Se ee Bbtetaduiie i r Dd etal LLY rel Vue ts ds Seed ten gs Senne tue db, jin aedsaetuaies nargieccerotonces benP bre edhe Suh dl ta nade Ga Mae-TEH Fas Nae 70 ihetas2 Pa tea fo faded Peeters? f Boretuds tenses were nv ity ef i MOS WP eens gonmavissesetesmraretéy bite Ein ate Het pera Sev eNsnd? AN GSaR AERA S Mm ha dV INS SLING Maes ety ect Smee “ Fate zs - . 4 ~ . 2 “ a : ete NTL SEES et hs ie henihietebatutecamenerecantate > , 2 = : oe a 9 dea RA Mae cboN ey eruheeen, nate Rie Gd tram yet nD fade fe Os hes Wy Aad amt ate hs Kee Lape, hc lia is Ret oe biatid bas fon Ba Tete Bulg he Bh stk = “ ss 3 = 5. . - , ¢ recuaien a bir erties ent Seu ha owton tg! Tate on ” ¢ res pase ‘ DINE MNCL a oO MN A ar Sons thabstw ce away be om shee me ay eens DeAnda aN ANA > - sma ban ns careny kohane colt ere nie ree a WAAL Te ee ere Se Wanee Rta aes a Bern Sc ieibaae? Parents Cepeda Bad ha Pua Yore Mn alts Bi IERLS wee eda eae one ar a 7 ne) ¢ la) rt - nfl a 2 ie et * ‘ay iH ist j a : a al ‘ + ‘ . - Adi sy i yy Ay el iy a vi ei jel vi nay ny y PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES February 15, 1977 Series 4, Volume 41, Number 1, 123 Pages, 45 figures THE OSTEOLOGY, CLASSIFICATION, AND RELATIONSHIPS OF THE EEL FAMILY OPHICHTHIDAE eo John E. McCosker Steinhart Aquarium, California Academy of Sciences San Francisco, California 94118 CALIFORNIA ACADEMY OF SCIENCES/GOLDEN GATE PARK/SAN FRANCISCO February 15, 1977 Series 4, Volume 41, Number 1, 123 Pages, 45 figures THE OSTEOLOGY, CLASSIFICATION, AND RELATIONSHIPS OF THE EEL FAMILY OPHICHTHIDAE By John E. McCosker Steinhart Aquarium, California Academy of Sciences San Francisco, California 94118 ABSTRACT. A classification of the genera of the apodal family Ophichthidae is proposed on the basis of internal and external morphology, with particular emphasis on osteological characters, Specimens of 89 ophichthid species from 44 genera and comparative material from ten other apodal families were prepared for osteological examination, usually by a trypsin-based staining and clearing technique, and critically compared. Forty-nine ophichthid genera are recognized and are distributed among six tribes in two subfamilies. Diagnostic characters for the recognition of genera include the shape and condition of elements of the gill arch and hyoid arch, number and placement of branchiostegal rays along the hyoid, sus- pensorium elements, neurocrania, dentition, pectoral girdle elements, cephalic pore patterns, lateral line ossification, fin placement, and morphometric characters. The Ophichthidae are defined on the basis of their numerous overlapping branchiostegals, supraorbital canals united by a transverse commissure through the fused frontals, first and second epibranchial interconnections, absence of a palatine, and the separation of the pterygoid from the vomer. A monophyletic origin of the family from a congrid-like ancestor is proposed. An evolu- tionary history of the Ophichthidae is suggested, in which the subfamily Myrophinae has separated into two tribes and the subfamily Ophichthinae has radiated into four tribes. The validity of the family name Ophichthidae is discussed. The family names Ophisuridae, Myridae, Myrophidae, Muraenichthyidae, Echelidae, Neenchelidae, Aoteidae, Acanthen- chelyidae and Sphagebranchidae are synonyms of the name Ophichthidae. A comparison is made between an ophichthid classification based primarily on osteology and the previous classification, based primarily on external morphology. The results of two computer-pro- grammed classification schemes of species relationships within a single tribe are compared with a classification developed using traditional methodology. Alternate hypotheses are pro- posed to explain the log-normal inverse relationship between genera and the distribution of species among genera in the Ophichthidae. CALIFORNIA ACADEMY OF SCIENCES/GOLDEN GATE PARK/SAN FRANCISCO PROCEEDINGS CALIFORNIA ACADEMY OF SCIENCES TABLE OF CONTENTS Page LOST? OOTP [PUGS ae a a Ee oes ee 5 LUST OE WBA Se an ee a ee ee Se ee eee ee 6 PNGIIN ON IEEID GNA EINGTIS seat oe eee a ere ee ee A ees ee ee 7 DONTUTRCC ET SATION i Sek ni on a Ro gn oe 9 CGT rel etcetera re Ne ON RR a RPS 2 a ei mn eas Le et 9 ISKOnVRO te © Piicehttni clan Glass iit Gell inp eeaea eee eEE 10 Valli ditwaotastineratimilvadNalnn ees sons see ee ee Se Ses 10 @steologicaleStuclesmotmthie © point Glivtl nical esses eee ee ee 11 Fetal! SwAROMMNVinMNY Cn We (VOM AMM MCNS: csc ses pence aracnoeasaeiteleennassecnqeecereeenens 12 SVMOPSISMm Olen © oli Cine en Glas SI fill Galt © meee eae 13 VAT EAT at @) DB) S prereset eect ae eee cen kN dot AU cers. Seta we ee Lee 2 eens Neuere ae ee Ree 14 AXON OMiLGe NIETO Seren 2: ae eee ee Seen ne te ee eee 14 ING) ON RERITE NTI TVS mae se ee ip a eg al ak ee ie mere ae ree Ew Be Terre ee eee nr do 14 INMare iia Summ sxcel I © Cl te sects cers re cor Sk nee nce Ran ae oS cee ne aE 115 Statist calle NA Cth © CS pe ccseeretas ee. elec nee cece Se ee oe ee ence a eee 16 OSTIEOLOGYAAN DOIN Gail @ INAS AINA GTI @ NAN eee ee ea 17 INQUITROXCTITUI I eI Ns A ep ae Ae A es eR, SRT WR EB Oe tare SP ae etek ew OD 18 SUSDEMSOnUin Mpa) Cala Stes eee eee ees oe eee ei ewe td eae ee tala eee ed eee 24 (OVOVSTRCUNLETE SYSU TRS eee a Ea en SP RM oe CO ee te AY 26 FLAVONIG!" AVC) OF IEE UI a a Ea REE ele Fa he ee a ne ee A Gee be ee ee ee 28 GrTTBRPAN Gin Steet acta nee cere ne eke Ese ee SE ee we eet ne nd Pes ee SK NE 32 Recto all Gite) Cie sere tek oom crete Bec Mee EN Ot ARG An ony ec ezcninrt se ceed ewes MAUR 33 TRAN pall Same SV-S1CG tae see nce eaten Noe sec Se eee OLDEN Ee EP RCE NN ied SO SO ROE ee ER 36 PSE) SIRS EXC CYS Bae ee tea a aS te Ee EN Re eh Ses eR erent 42 (@alliclalleeS Kel ete ri ee ne ee cee ce eee ee eae ee eo 45 NSCS TNT M CO NY ee a ee ee ee te eee 48 TAUOINIOUNAN LBS ater AEE RE re ONE rns = ed AER Oe 48 OsteolosicalsDetinitionmotmtiney. © fo lnt cht lnc ele een eee 49 AnalviticalakeyatonthenGemetanot © pln htin cl ele eee seen eee en eens 50 | EIUTOY Se GEA SH ST ae eS ee RES ONE en Oe eee Ne ee eer ne ee 56 Subfamilial and Tribal Diagnoses and Generic Descriptions ........................---------------- 57 S Ura fala Lame VAN @ PIMNIN a Cisse sete Ea cern cea eee cen ie aan See seed Sco Semen 57 Tribe Benthenchelyini (Genera are listed alphabetically within each tribe) -....... 57, Tithe cae WANT POINT lies sees ee eee ie ead ee Re Se ect eine Ree Ve ee 57 Sulfate @ pln clnthinaes<-cee.ccese ac cece ee oe cee en ers o ee Se cere ec ese ee 62 al ftai to eee @all Hehe bn rps ae ee ee ee ee ee ee DE Eo 62 inibemspiare bray Ghitiniee see ere te see tesa 2 ctor ee ie cael 2 eee 64 Miriam aS Gel ml Glinitih i Id gaa ea ae ee eke Fe OF We cee ee Nt eee eee 70 Wtloves CO fool pitied ni nin] Seen aes ene Bt ae eee eo LE ne ae ee Perea eee 73 GompanisonmwithmeneviOUSn Glass Calti@ mSy esses eee eee eye 85 ENWOILUNTIOUN (Ol TRIE COPRIICTRINRIIOYNE aoc ercnes ase ecneceenc ss Sesgee epee perpen Bop ee eS DO SEE EEE 85 Relat Oms tio m CO mOthne ray A tne Ulli rrr Siemens eee eae era nnn 85 EWOMUUNCO LH) AOMATUL MMM) tA COV TCT nyt CLS) ee 86 MMVIROFOLMIID ine! TEMG MEMEIMITDN — co sececccenne conecee eee seece ener ecm eee Rs aeneaerecerer pRoDAD HOSS EEEOSECEEEBEES 88 COWS RTSI LU MMUMLAN = ack eas es eer eee NO oe cer yt UC PCR er eee eee eee ete sks 89 Sita ge bral clini iif eaten ee eee ene ete noo ee eee ee tee ER ree 91 Beals earl GIG ip meee erecta meme nc tines ete eee Pi hy Chel Cle a cee eat Wee A he ad 94 Gallechyel\itriiiameteen erate eee hes Nae Deer: oe a Td oo ea ee Be ete 94 ZOOGEOGRAPHY AND COMMENTS ON OPHICHTHID SPECIATION ..............-.--...--------- 96 [LUPTTELRSVN TIRE (CTE Yo rey a AS ace aE ae Sa sad cn ER SR Ie a, ue eee oe, 101 AB KES eee ete ir, ener A RG Ser ee oe Ao mW ee a eee Har ek ee Sea 108 PROCEEDINGS CALIFORNIA ACADEMY OF SCIENCES Figure tN Eee EN NOWP WN RAOUOANDUARWH at Newoc) NONNNNON NN WN KN OOAON Da FRWH = CO HRHPA HR PWWWW WWW WwW Ww PWN RHFPOTOUPFN DU HLWNH = OC a on LIST OF FIGURES An Early Illustration of an Ophichthid, presumably Ophisurus serpens .................- NedrocraniUimmotm © polaiclatlnusweZ Ojo no Gliliga seen eee en NetnocraniUnn nots © a hniGiathtiSiez Op lO Ginliigesesae senses tenner Neuro Granite Bem thiernGin@ isin Gel itil irene eae eee ne en cE ES NG unOGraniUinmn iO fil VV{i@ Toil Sav cli ames ees En oo Neuro cranium One Meira eniliGithnySirGinilenmS 1S. esssssee teen ence en ee ee NeurocranUmmotm @alll@Gineliysiiimalrinp © nali Us apenas eee ane nee ee ee eee Neuro craniUinmoteaGiitinyaoUrsies eC] a\Gii@ [a Simsesseeere eee arenes eee NeURocAniuinn @ir SWiGiodminWS [OOMMIUS ccecee eer cece eee eee rsa eeeeeSeeeeee INEUKOcraniUMmOneBascaniGhth/Sm palmar C SiS esse tes tes scee eee eames ae NGUROGMIUIIN Cir IMAM AIS: SSVSTRUITWIS xcesseeccese cree ese c cence eeee coe cer eno a cea eeE Ree Re eem = HES aon nEnee @xoliths of Several’ @phichtinid Species EA ae Boe rin Se See Fleadmskeletommons © piGhitinUsiez OO © Cilia eeeeeea eee ee neon E Suspensonum) and jawsnot Ophichthus Zophochiin sss RightsPostonbital/ Series) of BrachySOmoOpinis) SAUMOPSIS, cess eee ee een Maxilla-VomensApposition: of Severall @yolniehnthyics wes cee eee Hyoid Arch and Branchiostegals of Ophichthus zophochir, an Ophichthine, ANG eel ea Gitay Sin Gln C1aSTSPaclan NAW/© [Ohi Cleese eee ec GillArcheskeletonmotm© piiiGitinustez.o in @ Cin aes ene ee Pectoral Girdle of Various Representative Ophichthines .................--------------0---00--- Pectoral Girdle of Various Representative Myrophines .............-..------:---------0000100+++= GephaliesRateralissSystembpanGeASSOGlatec | BOnes messes eee Lateral Line Ossicles of Representative Ophichthines .......................2.-.0--00ecceeeeeeeeeeeee Lateral Line Ossicles of Representative Myrophines .................-.-.--------------------------+ Cephalic Pore and Surface Sensory Papillae Development in Two Ophichthids .... Anteriormost Five Vertebrae of the Type Genera of the Tribes of Ophichthids .... Trunk and Caudal Vertebrae of Ophichhtus zophochir ...................-.--.-..2200--+-------+- Caudal Skeleton of Ophichthus zophochir ................-.......------- sate 5, SONIA ei et Oe GaicaleSkeletonmo te Vino phils aval Cymer Comparative Anatomy of Congrid and Ophichthid Digestive Tract and (GES. “BB V0 Kol eir Re ee ok A a, a as ae emetic a ek dill pe gee 41 hw Wn BO re aa Vomer, Maxillae, and Pterygoid of Ahlia egmontis and Myrophis vafer ................ Diagrammatic Representation of a Species with Well Developed Head Pores ...... Representation of Underside of Callechelyin Smouts .................22.--200000-222222000e--eee-e= Diagrammatic Representation of Posterior Trunk Vertebrae ................220.-00.---------- Diagrammatic Representation of Head and Pectoral Fins of Two Ophichthins ...... Diagrammatic Representation of an Ophichthin with a Fringed Upper Lip —.......... @©penculaneSenlesoteMiyropmismVvaret ss. = eee ee en Pe ee Proposed Evolutionary Relationships of Ophichthid Tribes —.................0222000.-...---- PROPOSE MEV. UETO MO itiatin Cal VAN TO [olin ele eee eee Nn RrOPOSEEEVO]UTO MMO Tate © knit Ghat nna eee ee ae ee sue en pS BLODOSEASEVOUTLONMOfatheMs Pliage ali Glniliiimeesemees seers eee BrODOSCGMEVOUTONMO fe tiveN Bas Gali Glitnir eres eee PROPOSCU MEV OUT ONMmOl mtn em alll Ec lne lly, iinet eee Phenogram of the Relationships of the Callechelyini, Using Program WVGM ...... Interrelationships of Species Groups of the Tribe Callechelyini, as Defined by PORN LEO © OI ES cee ne Sen salar Sean re Oe MN MeN RINE NTS ay eee eeaito Distribution of Species Among the Genera of the Ophichthidae, Gobiesocidae and _ Salariini 6 SERIES 4, V.41, #1 McCOSKER — EELS LIST OF TABLES Table Page PD eMtitOnmOimthem Gem enaeotin© plat cliche epseceee ee ease eee tenes eee snes 108 2 Number and Location of Branchiostegal Rays of the Species of the Ophichthidae .... 109 See GilleArcheGonelitio nny th yey © po ai ltl rea Crees eee etme ns enn seen 110 aL Gill “Aely Geommvaliittiormy flim: me IMAGO Son cere eee ee CeCe 111 5 Lateral Line and Cephalic Pore Conditions in Ophichthine Genera and Subgenera .. 112 6 Wereorll Coume Or Waltouws Wyolnichntimicl SlxXheS .sceesecs access cage ceccenennonccosenseseeseons Ms 7. ‘Gharacteristics of the Ophichthidae and Related Eel Families... 116 8 Morphological and Meristic Characters of the Species of the Callechelyini —........... 7, 9 Characteristics of the Species of Callechelyini Used in Programs REGROUP PATI) GR VIN Gt ANS ose ta steel NC 2 Sr ec SA Sao J we nyu ee Se, Seen ne 118 10. Wistribution“of Certain Ophichthid) Gemena: .....22. 2 cen2-05s2-cece ace oe cece 119 PROCEEDINGS CALIFORNIA ACADEMY OF SCIENCES ACKNOWLEDGMENTS The majority of this work is from my doctoral dissertation done at the Scripps Institution of Oceanography, University of California, San Diego, under the direction of Richard H. Rosen- blatt and Carl L. Hubbs. | sincerely thank Profes- sor Hubbs for his advice and guidance through the capricious nuances of zoological nomencla- ture, and Professor Rosenblatt for his invaluable advice, encouragement, and patience throughout the duration of my graduate studies. | wish to express my gratitude to the following individuals who have made various specimens available: Marie-Louise Bauchot, Paris Museum; Jacques Blache, Centre ORSTOM,; James E. Bohlke, Academy of Natural Sciences of Philadelphia; Peter H. J. Castle, Victoria University of Welling- ton, New Zealand; Lev Fishelson, Hebrew Uni- versity; John E. Fitch, California Department of Fish and Game; Warren C. Freihofer, then of Stanford University; Robert H. Gibbs, Jr. and Robert H. Kanazawa, National Museum of Natural History; William A. Gosline, then of University of Hawaii; Naercio A. Menezes, Universidade de Sao Paulo; Hans Nijssen, Zodlogisch Museum Amsterdam; John R. Paxton and Douglass F. Hoese, Australian Museum; John E. Randall, Bernice P. Bishop Museum; Tyson R. Roberts, Museum of Comparative Zoology, Harvard Uni- versity; C. Richard Robins, University of Miami Marine Laboratory; Margaret M. Smith, Rhodes University; Enrico Tortonese, Museo Civico di Storia Naturale, Genova; Boyd W. Walker and John Bleck, University of California, Los Angeles. | am particularly grateful to William N. Esch- meyer and the staff of the California Academy of Sciences for making the extensive and critical material from the George Vanderbilt Collections available to me. Thanks are also due to the following individ- uals: Ira Rubinoff and the staff of the Smithsonian Tropical Research Institute for assistance during my tenure as a Smithsonian pre-doctoral research fellow; Edward W. Fager and John H. Wormuth for assistance with computer programs; students and colleagues at Scripps Institution, and in par- ticular Joseph F. Copp and Donald M. Dockins, for aiding in numerous ways; Richard H. Rosen- blatt, Carl L. Hubbs, Robert R. Hessler, and Peter Paul Vaughn for their critical reading of my dis- sertation; and my wife, Sandra, for her help and encouragement. SERIES 4, V.41, #1 McCOSKER — EELS SERPENT MARIN. UMMA IFEOES 7570, i 4 VILL —oe wo MCU TET poe Ui eS: / ye tH LCC Gay, Sk < RAY 1 "ag, —\ IE eS DL uyetae 4 = gat nay sete Se SR Lhe. cAS-. ‘J S % ” Lo AS 5 2 % WR J, = P ; 4 SR 2 3 W pp ee , wee Zz E Wir rss ‘WS if? E oe, A > ince s os VAM \S= OB Ps 1 Fy Wi If gE 5 =.