"•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!) 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 £ (O, °C 4-> rt 1/1 1T3 V) A O> > CO 3 A REVISION OF THE LARGE AFRICAN BARBUS 67 '•' '.*'.-•' :> £ (50 a c i I M 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" > looooooooooooo OOOOOOOOOOOOO R-r £ooroMo»t^ooo^"-it^M'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 ,. — 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 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 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