in
shape, the distal side convex but flat-
tened medially (Fig. 8i)
Arm spines four on proximal segments,
rarely five at d.d. 8 mm
All tentacle pores with two scales, the
one on the ventral arm plate long and
narrow (Fig. 8g)
The eight discs of form B agree fairly well with that of the holotype of Amphioplus
pectinatus. The scaling is somewhat finer but this is probably accountable to the
fact that they are larger with diameter up to 8 mm. None of them show a primary
rosette so it is possible that all may be regenerated rather than original discs, the
species is clearly vulnerable to shedding of the disc. Distal to the radial shields on
the outer end of the genital plate they all have the spinose projections which prompted
the specific name ; these may be arranged irregularly but in one or two cases seem
to be in paired vertically aligned combs (Fig. 8f, lower).
The single disc found in the samples of form A contrasts in having extremely
small but very thick and projecting scales recalling those of the holotype and only
known specimen of Amphiura acutisquama A. M. Clark, 1952, from the west coast of
South Africa. However, the present disc is also unusual in having the six plates of
ECHINODERMS FROM SOUTHERN AFRICA 459
the primary rosette showing as conspicuously larger than the central scales. In
addition it has spinose projections distal to each radial shield, though these are not
so well developed as in the discs from the samples of form B. The ratio of radial
disc radius to radial shield length is 2-7 : i, though if the ratio is calculated from half
the disc diameter its value comes to 3-2 : i since the radial radius is markedly less
than the interradial one owing to bulging of the disc. Length : breadth of the radial
shields is between 4-0 and 4-5 : i. These compare with ratios of radial radius : shield
length of 2-0-2-5 : i, the shields being relatively larger, on the discs of form B, with
shield length : breadth 5-6:1. In spite of these differences, the coincidence of the
development of the spinose processes on the genital plates indicates a close relation-
ship. Also, in the type material of Amphioplus pectinatus the radius : shield length
is about 3 : i and the scales are thick, though coarse.
Amphioplus sp. indet. A (aff. A.falcatus)
Fig. 9
MATERIAL. TRA 104!., 34°3i'S : i9°2i'E (near Hermanus), 22 metres ; i discless
specimen.
MB 366, 34°o8'52"S : 22°og'ig"E (Mossel Bay), 11-5 metres ; i discless specimen.
MB 376, 34°09'i8"S : 22°io'o2"E, 31 metres ; arms.
MB 7iF, 34°o8'46"S : 220O7'20"E, 12 metres, i discless specimen.
MB 75F, 34°o8'4o"S : 22°O7'26"E, 15-5 metres ; i discless specimen.
LIZ 13?, 33°58-2'S : 25°38-8'E (Port Elizabeth), 7-8 metres ; 2 discless specimens.
NAD 63Q, 29°2i-6'S : 3i°35'7'E (NE of Durban), 57 metres ; i discless specimen
['Anton Bruun' st. 391 J],
DESCRIPTION. All these incomplete specimens appear to belong to an undescribed
species. Without knowledge of the disc it seems unwise to give them a specific name.
The torn edges of some of the discs still attached to the oral shields indicate that
scaling was present ventrally, though probably tenuous.
The largest specimen, from LIZ 13?, had d.d. c. 10 mm, judging from the extent
of the missing dorsal arm plates basally. The oral frame measures 2-6 mm from the
distal edge of one oral shield to the distal edge of the first ventral arm plate opposite.
The oral shields are elongate hexagonal, broadest in the middle with the proximal
and distal lobes broadly truncated. The adoral shields have a distal lobe between
the oral shield and the adjacent lateral arm plate but are very short interradially
and widely separated. The oral plates are sunken distally but sharply elevated
proximally, increasing the vertical height of the jaw. Of the four oral papillae, the
infradental is squarish but rounded at the corners. The second and third papillae
both arise from the side of the oral plate and are finger-like in shape, about three
times as long as broad. The short rounded fourth papilla arises from a proximal
extension of the first ventral arm plate, not from the adoral. The oral tentacle
scale is inset in the oral slit and is also finger-like but smaller than the second oral
papilla which may hide it. The oral formula is m,mm, ,m + t.
460 AILSA M. CLARK
FIG. g. Amphiopus sp. A (aff. A. falcatus). LIZ 13?. D.d. c. 10 mm. a. Two jaws
and the first two arm segments, b and c. Dorsal and ventral views of fifteenth arm
segment. The scale measures i mm.
The arms increase in width from a minimum of 0-75 mm proximally to i-o mm at
about the sixtieth segment, by which they are all broken ; they were probably
relatively long. The dorsal arm plates are approximately elliptical, though slightly
broader towards the proximal end ; most of them are just contiguous or very slightly
overlapping. The ventral arm plates are at first squarish, the proximal lobe being
broadly truncated and the sides only slightly tapering ; they soon become almost
hexagonal with the proximal lobe more prolonged and the distal tapering more
markedly. The distal edge of the first two or three plates is slightly convex but it
soon becomes straight and then develops a small notch in the middle. There are four
arm spines on the entire length of arm remaining, all somewhat flattened but the
uppermost one more so ; all taper more abruptly near the tip. The proximal spines
are all slightly longer than the corresponding segment and those on the broader part
of the arm are half again as long as the segment. There are no proper tentacle
scales but in the skin covering many of the pores there are single or sometimes double
calcareous accretions, well isolated from the ventral and lateral arm plates, which
may be homologous with tentacle scales.
The second specimen from sample LIZ 13? is much smaller. Its oral shields are
much shorter, broader than long, though the adorals are again widely separated.
The arm plates are relatively narrower, as would be expected, the ventral arm plates
longer than broad. The tentacle pores appear quite naked.
The other oral frames also show some variation in the shape of the oral shields,
which may be circular, octagonal or pentagonal. There may be a small gap in the
row of oral papillae between the first and second ones and in one specimen the two
elongated middle papillae are flattened and slightly fluted. Distal portions of arms
are unusual in that, although the arms taper, the segments remain broad and short
instead of attenuating and the arm becomes very compact. The number of spines
drops to three and the lowest one becomes relatively longer than the other two, up
to twice as long as the segment. Some of the longer arm spines may have a very
small terminal hook or a few divergent thorns.
The spiniform middle oral papillae, the broadly separated adoral shields, the
ECHINODERMS FROM SOUTHERN AFRICA 461
reduction of the tentacle scales and the distally compact arms easily mark off this
species from the other South African amphiurids.
? Amphioplus sp. indet. B
MATERIAL. FED 58, 24°46'S : 35°i8'E (S of Inhambane, Mozambique), no
metres ; i specimen ['Anton Bruun' st. 37iF].
D.d. is only 3-3 mm in this specimen and only part of one arm remains attached to
the disc. In many ways it agrees superficially with Amphipholis similis, as figured
by Mortensen, but the third oral papilla is only about twice as broad as long and is
also very thick, while in some series there appears to be a fourth papilla behind the
third one, as in young Amphioplus (Lymanella} integer. The shapes of the oral
shields and arm plates agree with A. similis as well as with A. integer but the disc
scaling is very thick, much more so than in the present material of either of the species
named. There is no distinct rosette but this could be due to regeneration of the disc.
Also the radial shields are all less contiguous than in the two species named, four pairs
being separated for at least their proximal thirds and the fifth only touching at their
distal ends. On the first two free arm segments there are four arm spines, then
three. The two fairly large tentacle scales agree with those of A. similis but those
of A . integer are even larger and usually somewhat angular.
If the oral papillae are only transitionally three, this with the broadness of the
third and distalmost and the straightness of the papilla series suggests affinity with
Amphioplus (Lymanella). However, the more or less separated radial shields and
the somewhat smaller tentacle scales and fourth arm spine disagree with this.
Apart from South African species, this specimen also has some resemblance to
Amphioplus platy acanthus Murakami, 1943, from the Caroline Islands, but that has
the fourth oral papilla fully exposed at d.d. 3 mm. Amphioplus (Amphichilus]
ochroleuca (Brock, 1888) also has thick disc scales and only three oral papillae fully
developed but the third of these is no larger than the second and there is no sign of
a fourth papilla underlying it, even at a larger size.
Amphiodia sp. aff. A. microplax Burfield
? Amphiodia sp. Balinsky, 1958 : 10.
MATERIAL. NAD 6;H, 29°2i-6'S : 3i°357'E (NE of Durban), 57 metres ; i
discless specimen ['Anton Bruun' st. 3QiH].
NAD 78G, 29°i9'i'S : 3i°26-8'E, 38 metres ; i discless specimen [St. 392 J].
NAD 856, 29°io'S : 3i°5i'E, 42 metres ; 3 discless specimens [St. 356G].
These incomplete specimens have the oral shields relatively long and only a single
tentacle scale. This accords with Amphiodia (Amphispina} microplax Burfield,
known from the Red Sea and Persian Gulf, which also frequently loses its disc so
that the presence of the modified marginal disc scales prolonged into spinose pro-
cesses, characteristic of the subgenus Amphispina, cannot be ascertained. The other
Amphiodia known from the western Indian Ocean - A. dividua Mortensen - has two
tentacle scales, apart from being fissiparous and six-armed.
462 AILSA M. CLARK
Ophionephthys lowelli* sp. nov.
Fig. 10
MATERIAL. SCD 748, 32°33'S : 28°38'E (NE of East London), 55 metres ; 2
specimens with discs, 2 loose discs and 21 discless specimens.
SCD 82?, 33°03'S : 27°54'E (East London), 51 metres ; i specimen.
DESCRIPTION. The holotype has the disc partly detached from the oral frame
and rather shrunken. Judging from the extent of the incomplete proximal dorsal
arm plates, its d.d. was c. 7-5 mm ; the diameter of the oral frame is 2-3 mm. The
arms are all broken off short ; judging from the remaining material they were
extremely long, well over ten times the d.d.
FIG. 10. Ophionephthys lowelli sp. nov. a, b. Holotype. SCD 748. a. Dorsal view of
part of disc (somewhat reconstructed) and base of one arm from which the disc has
shrunk back revealing segments with the dorsal arm plates divided by erosion of the
median part. b. Two jaws and the first two arm segments, c-f. Paratype. c. Jaw with
oral shield truncated proximally. d-f. Dorsal, ventral and side views of twentieth arm
segment. The scale measures i mm.
The upper side of the disc was probably covered with continuous extremely fine
scaling, except towards the periphery where it turns brown when partially dried.