= Zé ORT) aN IX: = Aa Bs : 5 A z N ~ 2 ° as 7 Se ; \ ye £[F aN LE AG * Za w= BS Af, e Ffg= \\ = FX SS — ' {! 1, 4, serpens, An early illustration of an ophichthid eel, presumably Ophisurus Figure 1. from Mattioli’s Commentaires (1568). PROCEEDINGS CALIFORNIA ACADEMY OF SCIENCES THE OSTEOLOGY, CLASSIFICATION, AND RELATIONSHIPS OF THE EEL FAMILY OPHICHTHIDAE John E. McCosker INTRODUCTION General The Ophichthidae comprise a large family of mostly fossorial eels limited to continental shelf depths of all tropical and subtropical oceans. The Ophichthidae contains more than 200 species here distributed among 49 genera, representing perhaps the greatest diversity of anatomical spe- cializations within a single apodal family. The resemblance of many ophichthids to snakes has earned them the common name of ‘‘snake-eels,’” and along with the morays, have evoked many sea-serpent legends among tropical coastal peo- ples. Perhaps the earliest illustration identifiable as an ophichthid (fig. 1) was that of Mattioli (1568) p. 388, probably based on an adult Ophi- surus serpens. The first described ophichthid species, Muraena ophis, was the second apodal fish in Linnaeus’ Systema Naturae (1758). The taxonomic treatment of the Ophichthidae, and of apodal fishes in general, has been in con- stant flux since the eighteenth century, and only within the past two decades has a coherent con- cept of the Ophichthidae become realized. Pre- vious theories of the interrelationships of ophich- thid genera have been based on such trivial char- acters as fin position, dentition types and colora- tion. It is now generally accepted that the osteol- ogy more conservatively reflects the phylogeny of a group and is less affected than external morphology by minor evolutionary adaptations. The unsatisfactory state of the classification of the ophichthids has been recognized by numer- ous authors (Myers and Storey, 1939; Myers and Wade, 1941; Gosline, 1951a; Smith, 1964; Rosen- blatt and McCosker, 1970; Robins and Robins, 1971; Castle, 1972) most of whom have con- cluded that an osteological revision of all in- cluded genera was necessary for a proper under- standing of the family. The aim and scope of this study have been directed to that objective. 10 History of Ophichthid Classification The 218 years since Linnaeus’ (1758) descrip- tion of the first ophichthid species have witnessed a plethora of generic and specific names applied to the Ophichthidae. The first generic name properly applied to an ophichthid was Ophich- thus (Ahl, 1789), which should more properly have been written “Ophichthys’”. The emenda- tions of Ahl’s generic spelling by subsequent authors have resulted in confusion over the gen- eric and familial names. This problem was briefly treated by Gosline (1951a: p. 298) and is further illuminated in the following discussion. The treatment of apodal taxonomy was in con- stant flux during the eighteenth and nineteenth centuries, which accounts for much of the lack of uniformity in nomenclature. Swainson’s (1838) treatment of the eel-like fishes was based largely on Cuvier’s work (1817), and was the first at- tempt at an arrangement of the eels into family groups. Swainson (p. 215) designated the ‘“Mur- aenidae (as) having two branchial spiracles in their ordinary position, and the Sphagebranch- idae, or sea eels, where the branchial spiracles are either close together or united into one.” Confusion ensues on the following page where the family name Gymnarchidae is apparently con- sidered synonymous with the Symbranchidae, and further evidenced in his discussion (p. 218) of the gill openings, by his statement that ‘‘among the Gymnarchidae, or sea eels, for instance, they are close together and united under the throat as in Sphagebranchus.”” Swainson divided the Mur- aenidae into two subfamilies, the Anguillinae and the Muraeninae, both of which contained species now known to be ophichthids. McClelland (1844) realigned the apodal classifications of Swainson and Cuvier and created the family Ophisuridae to include eels with a rayless caudal contain- ing the genera Leptognathus (=Ophisurus), Ophi- surus, and Ophithorax (=Ophichthus). Kaup (1856a,b) disregarded most of McClelland’s classification without comment, but did retain the name Ophisuridae. Kaup divided the apodal fishes into two ‘sections’. These were the Cryp- tomycteres (containing only the Ophisuridae) which included those eels with labial nostrils, and the Phaneromycteres which contained all other apodal families. The ophisurids were divided into three subfamilies, the Ophisurinae, the Sphage- branchinae, and the Myrophinae (containing Myrus, Myrophis, and Muraenichthys). Bleeker, in his Systema Muraenorum Revisum (1865), rec- ognized the family Ophisuroidei and considered SERIES 4, V. 41, #1 McCOSKER — EELS the Myrophinae (as Myriformes) to be a subfamily of the family Congroidei. Ginther (1870) con- siderably revised previous classification by plac- ing the majority of the known eels into a single family, the Muraenidae, which he divided into ten ‘‘Groups’’. The Ophisuridae of earlier authors was divided into two groups, the Ophichthyina containing those species with a rayless caudal [comprising the genera Liuranus (sic) with a single species and Ophichthys with at least 78 species], and the Myrina containing those with a rayed caudal (comprising Myrus, Myrophis, Para- myrus, Chilorhinus, and Muraenichthys). Gwtn- ther’s groups were elevated to family rank by Jordan and Davis (1891). The Myrinae became the family Echelidae (Jordan and Davis considered Myrus a synonym of Echelus) and the Ophich- thyina of Giinther (actually, the Ophisuroidei of Bleeker) became the Ophisuridae. Uncertainty concerning the synonymy of Myrus and Echelus resulted in the changing of the name Echelidae to Myridae by Jordan and Evermann (1896) and by Jordan and Snyder (1901). The family name Ophichthyidae, derived from Gunther's Ophich- thyina, first appeared in Jordan and Evermann (1896). The authors rejected the name Ophisuri- dae and considered Ophisurus a synonym of Ophichthus. Validity of the Family Name Confusion relating to the spelling of the Ophichthyidae with a “y’’ relates to the correc- tion by earlier authors of Ahl’s (1789) spelling of Ophichthus. The generic name is from the Greek re PF , meaning snake, and ixvovys, mean- ing fish, and would more correctly have been written Ophichthys. Bleeker, Gunther, and other classicists emended Ahl’s generic spelling (see Jordan and Gilbert, 1882), but Jordan and his later co-authors returned to Ahl’s original spell- ing. The retention of the original spelling of the generic name but the usage of the emended fam- ily name is illogical, and according to the Inter- national Code of Zoological Nomenclature, in- correct, as Gosline (1951a) has pointed out. Most recent authors, with few exceptions, have used “Ophichthidae”’ and ‘“Ophichthus”. A serious difficulty however exists, concerning the earlier family names proposed by Swainson and by McClelland. The inconsistencies in Swain- son’s usage of Sphagebranchidae would invalidate it as a family name. Although Sphagebranchus Bloch (1795) was then a valid genus and properly an ophichthid, Gymnarchus, a gymnarchid, was PROCEEDINGS CALIFORNIA ACADEMY OF SCIENCES later shown not to be an apodal fish. The prob- lem of recognizing McClelland’s usage of Ophi- suridae was avoided by Gosline who stated that “whether or not Ophisurus is a valid genus is a moot nomenclatorial question. Consequently | prefer not to use for this family, at the present time, the little-known and possibly invalid name Ophisuridae.”’ The logical solution to this nomen- clatural dilemma seems to be the invoking of the plenary powers of the International Commission on Zoological Nomenclature. The suppression of the rarely used name Ophisuridae in favor of the universally recognized name Ophichthidae would clearly be in the interests of stability. Osteological Studies of the Ophichthidae Regan (1912) was the first to attempt an osteo- logical definition of the Ophichthidae. Most im- portantly, his emphasis on the fused frontals of the ophichthids, congrids, and relatives has re- mained as a fundamental character in our con- cept of eel evolution. The other osteologicai characters identified by Regan (‘‘caudal vertebrae with transverse processes” and ‘‘maxillaries articu- lating with ethmoid near the end of the snout’) only described certain members of the family. His separation of the congrids from the ophichthids, on the basis of their long and slender rather than vestigial neural spines, was also an important ob- servation. Trewavas’ (1932) apodal classification scheme followed Regan’s characterization of the ophichthids. For nearly two decades, subsequent ophichthid studies dealt only with superficial characters. An exception was that of Myers and Storey (1939) who noticed the overlapping of the branchiostegal rays in ophichthid species. They pointed out that these rays are similar to the “jugostegalia’’ described by Parr (1930) in echelid eels. On that basis, and other external morpho- logical similarities, Myers and Storey suggested that the Echelidae might be merged with the Ophichthidae. Gosline (1950, 1951a, 1951b, 1952), in a series of papers, analyzed the species referred to the Echelidae and the Ophichthidae. His osteological study of Kaupichthys diodontus (1950) demon- strated that its osteology precluded its inclusion in the same family with Muraenichthys cookei. He found (p. 312-314) that K. diodontus differs in having sutured frontals and non-overlapping branchiostegal rays, as well as several other char- acters which later proved to be non-definitive. Lacking a specimen of Echelus myrus, Gosline was unable to further define the Echelidae, but suggested that species of Myrophis and Muraen- 11 ichthys might be referred to the Ophichthidae, to comprise the subfamily Myrophinae, on the basis of their “basket-like arrangement of the numer- ous, long branchiostegal rays.” In a following paper, Gosline (1951b) described the osteology of Chilorhinus brocki (=C. platyrhynchus) and related it to Kaupichthys diodontus. He suggested that the external similarities of species of Kaup- ichthys, Chilorhinus and ophichthids are ‘‘the re- sult of parallel evolution and not of close genetic relationship.” His prediction that Echelus, once examined on an osteological basis, would prove to be confamilial with Kaupichthys was incorrect (Gosline, 1952; Bohl ke, 1956a). Gosline (1951a), in a more comprehensive paper, prepared the first diagnostic treatment of the Ophichthidae. His study, however, was limi- ted to those species occurring in the Hawaiian area and thus did not include several critical genera. He compared the Ophichthidae with the Congridae (primarily Conger), and concluded that the ophichthid conditions are derived from, and more advanced than, those of their more primi- tive congrid ancestors. Two subfamilies within the Ophichthidae were recognized (the Myro- phinae and the Ophichthinae) although osteo- logical differences other than the caudal skeleton were not defined. In a subsequent paper, Gosline (1952) described the morphology of Echelus myrus in detail and concluded that it was refer- able to the ophichthid subfamily which contained Myrophis and Muraenichthys. On that basis he changed the subfamilial name of the Myrophinae to Echelinae. The results of that study did not alter his earlier (1951a) diagnosis of the family. Subsequent studies dealing with ophichthid osteology followed Gosline’s (1951a) general defi- nition of the family, but amended his diagnosis to include genera that he had not examined. Bohlke (1960) added Pseudomyrophis, and provi- sionally Neenchelys, to the Ophichthidae.In do- ing so, he expanded the familial diagnosis to allow the following: posterior nostrils either lat- eral or labial; maxillary articulation variable in position along the ethmoid; pharyngeal openings of the branchial clefts may be reduced; trans- verse processes of the caudal vertebrae either present or absent. Nelson’s (1966a) analysis of apodal gill arch conditions found most ophichthids to be ‘‘dis- tinguished in having the proximal ends of the dorsal parts of the first and second arches con- nected through a continuous cartilage, a peculiar- ity not present in any other of the eel families studied.’’ Certain generic lineages were identified 12 on the basis of their gill arch configurations. Nel- son's (1966b) study of the osteology of Neen- chelys buitendijki confirmed Bohlke’s earlier sup- position of its placement within the Ophichthidae. Nelson separated the ophichthids from the con- grids in the following manner: posterior nostril usually opening on the ventral surface of the upper lip; tongue adnate; branchiostegal rays overlapping along the midventral line; supraorbi- tal canals united by the transverse frontal com- missure; neural spines absent. Castle’s (1972) osteological study of Benthenchelys cartieri sum- marized the diagnoses from Gosline’s, BOhlke’s, and Nelson’s earlier works, but did not contribute to or amend their diagnoses. Familial Synonymy of the Ophichthidae The Ophichthidae, as currently recognized, in- cludes several families which were until recently considered distinct. The basis and validity of studies resulting in these actions are discussed below. The family Neenchelidae was erected by Bamber (1915) to contain Neenchelys microtretus, a new genus and species from the Red Sea. It was considered to be closely related to the Mur- aenesocidae as defined by Regan (1912). A sec- ond neenchelid, N. buitendijki, was described by Weber and de Beaufort (1916) from the Indo- Australian archipelago. The family received no further definitive treatment until Bohlke (1960), on the basis of Bamber’s description, suggested that Neenchelys may be related to species of Pseudomyrophis and thus properly be considered an ophichthid of the subfamily Echelinae (sensu Gosline, 1951a, 1952). The collection of adequate material of N. buitendijki and its osteological ex- amination by Nelson (1966b) supported Bohlke’s prediction. Nelson’s (1967) examination of the holotype of N. microtretus confirmed the recog- nition of the Neenchelidae as ophichthids in the subfamily Echelinae (herein considered as Myro- phinae). The Acanthenchelyidae also belongs in the Ophichthidae. Family recognition was short-lived, consisting of Jordan, Evermann, and Clark’s (1930) elevation of Acanthenchelys Norman. (errone- ously attributed to Regan) to family status, but this was largely ignored by later authors. Randall and Robins (1966) relegated Acanthenchelys to the synonymy of Ophichthus, an action which is followed here. The Aoteidae are provisionally included in the Ophichthidae. The Aoteidae were first recognized SERIES 4, V.41, #1 McCOSKER — EELS as ophichthids by Castle (1967), who referred the single aoteid species to the genus Muraenichthys. The family Echelidae (=Myridae, Myrophinae, Myrophidae, and Muraenichthyidae), has been a catch-all group with a checkered history. The echelids were considered congrid or muraene- socid relatives by most nineteenth century au- thors. Bleeker (1865), for example, considered the Myriformes (containing Myrophis, Echelys, and Muraenichthys) to be a subfamily of the Congroidei. Kaup (1856a,b) was exceptional in placing considerable importance on the