However, there is a conspicuous row of enlarged scales extending interradially from
the abradial edge of each radial shield at about half-way along its length. These
belts do not appear to meet interradially but the shrinkage of the discs renders this
uncertain. The radial shields have length : breadth 1-9:0-5 mm = nearly 4 : i and
* Named after Lowell P. Thomas, whose fine work on the amphiurids of Florida (1962) includes a
redescription of the type-species of Ophionephthys, O. limicola Liitken.
ECHINODERMS FROM SOUTHERN AFRICA 463
are contiguous in the distal half though narrowly separated proximally, at least in
the preserved condition. The scaling at their proximal ends is a little more sub-
stantial than elsewhere. The ventral skin of the disc bears only scattered rudi-
mentary scales and is more or less transparent.
The oral shields are as long as broad ; the proximal lobe is almost semicircular
but for a slight medial flattening and the distal lobe is distinctly angular. The
adoral shields are widely separated interradially but have a broad distal lobe between
the oral shield and the first lateral arm plate on each side. The oral plates are fairly
long and sunken, without a proximal peak and the broad infradental papillae, as
preserved, project almost vertically from their apical ends. There are two spiniform
oral papillae on each side of the jaw, even the distal one distinctly separated from the
adoral shield ; their tips are rugose and they also tend to project vertically. There
is a small spiniform oral tentacle scale inset in the slit close to the teeth. The oral
formula is m,mm,o + t.
The dorsal arm plates are bucket-shaped, tapering distally and relatively narrow,
as long as broad or longer. The ventral arm plates are similar in shape, broadly
overlapping, the distal edge convex. The lateral arm plates bear four spines, the
lowest one about equal in length to the segment, the rest slightly shorter. There is
one fairly long tentacle scale on each pore, equal in length to about half the length of
the ventral arm plate.
VARIATIONS. The other oral frames show considerable variation in the shape of
the oral shields, which may be rhombic but more often have the proximal lobe very
reduced and flattened, with the adoral shields more widely separated than in the
holotype. There are usually five arm spines on the proximal segments, of which the
lowest slightly exceeds the segment length beyond the base of the arm and all are
somewhat flattened and blunt-tipped ; under a high magnification they appear very
finely ridged.
Although there is some variation in the size of the oral papillae, their number
seems fairly constant ; only in one specimen do most of the jaws lack the second
papilla. None were seen to have a fourth papilla on the adoral shield.
AFFINITIES. The combination of normally two oral papillae on each side of each
oral plate with an oral tentacle scale in addition and the presence of belts of enlarged
scales contrasting with extremely reduced scaling elsewhere on the disc indicate
that this species is congeneric with Ophionephthys limicola Liitken, the type and only
species of that genus since I reviewed it in 1970. The only discrepancy from my
diagnosis of the genus is the minor one of the somewhat more extensive, though still
very reduced, disc scaling. 0. limicola is known from the West Indies. Geographi-
cally much closer and morphologically nearly as similar is the species from Madagascar
recently described by Cherbonnier (1972) as Amphioplus polymorphus, which likewise
has belts of enlarged disc scales extending interradially from the radial shields.
Cherbonnier made no comparison of A . polymorphus with 0. limicola, only with those
species which Fell had referred to Ophionephthys in 1962 and I had transferred to
Amphioplus, presumably because A. polymorphus seems consistently to have a
fourth oral papilla present on the adoral shield, although its jaws are otherwise rather
464 AILSA M. CLARK
variable. Cherbonnier compared it also with the West Indian A mphioplus coniortodes
H. L. Clark, 1918, which, incidentally, is sympatric with 0. limicola. It is noteworthy
that the two West Indian species both have almost keyhole-shaped oral shields
proximal to which the adorals meet broadly, whereas both Ophionephthys lowelli and
A . polymorphus have the proximal lobe of the oral shields more or less reduced and
the adorals widely separated. The relationships of all four will bear future study.
Apart from the difference in the oral and adoral shields, Ophionephthys lowelli
differs from 0. limicola in the spiniform shape of the second and third oral papillae,
the more elongate dorsal arm plates and the blunter arm spines, though the more
extensive disc scaling of the holotype may be subject to variation, as in Amphioplus
polymorphus, which likewise has the papillae on the side of the oral plates elongated
and the arm spines blunted. Clearly Ophionephthys lowelli occupies an intermediate
position.
Amphilepis scutata Mortensen
Amphilepis scutata Mortensen, iQ33a : 372-373, fig. 76.
MATERIAL. NAD 3iX, 29°37'5'S : 3i°33'E (NE of Durban), 175-200 metres ;
i specimen ['Anton Bruun' st. 390!!].
D.d. is only 3-3 mm compared with 5 mm in the holotype. The radial shields are
relatively smaller, more widely separated interradially and more tapering proximally
than shown in Mortensen's figure, so that the widest part is near the distal end, not
in the middle. Even the first two ventral arm plates are not contiguous at this size.
Family OPHIAGTIDAE
Ophiactis plana Lyman
Ophiactis plana Lyman, 1869 : 330-331 : H. L. Clark, 1923 : 333 : Mortensen, i933a : 345-346,
fig. 57 ; H. L. Clark, 1939 : 76-77.
? Ophiactis lymani Ljungman, 1871 : 629 ; Mortensen, 19335 : 442-449, figs. i5a-d, i6d, e ;
Balinsky, 1957 : 14.
? Ophiactis parva Mortensen, 1926 : 123-124, fig. 12 ; Balinsky, 1957 : X5-
Ophiactis flexuosa (part) Lyman, 1882 : 116 ('Challenger' st. 142).
MATERIAL. SST iaL, M, 35°22'S : 22°3i'E (SSW of Knysna), 200 metres ; 3
specimens.
SCD I55K, 34°O3'S : 25°59'E (Algoa Bay), 85 metres ; i specimen.
SCD 1770, 34°2o'S : 23°3i'E (SE of Knysna), no metres ; i specimen.
SCD 2030, 34°5i'S : 23°4i'E, 184 metres ; i specimen.
SCD 24gE, 34°48'S : 23°39'E, 146 metres ; 2 specimens.
NAD I5N, 3O°47'S : 30°27'5'E (S of Durban), 36 metres ; i specimen.
NAD 27F, 29°53'5'S : 3i°577'E (E of Durban), 71 metres ; 3 specimens.
NAD 8iR, 29°ii-5'S : 3i°37'E (NE of Durban), 18 metres ; 2 specimens ['Anton
Bruun' st, 3566].
DBN 238K, Durban Harbour entrance, LWST ; 2 specimens.
ECHINODERMS FROM SOUTHERN AFRICA 465
Most of these specimens are very small, d.d. <3 mm. The only one larger than
this lacks spinelets on the disc but these are present in small numbers in one or two
of the other specimens and show a small hyaline terminal point, much as in Morten-
sen's figure (i933b) from the specimens from St Helena which he refers to Ophiactis
lymani. In addition the South African specimens resemble the West Indian holo-
type of 0. lymani in having slightly concave distal edges to the ventral arm plates,
unlike Mortensen's specimens. Lyman described the distal edges of these plates in
0. plana as simply 'curved', presumably convex since he did not qualify it with
're-entering'.
In his St Helena paper Mortensen (ig33b) made no reference to Ophiactis plana.
Both Mortensen (ig33a) and Madsen (1971) have commented on the close simi-
larity between the West African Ophiactis luetkeni Marktanner-Turneretscher and
the South African 0. carnea Ljungman, both with five arms and more or less ovate
dorsal arm plates. However, the parallel resemblance between the geographically
corresponding 0. lymani and 0. plana, both with six arms and fan-shaped dorsal
arm plates, escaped them. The naming of South African specimens as Ophiactis
plana follows from Lyman's comment (1882) that some collected on the Agulhas
Bank by the 'Challenger' were 'scarcely to be distinguished' from 0. plana though,
for some reason, he referred them provisionally to 0. flexuosa, simultaneously
described from north of New Zealand. Mortensen accepted H. L. Clark's usage of
the name 0. plana for South African material as he assumed that Clark would have
compared them with Lyman's type material, also in the Harvard Museum, from the
West Indies.
Surprisingly, Lyman (1869) made no mention of the arm number, when describing
the type material of Ophiactis plana, or again in his 'Challenger' report where he gave
a synopsis of the species of Ophiactis. Only his comment about the six-armed
South African specimens resembling 0. plana implies that this number was also
found in the type material. H. L. Clark also gave little weight to the number of
arms as a specific character in the genus Ophiactis and in 1915 made no comment
when he illustrated as 0. plana a six-armed specimen from off Georgia. Mortensen
(1933 a) assumed that this was one of the type specimens but they were from the
Florida Strait area.
Mortensen (1940) contested H. L. Clark's provisional reference of Ophiactis parva
from the Suez Canal to the synonymy of 0. plana in 1939 but the diminutive type
material shows little in the way of definitive characters to justify distinguishing it
specifically in my opinion.
REGENERATION. The specimen forming sample SST I2M is extraordinary in that
it has five arms altogether with one old arm and four regenerating.
Family OPHIOTRICHIDAE
Ophiothrix (Acanthophiothrix) proteus Koehler
Fig. na, b
Ophiothrix comata : Koehler, 1898 : 105, pi. 2, figs. 11-14. [Non O. comata Muller & Troschel,
1842.]
466 AILSA M. CLARK
Ophiothrix proteus Koehler, 1905 : 100-101 ; 1922 : 260-261, pi. 36, figs. 3, 4, pi. 101, fig. 3.
Placophiothrix proteus : H. L. Clark, 1939 : 86 ; Balinsky, 1957 : 21.
Ophiothrix (Acanthophiothrix) proteus : A. M. Clark, 1967 : 643, 648.
MATERIAL. NAD 28, 3O04y-i'S : 3O°29'i'E (S of Durban), 44 metres ; i specimen.
NAD 526 and 556, 29°2g'S : 3i°45'E (near Tugela River mouth), 86 metres ; 3
specimens ['Anton Bruun' st. 3916, C].
NAD 58A, 29°26'S : 3i°46'E, 77 metres ; 3 specimens [St. 39iF].
NAD 63(3, 29°2i-6'S : 3i°357'E, 57 metres ; 5 specimens [St. 391 J].
FED igH, 25°07'S : 34°34'E (between Delagoa Bay and Inhambane, Mozambique),
112 metres ; i specimen [St. 372L].
AFFINITIES. Although similar to Ophiothrix aristulata in the bare radial shields,
long tapering arm spines, slender disc spines and light mid-line to the arms dorsally,
these specimens have the pair of dark lines on the arms much better denned, the light
mid-line between them being raised on a sharp continuous crest while the plates are
broadly truncated at the distal end. In fact the largest specimens seen, d.d. 15-
17 mm, have the dorsal arm plates broader than long, length : breadth i : c. 1-7,
contrary to my diagnosis of the subgenus Acanthophiothrix. Nevertheless, their
truncated shape allies the species with the other members of the subgenus, few of
which reach such a relatively large size.
[See also the discussion under the heading of Ophiothrix aristulata.]
Ophiothrix (Ophiothrix) aristulata Lyman
Fig. nc, d
Ophiothrix aristulata Lyman, 1879 : 50-51, pi. 15, figs. 421-424 ; 1882 : 223-224, pi. 21, figs.
9-12 ; Koehler, 190413 : 151 ; (?) 1922 : 205-208, pi. 35, figs. 1-3, pi. 97, fig. i ; A. M. Clark,
1967 : 640.
Placophiothrix aristulata : H. L. Clark, 1939 : 85.
MATERIAL. WCD 2i8Z, 34°43'3'S : i8°i2-5'E (SW of Cape Point), depth ? ;
2 specimens.
AFR 7610, 30°i3'S : I5°I7-6'E (W of Hondeklip Bay), 260 metres ; 5 specimens
[including three in the British Museum collections, no. 1949.9.27.119].
AFFINITIES. In splitting off Acanthophiothrix from Placophiothrix in 1967, I
referred Ophiothrix aristulata back to Ophiothrix sensu stricto. However, an attempt
to key it out from 0. (Acanthophiothrix) proteus, collected off Natal and Mozambique
revealed its intermediate position, emphasizing the infra-generic rank of Acantho-
phiothrix.
The dorsal armament of the disc, consisting exclusively of more or less rugose
spines in the holotype (d.d. 14 mm) and other similarly large specimens, seemed at
first sight to provide a reliable distinction of Ophiothrix aristulata from 0. (Acantho-
phiothrix) proteus in which trifid stumps normally predominate over fine spinelets
and often exist alone. However, the armament of a small paratype of 0. aristulata
from the Agulhas Bank with d.d. 7mm also has peripheral trifid stumps while
conversely several large specimens of 0. (Acanthophiothrix} proteus taken off Natal
ECHINODERMS FROM SOUTHERN AFRICA 467
a
FIG. ii. Dorsal views of proximal arm segments, a and b. Ophiothrix (Acanthophiothrix)
proteus Koehler. a. NAD 63Q. D.d. c. 16 mm. b. John Murray Expedition st.
MB IIC, Arabian coast, d.d. 9 mm. c and d. Ophiothrix (Ophiothrix} aristulata Lyman.
c. AFR y6iG. D.d. 16 mm. d. 'Challenger', Agulhas Bank, d.d. c. 10 mm. The scale
measures i mm.
by the Ecological Survey (d.d. c. 15 mm) all have numerous well-developed spinelets
or spines on the centre of the disc, though intermingled with trifid stumps, which
again predominate peripherally. Koehler (1922) has also commented on the varia-
tion in disc armament of 0. aristulata, though I think it possible that not all his
specimens were conspecific.
The shape of the arm spines appeared also distinct when comparing the John
Murray Expedition specimens, which H. L. Clark (1939) referred to these two
species, since those he named Placophiothrix aristulata have the longer spines on the
proximal half of the arms at most four times the segment length with relatively coarse
thorns numbering c. 14 each side, whereas those of proteus have spines up to six times
the segment length with more than 20 fine serrations each side. However, the holo-
type of Ophiothrix aristulata itself has the spines up to 4-6 times the segment length
(according to Lyman) and the number of thorns or serrations may also exceed 20.
Two of the specimens from AFR 761 with d.d. c. 16 mm have unusually broad
dorsal arm plates, breadth : length nearly 2 : i, on the more proximal segments and
they are contiguous for more than a third of their breadth.
I do not agree with Lyman's statement that the dorsal arm plates have a median
ridge giving a carinate appearance. In all the specimens I have studied each plate
has a rounded median distal convexity but proximally is flatter medially, in marked
contrast to the continuous sharp keel on the arms of Ophiothrix (Acanthophiothrix)
Proteus. More material from South Africa is needed to show if this character is
consistent.
Ophiothrix aristulata has an unusually wide geographical range, being also reported
from the Philippines to south-east Australia.
Ophiothrix fragilis (Abildgaard in O. F. Miiller)
Asterias fragilis Abildgaard in O. F. Miiller, 1789 : 28, pi. 98.
468 AILSA M. CLARK
Ophiothrix fragilis : J. Miiller & Troschel, 1842 : no ; H. L. Clark, 1923 : 337 ; Mortensen,
J933a : 338 ; Madsen, 1971 : 213-214, fig. 360.
Ophiothrix triglochis Miiller & Troschel, 1842 : 114 ; Koehler, i9O4a : 81-84, figs. 41-45 ; 1908 :
635 ; H. L. Clark, 1923 : 337-339.
Ophiothrix fragilis var. pentaphyllum : Koehler, 1908 : 635.
Ophiothrix aristulata : Doderlein, 1910 : 254, pi. 5, figs. 4, 5. [Non O. aristulata Lyman, 1879.]
MATERIAL. Forma Pentaphylla : Latitude/longitude squares 26/15, 321I7>
32/18, 33/17, 34/18, 34/21, 33/25, 34/25, 33/27, 32/28, 29/31 ; shore-i48 metres.
Forma triglochis : 33/17, 33/18, 34/18, 34/21, 33/26, 34/21, 34/22, 34/23, 33/25,
34/25, 33/26, 33/27, 32/28, 30/30, 29/31, 28/32 ; 7-183 metres.
SYNONYMY. The type material of Ophiothrix triglochis from Port Natal had the
disc armed only with short bifid or, probably more often, trifid stumps, also covering
the radial shields, though more sparsely. However, many superficially similar
South African specimens actually have the stumps 'multifid', averaging about five
points ; alternatively the stumps may be more or less elongated. H. L. Clark (1923)
comments 'in typical triglochis there are no disc spinelets among the stumps and the
radial shields are more or less covered by the latter'. The dorsal arm plates usually
have the distal edge simply convex or with a rounded angle, though sometimes with
a median distal bump. Such stump-covered South African specimens therefore
approximate either to Koehler's form b of Ophiothrix fragilis variety echinata (1924)
armed with trifid or quadrifid stumps only, or to his variety lusitanica, with quadri-
or quinti-fid stumps. These two northern varieties or forms are both common on the
Mediterranean coasts of France, Spain and Portugal, though lusitanica also occurs
in the Atlantic, as its name suggests. In British waters the common form of
Ophiothrix fragilis is pentaphylla Pennant, also found in France, with spines as well
as stumps on the disc, the radial shields normally relatively large and bare and the
dorsal arm plates rhombic with a well-marked distal angle and their surface carinate.
However, throughout the European range of 0. fragilis may be found specimens
intermediate in some or all of these characters, for instance with bare radial shields
but an armament solely of stumps on the scales or conversely with more or less
numerous stumps on the shields but disc spinelets present on the scales. The form
and number of the arm spines and the relative arm length also vary.
H. L. Clark (1923) named as Ophiothrix triglochis some specimens from False Bay
with spinelets among the disc stumps, in one even replacing the stumps and contrast-
ing with bare radial shields. These clearly approximate to some British specimens of
0. fragilis. Both H. L. Clark and earlier (1908) Koehler identified some specimens of
Ophiothrix from Saldanha Bay with coarse-spined discs and bare radial shields as
0. fragilis, Koehler referring them to the variety pentaphyllum. However, among
the similarly spinose specimens taken by the Ecological Survey are some with the
disc spines very stout, mounted on low but distinct tubercles on the scales, as Koehler
(1924) described for the Mediterranean 0. quinquemaculata (Delle Chiaje) in his key to
the European species of Ophiothrix. Recently, however, Guille (1964) has shown
that 0. fragilis may also encompass 0. quinquemaculata.
In view of the immense variability of Ophiothrix fragilis in the north-east Atlantic,
it seems pointless to distinguish 0. triglochis specifically in the south-east Atlantic.
ECHINODERMS FROM SOUTHERN AFRICA 469
Although the geographical ranges of the two almost coincide in South Africa it
may be useful to retain the name triglochis for specimens without disc spines.
Ophiothela nuda (H. L. Clark) new comb.
Ophiopsammium nudum H. L. Clark, 1923 : 341-342.
Ophioteresis beauforti Engel, 1949 : 140-143, figs, i, 2.
Ophiothela beauforti : Balinsky, 1957 : 22~24> pi- 4> fig- 16 ; A. M. Clark & Rowe, 1971 : 117.
MATERIAL. NAD 2T, 30^7 -i'S : 30°29-i'E (S of Durban), 44 metres ; 2 speci-
mens.
NAD I5R, 30°47'S : 3O°27'5'E, 36 metres ; 4 specimens.
NAD 63Q, 29°2i-6'S : 3i°357'E (near Tugela River mouth), 57 metres : i speci-
men ['Anton Bruun' st. 391 J].
FED loW, 24°46'S : 34°5o'E (S of Inhambane, Mozambique), 22 metres ; 8 speci-
mens [St. 3726].
FED I7H, 24°53'S : 34°56'E, 55 metres ; 14 specimens [St. 3726].
MOR 458, Morrumbene estuary, N of Inhambane (23°S : 35°E) ; i specimen.
VARIATION. Two of the four specimens from sample NAD 15 have some coarse,
spaced granules towards the periphery of the disc dorsally and on the arms ; the
other two are naked except for some peripheral and ventral disc spines. Of FED 10,
seven specimens have a continuous coat of coarse granules, one of them with an
additional cluster of rough spines in the middle of the disc, but the eighth specimen
is almost completely bare on both disc and arms. About half of the 14 specimens
of FED 17 have a few median disc spinelets but only one has granules on the arms,
mainly sited over the vertebrae. MOR 45 has the disc mainly bare above but many
coarse granules near the periphery, which graduate into rugose spines. The arms
are covered with a continuous coat of coarse granules for the proximal half but be-
come naked distally.
NOMENCLATURE. The type locality of Ophiopsammium nudum is near the Ton-
gaat River mouth, Natal, while that of Ophioteresis beauforti is further north in
Mozambique. The holotype of 0. nudum has some inconspicuous disc and arm
granules dorsally, despite its specific name. In view of the notes above and previous
comments by Balinsky and by me on the variability of 0. beauforti, clearly this
nominal species and 0. nudum cannot be specifically distinguished. The type-species
of Ophiopsammium, 0. semperi Lyman, has a continuous coat of extremely fine
granules all over and I am convinced that 0. nudum is much more closely related to
Ophiothela mirabilis (Verrill), the type-species, and other members of the genus
Ophiothela with their relatively coarse granulation, varying in both density and size.