labial nostril condition, and in allying the Myrophinae with the Ophisurinae as a single unit which ex- cluded the congroids and relatives. Twentieth century authors considerably expanded the Myro- phidae (which was to become known as the Echelidae, fide Jordan and Evermann, 1896) to include as many as 22 genera at various times (Schultz and Woods, 1949). The dissection of this large and cumbersome family was initiated by Myers and Storey (1939), and was followed by Schultz and Woods (1949) and by Gosline (1950, 1951a, b, 1952). Myers and Storey noted the pres- ence of accessory branchiostegal rays (the “‘jugo- stegalia’’ of Parr, 1930) in both ophichthids and echelids, but were hesitant to merge the families without an extensive anatomical examination. Gosline (1951a) established the similarities of the two families on an osteological basis and included the genera Myrophis and Muraenichthys in the Ophichthidae to comprise the subfamily Myro- phinae. His subsequent (1952) osteological ex- amination of Echelus myrus resulted in its inclu- sion into the family, and the replacement of the name Myrophinae with Echelinae. The present study has demonstrated that Gosline was correct in considering the Myrophines to be ophichthids, but erred in including Echelus with the Myro- phinae. Although certain authors have continued to recognize a family Echelidae for the forms with caudal rays (e.g., J. L. B. Smith, 1962; Blache, 1968), no convincing arguments have been pro- posed which would merit familial separation. Blache’s (1968: 1501) continued usage of the name Echelidae, with the justification that “.,.nous sommes également, tout a fait, partisan de cette position et nous ne conservons ici, la famille des Echelidae, que pour des raisons arti- ficielles de commodité taxonomique,” is both illogical and incorrect. Further substantiation of the inclusion of the Echelidae with the Ophichthidae is evidenced in the leptocephalus larval stage. (The leptoce- phalus of Neenchelys has not been identified.) PROCEEDINGS CALIFORNIA ACADEMY OF SCIENCES Eel leptocephali display evolutionarily conserva- tive features that could prove useful in phylogen- etic investigations (Castle, 1965, 1967), yet the problem of generic and specific identification still remains and has precluded their usage in this study. It is important to note, however, that the morphology of ophichthid leptocephali re- flects the family grouping fairly well. Castle (1965) and D. Smith (unpub. MS) have diagnosed the ophichthid leptocephalus as moderately elongate when full grown, characteristically possessing gut thickenings or loops which usually accompany swellings of the pronephric ducts, conspicuously pigmented with patches of chromatophores which occasionally occur on the head, at various points along the gut, and often on the myosepta, lateral caudal midline, and dorsal and anal bases, and possessing a moderate to blunt tail. Castle (1965: 98) has stated that ‘‘the pectoral remains obvious throughout larval life, probably even in those ophichthids which show loss of the pectoral in the juvenile and adult.’’ Phylogenetic implications at the generic level would therefore be further evidenced in this conservative larval condition if Castle’s suggestion proves correct. For example, his tentative identification of Leptocephalus Mur- aenichthys sp. is based on a leptocephalus pos- sessing a short, rounded pectoral fin (Castle, 1965: figs. 2F, G), not unlike that of Myrophis (Eldred, 1966; Castle, 1965: figs. 3e-f). In contrast to the above mentioned families, the Macrocephenchelyidae was incorrectly synony- mized with the Ophichthidae. This family, known only from the holotype and damaged paratype of Macrocephenchelys brachialis Fowler, was syn- onymized without comment with the Ophichthi- dae by McAllister (1968: 85). Robins and Robins (1971) have re-erected the family on the basis of a thorough osteological examination of the para- type. They have shown its affinities to be with the Congridae and referred it to the superfamily Congroidea. Macrocephenchelys displays several characters quite divergent from the Ophichthidae, including the extensive ossification of the bran- chial apparatus (yet there is no lower pharyngeal tooth plate), the absence of the transverse frontal commissure of the cephalic lateralis system, and the presence of eight stout branchiostegal rays and a complete palatopterygoid arch. Synopsis of Ophichthid Classification The listing of nominal taxa below summarizes the taxonomic conclusions of this study. Full de- scriptions of new taxa and complete generic syn- Onymies are presented later in this paper. 13 Subfamily Myrophinae Tribe Benthenchelyini Benthenchelys Fowler 1934 Tribe Myrophini Ahlia Jordan and Davis 1891 Muraenichthys Bleeker 1853 Subgenus Muraenichthys Bleeker 1853 Subgenus Scolecenchelys Ogilby 1897 Myrophis Lutken 1851 Neenchelys Bamber 1915 Pseudomyrophis Wade 1946 Schismorhynchus McCosker 1970 Schultzidia Gosline 1951 Subfamily Ophichthinae Tribe Callechelyini Aprognathodon Bohlke 1966 Callechelys Kaup 1856 Letharchus Goode and Bean 1882 Leuropharus Rosenblatt and McCosker 1970 Paraletharchus McCosker 1974 Tribe Sphagebranchini Achirophichthys Bleeker 1865 Apterichtus Duméril 1806 Caecula Vahl 1794 Cirricaecula Schultz 1953 Hemerorhinus Weber and de Beaufort 1916, incertae sedis Ichthyapus de Barneville 1847 Lamnostoma Kaup 1856 Stictorhinus Bohlke and McCosker 1975 Yirrkala Whitley 1940 Tribe Bascanichthyini Allips McCosker 1972 Bascanichthys Jordan and Davis 1891 Caralophia Bohlke 1955 Dalophis Rafinesque 1810 Ethadophis Rosenblatt and McCosker 1970 Gordiichthys Jordan and Davis 1891 Leptenchelys Myers and Wade 1941 Phaenomonas Myers and Wade 1941 Tribe Ophichthini Aplatophis Bohlke 1956 Brachysomophis Kaup 1856 Cirrhimuraena Kaup 1856 Subgenus Cirrhimuraena Kaup 1856 Subgenus Jenkinsiella Jordan and Evermann 1905 Echelus Rafinesque 1810 Echiophis Kaup 1856 Elapsopis Kaup 1856 Evips McCosker 1972 Leiuranus Bleeker 1853 14 Malvoliophis Whitley 1934 Myrichthys Girard 1859 Mystriophis Kaup 1856 Ophichthus Ahl 1789 Subgenus Ophichthus Ahl 1789 Subgenus Microdonophis Kaup 1856 Subgenus Centrurophis Kaup 1856 Subgenus Coecilophis Kaup 1856 Ophisurus Lacépéde 1800 Phyllophichthus Gosline 1951 Pisodonophis Kaup 1856 Pogonophis Myers and Wade 1941 Quassiremus Jordan and Davis 1891 Scytalichthys Jordan and Davis 1891 Xyrias Jordan and Snyder 1901 METHODS Taxonomic Methods Osteological examinations, whenever possible, were based on entire stained and cleared speci- mens. Rare specimens and holotypes were studied by gill arch removal and radiographic examina- tion. Radiographs were prepared using a General Electric 40 KV x-ray unit and Kodak Industrial Type M film. Radiographs were either examined under a dissecting microscope or from photo- graphic enlargements. Stained and cleared gill arches or entire specimens were prepared using the trypsin-preparation method of Taylor (1967) and, in certain instances, the modifications of Miller and Landingham (1969). Neurocrania were prepared by dissection, soaking in a 5-7 percent potassium hydroxide (KOH) solution to remove the flesh, and staining in an alizirin bath. Sutures along the dry skulls became more apparent dur- ing examination when painted with pure glycerin using a fine camel’s hair paint brush. Certain skulls were disarticulated in a 7-10 percent KOH solution to better identify certain sutures. Draw- ings were made using a camera lucida attach- ment on a Wild dissecting microscope. Gill arch terminology is that of Nelson (1969). Bone terminology follows that of Asano (1962) with certain modifications that are identified in the section dealing with bone complexes. The following measurements, used in the gen- eric key and descriptions, are defined as follows: Head length. Measured from the snout tip to the posterodorsal point of the gill opening. Trunk length. Measured from the posterodorsal point of the gill opening to mid-anus. Tail length. Measured from mid-anus to the tail tip. SERIES 4, V.41,#1 McCOSKER — EELS Inclination of the suspensorium. The suspen- sorium is considered to be ‘anteriorly inclined’’ if the angle formed by the midlines of the hyo- mandibular and the mandible (when the mouth is closed) is greater than 90°. If the angle is less than 90° the suspensorium is considered to be “nosteriorly inclined’. This measurement is some- what subjective, and made either from radio- graphs or observations of stained and cleared specimens under the dissecting microscope. All fish lengths are listed as total lengths. Gen- eric descriptions and diagnoses were based on adults unless otherwise stated. ABBREVIATIONS Anatomical Abbreviations A-anus; an-anterior nostril; AR- anal fin ray; B, - first basibranchial; BO - basioccipital; BR - branchiostegal ray; BS-basisphenoid; Cy, - first ceratobranchial; CE-centrum; CH - ceratohyal; Cl-cleithrum; Co-coracoid; CTP-transverse pro- cesses of caudal vertebrae; CX - cartilaginous ex- tension of terminal vertebra; D-dentary; DFO - dorsal fin origin; DR - dorsal fin ray; E- eyeball; E, - first epibranchial; EH - epihyal; EN - epineur- al; EO - epiotic; ET - ethmoid portion of premax- illoethmovomer; EX - exoccipital; F - frontal; GB - gas bladder; GH - glossohyal; GO - gill opening; H - heart; H: - first hypobranchial; HA - haemal arch; HH-hypohyal; HY-hypural; HYM - hyo- mandibular; I - intestine; I, - second infrapharyn- gobranchial; IM - intramuscular bone; io - infra- orbital pore; IO - interopercle; L/D - relation of length to depth; LL- lateral line; LP - lower pharyngeal tooth plate; MX - maxilla; N - nasal; NA - neural arch; NS - neural spine; OP - opercle; OR - orbit; P- parapophysis; PA - parietal; PAS - parasphenoid; PD- pneumatic duct; PG - ptery- goid; PL- pleural rib; pm - preoperculomandibu- lar pore; PO-preopercle; pop - preopercular pore; por-postorbital pore; POR - postorbital; PR - pectoral fin rays; Pt-pterygiophore; PT - pterotic; PTS - pterosphenoid; Q-quadrate; S- stomach; SA- sagitta; Sc-scapula; SCI - supra- cleithrum; so - supraorbital pore; SO - supraocci- pital; SOC -supraoccipital crest; SOP - subop- ercle; sp-surface sensory papillae; SP - sphen- otic; stp-supratemporal pore; tp-temporal pore; UH - urohyal; UP - upper pharyngeal tooth plate; V - vertebra; VO - vomer. Distributional Abbreviations EA -eastern Atlantic; EP-eastern Pacific; IP - central and Indo-west Pacific; M - Mediterranean; WA - western Atlantic. PROCEEDINGS CALIFORNIA ACADEMY OF SCIENCES Institutional Abbreviations The following abbreviations are used in reference to material examined: ANSP - Academy of Natural Sciences of Philadel- phia; BPBM - Bernice P. Bishop Museum; CAS - California Academy of Sciences; DANA - Carls- bergfondets DANA-Ekspeditioner, Marinbiologisk Laboratorium Charlottenlund Slot, Denmark; IA - Australian Museum at Sydney; LACM - Los An- geles County Museum; MCZ- Museum of Com- parative Zoology, Harvard University; MNHN - Muséum National d’Histoire Naturelle, Paris; MSNG - Museo Civico di Storia Naturale, Genoa, Italy; MZUSP - Museu de Zoologia, Universidade de Sao Paulo, Sao Paulo, Brazil; RU - Rhodes University, Grahamstown, South Africa; SIO - Scripps Institution of Oceanography; SU - Stan- ford University, also listed as SNHM for Stanford Natural History Museum, specimens now de- posited at the CAS; TABL- Southeast Fisheries Center, Miami, Florida; UCLA - Fish Collection, Department of Zoology, University of California at Los Angeles; UMML- Rosenstiel School of Marine and Atmospheric Science of the Univer- sity of Miami; USNM - National Museum of Nat- ural History, Smithsonian Institution, Washing- ton, D.C.; ZMA- Zodlogisch Museum Amster- dam. Material Examined Listed below is the material utilized for osteo- logical examination in this study. The specimens are grouped by tribes and listed alphabetically within each tribe. Following each specific name is the museum abbreviation, museum catalogue number, number of specimens, range of the total length(s) of the specimen(s) involved. Abbrevia- tions are: CS, stained and cleared by the Taylor (1967) trypsin technique; GA, gill arches re- moved, stained and cleared; H, hyoid removed; S, skull preparation; X, radiograph. Specimens utilized only for vertebral counts are not in- cluded in this listing. Ophichthidae - Benthenchelyini. Benthenchelys cartieri, DANA 3735, 2(105-115mm), CS. Ophichthidae - Myrophini. Ahlia egmontis, SIO 67-87, 1(268), S; SIO 71-266, 1(337), CS. Muraen- ichthys chilensis, SIO 65-645, paratype, 1(248), CS; SIO 65-655, paratype, 1(276), CS, 1(292), S. Muraenichthys gymnopterus, SIO 69-276, 1(129), S, 1(144), CS. Muraenichthys gymnotus, SIO 69- 266, 1(244), CS. Muraenichthys macropterus, SIO 69-277, 1(181), CS. Myrophis plumbeus, SIO 69- 15 371, 1(182), CS. Myrophis uropterus, CAS 13971, 1(159), CS; BPBM 27209, 1(182), CS. Myrophis vafer, SIO 68-242, 1(193), CS, 2(265-325), S. Pseu- domyrophis micropinna, SIO 60-72, 1, head and trunk only, CS. Pseudomyrophis nimius, ANSP 110150, 1(350), CS. Schismorhynchus labialis, CAS 24687, 5(114-137), CS. Schultzidia johnstonensis, SIO 69-267, 1(138), CS. Ophichthidae - Callechelyini. Aprognathodon platyventris, SIO 68-393, paratypes, 2(312-330), CS. Callechelys bilinearis, SIO 70-376, 1(260-+), CS. Callechelys cliffi, SIO 61-247, 1(218), GA, H, X; SIO 65-281, 1(298), GA, H, X. Callechelys eris- tigmus, SIO 65-185, paratype, 1(552), GA, H, X; SIO 65-354, paratype, 1(431), CS. Callechelys gal- apagensis, UCLA 64-40, paratype, 1(767), GA, H, X. Callechelys holochromus (holotype of Crypto- pterygium holochroma), USNM 154994, 1(801), X. Callechelys luteus, SIO 68-497, 1(1038), H, X. Callechelys marmoratus, SIO 69-269, 1(286), CS, 1(340), S. Callechelys melanotaenius, SIO 69-269, 1(401), CS. Callechelys muraena, TABL Oregon 2819, 1(235), GA, H, X. Callechelys nebulosus, SIO 71-197, 1(283), CS. Callechelys springeri (hol- otype of Gordiichthys springeri), USNM 121604, 1(372), X. Callechelys striatus SIO 71-165, 1(430), H, X. Letharchus velifer, holotype, USNM 31458, 1(396), X. Letharchus rosenblatti, SIO 67-40, para- type, 1(248), CS. Leuropharus lasiops, holotype, SU 57313, 1(174), GA, X. Paraletharchus opercul- aris, UCLA 64-38, 1(435), GA, X. Paraletharchus pacificus, SIO 65-321, 2(276-369), CS. Ophichthidae - Bascanichthyini. Allips concol- or, holotype, CAS 13967, 1(375), GA, X. Bascan- ichthys panamensis, SIO 71-98, 1(425), CS; SIO 71-224, 1(295), CS, 1(510), S. Caralophia loxochila, SIO 70-228, 1(445), CS; SIO 70-376, 1(238), CS. Dalophis imberbis, SIO 72-290, 1(440), GA, X. Ethadophis byrnei, holotype, SIO 67-31, 1(508), GA, X. Ethadophis merenda, holotype, SIO 65-47, 1(530), GA, X. Leptenchelys vermiformis, holo- type, USNM 101785, 1(115), X. Phaenomonas cooperae, CAS 13964, 2(451-549), CS. Phaenom- onas pinnata SIO 65-348, 1(375), CS, 1(375), S. Ophichthidae - Sphagebranchini. Apterichtus caecus, MSNG 41058, 1(435), X. Apterichtus flavi- caudus, SIO 69-364, 1(300), CS. Caecula ptery- gera, USNM 206375, 1(232), CS. Cirricaecula johnsoni, paratype, USNM 141189, 1(325), CS. Ichthyapus ophioneus, SIO 70-376, 1(337), GA, X. Ichthyapus selachops, SIO 61-232, 1(400), CS, 1(400), S; SIO 65-343, 1(234), CS. Ichthyapus vul- turis, holotype, ZMA 104.153, 1(240), X; SIO 69- 16 366, 1(446), GA, X. Lamnostoma kampeni, SU 24593, 1(435), GA, X. Lamnostoma orientalis, CAS 