A young specimen from NAD 15 with d.d. only 2 mm shows that this species, like
several other ophiotrichids, goes through an Ophiopteron-stage with the arm spines
webbed.
470 AILSA M. CLARK
Ophiothela danae Verrill
Ophiothela danae Verrill, 1869 : 391 ; Koehler, 1922 : 297-298, pi. 59, figs. 1-3, pi. 103, fig. i ;
Mortensen, i933a : 342 ; A. M. Clark in Clark & Rowe, 1971 : 116, pi. 14, fig. 5.
? Ophiothela dividua von Martens, 1879 : 127-130, figs. 1-4 ; Balinsky, 1957 : 22-
MATERIAL. NAD i5Q, 30°47'S : 3O°27'5'E (S of Durban), 36 metres ; 2 specimens.
JAN 27, Jangamo Reef, Mozambique (c. 24°S : 35°E) ; 5 specimens.
SYNONYMY. In 1933 Mortensen could only put forward the colour pattern of the
type material of Ophiothela dividua, usually with two dark blue bands across each
pair of radial shields, to try and distinguish between it and the similarly fissiparous
0. danae. In 1971 I noted that the two will probably prove to be synonymous.
Unfortunately only three of the present seven specimens have any colour left. One
from Jangamo Reef has traces of blue bands across the arms, each about one segment
wide and separated by three to five pale segments. Of the two specimens from
Natal, one has some large dull red patches, not bands, and the other has fine linear
blue-black markings running parallel to the edges of the radial shields (i.e. more
longitudinal than transverse) and defining the distal edges of the dorsal arm plates.
Koehler (1922, pi. 59, fig. 3) shows a specimen of 0. danae with slightly coarser dark
lines across the arms and irregularly on the disc with one pair of radial shields
marked with a transverse band near their distal ends but the others mostly with
longitudinal bands on or near the shields. Koehler gives the general colour as pinkish
grey 'with dark blue striae on the dorsal surface of the disc and on the arms'. Clearly
there is considerable variation in the precise pattern.
Nor is there a reliable zoogeographical distinction. Von Martens said that the
type material of Ophiothela dividua came allegedly (angeblich) from Algoa Bay, on
the south coast of South Africa. However, since the specimens were evidently
epizoic on the tropical Indo-Pacific gorgonian Melitaea ochracea, of which Natal
would be expected to be the southern limit, I suspect that they may have been
rather from Delagoa Bay at the southern end of Mozambique. Balinsky has re-
corded a single specimen from Inhaca, Delagoa Bay, as 0. dividua since its colour
pattern matched that of the types. Mortensen's specimens of 0. danae were from
the vicinity of Durban like the southernmost of the Survey's specimens.
Family OPHIOGOMIDAE
Ophiopsila seminuda A. M. Clark
Fig. 12
Ophiopsila seminuda A. M. Clark, 1952 : 218-219, fig- 3-
MATERIAL. FAL428C and 4346, 34°i5-2'S : i8°33-2'E (False Bay), 22 metres ;
ii specimens.
FAL 467?, 34°07'S : i8°42'E, depth ? ; i specimen.
SST 22Z, 35°o6'S : 22°i5'E (S of Mossel Bay), 120 metres : i specimen.
LIZ 19!!, 33°58-5'S : 25°42'E (Port Elizabeth), 27 metres ; i specimen.
SCD 2726, 34°23'S : 25°54'E (S of Port Elizabeth), 181 metres ; i specimen.
ECHINODERMS FROM SOUTHERN AFRICA
FIG. 12. Ophiopsila seminuda A. M. Clark, a-e. SCD 2Q4N. D.d. 9-10 mm. a. Dorsal
view of part of disc. b. Two jaws and the first two arm segments, c. Dorsal view of
fifth free arm segment, d and e. Dorsal and ventral views of thirty-fifth free arm
segment, f. FAL 4346. Isolated dental plate. The scale measures i mm for a-e and
0-3 mm for f.
SCD 2Q4N, 33°04'S : 27°57'E (near East London), 84 metres ; i specimen.
SCD 3260, 34°27'S : 25°57'E, 172 metres ; i specimen.
VARIATION. Although the holotype of Ophiopsila seminuda gave the appearance
of lacking disc scales ventrally, most of the better-preserved specimens since collected
have the under-side of the disc opaque with very fine scales embedded in the rather
pustular skin ; only near the genital slits is the scaling barely distinguishable.
In most specimens the radial shields taper proximally, or at least the disc scaling
encroaches more on to their surface than in the holotype. They are usually well
separated but in the smallest specimen, SST 22, d.d. only 1-8 mm, they approximate
distally.
The armament of the jaws also varies, with additional papillae developing in larger
specimens. At d.d. S-io mm there are usually four to seven tooth papillae at the
apex of each jaw superficial to the lowest tooth. The holotype at d.d. 4 mm has
only three papillae, the median third one smaller than the other two, which make
an almost symmetrical pair on most jaws. The smallest specimen now present,
d.d. <2 mm, lacks even the third papilla and looks very amphiurid-like. In addi-
tion to the two flat, rounded, distal oral papillae each side observed in the holotype,
472 AILSA M. CLARK
most of the larger specimens show a much smaller papilla (or scale) distally between
the enlarged outer papilla and the ventral arm plate.
The oral shields have a more or less well-developed distal lobe and may even be as
long as broad.
The dorsal arm plates broaden somewhat on the stoutest part of the arms beyond
the disc and become as broad as long, or broader by about the thirtieth segment,
transverse oval in shape.
The arm spines number up to 10 in a specimen with d.d. 8 mm and up to nine at
d.d. 7 mm. One with d.d. only 5-5 mm also has nine spines on a few segments,
though the smaller holotype has only seven. The spines are all spatulate in shape
with broad round tips. The lowest and longest is usually about one-and-a-half
times the segment length. The length of the long inner tentacle scale ranges from
equal to the segment length in smaller specimens up to twice as long in the largest
ones.
AFFINITIES. In spite of the discovery of specimens with more arm spines and
ventrally scaled discs, this species still differs from Ophiopsila aranea (known from
Europe to tropical West Africa) in the fineness of this scaling as well as in the fewer
tooth papillae at a comparable size, though clearly the two are fairly closely related.
JAWS. A drawing of a dissociated dental plate is given here since no illustra-
tions of the dental and oral plates of Ophiopsila were included by Murakami in his
survey of the ophiuroids (1963), though he described those of a new species of the
genus which he subsequently named Ophiopsila squamifera. Judging from his
description, the dental plates of 0. seminuda are very similar to those of his Japanese
species and similar to thos'e of Ophiocoma and Ophiomastix which he illustrated,
though rather smoother.
Ophiopsila bispinosa sp. nov.
Fig. 13
MATERIAL. NAD 35X, 2Q°35'S : 3i°38'E (near Tugela River mouth), 150 metres ;
i broken specimen ['Anton Bruun' st. 39oL].
NAD 466, 29°35'S : 3i°42'E, 138 metres ; i specimen, the holotype [St. 3908].
NAD 520, 29°29'S : 3i°45'E, 86 metres ; i specimen [St. 3916].
NAD 58A, 29°26'S : 3i°46'E, 77 metres ; i specimen [St. 39iF].
DESCRIPTION. The holotype is the least poorly preserved of all the specimens but
even it has the disc broken, probably regenerated. The original d.d. was probably
c. 9 mm. The arms are broken. The oral frame (from the distal edge of one oral
shield to the distal edge of the first ventral arm plate opposite) measures 3-3 mm in
diameter. [This compares with an oral frame diameter of 3-5 mm in the specimen
of Ophiopsila seminuda with d.d. c. 8 mm.] The disc is very poorly calcified, mainly
bare or with scales so minute and thin as to be transparent except around the radial
shields. The shields have length : distal breadth 1-5 : 0-33 mm.
ECHINODERMS FROM SOUTHERN AFRICA
473
FIG. 13. Ophiopsila bispinosa sp. nov. a-f. Holotype. Oral frame diameter 3-3 mm.
NAD 460. a. Dorsal view of part of disc and the base of one arm. b. Two jaws and the
first two arm segments, c and d. Dorsal and ventral views of the twelfth free arm
segment, e and f . Sixth and twenty-fifth ventral arm plates, g- j . Paratype. NAD 35X.
Oral frame diameter 3-5 mm. g-i. Dorsal, ventral and side views of arm at about the
twentieth free segment, j. A particularly broad lowermost spine. The scale measures
I mm.
The oral shields are longer than broad, spearhead-shaped with the proximal angle
rounded and the distal lobe narrower. The adorals are contiguous or closely
approximating interradially and have a distal lobe between the oral shield and the
first lateral arm plate each side. The tooth papillae at the apex number only three
to five, two of them forming a symmetrical amphiurid-like, though pointed, pair on
four out of the five jaws. The distal oral papillae number two in five positions and
three in the other five ; all of them are conspicuously spiniform, broad medially,
slightly flattened and tapering to a more or less blunted point. The innermost one
arises partly or completely from the oral plate but leaves a diastema between it and
the tooth papillae, revealing the long spiniform oral tentacle scale within the slit.
In addition, on the outer corner of the adorals is a small rounded supernumerary
papilla, inset slightly into the oral slit and partly abutting on the first ventral arm
plate.
32
474 AILSA M. CLARK
The basal dorsal arm plates are very ill-defined but may be regenerating ; they
are about as broad as long but by the twelfth free segment have become slightly
broader than long and deep fan-shaped with the two proximal sides almost straight;
they are widest proximal to the middle of their length and have a slight angle in the
middle of the distal curve. They are contiguous for less than half their breadth.
The proximal ventral arm plates are broader than long since each has a prolonged
lateral angle distal to the tentacle pore, though this is progressively reduced beyond
the basal segments. The middle of the distal edge is distinctly concave. By the
middle of the arm the ventral arm plates have become longer than broad and
consecutive ones are spaced from each other, exposing a membranous space between
the lateral arm plates medially.
The lateral arm plates bear eight spines for about the first 12 free segments, then
nine for two or three segments, then ten for about two segments before dropping to
nine again. The upper and middle spines are markedly flattened in approximately
the vertical plane, presenting their maximum area when looking along the length of
the arm, when standing erect. However, the more ventral ones are obliquely
flattened and the lowermost, when pointed downwards, is flattened laterally. The
lowest is the longest spine, up to 1-3 mm long, compared with a corresponding
segment length of 07 mm. The second from lowest is also relatively long, then come
four or five smaller spines, while the uppermost ones are longer again. The spines
are hollow centrally but the cavity is narrow, owing to the considerable flattening.
The inner tentacle scale is sword-like and may exceed 0-8 mm in length. As
preserved, it normally crosses its partner over the distal end of the ventral arm
plate. The outer tentacle scale is also flattened and spiniform but only about half
as long.
PARATYPES. The larger of the three paratypes (sample NAD 35) (fig. I3g-j) has
only a remnant of the disc attached to the oral frame, which measures 3-5 mm in
diameter. The distal oral papillae usually number three each side and all are
pointed and slightly flattened ; the outer one (or two) may be slightly broadened
medially but markedly tapering in the distal third. The arm spines appear less
markedly flattened than in the holotype, very slender and tapering ; the uppermost
are almost as markedly elongated as the lowermost.
The two other paratypes both have the oral frame 2-2 mm in diameter and are
remarkable in having six arms ; because of this the jaws are unusually narrow with
only two to four tooth papillae and the oral shields are even longer than in the
holotype. One has some quite large but extremely thin disc scales dorsally. Their
arm spines number up to eight.
AFFINITIES. Ophiopsila bispinosa resembles 0. seminuda in the relatively small
number of tooth papillae but carries the reduction of the disc scaling even further.
It is most easily distinguished by the spiniform distal oral papillae and the more
pointed arm spines, also the spiniform second tentacle scale. There may also be a
geographical distinction, the new species being found so far only in Natal whereas
0. seminuda has not been recorded from north-east of East London. The other
neighbouring species is Ophiopsila paucispina Koehler, from northern Mozambique,
ECHINODERMS FROM SOUTHERN AFRICA 475
which differs not only in the few arm spines (only four at d.d. n mm - actually five
according to Koehler in 1930) but also in the oral shields being broader than long,
the tooth papillae forming a cluster of four rows and the distal oral papillae being
rounded. Morphologically, the closest species of Ophiopsila is probably 0. timida
Koehler (1930) from the East Indies. This also appears to have tenuous disc scaling,
spiniform distal oral papillae (but only two in each series), the outer tentacle scale
also spiniform and a similar number of arm spines, eight at d.d. 5 mm. However,
in 0. timida the spines are cylindrical with rugose blunt tips, the oral shields are as
much as two-and-a-half times as long as broad and there is a large cluster of tooth
papillae with usually four superficial ones, even at a smaller size.
Remarkably reminiscent of Ophiopsila bispinosa, at least in the shape of the arm
spines and the bar-like separated radial shields, is Amphiura linearis Mortensen, also
from Natal, which is unusual for an Amphiura just because of the separated shields.
The types were in poor condition, even the larger with d.d. only 3-5 mm. However,
the disc is evidently covered with fine scales and the distal oral papillae and the
tentacle scales, though pointed, are single.
Family OPHIURIDAE
Ophiura trimeni Bell
Ophiura trimeni Bell, 1905!) : 257, pi. i, figs. 3, 4 ; H. L. Clark, 1923 : 360-361 ; Mortensen,
i933a : 384-385. fig- 84.
Gymnophiura novembris Hertz, 1927 : 72-73, pi. 6, figs. 9, 10 ; Mortensen, I933a : 393-394,
fig. 89.
MATERIAL. AFR 83oA, 35°n'S : i8°42'E (S of False Bay), 315 metres ; i poor
specimen.
WCD 76D, 34°i7'S : I7°53'E (W of Cape Point), 320 metres ; 5 specimens.
WCD 1050, 33°07'S : I7°33'E (W of Saldanha Bay), 183 metres ; 8 poorly pre-
served specimens.
TRA 51 A, West of Dassen Island (33°S : I7°E), 356 metres ; i specimen.
TRA 736, 32°o6'S : i6°37'E (W of Lambert's Bay), 310 metres ; 4^ poor speci-
mens.
LET 27L, 68C and 72F, off Lambert's Bay (32°S : i6-i7°E), 280-400 metres ;
i poor specimen and 7 small ones.
SYNONYMY. The type and only recorded locality of Gymnophiura novembris is
35°09'S : i8°32'E, on the Agulhas Bank very close to sample AFR 83oA, in 564
metres. The largest syntype has d.d. only 5 mm. Mortensen figured one of the
syntypes, showing the same unusual membranous areas bordering the first two dorsal
arm plates laterally that occur in specimens which have arm combs present and thus
run down to Ophiura trimeni. He pointed out the great similarity between the two.
In fact, the arm combs were overlooked by Bell in the type of 0. trimeni and they
are certainly difficult to see or absent in smaller specimens, observation not being
helped by the poor condition shown by most preserved individuals of this species.
H. L. Clark identified as 0. trimeni specimens from both south-east and north-west
476 AILSA M. CLARK
of the type locality of G. novembris. Clearly the two nominal species are sympatric
and I cannot see any way in which G. novembris can be distinguished. It is accord-
ingly now referred to the synonymy of 0. trimeni.
Amphiophiura trifolium Hertz
Amphiophiura trifolium Hertz, 1927 : 78-79, fig. 3, pi. 6, figs. 14, 15 ; H. L. Clark, 1939 : 108-109.
MATERIAL. ABD yC, 24°i2'S : 36°oi'E (SE of Inhambane, Mozambique),
1140 metres ; 4 specimens ['Anton Bruun' st. 36gJ].
ABD loB, 27°09'S : 34°09'E (E of Natal-Mozambique border), 1335 metres ;
3 specimens [St. 374D].
RANGE. These two records extend the known range of the species southwards
from off Mombasa and Somalia in East Africa ; it has also been recorded from the
Maldive area. The minimum depth known is also extended to 1140 metres, other
depths ranging from 1289 to 2727 metres.
Ophiomisidium pulchellum (Wyville Thomson)
Ophiomusium pulchellum Wyville Thomson, 1877 : 67-69, figs. 18, 19 ; Lyman, 1882 : 96-98,
pi. 3, figs. 1-3.
Ophiomisidium pulchellum Koehler, 1914 : 37 ; H. L. Clark, 1923 : 356-357.
MATERIAL. SST i2j, 35°22'S : 22°3i'E (S of Mossel Bay), 200 metres ; 6 speci-
mens.
The size ranges from 2-5 mm to 4-1 mm d.d. The smallest specimen has only
three pairs of tentacle pores and only the largest has appreciable knobs in the middle
of each primary disc plate. The holotype has d.d. 4-5 mm and is the largest recorded.
Family OPHIOLEUGIDAE
Ophiocirce inutilis Koehler
Fig. 14
Ophiocirce inutilis Koehler, 19045 ; 13-14, pi. 3, figs. 4, 5 ; 1922 : 436 ; H. L. Clark, 1939 : 131-
132.
MATERIAL. FED 2], 24°46'S : 35°2o'E (S of Inhambane, Mozambique) ; 132
metres ; 5 specimens ['Anton Bruun' st. 37iE].
RANGE. This record provides an extension of range southwards from the Zanzibar
area.
The specimens agree with Koehler 's description except that the better preserved
ones have only a small indentation in the disc opposite each arm.
Since Koehler's figures are very diagrammatic, fresh drawings are given here.
ECHINODERMS FROM SOUTHERN AFRICA
477
FIG. 14. Ophiocirce inutilis Koehler. FED 2j. D.d. 6-2 mm. a. Dorsal view of part of
disc and the base of one arm. b. Two jaws and the base of one arm with adjacent
interradii ; the fingerprint-like texture of the lateral arm plates indicated by dashes.
The scale measures i mm.
Ophiopallas paradoxa Koehler
Fig. 15
Ophiopallas paradoxa Koehler, 190^ : 12-13, pi- 3. ngs- I~3 '• *922 '• 436-437, pi. 79, figs, i, 2.
Ophiopallas paradoxa altera Hertz, 1927 : no, pi. 9, fig. 5.
FIG. 15. Ophiopallas paradoxa Koehler. 'Albatross' st. 5541, Philippine Is., 400 m.
U.S.N.M. no. 41347. D.d. 6 mm. a. Dorsal view of part of disc and the base of one
arm. b. Two jaws and the base of an arm with adjacent interradii (some of the granules
probably rubbed off), c. Side view of base of an arm. The scale measures i mm.
478 AILSA M. CLARK
MATERIAL. ABD 8R, 24°40'S : 35°28'E (S of Inhambane, Mozambique), 347
metres ; i specimen ['Anton Bruun' st. 3706].
RANGE. This record provides an extension of the range to Mozambique. Un-
fortunately the specimen is not in good condition and does not help to settle whether
Hertz' subspecies altera from off Dar-es-Salaam is valid or distinguished only by
immature characters. Her type has d.d. only 4 mm whereas in the holotype of
Ophiop alias paradoxa it is 8 mm. This may well account for the smaller number of
oral papillae and the relatively narrower dorsal arm plates.
The figure given here is of a specimen from the Philippines, named by Koehler,
which is in better condition than the present one.
INCERTAE SEDIS
Anamphiura valida H. L. Clark
Fig. 1 6
Anamphiura valida H. L. Clark, 1939 : 70-72, figs. 26A, 27.
MATERIAL. ABD 150, 29°42'S : 3i°38'E (near Tugela River mouth), 350 metres ;
2 specimens ['Anton Bruun' st. 3goE],
DESCRIPTION. This peculiar species is known only from H. L. Clark's description
of the holotype, d.d. 5 mm, and I think a supplementary account of the present
material is worth while.
The larger of the two specimens has d.d. 3-6 mm and arm length c. 9 mm. The
primary rosette is perfectly regular, as in the three paratypes. The radial shields
are slightly convex, joined for just over half their lengths but cut away in the distal
third with a slight concave adradial edge to make a notch above the base of the arm
in which the very narrow swollen first dorsal arm plate appears below. At least one
pair of radial shields is distinctly asymmetrical with one shield larger than the other.
The distal ends of the genital plates projecting beyond the radial shields each bear
two conical spinelets, much shorter than the scattered disc spinelets and resembling
an ophiurid arm comb. H. L. Clark's figure shows only one such spinelet on each
plate in the holotype. The jaw structure is variable with two to four small tapering
apical papillae, not one jaw showing the symmetrical pair of infradental papillae
characteristic of the family Amphiuridae, to which H. L. Clark referred Anamphiura.