13959, 1(205), CS; CAS 13968, 1(229), CS. Stictor- hinus potamius, MZUSP 8959, paratype, 1(289), CS. Yirrkala kaupi, SU 26827, 1(345), GA, X. Yir- rkala lumbricoides, CAS 13969, 1(346), CS; para- type of Y. chaselingi, 1A 16190-601, 1(560), X. Yirrkala misolensis, CAS 13965, 1(335), CS. Yir- rkala tenuis, SIO 71-165, 1(370), CS. Yirrkala sp., BPBM 11858, 1(306), CS. Ophichthidae - Ophichthini. Aplatophis chau- liodus, UMML 27209, 1(165), CS. Brachysomophis sauropsis, SIO 69-267, 1(323), CS; SIO 69-271, 1(197), CS. Cirrhimuraena macgregori, SIO 68- 434, 2(291-317), CS. Cirrhimuraena taeniopterus, CAS 13962, 1(445), GA, H, X. Elapsopis cyclor- hinus, SIO 69-267, 1(268), CS, 1(475), GA. Eche- lus myrus, SIO 69-369, 1(449), dissected. Echelus pachyrhynchus, SIO 69-370, 1(355), CS. Echiophis sp., UMML 29144, 1(298), CS. Evips percinctus, holotype, CAS 13966, 1(125.5), GA, X. Leiuranus semicinctus, SIO 61-132, 1(195), CS; SIO 69-268, 1(243), CS; SIO 69-273, 1(344), S. Malvoliophis pinguis, |A 3646, 1(470), CS. Myrichthys colub- rinus, SIO 69-272, 1(345), CS. Myrichthys macu- losus, SIO 68-497, 1(390), CS; SIO 69-272, 1(310), CS. Myrichthys xystrurus, SIO 65-335, 1(420), S; SIO 65-354, 2(243-268), CS. Myrichthys sp., SIO 34-371, 1(386), CS. Ophichthus altipinnis, CAS 14647, 1(915), GA, X. Ophichthus cephalazona, SIO 69-279, 2(230-330), CS. Ophichthus erabo, CAS 13960, 1(480), GA, H, X. Ophichthus ophis, SU 51724, dissected. Ophichthus rutidoderma- toides, CAS 28727, 1(330), CS. Ophichthus tri- serialis, SIO 61-193, 1(230), CS; SIO 69-252, 1(800), S, prepared skeleton. Ophichthus zopho- chir, SIO 60-304, 1(217), CS; SIO 65-166, 1(310), S, 1(340), S. Ophisurus serpens, RU 76-78, 1(325), CS. Phyllophichthus xenodontus, SIO 69-273, 2(270-305), CS. Pisodonophis boro, SIO 69-281, 1(410), CS. Pisodonophis cancrivorus, SIO 69- 307, 1(345), CS. Pisodonophis cruentifer, MCZ 34529, 1(235), CS. Pisodonophis daspilotus, SIO 72-73, 1(251), CS. Pogonophis fossatus, SIO 61- 227, 2(232-249), CS. Quassiremus evionthas, UCLA 64-19, 1(283), GA, X. Quassiremus nothochir, SIO 65-334, 2(271-342), CS. Scytalichthys miurus, CAS 13970, 1(235), CS. Xyrias revulsus, holotype, SU 6476, 1(890), GA, X. Comparative Material - Non-ophichthids. The following material was stained and cleared with two exceptions. These, Gymnothorax mordax and Muraenesox coniceps, were examined from pre- pared skeletons. SERIES 4, V. 41, #1 McCOSKER — EELS Anguillidae. Anguilla rostrata, SIO 69-254, 1(210). Muraenidae. Anarchias galapagensis, SIO 65- 345, 1(130). Echidna nebulosa, SIO 59-8, 1(178). Enchelycore bayeri, CAS GVF 1957-18, 1(255). Gymnothorax castaneus, SIO 65-291, 1(140). Gymnothorax mordax, SIO skeletal collection, 1(ca. 1 meter). Gymnothorax panamensis, SIO 61- 239, 1(180). Gymnothorax schismatorhynchus, CAS GVF 1958-13, 1(265). Muraena lentiginosa, SIO 65-354, 1(165). Uropterygius necturus, SIO 65-302, 1(175). Simenchelyidae. Simenchelys parasiticus, SIO 68-479, 1(295). Derichthyidae. Derichthys serpentinus, SIO 60- 239, 1(140). Serrivomeridae. Serrivomer sector, SIO 63-374, 1(305). Nemichthyidae. Nemichthys scolopaceous, SIO 65-243, 1(440). Heterenchelyidae. Pythonichthys asodes, para- type, UMML 23481, 1(290). Muraenesocidae. Muraenesox coniceps, SIO skeletal collection, a large adult. Moringuidae. Moringua ferruginea, SIO 68-531, 1(305). Congridae. Ariosoma_ gilberti, SIO 62-709, 3(175-190). Conger cinereus, SIO 68-531, 1(210). Gorgasia punctata, SIO 62-270, 1(365). Taenio- conger sp., SIO 62-42, 1(235). Xenocongridae. Chlopsis apterus, LACM 32555, 2(152-155). Kaupichthys hyoproroides, SIO 67-45, 2(150-190). Statistical Methods Comparisons of vertebral means and variances were made with a standard ‘t’” test. Two com- puter-programmed grouping techniques were used to compare inter- and intra-generic rela- tionships within the Callechelyini. These were modifications of a clustering technique, the weighted variable group method (WVGM)_ of Sokal and Michener (1958), and of a recurrent group analysis (REGROUP) devised by Fager (1957). Each will be discussed briefly, although the original sources should be referred to_ if further information is desired. Wormuth’s (1971) modification of WVGM, used by him in ommastrephid squid taxonomy, was used to identify the interspecific relation- ships of 18 of the 21 species of the Callechelyini. Cluster analysis is a technique widely used by numerical taxonomists, wherein each species (or individual) is termed an operational taxonomic PROCEEDINGS CALIFORNIA ACADEMY OF SCIENCES unit (OTU). The program, as described by Wor- muth (1971), operates in the following manner: Initially a matrix of m characters by n OTU’s is set up and the values in each row are standardized using row means and variances. From the standardized matrix (mxn) a product moment correlation coefficient is calculated. This matrix expresses the rela- tionships between all possible pairs of OTU’s quantitatively. At this point a cluster- ing procedure is employed to extract a graphic representation of the information contained in the correlation matrix. Any pair of OTU’s which has a higher correlation with each other than either has for any other OTU is put together as a group. An average correlation coefficient is computed for each group and it is, thenceforth, treated as a single OTU. In the WVGM a variable num- ber of new groups are formed on each cycle. At the end of each cycle, correlation coeffi- cients are recomputed based on the previous matrix. On any single clustering cycle two alternatives are available. One permits only groups of two OTU’s to form prior to recom- putation of the correlation matrix if their incorporation lowers the overall group cor- relation value by less than a_ preselected amount. As the results of both alternatives over a number of trials were very similar, the latter option was selected for its shorter computation time. The levels at which groups are formed are plotted. The graphi- cal representation of the results is termed a phenogram. Data and characteristics used in program WVGM for the Callechelyini are presented in tables 8 and 9. Recurrent group analysis was devised by Fager (1957) to identify communities of species on the basis of their co-occurrence in samples (Fager and McGowan, 1963; Fager and Longhurst, 1968) and later modified for taxonomic purposes by Ebeling and Weed (1963) and by Fager (1969). For each species pair, the program calculates an index of affinity. Fager and McGowan (1963) state that this index does not follow the hyper- geometric distribution exactly, and have there- fore replaced it with the geometric means of the proportion of common characteristics, corrected for the number of characteristics recorded for the species, such that: =u Be l= J AxB where | = index of affinity, ) = number of com- mon characteristics, A and B = total characteris- tics recorded for species A and species B, and 17 where B = A. All characteristics are equally weighted. From the matrix of species pairs which is generated the largest possible group of species is selected. A “breakpoint” is selected such that pairs of species in which | is greater than or equal to that point are considered to show af- finity. For example, the selection of 0.500 as a breakpoint would group species which share somewhat more than “half” their characteristics. In this study, breakpoints of 0.500 and 0.600 were utilized, the latter appearing to give more reasonable groups without involving excessive al- ternative and unnatural groups. Once selected, the largest group is removed and the procedure is continued until all possible groups are formed. Where two or more groups of equal size are pos- sible, the program selects the one for which the sum of species pairs’ affinity indices is largest. Intergroup relationship can be calculated using the sum of characteristics shared by members of each group as a fraction of the total possible connections (see fig. 44). Inter-group similarity is therefore directly related to this fraction. It should be noted that the computer programs used are each affected by the amount of, and manner in which, data are presented. The sensi- tivity of each program increases with increased data input. REGROUP is particularly insensitive to continuous data (e.g., vertebral and branchio- stegal ray numbers, body proportions) and treats each data interval equally. WVGM, by contrast, takes account of continuous data in calculating a correlation coefficient, but is somewhat more insensitive to dichotomous and_ trichotomous data. The characteristics used in each program are identified in tables 8 and 9. OSTEOLOGY AND FUNCTIONAL ANATOMY In the following section the osteology of Oph- ichthus zophochir is described and illustrated in detail. The bone complexes are treated sepa- rately, each beginning with a description of the condition of O. zophochir and followed by a discussion of variations and_ specializations among other genera within the family. Also in- cluded in this section are discussions of the oto- liths and of the gas bladder and digestive tract conditions of certain ophichthids. Ophichthus zophochir was selected as the spe- cies with which other ophichthids are compared because it possesses the majority of ophichthid anatomical characters in a rather generalized state. This is not meant to imply that O. zopho- chir is the most primitive ophichthid, but rather 18 that by being generalized, and not specialized by means of extreme anatomical reduction, it provides a framework for comparison and discus- sion. Neurocranium The neurocranium of O. zophochir is com- pletely ossified and well fused along the cranial sutures. The skull is stout, rather elongate, and truncate posteriorly. The neurocranium of an adult O. zophochir, like that of most anguilli- forms, is small in relation to the total length of the fish. It occupies 5.5 percent of the TL, yet the neurocrania of more elongate ophichthids, such as Phaenomonas cooperae, occupy as little as 1.3 percent. Various aspects of the neuro- SERIES 4, V.41, #1 McCOSKER — EELS cranium of O. zophochir are illustrated in figures 2 and 3. Described below are the elements com- posing the neurocranium. Premaxilloethmovomer. The premaxillae, eth- moid, and vomer are fused into a single com- plex (PEV) articulating posterodorsally with the frontal and posteroventrally with the parasphe- noid, and forming the anterior margin of the orbit. The anterior portion of the PEV, the pre- maxillae, has been shown to be separate from the vomer in other eels, including Anguilla an- guilla (Norman, 1926), Derichthys serpentinus (Beebe, 1935), and Coloconger scholesi (Chan, 1967), yet in ophichthids there are no distinct sutures separating the elements, and their precise limits can only be determined ontogenetically. The premaxillary portion is expanded in most ke : rhs Figure 2. Neurocranium of Ophichthus zophochir, SIO 65-166. Upper, dorsal view; lower, left lateral view. Scale represents 1 mm. Stippled lines represent cephalic lateralis canals. Abbreviations are: BO, basioccipital; BS, basisphenoid; E, ethmoid portion of pre- maxilloethmovomer; EO, epiotic; EX, exoccipital; F, frontal; N, nasal; OR, orbit; PA, parietal; PAS, parasphenoid; POR, postorbitals; PRO, prootic; PT, pterotic; PTS, pterosphenoid; SO, supraoccipital; SP, sphenotic; VO, vomer. Figure 3. Neurocranium of Ophichthus zophochir, SIO 65-166. Upper, ventral view; lower, posterior view. Scale represents 1 mm. Abbreviations are as in Figure 2. Sagitta (SA) is outlined by stippled line. PROCEEDINGS CALIFORNIA ACADEMY OF SCIENCES Figure 4. Neurocranium of Benthenchelys cartieri, DANA Sta. 3735. Scale represents 1 mm. Figure 5. Neurocranium of Myrophis vafer, SIO 68-242. Scale represents 1 mm. Figure 6. Neurocranium of Muraenichthys chilensis, SIO 65-655. Scale represents 1 mm. Figure 7. Neurocranium of Callechelys marmoratus, SIO 69-629. Scale represents 1 mm. a9 SERIES 4, V.41,#1 McCOSKER — EELS Figure 8. Neurocranium of Ichthyapus selachops, SIO 65-232. Scale represents 1 mm. Abbreviations are: BS, basisphenoid; OR, orbit. Figure 9. Neurocranium of Stictorhinus potamius, MZUSP 8959. Scale represents 1 mm. Abbreviations are: BS, basisphenoid; OR, orbit. Figure 10. Neurocranium of Bascanichthys panamensis, SIO 71-224. Scale represents 1mm. Figure 11. Neurocranium of Myrichthys xystrurus, SIO 65-335. Scale represents 1 mm. Abbreviation PTS is for pterosphenoid. PROCEEDINGS CALIFORNIA ACADEMY OF SCIENCES ophichthids (extremely so in Benthenchelys, fig. 4) and is toothed in all genera except Aprognath- odon. It narrows posteriorly to become the vomer, which is toothed in most ophichthids (exceptions are Schultzidia, Phyllophichthus, Leiuranus, and Leuropharus). The premaxillary and ethmoid dentition, here termed intermaxil- lary, is continuous with that of the vomer in O. zophochir. A gap separates the intermaxillary and vomerine dentition of many ophichthid genera, and appears to be a useful character to indicate relationship (table 1). The vomerine dentition does not continue onto the parasphenoid. The ethmoid portion forms the anterior margin of the orbit and the medial margin of the nasals. The ethmoid is perforated anterolaterally by the lat- eral commissure of the first cranial nerve. Nasals. The nasals of O. zophochir are paired, thin, laminar, and cartilaginous along their ex- ternal edges. The anterior portion of the supra- orbital cephalic lateralis nerve tract passes through the canal along the median edge of the nasal. The nasals of the Myrophinae are either cartilaginous or absent. Nasal development among the Ophichthinae is variable. In certain long-jawed ophichthines (including Brachysomo- phis, Scytalichthys, and Aplatophis) the nasals are either rudimentary or absent. Nasal cartilage ex- tends from the anterior edge of the PEV; its de- velopment is quite variable, generally consisting of two short lateral rods; in certain genera it is fused centrally. Parasphenoid. The parasphenoid (PAS) is a long, narrow, toothless bone, anteriorly overlying the vomer and forming the ventral margin of orbit. Centrally, it is spread laterally to form the anterior floor of the cranium, narrows posteri- orly, and splits into two short prongs. It forms the ventral margin of the orbit, and borders the orbitosphenoid, pterosphenoid, prootic, and basioccipital. Prootic. The paired prootics (PRO) combine with the paired basioccipitals and pterotics to form the otic bulla. They are small, nearly rec- tangular, and highly perforated with numerous Openings for the passage of nerves and blood vessels. Through the most conspicuous foramen passes the hyomandibular trunk of the facial nerve (VII). Smaller foramina exist for the pas- sage of the orbital artery and the jugular vein. The major axis of the PRO is horizontal. The PRO are bordered by the parasphenoid medially, the pterosphenoid anteriorly, the sphenotic an- terolaterally, the pterotic dorsolaterally, and the basioccipital posteriorly. 