The abradial flanges of the oral plates each side are more nearly superficial than in
most amphiurids with the free proximal end of the oral tentacle scale level with the
apical papilla on one jaw and only slightly inset on the others. The distal papillae
number only two, the outer one wider, and are quite separate. H. L. Clark found
three in the holotype and his figure shows them incorrectly as a single tri-lobed unit ;
though clustered together they are not fused. The teeth are rounded or slightly
squared off but relatively narrower than is usual in amphiurids. The ventral arm
plates are quite unlike the usual amphiurid shape, having a prolonged lateral angle
each side distal to the tentacle pore meeting the lateral plate so as to encircle the
ECHINODERMS FROM SOUTHERN AFRICA 479
b
FIG. 16. Anamphiura valida H. L. Clark. ABD I5D. D.d. 3-6 mm. a. Dorsal view of
part of disc and the base of one arm. b. One jaw, the distal oral papillae missing on the
left side. c. Ventral view of third arm segment. The scale measures i mm.
pore. The alignment of the two tentacle scales found on the first five or six segments
is also unlike that of any amphiurid, the smaller scale arising on the edge of the ventral
arm plate from about midway along its length, not at the proximal end, while the
larger rounded scale is hinged almost opposite to it on the lateral plate. The three
upper spines of the five on the first and second arm segments are markedly
flattened and expanded ; those of the first segment at least make a continuous flange
on the adradial side of the genital slit. The entire surface has a distinctly granular
texture.
AFFINITIES. The relatively huge primary rosette, presence of arm combs of a
kind, apical jaw structure and form of the ventral arm plates and tentacle pores, all
seem to me very unamphiurid-like but show affinities in some ways with the Ophiuri-
dae, in others with the Ophiactidae. However, the notch between the distal ends
of the radial shields, the spines on the genital plates distal to the radial shields, the
flattened arm spines opposite the genital slits, the irregular papillae at the apex of
the jaw, the shape of the ventral arm plates with latero-distal angles meeting the
lateral plates distal to the pore and the position of the tentacle scale on the ventral
plate (relatively short through the scale is) all agree with the anomalous genus
Amphilimna, recently referred from the Amphiuridae to the Ophiacanthidae by
Thomas (1967). Hopefully, the main 'Anton Bruun' collections include further
material of this sample for anatomical study to help decide the best systematic
position for this genus.
RANGE. This record extends the known range of the species south from the
Zanzibar area.
Class ECHINOIDEA
Family ECHINIDAE
Dermechinus horridus (A. Agassiz)
Echinus horridus A. Agassiz, 1879 : 203 ; Doderlein, 1906 : 220, pi. 28, figs, i, 2, pi. 35, figs. 2, 3,
pi. 47, figs. 10, ii.
Dermechinus horridus : Mortensen, 1943 : 112-117, figs. 46-49, 5oa, pi. 19, figs. 6-10, pi. 20,
figS. 1-3, pi. 56, figS. 22, 23, 29-31.
480 AILSA M. CLARK
MATERIAL. WCD 2igF, 34°42-8'S : i8°o8-8'E (SW of Cape Point), depth ? ;
i specimen.
This individual is unusual in having the peristome larger than the apical system
by 4-5/4-0 mm. However, it has horizontal diameter only 15 mm, whereas Morten-
sen's smallest specimen measured has h.d. 39 mm. The test is coloured vermilion
and white, the very edges of the apical system opposite the interambulacrals are
densely red and also the median parts of the interambulacrals but the ambulacrals
and adjacent parts of the interambulacrals are white. All the spines are vermilion.
The pores and tube feet are very small. Pedicellariae appear rare but possibly
many have been lost. A globiferous one has two teeth each side and a tridentate
has spoon-shaped valves but constricted necks.
Family BRISSIDAE
Spatagobrissus mirabilis H. L. Clark
Spatagobrissus mirabilis H. L. Clark, 1923 : 402-404, pi. 23 ; Mortensen, 1951 : 492-494, figs.
258-260, pi. 28, figs. 10-12, pi. 60, fig. 8.
MATERIAL. FB 1067, 34°O9'5'S : i8°24-25'E (False Bay), 26-5 metres ; i specimen.
Length : breadth is 28/23 mm = 1-22 : i. Height 15 mm.
The peripetalous fascicle is almost oval but well separated from the posterior end
by 6-7 mm. It is more like the holotype than the specimen figured by Mortensen,
which was rounder. The genital pores are undeveloped, as in the specimen of similar
size in the British Museum collection. The test is broken in the subanal area.
The holotype has test length 112 mm and length : breadth 1.18 : i.
APPENDICES
[Specimens from outside southern Africa]
I. VEMA SEA-MOUNT (see Berrisford, 1969)
This is in approximately 32°S : 8°E, c. 450 miles W of Lambert's Bay ; 54-61 metres.
Comanthus wahlbergi (J. Miiller)
VEM 3F, 6A, loA and I9W ; 19 specimens.
Ophionereis porrecta Lyman
VEM I5"V ; i specimen. The middle arm spines are unusually enlarged at about a
third of the length from the disc.
Ophionereis sp. indet.
VEM I5T ; i specimen, small and in poor condition. This might be 0. dubia but
the probability of both species having reached the sea-mount from Africa is small.
Ophiarachnella capensis Bell
VEM 3E, I5U ; 3 specimens. These have d.d. 19, 17 and 4 mm whereas in the
largest mainland specimen recorded (H. L. Clark, 1923) it is 15 mm and the largest
Survey specimen has d.d. 13 mm. The two largest Vema specimens have up to
seven arm spines, equal in length to about half the segment length, except the longer
ECHINODERMS FROM SOUTHERN AFRICA 481
lowest one ; the middle ones are less conical than in mainland specimens. Also the
supplementary oral shields are relatively larger, one or two of them almost equal
in area to the main shield, the two together forming a continuous oval in the largest
specimen, which also shows a pair of pores between the first and second ventral arm
plates not seen in the 17 mm specimen. I think that these small differences are
accountable to the greater size ; certainly the trend towards relatively larger sup-
plementary oral shields is shown in mainland specimens with d.d. c. 13 mm.
II. WALTER'S SHOAL
This is in approximately 33°S : 44°E, S of Madagascar and c. 1000 miles E of East
London ; 38-46 metres.
Comanthm wahlbergi tenuibrachia A. M. Clark
WSS lU ; 7 specimens. ['Anton Bruun' st. 38iA-C. See A. M. Clark, 1972 : 78-81.]
Ophiocoma sp. aff. 0. pica Miiller & Troschel
WSS iV ; i specimen.
III. MADAGASCAR
Amphioplus (Lymanella} hastatus (Ljungman)
MDD iB, 23°i9'S : 43°36'E (near Tulear), 82 metres ; i complete and 2 discless
specimens ['Anton Bruun' st. 363 W].
Amphioplus sp. C
MDD 2C, 23°2o'S : 43°36'E (near Tulear), depth ? ; i complete specimen, i with disc
detached and i with disc missing.
D.d. of the intact specimen is 5-0 mm. It is covered with fine smooth scales except
that the uppermost ventral scales have their edges prolonged into spinose projec-
tions, as in Amphioplus (Lymanella) furcatus and Amphiodia (Amphispina) microplax.
In addition, there is a triple spinose projection distal to each radial shield as in
Amphioplus pectinatus, to which it has a considerable resemblance. The radial
shields are narrow, c. i-o mm long and are contiguous for about the distal third,
separated proximally by a wedge of scales. The primary plates are just distinguish-
able, though widely spaced among the disc scales.
The oral shields are elongated rhombic, widest in the middle, with the proximal
and distal lobes both broadly rounded. The adoral shields meet interradially. The
oral plates are rather short. The four rounded oral papillae are distinctly arranged
in a straight line, as in the subgenus Lymanella, rather than the concave row of most
species of Amphioplus sensu stricto. There is a very small gap between the infradental
and second papillae, through which the small oral tentacle scale can be seen. Also
the third and fourth oral papillae tend to gape around the second oral tentacle. The
fourth papilla is based at least partly on the first ventral arm plate. The oral
formula is m,mnji nji + 1.
The dorsal arm plates are fan-shaped, broadest distally, the rounded distal edge
sometimes with a slight median angle ; they are mostly not quite contiguous. The
proximal ventral arm plates are almost square but the shape becomes more pentagonal
482 AILSA M. CLARK
with the proximal lobe prolonged. The sides are almost parallel, slightly excavated
for the pores and the distal edge is straight or slightly concave. The proximal lateral
arm plates carry four slender pointed spines, giving way distally to three. There are
two relatively small tentacle scales, the inner one extending less than half the length
of the side edge of the ventral arm plate.
The detached disc is similar to that of the intact specimen, with the radial shields
distally contiguous and spinose processes present along the disc margin and outside
the radial shields. The oral frames show minor differences in the shapes of the oral
shields but again the alignment of the oral papillae in a straight line is noticeable.
The combination of these two characters probably justifies the distinction of a new
species. Since it is likely that there is more abundant and better preserved material
with the balance of the 'Anton Bruun' collections, I do not think it desirable to name
it in the present context.
IV. SOUTH-WEST INDIAN OCEAN SUBMARINE RIDGE
A single dredging station on about the highest point of the ridge at 36°48'S : 52°o8'E
(about mid-way between Port Elizabeth and Amsterdam Island, or north of the
Crozet Islands), 400 metres.
Crotalometra magnicirra (Bell) 5 specimens
Glyptometra sclateri (Bell) 4 specimens
Juvenile indeterminable Antedonidae 3 specimens
Democrinus chuni (Doderlein) i specimen
Porphyrocrinus polyarthra A. M. Clark. 2 specimens
[See A. M. Clark, 1973.]
Odinia clarki Koehler 5 broken specimens
Otherwise known from the Maldive area.
Ophiomyxa vivipara capensis Mortensen 4 specimens
Ophiomitrella corynephora H. L. Clark 3 specimens
Ophiomitrella sp. i specimen
Like the three 0. corynephora, this was clinging to a gorgonian. It differs from them
in the much shorter disc armament, the stumps barely longer than broad. The arm
spines do not exceed the segment in length, as in 0. hamata, from which it differs in
the spines not being distinctly rugose and again probably in the shorter disc spines.
Unfortunately the size is small, d.d. only 4-5 mm.
Juvenile indeterminable Ophiacanthidae 3 specimens
Ophiomusium sp. 5 specimens
All these specific identifications provide extensions of range to this remote part of
the southern Indian Ocean, except for the stalked crinoid Porphyrocrinus polyarthra
of which this is the first record.
The fact that five of the six other species are common to South Africa, having
spanned the enormous gulf between, is of particular interest, though of course larval
transport from South Africa must be facilitated by the eastwards flowing current.
ECHINODERMS FROM SOUTHERN AFRICA 483
SUMMARY OF TAXONOMIC CHANGES
NEW SPECIES
Amphilimna cribriformis (Ophiuroidea) Family Ophiacanthidae
Ophionephthys lowelli (Ophiuroidea) Family Amphiuridae
Ophiopsila bispinosa (Ophiuroidea) Family Ophiocomidae
OTHER TAXONOMIC CHANGES
Astropecten phragmorus Fisher - reduced to a subspecies of A. poly acanthus Muller &
Troschel.
Ceramaster chondriscus H. L. Clark, 1923 - referred to synonymy of C. patagonicus
euryplax H. L. Clark, 1923.
Hacelia superba var. capensis Mortensen - raised to specific rank.
Asterina dyscrita H. L. Clark - referred to Patiriella.
Echinaster reticulatus H. L. Clark - referred to Henricia.
Perissasterias obtusispina H. L. Clark, 1926 - probably a synonym of P. polyacantha
H. L. Clark, 1923.
Amphiura adjecta and A. compressa Mortensen, 1933 - referred to synonymy of A.
capensis Ljungman, 1867.
Amphiura kalki Balinsky, 1957 - referred to synonymy of A. Candida Ljungman,
1867.
Amphioplus (Lymanella) furcatus Mortensen - restored from the synonymy of A.
hastatus.
Ophiothrix triglochis Muller & Troschel - reduced to a form of 0. fragilis
(Abildgaard).
Ophiopsammium nudum H. L. Clark, 1923 - referred to Ophiothela and with synonym
Ophiothela (formerly Ophioteresis) beauforti (Engel, 1949).
Gymnophiura novembris Hertz, 1927 - referred to synonymy of Ophiura trimeni
Bell, 1905.
REFERENCES
AGASSIZ, A. 1879. Preliminary report on the Echini of the exploring expedition of H.M.S.
'Challenger'. Proc. Am. Acad. Arts Sci. 14 : 190-262.
BALINSKY, J. B. 1957. The Ophiuroidea of Inhaca Island. Ann. Natal Mus. 14 : 1-33,
7 fig5-. 4 Pls-
- 1969. See : Macnae & Kalk.
BELL, F. J. 1904. The Echinoderma found off the coast of South Africa, i. Echinoidea.
Marine Invest. S. Afr. 3 : 167-175.
i9O5a. The Echinoderma found off the coast of South Africa. 2. Asteroidea. Marine
Invest. S. Afr. 3 : 241-253.
- 19055. The Echinoderma found off the coast of South Africa. 3. Ophiuroidea. Marine
Invest. S. Afr. 3 : 255-260, pi. I.
- igo5c. The Echinoderma found off the coast of South Africa. 4. Crinoidea. Marine
Invest. S. Afr. 4 : 139-142, pis. 2-4.
484 AILSA M. CLARK
BELL, F. J. 1909. Report on the Echinoderms (other than Holothurians) collected by Mr J. S.
Gardiner in the western parts of the Indian Ocean. Trans. Linn. Soc. Lond. (Zool.) 13 (i) :
17-23, pi. 3.
BERRISFORD, C. D. 1969. Biology and zoogeography of the Vema Seamount : a report on the
first biological collection made on the summit. Trans. R. Soc. S. Afr. 38 : 387-398,
i fig.
CHERBONNIER, G. 1962. Ophiures. R6s. scient. Exped. oceanogr. Beige Eaux C6t. Afr. Atlant.
Sud 3 (8) : 1-24, 7 pis.
1972. Amphioplus polymorphus n. sp., nouvelle espece d'Ophiure (Echinoderme) des
c6tes malgache. Bull. Mus. Hist, nat., Paris (Zool.) 24 : 285-290, i fig.
CLARK, A. H. 1911. The recent crinoids of the coasts of Africa. Proc. U.S. natn. Mus.
40 (1808) : 1-51.
1937- Crinoidea. Scient. Rep. John Murray Exped. 4 : 88-108, i pi.
1947- A monograph of the existing crinoids. i. The Comatulids. 4b. Superfamily
Mariametrida (concluded - the family Colobometridae) and Superfamily Tropiometrida
(except the families Thalassometridae and Charitometridae) . Bull. U.S. natn. Mus.
82: vii + 473. 43 pis.
1948. Two new starfishes and a new brittle-star from Florida and Alabama. Proc. biol.
Soc. Wash. 61 : 55-64, pi. 2.
1950. A monograph of the existing crinoids. i. 4C. Superfamily Tropiometrida (families
Thalassometridae and Charitometridae). Bull. U.S. natn. Mus. 82 : vii + 383, 32 pis.
1952. A new species of the crinoid genus Cyclometra from South Africa. Trans. R. Soc.
S. Afr. 33 : 189-192, I fig., pis. 15, 16.
CLARK, A. M. 1951. On some echinoderms in the British Museum (Natural History). Ann.
Mag. nat. Hist. (12) 4 : 1256-1268, 4 figs., pi. 22.
1952. Some echinoderms from South Africa. Trans. R. Soc. S. Afr. 33 : 193-221, 3 figs.,
pi. 17.
1967. Notes on the family Ophiotrichidae (Ophiuroidea) . Ann. Mag. nat. Hist. (13) 9 :
637-655, i fig., 2 pis.
1970. Notes on the family Amphiuridae (Ophiuroidea). Bull. Br. Mus. nat. Hist. (Zool.)
19 : 1-81, ii figs.
1972. Some crinoids from the Indian Ocean. Bull. Br. Mus. nat. Hist. (Zool.) 24 : 73-156,
17 figs.
1973- Some new taxa of recent stalked Crinoidea. Bull. Br. Mus. nat. Hist. (Zool.)
25 : 265-288, 6 figs., 2 pis.
CLARK, A. M. & ROWE, F. W. E. 1971. Monograph of Shallow-water I ndo-West Pacific Echino-
derms. London : vii + 238, 100 figs., 31 pis.
CLARK, H. L. 1921. The echinoderm fauna of Torres Strait : its composition and its origin.
Pap. Dep. mar. Biol. Carnegie Instn Wash. 10 : vii + 223, 38 pis.
1923. The echinoderm fauna of South Africa. Ann. S. Afr. Mus. 13 : 221-435, 4 figs.,
pis. 8-13.
1924. Echinoderms from the South African Fisheries and Marine Biological Survey.
1. Sea-urchins (Echinoidea) . Rep. Fish. mar. biol. Surv. Un. S. Afr. No. 4 : Spec. Rep. No. i :
1-33, 4 P^
1926. Echinoderms from the South African Fisheries and Marine Biological Survey.
2. Sea-stars ( Asteroidea) . Rep. Fish. mar. biol. Surv. Un. S. Afr. No. 4 : Spec. Rep. No. 7 :
1-34, 7 Pis-
1939. Ophiuroidea. Scient. Rep. John Murray Exped. 4 : 29-136, 62 figs.
DARTNALL, A. J. 1971. Australian sea-stars of the genus Patiriella (Asteroidea, Asterinidae) .
Proc. Linn. Soc. N.S.W. 96 : 39-49, i fig., 2 pis.
DAY, J. H., FIELD, J. G. & PENRITH, M. J. 1970. The benthic fauna and fishes of False Bay,
South Africa. Trans. R. Soc. S. Afr. 39 : 1-108.
DODERLEIN, L. 1906. Die Echinoiden der deutschen Tief see-Expedition. Wiss. Ergebn. dt.
Tiefsee-Exped. 'Valdivia', 5 : 61-290, pis. 9-50.
ECHINODERMS FROM SOUTHERN AFRICA 485
DODERLEIN, L. 1910. Asteroidea, Ophiuroidea, Echinoidea. In Schultze, L. Forschungsreise im
westlichen und zentralen Siidafrika. 4 (i) . Denkschr. med-naturw. Ges.Jena, 16 : 245-258, 2 pis.
1911. Uber japanische und andere Euryalae. Abh. bayer. Akad. Wiss. 2. Suppl. 5 :
1-123, 52 figs., 9 pis.
1917. Die Asteriden der Siboga-Expedition. i. Die Gattung Astropecten. Siboga Exped.
46a : 1-190, 20 figs., 17 pis.
DOWNEY, M. E. 1971. Two new species of the genus Tamaria (Echinodermata : Asteroidea)
from the tropical western Atlantic. Proc. biol. Soc. Wash. 84 : 43-50, 3 figs.
ENGEL, H. 1949. Ophioteresis beauforti nov. spec. Bijdr. Dierk. 28 : 140-143, 2 figs.
FELL, H. B. 1962. A revision of the major genera of amphiurid Ophiuroidea. Trans, R.
Soc. N.Z. (Zool.) 2 : 1-26, i pi.
FISHER, W. K. 1913. Four new genera and fifty-eight new species of starfishes from the
Philippine Islands, Celebes and the Moluccas. Proc. U.S. natn. Mus. 43 : 599-648.
1919. Starfishes of the Philippine Seas and adjacent waters. Bull. U.S. natn. Mus.
100(3) 11-546, 156 pis.
1928. Asteroidea of the North Pacific and adjacent waters. 2. Forcipulata (part). Bull.
U.S. natn. Mus. 76 : 1-245, 81 pis.
1940. Asteroidea. Discovery Rep. 20 : 69-306, 23 pis.
GISLEN, T. 1938. Crinoids of South Africa. K. svenska VetenskAkad. Handl. (3) 17 (2) : 1-22,
26 figs., 2 pis.
GRAY, J. E. 1840. A synopsis of the genera and species of the class Hypostoma (Asterias,
Linn.). Ann. Mag. nat. Hist. 6 : 175-184 ; 275-290.
1847. Descriptions of some new genera and species of Asteriadae. Proc. zool. Soc. Lond.
1847 : 72-83.
GUILLE, A. 1964. Contribution a l'6tude de la systematique et de Ideologic d'Ophiothrix
quinquemaculata d. Ch. Vie Milieu, 15 : 243-308, 13 figs., 6 pis.
HERTZ, M. 1927. Die Ophiuroiden der deutschen Tief see-Expedition. Wiss. Ergebn. dt.
Tiefsee-Exped. 'Valdivia' , 22 (3) : 59-122, 7 figs., pis. 6-9.
JOHN, D. D. 1948. Notes on asteroids in the British Museum (Natural History). I. The
species of Astropecten. Novit. zool. 42 (3) : 485-508, 4 pis.