2il Basioccipital. The median basioccipital (BO) is a small, irregularly shaped bone which forms the posterior portion of the otic bulla. Its major axis, in contrast to the prootics, is vertical. It is bordered medially by the parasphenoid, anteri- orly by the prootics, and dorsally by the pterotics and exoccipitals. Pterotic. The paired pterotics (PT) are elongate, narrow anteriorly, and broadly flared laterally and posteriorly, forming the lateral edge of the roof of the cranium. The cephalic lateralis nerve tract passes through the PT and opens anteriorly in the frontal and posteriorly at the posterior PT margin. Anteromedially the PT are bordered by the frontal, followed medially by the parietals and epiotics, anterolaterally by the pterosphe- noid and sphenotic, ventrolaterally by the prootic and basioccipital, and posteriorly by the exoc- cipitals. Posterior to the sphenotics, the PT forms a sheet-like eave extending beyond the body of the cranium. Exoccipital. The paired exoccipitals (EX) form the dorsal and lateral margins of the foramen magnum. They are sutured along their dorsal midline, and extend posterodorsally and laterally as a semicircular sleeve around the foramen magnum. Ventrally, they contact the basioccipi- tal, and dorsally the supraoccipital. The foramen of the tenth cranial nerve opens posteriorly along the ventral EX-BO border, with the ninth opening lateroventrally from a foramen slightly antero- lateral to that of the tenth. Supraoccipital. The supraoccipital (SO) is single, small, square to subrectangular in shape, and lies along the posterodorsal cranial midline. In O. zophochir the narrow median crest of the SO extends posteriorly as a small point. The condi- tion is typical of many ophichthids, although in certain generic groups, particularly Callechelys and related genera, they are rounded along the posterior SO margin. The SO extends anteriorly beneath the parietals, and is bordered laterally by the epiotics, posteriorly by the exoccipitals, and anteriorly by the prootics in most genera. In Schultzidia the SO has surfaced and has sepa- rated the parietals, extending anteriorly to con- tact the frontal. Epiotic. The epiotics (EO) are paired, thin, and subrectangular. In O. zophochir they are bor- dered anteriorly by the parietals, laterally by the pterotics, and medially by the supraoccipital. Posteriorly their major axis is changed from hori- zontal to vertical in forming the dorsolateral mar- gin of the posterior cranial face bordering the exoccipitals. The EO, like the supraoccipital and the pterotics, forms a narrow sheet-like eave along their posterior margins. Parietal. The paired, thin, subrectangular pari- etals (PA) overlay the posterior margin of the frontal. They are bordered anteriorly by the frontal, laterally by the pterotics, posterolaterally by the epiotics, posteriorly by the supraoccipital, and fused medially. In certain ophichthids which have a prominent supraoccipital crest the median ridge development begins along the parietal mid- line and gradually increases to its posterior pro- jection. Castle (1972: fig. 10) has erred in illus- trating the PA of Benthenchelys as contacting the sphenotic; in no ophichthid did | find this juncture. Frontal. The frontal (F) is a single long element which, along with the epiotics and_parietals, forms the roof of the cranium. Ontogenetically, the frontal is presumably formed from the fusion of paried lateral elements, but in juveniles and adults there is no evident suture. In O. zophochir the F is ridged posteriorly along the dorsal mid- line. In Aplatophis this ridge is developed as a sharp crest. Several nerve tracts pass through the F, including the anterior tract of the cephalic lateralis nerve and the transverse frontal com- missure, which is unique to the Ophichthidae. The frontal is deeply split anteriorly by the in- sertion of the ethmoid portion of the PEV in some genera, and bordered anteroventrally by the orbit and orbitosphenoid, ventrolaterally by the parasphenoid, laterally by the pterotic, and posteriorly by the parietals. The dorsalmost post- orbital of certain species of Ophichthus, Echio- phis, and Brachysomophis is weakly sutured to the frontal at the level of the transverse com- missure. Basisphenoid (orbitosphenoid of others). The basisphenoid (BS) is a small, unpaired median bone with two lateral wings which forms the posteroventral margin of the orbit. It is bordered dorsally by the frontal, posteriorly by the ptero- sphenoids, and ventrally is supported by the parasphenoid (a myodome is not present). The BS in certain genera with elongate and depressed neurocrania has become narrow and elongate, as can be seen in a comparison of Ophichthus (fig. 2), Ichthyapus (fig. 8), and Stictorhinus (fig. 9); Robins (1971: 164-165) has noted that the use of the term “orbitosphenoid” in other eel studies (including Gosline, 1950, 1951, 1952; Regan, 1912; Robins and Robins, 1967; Trewavas, 1932; and others) actually pertained to the BS. Chaba- naud (1936) stated that the teleostean BS is not SERIES 4, V.41, #1 McCOSKER — EELS homologous with that of higher vertebrates and proposed the name ‘‘porpitual’” for that bone in teleosts. Springer (1968: 43-44) agreed with Chabanaud’s conclusions but conserved the name “basisphenoid’’ because of its widespread usage in ichthyology, an act with which | fully agree. Pterosphenoid (=alisphenoid). The small paired pterosphenoids (PTS) form the anterodorsal roof of the cranium. In O. zophochir they are con- cave, turning evenly from a longitudinal axis (along the margin of the frontal) to a nearly transverse axis which abuts the anterior margin of the sphenotic. They are bordered anteriorly by the frontal, dorsally by the pterotics, pos- teriorly by the prootic and sphenotic, and medi- ally by the parasphenoid. The conspicuous fora- men along the PTS-pterotic border is the an- terior opening of the trigemino-facialis chamber. Otoliths. Ophichthid otoliths, like those of most anguilliforms, are small, and hence have received little attention either on a descriptive or a comparative basis. Studies are limited to those of Frost (1926), which included illustrations of the sagittae of Myrus vulgaris (=Echelus myrus), Ophichthus gomesii, and Pisodonophis boro, and the photograph of the sagitta of Myrophis lepturus in Kotthaus (1968). The sagitta (largest of the three otolith pairs) of twelve ophichthid species were compared in this study. The asteris- cus and lapillus were too small to be of com- parative value. Ophichthid otoliths are ovate and biconvex, with a shallow sulcus on the medial surface. They are particularly distinctive in having a shallow ostial channel which opens anteriorly rather than turning dorsally and opening from the sulcus, as is typical of the Congridae (cf. Friz- zel and Lamber, 1962: fig. A; Frost, 1926: figs. 10-11, 15). The sagittae of Ophisurus and Oph- ichthus (fig. 12) possess a short anterior rostrum and are more elongate than those of Ethadophis and Myrophis, which are roughly circular. Sagit- tae of Myrichthys and Echiophis are intermediate in shape. Dentition. Dentition has often been used as a principal character to define and differentiate ophichthid genera. The location and shape of teeth was found in this study to be an important indication of relationship. The dentition of O. zophochir represents the generalized ophichthid condition in being multiserial, conical, and on all tooth bearing bones (vomer, ethmoid, pre- maxilla, maxilla, and dentary). Considerable vari- ation exists within the family, including the elongate fanglike dentition of Aplatophis, the molariform or granular dentition of Myrichthys, PROCEEDINGS CALIFORNIA ACADEMY OF SCIENCES Figure 12. Otoliths (medial face, left sagitta) of several ophichthid species. Anterior end up. Sulcus outline inked in. Scale represents 2 mm. A. Ophichthus triserialis B. Ophichthus zophochir C. Ophisurus serpens D. Echiophis intertinctus E. Ethadophis merenda (orientation uncertain) > Be Ge ae Myrichthys xystrurus Echelus pachyrhynchus Aplatophis chauliodus Myrophis vafer i) W SERIES 4, V. 41, #1 McCOSKER — EELS -—H Figure 13. Head skeleton of Ophichthus zophochir, SIO 60-304. Scale represents 1 mm. Refer to Figures 2-3 (neurocranium), 14 (suspensorium), 17 (hyoid), and 19 (pectoral girdle) for names of bones. Pisodonophis and certain species of Muraenich- thys, the minute, nearly villiform dentition of Schultzidia, and the smooth toothless vomer of Leiuranus, Leuropharus, Phyllophichthus and Schultzidia. Characteristics of the dentition of ophichthid genera are summarized in table 1. Suspensorium and Jaws The conditions of the dentition, suspensorium, and jaws are directly related to the feeding habits of the various genera. In ophichthids, differences in feeding habits (e.g., major differences in prey items and adaptations of the predators to the different habitats) are greater between the spe- cies of different genera than between congeners. These differences are well evidenced in the form of the suspensorium and jaws of various ophich- thids. The juxtaposition of the neurocranium, suspensorium and jaws, pectoral girdle and hyoid apparatus of O. zophochir is illustrated in figure 13. All elements of the suspensorium and jaws are paired. Hyomandibular. The hyomandibular (HYM)_ is stout and shaped like an inverted right triangle. The HYM of strong-jawed piscivorous genera is generally strongly ridged for the attachment of the massive adductor mandibularis muscle. The dorsal surface of the HYM abuts the sphenotic and pterotic. A small irregular condyle along the antero-dorsal margin of the HYM fits into a shal- low socket formed along the sphenotic-pterotic suture. The large process on the posterior HYM margin adjoins the anterior process of the op- ercle. The ventral portion of the HYM contacts the quadrate. Quadrate. The small, stout quadrate (Q) is tightly sutured to the HYM. The vertical ridge along the outer face of the HYM is continuous along the Q. Ventrally the Q bears a broad rounded con- cave process that contacts the articular bone of the mandible. Articular. The wedge-shaped articular (AR) is narrowed anteriorly, and slides into a pocket within the dentary. A remnant of the corono- meckelian is present along the inner face of the AR of O. zophochir; its presence in other genera was not systematically determined. Posteriorly, a grooved socket in the AR meets the rounded socket of the quadrate. Dentary. The dentary (D) is the toothed bone of the mandible, joined by the articular pos- teriorly and adjoining its opposite member at the symphysis by a cartilaginous connection. Postorbitals. The three postorbitals (POR) of most ophichthids are separate, weak ossicles that surround the nerve tract connecting the supra- orbital and infraorbital pore tracts. The POR of O. zophochir however, are specialized by en- largement and fusion to form a continuous strut bracing the mandible and neurocranium. This specialization, also present in other species of Ophichthus and Echiophis, is extreme in Brachy- somophis (fig. 15). The antorbital cartilaginous strut which Gos- line (1952) described for Echelus myrus was not PROCEEDINGS CALIFORNIA ACADEMY OF SCIENCES Yat 4 flea Oh) Wy T N Figure 14. Suspensorium and jaws of Ophichthus zophochir, SIO 65-166. Scale repre- sents 5 mm. Abbreviations are: AR, articular; D, dentary; HYM, hyomandibular; IO, inter- opercle; MX, maxilla; OP, opercle; PG, pterygoid; PO, preopercle; POR, postorbitals; Q, quadrate; SOP, subopercle. A. Outer face B. Inner face 26 SERIES 4, V.41, #1 McCOSKER — EELS Figure 15. Right postorbital series of Brachysomophis sauropsis. Scale represents 1 mm. Abbreviations are: E, eyeball; MX, maxilla; POR, dorsal-most postorbital; io, infraorbital pore tract. found in other ophichthids studied (see Remarks on Echelus). A similar cartilaginous strut has been observed in other eels however, including the dysommid Atractodenchelys phrix and in the Syn- aphorbranchidae (Robins and Robins, 1970: 307). They felt “it (probably) represents either the pre- frontal or lateral ethmoid or their fusion.” Pterygoid (=palatopterygoid). Pterygoid (PG) development is variable within the family, al- though a clear pattern of relationship was not apparent. The PG is reduced to a narrow splint in most ophichthids. In O. zophochir it is thin, laminar, pointed anteriorly, and blunt posteriorly. It is held in place by the dermal layer and con- nective tissue, and contacts neither the quadrate nor the hyomandibular. The PG of all ophichthids tapers anteriorly, and often posteriorly, is largely cartilaginous in many species, and does not ap- pear to serve any distinct purpose. Ahlia is ex- ceptional in differing from the closely related species of Myrophis in the shape of its PG and in lacking vomerine teeth (fig. 30). A true pala- tine is not present in ophichthids, as shown by Robins and Robins (1971) in their discussion of the “palatopterygoid arcade.” Maxilla. The maxilla (MX) of all ophichthids is toothed, elongate, and possesses an anterior dorsal process which articulates with the PEV. Posteriorly, the maxillae of O. zophochir are truncate and do not extend beyond the articular. The generalized ophichthid condition however, is that of a toothless, elongate, ossified or cartil- aginous extension of the MX beyond and lateral to the articular. The location of maxillary articula- tion with the vomer is affected by the elongation of the snout and jaw, and is quite variable within the family. The condition is certain ophichthid genera is illustrated in figures 16 and 30. In concluding this section on the neurocran- ium and suspensorium and jaws, certain com- ments are in order relating to their specializa- tions and functional anatomy. Typical of the pis- civorous adaptations of species of Ophichthus, Brachysomophis, Echiophis and related species are the strengthening provided by the cranial vault, the elongate pterotics, the broad junction of the hyomandibular along the neurocranium, and the bracing of the maxillae to the frontal by means of the fused postorbitals. An analagous condition exists in the muraenid genus Gymno- thorax, in which a postorbital strut strengthens the jaws and suspensorium (cf. Burton, 1956: fig. 6). Other ophichthid genera, particularly among the Myrophinae and elongate ophich- thines, are adapted to diets of minute inverte- brate prey, and have extremely reduced neuro- crania, suspensoria, jaws, and dentition. Opercular Series The opercular series of ophichthids, and of anguilliforms in general, is greatly reduced. This reduction is apparently related to the increase PROCEEDINGS CALIFORNIA ACADEMY OF SCIENCES Figure 16. Maxillary-vomer apposition of several ophichthids. Benthenchelys cartieri enlarged 20 times, all others 9 times. A. Myrichthys xystrurus D. Callechelys marmoratus B. Ophichthus zophochir E. Muraenichthys chilensis C. Phyllophichthus xenodontus F. Benthenchelys cartieri 28 in number and importance of the branchiostegal rays as supporting elements for the branchial cavity (Greenwood, et al., 1966). Gosline (1959) has correlated the reduction of the opercular series, posterior displacement of the gill arches and pectoral girdle, and the separation of the pectoral girdle from the neurocranium with the peculiar branchial pump and circulatory mechan- ism of anguilliforms. Opercular reduction and deossification of cartilage, and the increase in the number of branchiostegals of ophichthids appear to be greatest in the smaller myrophines and elongate bascanichthyins and sphagebranch- ins. Within the Ophichthidae, the condition of O. zophochir closely approximates the general- ized (primitive) state (fig. 14). The opercle (OP), the largest of the series, is