KOEHLER, R. 1898. Echinodermes recueillis par ITnvestigator dans 1'Ocean Indien. 2. Les
Ophiures littorales. Bull, sclent. Fr. Belg. 31 : 54-124, pis. 2-4.
Ophiures nouvelles ou peu connues. Mim. Soc. zool. Fr. 17 : 54-119, 98 figs.
Ophiures de 1'Expedition du Siboga. i. Ophiures de mer profonde. Siboga
Exped. 45a : 1-167, 3^ pis.
1905. Ophiures de 1'Expedition du Siboga. 2. Ophiures littorales. Siboga Exped.
45b : 1-142, 18 pis.
1907. Revision de la collection des Ophiures du Museum d'Histoire naturelle de Paris.
Bull, scient. Fr. Belg. 41 : 279-351, 5 pis.
1908. Asteries, Ophiures et Echinides de l'Exp6dition antarctique nationale Ecossaise.
Trans. R. Soc. Edinb. 46 (3) 22 : 529-649, 16 pis.
1914. A contribution to the study of Ophiurans of the United States National Museum.
Bull. U.S. natn. Mus. 84 : 1-173, 18 pis.
- 1922. Ophiurans of the Philippine seas. Bull. U.S. natn. Mus. 100 (5) : x + 486, 103 pis.
1924. Les Echinodermes des Mers d' Europe, i. Paris : xiii + 36o, 9 pis.
1927. Revision de quelques ophiures de Ljungman appartenant au Musee d'Histoire
Naturelle de Stockholm. Ark. Zool. 19A (2) : 1-29, 4 pis.
LAMARCK, J. B. P. A. de. 1816. Histoire naturelle des Animaux sans Vertebres. Ed. I.
Paris 2 : 522-568 ; 3:1-59; 60-76.
LJUNGMAN, A. 1867. Ophiuroidea viventia hue usque cognita enumerat. Ofvers. K. Vetensk-
Akad. Fork. Stockh. 23 : 303-336.
1871. Forteckning ofver uti Vestindien af Dr. A. Goes samt under Korvetten Josefinas
Expedition : Atlantiska Oceanen samlade Ophiurider. Ofvers. K. VetenskAkad. Fork.
Stockh. 28 : 615-658.
486 AILSA M. CLARK
LYMAN, T. 1869. Preliminary report on the Ophiuridae and Astrophytidae dredged in deep
water between Cuba and the Florida Reef by L. F. de Pourtales. Bull. Mus. comp. Zool.
Harv. 1 : 309-354.
- 1878. Ophiuridae and Astrophytidae of the 'Challenger' Expedition, i. Bull. Mus. comp.
Zool. Harv. 5 : 65-168, 10 pis.
1879. Ophiuridae and Astrophytidae of the 'Challenger' Expedition. 2. Butt. Mus. comp.
Zool. Harv. 6 : 17-83, 8 pis.
1882. Ophiuroidea. Rep. scient. Results Voy. Challenger, Zool. 5 : 1-386, 46 pis.
MACNAE, W. & KALK, M. 1962. The fauna and flora of sand flats at Inhaca Island, Mofambi-
que. /. Anim. Ecol. 31 : 93-128, 5 figs.
1969. A Natural History of Inhaca Island, Mozambique. Revised ed. Johannesburg :
iii + i63, figs., ii pis. [Echinoderms by Balinsky, B. I. pp. 96-107.]
MADSEN, F. J. 1971. West African ophiuroids. Atlantide Rep. No. n : 151-243, 49 figs.
MARKTANNER-TURNERETSCHER, G. 1887. Beschreibung neuer Ophiuriden und Bemerkungen
zu bekannten. Annln naturh. Mus. Wien, 2 : 291-316, pis. 12, 13.
MARTENS, E. VON. 1879. Vorzeigung eines eigenthiimlichen sechsarmigen Schlangensterns,
Ophiothela dividua, sp. n. Sber. Ges. naturf. Freunde Berlin, 1879 : 127-130.
MORTENSEN, T. 1925. On some echinoderms from South Africa. Ann. Mag. nat. Hist.
(9) 16 : 146-154, 5 figs., i pi.
1926. Cambridge Expedition to the Suez Canal, 1924. 6. Report on the Echinoderms.
Trans, zool. Soc. Lond. 22 : 117-131, figs. 11-13.
I933a. Echinoderms of South Africa (Asteroidea and Ophiuroidea). Vidensk. Meddr
dansk naturh. For en. 93 : 215-400, 91 figs., 12 pis.
JQSSb. The echinoderms of St Helena (other than crinoids). Vidensk. Meddr dansk
naturh. Foren. 93 : 401-472, 29 figs., 3 pis.
1940. Echinoderms from the Iranian Gulf. Dan. scient. Invest. Iran. Part 2 : 55-112,
24 figs., 2 pis.
1943- A Monograph of the Echinoidea. III. 3. Camarodonta II. Copenhagen: 1-446,
215 figs., 66 pis.
1951. A Monograph of the 'Echinoidea. V. 2. Spatangoida II. Copenhagen : 1-593, 286
figs., 64 pis.
MULLER, J. & TROSCHEL, F. H. 1842. System der Asteriden. Braunschweig : xx+ 134,
12 pis.
MULLER, O. F. [with ABILDGAARD, P. C.]. 1789. Zoologica Danica. Ed. 3. Hauniae.
3 : iv + 7i, pis. 81-120.
MURAKAMI, S. 1963. The dental and oral plates of Ophiuroidea. Trans. R. Soc. N.Z. (Zool.)
4 : 1-48, i fig., 7 pis.
PERRIER, E. 1881. Description sommaire des especes nouvelles d'Asteries. Bull. Mus. comp.
Zool. Harv. 9 : 1-31.
RETZIUS, A. J. 1783. Anmaerkningar vid Asteriae Genus. K. svenska VetenskAkad. Handl.
4 : 234-244.
SIMPSON, J. C. & BROWN, R. N. R. 1910. Asteroidea of Portuguese East Africa collected by
Jas. C. Simpson. Proc. R. Phys. Soc. Edinb. 18 : 45-60.
SLADEN, W. P. 1889. Asteroidea. Rep. scient. Results Voy. Challenger, Zool. 30:1-935,
118 pis.
SMITH, G. A. 1927. On Asterina burtoni Gray. Ann. Mag. nat. Hist. (9) 19 : 641-645.
THOMAS, L. P. 1962. The shallow water amphiurid brittle stars (Echinodermata, Ophiuroidea)
of Florida. Bull. mar. Sci. 12 : 623-694, 23 figs.
1967. The systematic position of Amphilimna (Echinodermata : Ophiuroidea). Proc.
biol. Soc. Wash. 80 : 123-130, 9 figs.
THOMAS, L. P. & SCHOENER, A. 1972. Growth changes in Amphilimna olivacea (Lyman) and
the systematic status of Amphitarsus spinifer Schoener. Breviora No. 387 : 1-9, 2 figs.
THOMSON, C. WYVILLE. 1877. The Voyage of the 'Challenger'. The Atlantic. London.
1 : 1-424, 106 figs., pis. 1-14 ; 2 : 1-396, 62 figs., pis. 15-42.
ECHINODERMS FROM SOUTHERN AFRICA 487
VERRILL, A. E. 1869. On some new and imperfectly known echinoderms and corals. Proc.
Boston Soc. nat. Hist. 12 : 381-396.
1915. Report on the starfishes of the West Indies, Florida and Brazil. Bull. Labs nat.
Hist. St. Univ. la, 7 (i) : 1-232, 29 pis.
Miss A. M. CLARK
Department of Zoology
BRITISH MUSEUM (NATURAL HISTORY)
CROMWELL ROAD
LONDON, SWy
PLATE i
FIG. i. Astropecten granulatus natalensis John. Lectotype. B.M. reg. no. 1904.4.20.120.
FIG. 2. A . polyacanthus phragmorus Fisher. 1904.4.26.9. R 70 mm. [The locality of both
specimens is assumed to be Natal.]
Bull. Br. Mus. not. Hist. (Zool.) 26, 6
PLATE i
PLATE 2
FIG. i. Astropecten polyacanthus phragmorus Fisher. 1904.4.26.9. Enlargement ( x 5) of
proximal part of arm viewed obliquely, showing continuous series of superomarginal spines.
FIG. 2. A. polyacanthus polyacanthus Miiller & Troschel. B.M. reg. no. 1972.8.22.32, Assab,
Eritrea. R 105 mm. Similar view (x 2) showing the proximal gap in the spine series. (This
specimen has an unusually large number of supernumerary superomarginal spines.) [Photo-
graphed by Patricia Cunningham, the remainder by Mr P. A. Richens of the British Museum
Photographic Studio.]
Bull. BY. Mus. nat. Hist. (Zool.) 26, 6
PLATE 2
I
PLATE 3
FIGS, i and 2. Astrodadus euryale (Retzius). SCD 155?. Ventral and dorsal views.
Bull. Br. Mus. nat. Hist. (Zool.) 26, 6
PLATE 3
aft*. ^ \<*
A LIST OF SUPPLEMENTS
TO THE ZOOLOGICAL SERIES
OF THE BULLETIN OF
THE BRITISH MUSEUM (NATURAL HISTORY)
1. KAY, E. ALISON. Marine Molluscs in the Cuming Collection British Museum
(Natural History) described by William Harper Pease. Pp. 96 ; 14 Plates.
1965. (Out of Print.) £3.75.
2. WHITEHEAD, P. J. P. The Clupeoid Fishes described by Lacepede, Cuvier and
Valenciennes. Pp. 180 ; n Plates, 15 Text-figures. 1967. £4.
3. TAYLOR, J. D., KENNEDY, W. J. & HALL, A. The Shell Structure and
Mineralogy of the Bivalvia. Introduction. Nuculacea-Trigonacea. Pp. 125 ;
29 Plates, 77 Text-figures. 1969. £4.50.
4. HAYNES, J. R. Cardigan Bay recent Foraminifera (Cruises of the R.V. Antur)
1962-1964. Pp. 245 ; 33 Plates, 47 Text-figures. 1973. £10.80.
5. WHITEHEAD, P. J. P. The Clupeoid Fishes of the Guianas. Pp. 227 ; 72
Text-figures. 1973. £9.70.
6. GREENWOOD, P. H. The Cichlid Fishes of Lake Victoria, East Africa : the
Biology and Evolution of a Species Flock. Pp. 134 ; i Plate, 77 Text-figures.
1974- £3.75.
Printed in Great Britain by John Wright and Sons Ltd. at The Stonebridge Press, Bristol BS4 5NU