posterior to the hyo- mandibular and above the subopercle (SOP), with its ventral margin lying lateral to the dorsal margin of the SOP. The interopercle (lO) over- lays the SOP and is below the OP. The pre- opercle (PO), smallest of the series, overlays the (1O) and contains the tract of the preoperculo- mandibular nerve. The SOP and IO are the least ossified of the series, although the distal margins of most members of the opercular series are car- tilaginous in most ophichthids. Several specializations in each subfamily are useful indicators of phylogeny. For example, the SOP is produced posteriorly as a projection en- closing the ventral and posterior margins of the OP in species of Myrophis (fig. 33), Ahlia, Mur- aenichthys (Gosline, 1951a: fig. 3), Pseudomyro- phis, and Schismorhynchus. This SOP-OP morph- ology is typical of other eel families, including certain Congridae (Asano, 1962; Rosenblatt, 1967), Moringuidae (Trewavas, 1932; Smith and Castle, 1972), Xenocongridae (Gosline, 1950, 1951b; Robins and Robins, 1967), and Xenomy- stax atrarius (Peden, 1972). The opercular series of Callechelys and related genera is reduced and has a conspicuously fringed appearance along the margin. The opercular series of Stictorhinus, Apterichtus, Ichthyapus, and related genera are quite reduced, with elements absent in certain genera. Hyoid Apparatus The hyoid apparatus and the associated branch- iostegals provide fundamental characters which help to unify the subfamilies within the Ophich- thidae. In particular, the broad overlap along the ventral midline of the branchial basket is herein considered a major phylogenetic character of the SERIES 4, V.41,#1 McCOSKER — EELS family, not evidenced by homology or converg- ence in other eel families. The importance of this character, which later led to the combining of the Neenchelidae, Echelidae (in part), and the Ophichthidae, was recognized by Myers and Storey (1939), Gosline (1952), Bertin and Aram- bourg (1958), BOhlke (1960) and Nelson (1966b). The general usage of the term ‘‘epihyal” in the ichthvological literature has been incorrect. As Goodrich (1930: 405-406) has pointed out, the true epihyal is homologous with the hyomandib- ular, and the element incorrectly termed the epihyal represents the posterior ossified element of the ceratohyal. | am in agreement with Good- rich’s conclusions, however in view of its wide- spread usage in ichthyology, | have herein used the term ‘‘epihyal’”’ to represent the posterior of the two ceratohyal elements, and the term ‘‘cera- tohyal” for the anterior element. The following description of the hyoid appara- tus is based on that of O. zophochir (fig. 17A). The apparatus consists of the unpaired glossohyal (GH) and urohyal (UH), and paired upper hypo- hyals (HH), ceratohyals (CH), and epihyals (EH). The interhyal is absent. The outer posterodorsal margin of the EH is connected by cartilage to the inner face of the quadrate and provides sup- port for the branchial basket. The CH and EH are connected by a stout cartilaginous strut af- fording little flexibility along the arch; further strengthening is provided by the flanking spike- like posterior extension of the CH along the outer edge of the EH. The GH is grooved along the posterodorsal half and extends forward as a small cylindrical teat. The GH and paired upper HH are interconnected by a weak cartilaginous plate. The upper HH is fused along a broad and slightly flexible suture to the CH. A ventrolateral extension of the CH braces the HH. The UH, lying ventrad to the HH, GH and CH, is flattened anteriorly and has a short cartilaginous connec- nection from its anterior midpoint to the dorso- ventral margin of the GH. Posteriorly, the UH extends as a slender ossified spike surrounded by a membranous sheath and overlays the branchiostegal rays. All the branchiostegal rays are inserted on the external face of the arch, 6 on the CH, 1 on the cartilaginous interspace, and 19 on the EH. The inner 3 and outer 9 rays are closely grouped basally and the remainder are more widely spaced. Rays 6-16 are expanded slightly at their bases but filiform for the re- mainder of their length. The rays broadly overlap along the ventral midline in a characteristic man- ner. The distal 19 rays of the left EH and cartil- PROCEEDINGS CALIFORNIA ACADEMY OF SCIENCES = Figure 17. Hyoid arch and branchiostegals of Ophichthus zophochir, an ophichthine, and Muraenichthys chilensis, a myrophine. Dorsal view. Scale represents 1 mm. Abbrevia- tions are: BR, branchiostegal rays; CH, ceratohyal; EH, epihyal; GH, glossohyal; HH, hypo- hyal; UH, urohyal. A. Ophichthus zophochir B. Muraenichthys chilensis (posterior-most branchiostegals not illustrated) 29 30 aginous interspace are overlain by all the rays of the right arch, which are in turn overlain by the remaining left CH rays. This pattern of the left CH rays overlaying the right CH and EH rays which overlap the left EH rays is consistent within the Ophichthidae. The location and number of branchiostegal rays among the genera of ophichthids (table 2) and the proximity of the branchiostegal rays to the hyoid arch differs markedly in the ophichthid subfamilies. In the Myrophinae, which appear to follow the generalized anguilliform condition, the branchiostegal rays are attached to the outer face of the EH, often with a single ray on the CH (fig. 17B, Muraenichthys). The remainder of the branchiostegal rays, which will hereafter be referred to as the ‘accessory branchiostegal rays’, are unattached and basally lie well behind the hyoid arch. In the Myrophinae, these vary from as many as 13 pairs in Benthenchelys cartieri to 42 pairs in Myrophis vafer. In the Ophichthinae, all branchiostegal rays are at- tached to the outer face of the hyoid arch( fig. 17A, Ophichthus), although in some species the rays have secondarily become detached. The extreme development of this branchio- stegal apparatus is obviously a means of strength- ening the gill basket. The manner in which sev- eral ophichthines, particularly Ophichthus, Echio- phis, Brachysomophis, and Aplatophis have all the rays attached to the face of the hyoid is probably related to a diet of struggling prey, in contrast to the myrophine condition of free rays and their diet of comparatively weak prey. The reduction of the opercular apparatus and the posterior displacement of the entire gill arch complex necessitates a supplementary skeletal framework to prevent the gill basket from col- lapsing during the normal burrowing activities of ophichthids. The posterior displacement of the gill arches among anguilliform families is extreme in the Ophichthidae and the Moringuidae (see Nelson, 1966a: fig. 58), both of which comprise predominantly sand and mud-burrowing forms. The accessory branchiostegal apparatus of ophichthid and echelid eels was recognized by earlier workers, but Parr (1930) was the first to describe it and suggest its function. He created the term ‘“‘jugostegalia’’ for the accessory skele- ton of the gill cover in species of Myrophis. Be- cause of their number he did not consider the attached rays to be homologous with the bran- chiostegals. In that the accessory rays are un- differentiable from the branchiostegal rays, espe- cially in those ophichthine species in which all SERIES 4, V. 41, #1 McCOSKER — EELS rays are attached to the hyoid, | prefer not to use Parr’s term, jugostegalia. Problems occur in the consideration of jugostegalia as a separate en- tity, even if one assumes them to be homologous with the branchiostegals. For example, in follow- ing Parr’s concept of jugostegalia as those rays which are free from the hyoid, McAllister (1968: 85) stated that “Myrichthys (has) 28 (branchio- stegal rays) plus 4 jugostegalia...Myrophis 5 plus 34-36 jugostegalia.”” McAllister’s (p. 80) re- tention of the term jugostegalia ‘‘for those sec- ondarily multiplied, overlapping and free bran- chiostegals found in certain anguilliforms” is in- adequate as seen in the above usage. | therefore propose that a more flexible terminology be ap- plied to those rays, and prefer the term “‘acces- sory branchiostegal rays’’. The branchiostegal rays also reflect intragen- eric and intergeneric similarities through their basal thickening and secondary multiplication. The outermost rays of all myrophines and several ophichthines are proximally broadened (e.g., Muraenichthys, fig. 17B, and Neenchelys, Nelson, 1966b: fig. 2a). This condition is similar to the general anguilliform condition as displayed by congrids, muraenids, anguillids, and other apodal families. In other ophichthines, particularly Ophi- chthus, Aplatophis, Myrichthys, and Quassire- mus, the branchiostegals are filiform throughout. The basal splitting or secondary multiplication of the few inner and outermost rays is probably secondarily related to the number and spacing of the rays along the arch. For example, in species of Echiophis and Ophisurus the branchiostegals on the CH are paired, and are often fused at their bases. No trends were apparent in this con- dition within the Ophichthidae, hence it appears to be of limited value as a phylogenetic indi- cator. The urohyal of most ophichthines is produced posteriorly as an ossified spike. The UH of all myrophines is limited to an ossified basal plate with cartilaginous posterior filaments. The few ophichthines that lack the ossified spike are clearly derived from the generalized condition typified by Ophichthus, rather than from the somewhat similar myrophine condition. A cur- sory survey of other eel families indicates that the ophichthine condition is primitive in relation to the more specialized myrophine condition. A major subfamilial difference is also evi- denced in the ceratohyal. Without exception, the CH of the Ophichthinae is split into an elongate and pointed distal portion and a shorter, truncate medial portion which connects, by means of PROCEEDINGS CALIFORNIA ACADEMY OF SCIENCES Figure 18. Gill arch skeleton of Ophichthus zophochir, SIO 65-166. Dorsal view. Gill arches have been cut along the dorsal midline and spread laterally; left upper and lower pharyngeal tooth plates are removed to show underlying bones. Stippling indicates cartilage. Scale represents 1 mm. Abbreviations are: By, first basibranchial; Ci, first ceratobranchial; Ei, first epibranchial; H:, first hypobranchial; I2, second infrapharyngobranchial; LP, lower pharyngeal tooth plate; UPs, upper pharyngeal tooth plate. 32 cartilage, with the EH (fig. 17A). The myrophine CH is not split, but is rather a simple bowed bone, distal to, and terminating at about the midpoint, of the EH. The myrophine condition appears to involve a unique reduction, whereas the ophichthine condition is similar to that of the Congridae (Asano, 1962; D. Smith, 1971; Rosenblatt, 1967) and other eel families. The hypohyals of most ophichthines are like those of O. zophochir. In certain otherwise dis- similar genera, including Schismorhynchus, Apro- gnathodon, and Phyllophichthus, the HH are absent. Careful examination of the anterior end of the CH did not reveal a suture or line of fu- sion, so it might be assumed that the HH are lost altogether. Observation of an ontogenetic series might provide proof of fusion or absence. Gill Arches Certain elements of the anguilliform gill arch skeleton have been shown to be important in- dicators of phylogeny (Nelson, 1966a). The Ophichthidae differ from all other eel families in a combination of gill arch characters, includ- ing: a cartilaginous connection between the proximal ends of the dorsal part of the first and second gill arches (according to Nelson, peculiar to the Ophichthidae); first basibranchial either ossified or absent, all others cartilaginous, rudi- mentary, or absent; hypobranchials 1-2 ossified; second infrapharyngobranchial (ls) ossified. If one considers the anguillid or congrid gill arch conditions, that of numerous ossified elements with minor loss or reduction, to be primitive among the anguilliforms, then the ophichthids are considerably advanced in having several os- seus elements replaced with cartilage, and in having others reduced or entirely lost. Extending this supposition within the Ophichthidae, the Ophichthinae, and in particular the Ophichthini, are more primitive than the Myrophinae, which have lost the fifth ceratobranchial (C;) and have reduced or lost certain basibranchials (tables 3-4). The ophichthids are also specialized in having the gill arch skeleton displaced posteriorly in re- lation to the cranium. Among shallow-water eel families this condition is exceeded only in the Moringuidae (see Nelson, 1966a: fig. 58). Refer- ences to gill arches of ophichthid species are limited to Popta’s (1904) pioneering study of apodal gill arches (which treated Muraenichthys gymnopterus, Leiuranus semicinctus, Caecula polyophthalmus, Pisodonophis boro, and Myr- ichthys colubrinus), Nelson’s (1966a) detailed SERIES 4, V.41, #1 McCOSKER — EELS study, which included species from 18 ophich- thid genera, and Nelson’s (1966b) treatment of Neenchelys buitendijki. The following description of the gill arch sys- tem of Ophichthus zophochir (fig. 18) illustrates the presumably primitive condition within the Ophichthidae. Terminology of gill arch elements follows that of Nelson (1969: 480). The _ basi- branchials are single elements, not intercon- nected, lying along the ventral midline; all other gill arch elements are paired. The first basibran- chial is ossified, slender, and connected by cartil- age to the first hypobranchials. Basibranchials 2-4 are cartilaginous and connected to the adjoining hypobranchial pair. Hypobranchials 1-2 are ossi- fied and stout. Hypobranchials 3-5 are cartilagi- nous, with 4 and 5 fused. Ceratobranchials 1-4 are ossified and subequal. Ceratobranchial 5 is reduced to a slender filament which is fused for most of its length with the ventral surface of the ventral pharyngeal tooth plate. Epibranchials 1-4 are short, stout, and bear various processes for cartilaginous or ligamentous attachment. The first infrapharyngobranchial (l:), as in all eels, is ab- sent. The second connects to the first epibranch- ial by a cartilaginous strap; this condition, as mentioned earlier, is peculiar to all ophichthids. The third is “T’’ shaped and distally supports the third upper pharyngeal tooth plate (UP3). The upper pharyngeal plates are separated by a su- ture, the third being much smaller than the fourth (UPs). Each plate bears along one margin a single row of slightly retrorse conical teeth which grades to a fine-toothed pavement. /n situ, the tooth plates overlie each other, and surround the esophageal canal so that the corresponding toothed areas of the upper and lower plates are aligned. The third hypobranchial is cartilaginous in nearly all ophichthids. This specialization prob- ably functions to increase the flexibility of the gill arch skeleton and, in particular, to allow further anterior movement of the lower pharyn- geal tooth plates. For the same reason the fourth hypobranchial is never ossified. The species with ossified third hypobranchials, Dalophis imberbis, Aprognathodon platyventris, and Elapsopis cyclor- hinus, are highly specialized and not closely re- lated. Two examples of each of the latter two species were examined to rule out the possibility of anomalous specimens. Both specimens of Elapsopis had a well ossified Hs pair, and slender, nearly subequal tooth plates bearing conical biserial teeth. The two specimens of Aprognathodon were somewhat aberrant in the PROCEEDINGS CALIFORNIA ACADEMY OF SCIENCES Hs condition. In one, the left Hz was well ossi- fied but the right was cartilaginous. The other had small discontinuous pieces of bone lying within a cartilaginous matrix on both sides. In both specimens the short, stout Hi-2 pairs nearly met along the ventral midline, and the small tooth plates bore several slender pointed teeth. In these exceptional species there has probably been a return to an ossified Hs condition rather than a retention of the primitive ancestral oph- ichthid condition, possibly related to a special- ized diet which would require less flexibility in the gill arch skeleton. Microscopic examination and interpretation of the gill arch elements is often difficult, particu- larly in differentiating the rudimentary and cartil- aginous conditions of the basibranchials. Nelson (1966a: 393), for example, considered the fourth basibranchial condition of Leiuranus semicinctus and Machaerenchelys phoenixensis to differ, al- though most authors have regarded the latter species as a color variant synonymous with the former species. The interpretation of the UP3-UPs fusion also tends to be subjective and is not heavily relied upon. Since anomalies often occur in the gill arch skeleton, duplicate specimens were examined in this study when suspicious re- sults were encountered. Nelson (1966a) identified certain ophichthid lineages on the basis of the gill arch conditions. These comprised (a) those genera with a moder- ately well-developed series of basibranchials and an ossified fifth ceratobranchial, (b) those with C; reduced or cartilaginous, and (c) those with very reduced basibranchials and lacking C;, i.e., the Myrophinae. Nelson erred in considering Echelus myrus to be an echeline (=myrophine) but did recognize the generalized condition of its basibranchials and C;. My study has shown Nelson correct in his other interpretations and now places other genera within his general framework. The functional significance of the modification of the gill arches in apodal fishes was discussed by Nelson (1966a). The posterior displacement of the gill arches, the reduction and modification of osseus elements, and the absence of a firm interconnection with the cranium has_ trans- formed their function from prey catching (now left to the jaws and cranium alone) to one of moving large food items through a secondarily elongated pharynx. As Nelson (p. 404) has pointed out, ‘many of these same modifications have occurred independently among syngnathi- form fishes and symbranchiform fishes . . ., pos- 338) sibly also as a result of spatial separation of jaws and gill arches.” Pectoral Girdle The pectoral girdle varies considerably within the Ophichthidae, grading from a well developed to a reduced condition in both subfamilies. The primitive condition, represented by Ophichthus (fig. 19A) in the Ophichthinae and by Myrophis in the Myrophinae (figs. 20E-F), includes the re- tention of the cleithrum (Cl), supracleithrum (SCI), scapula (=hypercoracoid) (Sc), and cora- coid (=hypocoracoid) (Co), and in the Ophich- thinae, the actinosts. The postemporal is absent and the pectoral girdle is not attached to the cranium in anguilliforms (Gosline, 1971). The pec- toral fin is best developed in genera such as Ophichthus, Echelus, Echiophis, Mystriophis, Ap- latophis, Malvoliophis, Cirrhimuraena, Pisodono- phis, and Pogonophis in the Ophichthinae, and Myrophis and Ahlia in the Myrophinae. The pec- toral fin rays are in some cases (Ahlia and Myro- phis) multiply-branched. Pectoral rays vary in number from three to four weak stubs in Bas- canichthys panamensis to 18 rays in Pogonophis fossatus. The coracoid and scapula generally lie within a cartilaginous plate which provides a rigid sup- port for pectoral fin movement. When present, the pectoral fin base lies in a dorsoventral plane and is broad-based in relation to the fin length (extremely so in Myrichthys), offering little ro- tational movement. Observations of live Echio- phis sp. and Ophichthus triserialis indicate that those ophichthines use their pectoral fins to as- sist the body musculature in making short dart- ing movements during feeding. Observations of live Bascanichthys panamensis suggest that the highly reduced pectoral fin serves little, if any, locomotory function. In those forms lacking pec- toral fins, the cartilage, as well as the Co and Sc, is reduced or absent entirely. The trend in pec- toral girdle reduction is related to the burrow- ing habit of many ophichthids. The primary func- tion of the pectoral girdle as the structural sup- port for the pectoral fin has in burrowing forms become one of structural support for the poster- ior end of the branchial basket. In no species are all girdle elements absent, but in Cirricaecula (fig. 19N) and in species of Ichthyapus only the Cl remains, and in species of Muraenichthys the Cl is a thin curved bone which is nearly identical in appearance and in function to the last of the accessory branchiostegals. The SCI of several 34 Figure 19. \ SERIES 4, V.41, #1 McCOSKER — EELS Pectoral girdle of various representative ophichthine genera. All are shown in lateral view, right side, except Cirricaecula which is viewed ventrally. Abbreviations are: Cl, cleithrum; Co, coracoid; PR, pectoral rays; Pt, pterygiophores; Sc, scapula; SCI, supra- cleithrum. Sie G)) ott Ta SG) Sa) . Ophichthus zophochir . Pisodonophis boro . Elapsopis cyclorhinus . Myrichthys xystrurus Quassiremus nothochir Aplatophis chauliodus . Caralophia loxochila . Bascanichthys panamensis Phaenomonas pinnata Caecula pterygera . Lamnostoma orientalis Callechelys marmoratus . Aprognathodon platyventris . Cirricaecula johnsoni . Apterichtus flavicaudus Yirrkala tenuis PROCEEDINGS CALIFORNIA ACADEMY OF SCIENCES rc at ID Wy, Figure 20. Pectoral girdle of various representative myrophine genera. All are shown in lateral view, right side. . Muraenichthys gymnopterus Muraenichthys chilensis . Benthenchelys cartieri . Ahlia egmontis Myrophis vafer Myrophis uropterus _ Neenchelys buitendijki (from Nelson, 1966b: fig. 2c), pectoral fin rays not illustrated . Pseudomyrophis micropinna ae (@) art (eal (o) (@Wesh 36 species of Muraenichthys is merely a thin sliver (e.g., M. chilensis, fig. 20B) and it is lost alto- gether in Schismorhynchus. A peculiar Co and Sc condition exists in the Callechelyini and several of the Bascanichthyini, representing a unique ophichthid specialization which is not seen in other apodal fishes. It is nearly universal among eels for the Sc to be a nearly round bone (flattened slightly on the posterodorsal edge) lying above the Co (also nearly round but slightly flattened on the post- eroventral edge), one or both of which are fen- estrated. This is also the generalized ophichthine and myrophine condition. The Callechelyini possess either one or two small rod-shaped bones which are connected by cartilage and are oriented horizontally in the normal location of the Co (see Aprognathodon, fig. 19M and Cal- lechelys marmoratus, fig. 19L). Species of Phae- nomonas and Ethadophis, and the sphage- branchin Lamnostoma orientalis are similar in this condition. The homologies of these two rod- shaped bones are not entirely clear, but until further evidence is discovered, | will consider the anterior bone to be homologous with the Co and the posterior one to be homologous with the Sc. Pectoral girdle reduction in Myrich- thys provides an indictation of its generic homo- logies, particularly to species of Pisodonophis. The Sc in all species of Myrichthys is lost and the Co has lost its dorsal curvature (see fig. 19D). This condition is also that of Pisodonophis das- pilotus, a species clearly more similar to other Pisodonophis than to species of Myrichthys in other osteological characters. Further similarities in all species of both genera include the shape of the Cl and SCI. Other Pisodonophis examined have retained the Sc and a complete Co, a con- dition more like that of Ophichthus. The ancestral condition of those genera which entirely lack the Co and Sc is not indicated by remnants of cartilage or bone or transitional species, and is therefore indeterminable. The loss of the Co and Sc in Ichthyapus, Apterich- tus, Cirricaecula and Quassiremus may have been independent, although the first three gen- era are related on the basis of other characters. Lateralis System The apodal lateralis system and associated bones have been shown to be useful indicators of relationship, particularly within the Congridae (Asano, 1962; D. Smith, 1972). Asano showed that the number of pores within the cephalic SERIES 4, V.41, #1 McCOSKER — EELS canal vary within and between genera of Japa- nese congrids. The lateralis system within the Ophichthidae was found to differ in a character- istic manner at the subfamilial and tribal level. It is best developed in the Ophichthinae and re- duced in the Myrophinae. The ophichthid later- alis system has been described and illustrated on several occasions, usually on a species by species basis and not in a comparative manner. These works include: Allis (1903), Ophisurus serpens and Echelus myrus (as Myrus vulgaris); Gosline (1951a), Ichthyapus vulturis (as Caecula_ platy- rhyncha); Hopkirk (1965), Ophichthus zopho- chir; Nelson (1966b), Neenchelys buitendijki; Blache (1968), Echelus myrus, E. pachyrhynchus and Myrophis plumbeus; Blache (1971), Mystrio- phis rostellatus, M. crosnieri, and Echiophis in- tertinctus; and Blache and Cadenat (1971), My- richthys pardalis, Bascanichthys spp., and Cal- lechelys spp. The ophichthid lateralis system is divisible into seven canals: lateral line, supraorbital, infraorbi- tal, preoperculomandibular, temporal, supratem- poral commissure, and frontal commissure. The canals lie either within certain cephalic bones (frontal, pterotic, nasal, preopercle, postorbital, and dentary) or weakly ossified tubes (ossicles) which are broken at short intervals to provide flexibility. The Ophichthidae are distinctive in having the right and left sides of the cephalic lateralis system connected through the frontal and temporal canals. The following description of the lateralis sys- tem of Ophichthus zophochir, a species which shows little reduction and minor specialization, illustrates the general ophichthine condition (fig. 21). The supraorbital canal connects with the infraorbital and temporal series posteriorly, and with its opposite member dorsally across the transverse frontal commissure. A single median supraorbital pore is present. Three supraorbital pores are associated with the nasal, the central pore lying within that bone. The anteriormost ethmoidal pore is connected to the supraorbital canal by a short cartilaginous connection. There are six infraorbital pores. Four lie horizontally beneath the eye in an ossified canal posterodis- tally overlying the maxilla, followed by two verti- cal pores lying behind the orbit and connected to a canal passing through the three postorbital bones. The supraorbital canal connects posteri- orly with the temporal canal and commissure, passing through the frontal and pterotics. A single median supratemporal pore is flanked lat- erally by a pore on each side. The temporal PROCEEDINGS CALIFORNIA ACADEMY OF SCIENCES stp so Figure 21. Cephalic lateralis system and associated bones of Ophichthus zophochir, SIO 60-309. Stippling indicates cartilaginous canal section. Scale represents 1 mm. Abbrevia- tions are: an, anterior nostril; es, ethmoid section of supraorbital pores; FC, transverse frontal commissure; io, infraorbital pores; LL’, anteriomost lateral line pore; N, nasal bone; pm, preoperculomandibular pores; pn, location of posterior nostril; PO, preopercle; por, postorbital pores; POR, postorbital bones; so, supraorbital pore; stp, supratemporal pore. A. Right side view B. Dorsal view 37 38 canal extends posteriorly to the lateral line canal and ventrally to the preoperculomandibular canal. The preopercular section connects by an ossified tube to the preopercle, with two ventrolateral pores passing from the bone. A cartilaginous connection with the third preopercular pore joins the preopercular and mandibular sections. By overlying the quadrate and angular-articular juncture this cartilaginous connection provides the flexibility necessary during jaw movement. The seven pores comprising the mandibular series are unevenly spaced along the dentary. The lateral line (LL) canal extends posteriorly from the preoperculomandibular-temporal canal juncture, loops above the branchial basket, and continues midlaterally to within 0.2 head length of the tail tip. There are approximately 150 LL pores, 51 of them before the anal origin. Lateral line ossicles are moderately ossified, separated at each pore, and open along the distal margin (fig. 22F). On the medial face of all LL ossicles there is a centrally located opening for the nerve. Lateral line ossicle development is reduced pos- teriorly and absent before the last few caudal pores. Suprageneric relationships among ophichthid genera are indicated by repeating pore patterns in certain cephalic canals, particularly in the pre- opercular portion of the preoperculomandibular canal and in the temporal canal (table 5). The generalized ophichthine condition consists of three preopercular pores and a single temporal pore on either side of the median supratemporal pore (as in Ophichthus zophochir, Fig 21, and Ophisurus serpens, Fig. 24B). The third preoper- cular pore (pop*) is lost in certain Ophichthini and all Callechelyini and Bascanichthyini. Cer- tain Sphagebranchini are specialized in having a fourth preopercular pore (pop*) and a second temporal pore (tp’), as in Ichthyapus selachops (fig. 24A). This condition is not uniform through- out the Sphagebranchini, and may vary between and within populations, as evidenced by isolated populations of Ichthyapus vulturis (Randall and McCosker, 1975). The Myrophinae lack tp? and pop*. The pop® is present in Ahlia, Myrophis, Pseudomyrophis, Muraenichthys, and Schismo- rhynchus, but is absent in Neenchelys, Schultz- idia, and Benthenchelys (see Nelson, 1966a, figs. 21-25, in which Schismorhynchus was called Leptenchelys labialis, and Nelson, 1966b, fig. 1a). The lateral line ossicles also indicate relation- ship in their degree of ossification, separation at each pore, and the pore position along the canal. Most pores lie below the midline of the SERIES 4, V.41, #1 McCOSKER — EELS LL canal, although some genera are specialized in having the pores located centrally within the canal. Lateral line ossicles are nearly solid struc- tures in the Sphagebranchini and Callechelyini, less substantial in the Ophichthini, and most re- duced in the Myrophinae (figs. 22-23). Certain genera of the Ophichthini are special- ized in having a well-developed free sensory neuromast system along the sides and top of the head. The elaborate development of these sense organs has been generally overlooked. The neuromasts are not randomly scattered across the head region, but tend to follow distinct pat- terns. Nelson (1972) identified these lines of papillae in esocids as “‘pitlines’’, which are ap- parently homologous to the free sensory neuro- masts described herein. Following his terminol- ogy, the neuromast lines of Ophisurus serpens (fig. 24), beginning at the snout tip, include paired subnasal, antorbital, anterior, and cheek lines, and a single midline crossing the nape. Various degrees of development are also present in species of Ophichthus, Echelus, Pisodonophis, Quassiremus, Cirrhimuraena, Echiophis, Mystrio- phis, and Aplatophis. The neuromasts appear as minute papillae and are often difficult to discern due to skin rugosity and a waxy precipitate that forms on preserved specimens. Neuromast development is probably related to the soft bottom habitat occupied by these spe- cies. Schwartz and Hasler (1966) suggested that LL pore development of the mudminnow Umbra limi is reduced and free neuromasts are devel- oped in response to its habit of digging into soft mud substrate. In doing so, they suggested, the pores of the LL canal could become impacted with mud and severly impaired. The numerous and widely distributed superficial organs, how- ever would remain functional. Rosenblatt and Rubinoff (1972: 362) inferred a similar adapta- tion in a heterenchelyid eel in noting that “the absence of lateral-line pores in P. asodes indicate that it may be a burrower in soft mud or in the semiliquid mud-water interface.” This inverse relation between sensory neuromasts and LL canal development is further evidenced by the ophichthids in that (a) a general correlation ex- ists between neuromast development and a mud, rather than a sand, substrate occupied by the species involved, (b) free neuromast development is absent in the Sphagebranchini where cephalic pore development is greatest, and (c) conversely, the mud-dwelling species of Echelus display ex- treme pore reduction and moderate neuromast development. PROCEEDINGS CALIFORNIA ACADEMY OF SCIENCES me iii a — =' c a é a Paa65) SEC X Mer con SS (ESS Ese aie reenae | aly IE), ow 5 3 : Z D = r] = ASS = Fi noe . rac Os “A opens SE as ie E Figure 22. 3g) nga @ 4 son i —____—————“4 Fey ANTI i e \ (Rv Say go ————$——— L Q AMWASOMegET BAI ee, LIBERA WOR pgs —— s — ae. NOU; Lateral line ossicles of representative ophichthines. Distal view, right side of mid-trunk region. Scales represent 1 mm. Stippling indicates lateral line canal. . Aplatophis chauliodus . Quassiremus nothochir . Phyllophichthus xenodontus . Aprognathodon platyventris Myrichthys xystrurus Ophichthus zophochir . Cirrhimuraena taeniopterus . Paraletharchus pacificus Callechelys eristigmus Pisodonophis cancrivorus = 6 os) ol bal ) eyes oe K. Phaenomonas pinnata Ethadophis byrnei . Ichthyapus selachops . Yirrkala tenuis . Caralophia loxochila Bascanichthys panamensis . Allips concolor Cirricaecula johnsoni Lamnostoma orientalis iE M N O P. Q R. Se T. Elapsopis cyclorhinus 40 SERIES 4, V.41,#1 McCOSKER — EELS _ — hi : = TS OS TY eR a al era © A ey eS eS Casas 2S eS ee a eee : : | a =] = ae = — <- US o = o.2 EOS OSSS oP PSS ee oe 50° Ss inched oe be Co ft eS eS Ea aS Sue rs 09s ed De ee ee a = 4 —— ——= Cc G trey © Hs Ra — ee H Figure 23. Lateral line ossicles of representative myrophines. Distal view, right side of mid-trunk region. Scales indicate 1 mm. Stippling indicates lateral line canal. A. Pseudomyrophis nimius E. Pseudomyrophis micropinna B. Myrophis vafer F. Ahlia egmontis C. Benthenchelys cartieri G. Schismorhynchus labialis D. Muraenichthys chilensis H. Schultzidia johnstonensis PROCEEDINGS CALIFORNIA ACADEMY OF SCIENCES Figure 24. Cephalic pore and surface sensory papillae development in two ophichthids. Abbreviations are: an, anterior nostril; pop!, first preopercular pore; sp, surface sensory papillae (free neuromasts); stp, supratemporal pore; tp', first temporal pore. A. Ichthyapus selachops, SIO 65-290. Dotted lines represent contours, not surface sensory papillae. B. Ophisurus serpens, unnumbered Rhodes University specimen, adapted from Allis (1903), left lateral view. C. Ophisurus serpens, dorsal view. 41 42 Papillae also occur on the snout and tail tip of several ophichthids (cf. Rosenblatt and McCosker, 1970). Their function has neither been examined nor proposed, but might be inferred from ob- servations and analogous structures on other fishes. Tail tip papillae are generally small and may function as contact sensory devices in rela- tion to the tail-first digging behavior of most observed ophichthids. Certain of the snout pa- pillae are often larger (e.g., in Leuropharus lasi- ops and Evips percinctus) and may serve a gusta- tory as well as a contact sensory function. These papillae are best developed on snouts of several of the small-eyed permanent burrowing species (including species of Phaenomonas, Bascan- ichthys, and Ichthyapus). A gustatory function for similar papillae on the snout of a heterenchelid eel might also be inferred from Rosenblatt and Rubinoff’s (1972: 362) description of Python- ichthys asodes. They observed a reduction in the olfactory epithelium and the development of papillae on the jaws of this small-eyed species, and suggested this was related to a fossorial habit. Most species of ophichthids have not re- duced their olfactory epithelium, but probably encounter environmental problems similar to those faced by Pythonichthys in their modes of feeding. Axial Skeleton Regan (1912) considered the axial skeleton to be of major importance in separating eel famil- ies. He separated the Echelidae (considered by Regan to include Echelus, Ahlia, Myrophis, Par- amyrus, Chilorhinus, Muraenichthys, and Eomy- rust) and the Ophichthidae from the Congridae on the basis of the formers’ vestigial neural spines. He further separated the Echelidae from the Ophichthidae on the basis of the weaker ribs of the latter family. Gosline (1951a: 302-303) clarified Regan’s statements in his discussion of the ophichthid axial skeleton. Difficulties in the preparation and dissection of the anterior verte- brae have precluded their usage in this study in a systematically comparative manner. The following description is based on the axial skeleton of Ophichthus zophochir (figs. 25-26). The first vertebral (V) centrum (CE) is reduced and not fused to the skull (fig. 25A). Its neural arch (NA) extends posteriorly over the second V. A lateral flange on the CE is present on the second and following trunk vertebrae. The NA of V 1-5 are smooth. Along the midline of the NA of V 1-12 is a single longitudinal crest which SERIES 4, V.41,#1 McCOSKER — EELS is split at its posterior margin to form two short ridges (fig. 26A). Neural spines (NS) are undevel- oped on the trunk vertebrae but become weakly developed points posterior to the 6th or 7th caudal vertebra. Epineurals (EN), epipleurals (EP), and pleural ribs (PL) extend posteriorly from the NA and parapophyses (P), and are approximately 5-7 V in length. The EN and EP of most ophich- thids begin at the posterior margin of the neuro- cranium. The P of V 1-12 are posteriorly directed and increase gradually in length. At approxi- mately V 13 the P are symmetrical and shaped like normal isosceles triangles. Foramina exist on each P, slightly posterior to mid-centrum (fig. 26B). The P of V 13-45 (approximate) are homogenous in size and shape; the lateral processes of the following 5-6 V are reduced. The first caudal ver- tebra (at which point the haemal arch begins to form) differs markedly in having its P split, the upper portion directed laterally to become the first caudal transverse process (CTP) and the lower directed downward to become the haemal arch (figs. 26C-D). The CTP are sharp lateral pro- jections, incised at their midlines to the centrum, and continuing nearly to the caudal tip. The haemal arch closure occurs at approximately the 10th caudal V. The closure however, is incom- plete, and consists of the joining of the posterior ends of the haemal spines. Intramuscular (IM) bones replace the neural and pleural ribs in the caudal region. Differences in neural arch shape and sculptur- ing are evident in comparing the anteriormost five vertebrae of species of the type genus of each ophichthid tribe (fig. 25). The NA of V 1-5 of members of the Callechelyini can be distin- guished, at the tribal level, on that basis alone. Other characters from the axial skeleton were found to be useful indicators of relationship. The parapophyses of certain sphagebranchin genera, for example, were found to possess a marginal process which was lacking in related genera (fig. 33). Also, the CTP are lacking in most myrophins, yet in Muraenichthys and related genera the an- terior half of the column is similar to the ophich- thine column. Finally, the characteristics of the pleural ribs of Ahlia and Myrophis were found to differ from that of all other ophichthids in that they are limited to the anterior 15-20 vertebrae (see Remarks concerning Ahlia and Myrophis). Vertebral numbers have been shown to be useful characters for the separation of species and populations of apodal fishes. Their applica- tion to the separation of genera is somewhat difficult because of the high degree of overlap PROCEEDINGS CALIFORNIA ACADEMY OF SCIENCES Figure 25. Anteriormost five vertebrae of representative species from the tribes of ophichthids. All are shown in left lateral view. Scale represents 1 mm. Abbreviations are: Ce, centrum; EN, epineural; NA, neural arch; P, parapophysis; PL, pleural rib. A. Ophichthus zophochir D. Callechelys marmoratus B. Stictorhinus potamius E. Myrophis vafer C. Bascanichthys panamensis F. Benthenchelys cartieri 44 SERIES 4, V. 41, #1 McCOSKER — EELS CTP Figure 26. Trunk and caudal vertebrae of Ophichthus zophochir. Arrows point anteri- orly. Abbreviations are: Ce, centrum; CTP, transverse processes of caudal vertebrae; EN, epineural; IM, intramuscular bone; NA, neural arch; NS, neural spine; P, parapophysis; PL, pleural rib. A. B. Cc Anterior view of 14th vertebra. Ribs appear foreshortened due to viewing aspect. Ventral view of 14th-16th vertebrae. Dorsal view of last precaudal (51st) and anterior five caudal vertebrae (52nd-56th). Ribs and IM bones not illustrated. Ventral view of vertebrae illustrated in C. Ribs and IM bones not illustrated. PROCEEDINGS CALIFORNIA ACADEMY OF SCIENCES between genera, and a single mean value cannot be applied to a genus. Trends, however, are present within genera, and are probably relatable to the mode of life and associated anatomical specializations of the species involved. The spe- cies of Phaenomonas, Allips, and Bascanichthys, for example, tend to have increased vertebral numbers, primarily in the trunk region. An exam- ination of radiographs and gut contents of speci- mens of those genera disclosed the presence of copious sand and gravel particles in the gut and intestine, and the absence of any recognizable macroscopic animal material. On that basis as well as my observations of live specimens, | sug- gest that these eels indiscriminately eat their way through the substrate, digesting any utiliz- able organic material they encounter. In many ophichthids the length of the gut is increased by a loop that extends into the tail portion. The gut of species of Phaenomonas and Bascanichthys is straight, presumably to prevent blockage by sand particles passing through the lower tract. The in- creased trunk length, which is typical of these genera, perhaps reflects this problem, and may have been necessary to achieve this feeding mode. Certain callechelyins exhibit a similar in- crease in trunk vertebrae and a comparable life style. Vertebral number is also the basis of correlat- ing the pelagic leptocephalus with the trans- formed adult stage of each eel species. Included in table 6 are the vertebral numbers of eel spe- cies examined and radiographed in the course of this study as well as several literature records which are assumed to be correct in species identification. The literature concerning numbers of eel vertebrae has not been exhaustively searched in that errors may inadvertently have been introduced through improper identification. Caudal Skeleton Differences in the myrophine and ophichthine caudal fins seemed important enough to most earlier authors to recognize the lineages as dis- tinct families. The Ophichthidae of nineteenth century authors was indeed a unique and unified assemblage; primarily due to the conspicuously pointed tail tip. Important members, most not- ably the species of Echelus, were erroneously ex- cluded from the Ophichthidae because they possessed a weakly developed caudal fin. Gos- line (1951a: 303) noted the similarity in the oph- ichthine and myrophine caudal skeletons, but felt that the continuous median fin condition 45 merited subfamilial separation. His findings, to my knowledge, have not been questioned by subsequent authors. He stated that: osteologically, the difference between the tails of Muraenichthys and Cirrhimuraena is less than that between those of Cirrhimur- aena and Caecula platyrhyncha. Rudiment- ary rays are present around the tails of both Muraenichthys and Cirrhimuraena; they are embedded in flesh in Cirrhimuraena (as also in Myrichthys). In Caecula platyrhyncha, on the other hand, there are no rudimentary rays either around the tip of the tail or else- where. It is obvious from this discussion... that a separate family cannot be maintained for Muraenichthys on the basis of tail struc- ture. The findings of this study are in agreement with Gosline’s. Difficulties in the dissection and prep- aration of the caudal skeleton has precluded its wide usage in this study. Careful examination of certain species however (including Echelus my- rus, E. pachyrhynchus, Leptenchelys vermiformis, and Bascanichthys tenuis), has clarified their position within the family. The homologies of ossified elements within the apodal caudal skeleton are difficult to de- termine, and especially so in the case of the sharp-tailed ophichthins and sphagebranchins that have undergone major modification as an adaptation to rapid burrowing. The caudal tip of Ophichthus zophochir (fig. 27) is pointed, hard, and without visible caudal rays. The median fins submerge shortly (approximately one eye dia- meter) before the caudal tip. The underlying osteology is complex, as is illustrated in Figure 27. According to the terminology of Rosenblatt (1967), which was adapted from Nybelin (1963), there is but one weak centrum (CE). Fused to the CE is a pointed hypural (HY) and a much reduced neural arch (NA). Reduced caudal rays (CR) which lack basal elements are imbedded in the skin and weakly associated with the HY. The caudal skeleton of Myrophis vafer (fig. 28) differs somewhat from that of O. zophochir in possess- ing a short CE, two elongate HY, and a short cartilaginous extension posterior to each HY. The caudal rays of Myrophis are split anteriorly to receive the hypural plate. This is similar to the condition of caudal rays of xenocongrids (Robins and Robins, 1967), but appears to differ from the relatively unspecialized condition of Anguilla (Smith and Castle, 1972: fig. 19a). Blache’s illus- tration (1968: figs. 5, 10) of the caudal skeleton of Echelus indicates a caudal ray attachment sim- ilar to that of Myrophis. Present in most apodal 46 SERIES 4, V.41, #1 McCOSKER — EELS Figure 27. Caudal skeleton of Ophichthus zophochir, SIO 65-166. Scale represents 1 mm. Abbreviations are: AR, anal ray; B, basal element of pterygiophore; CE, centrum; CR, caudal ray; DR, dorsal ray; HA, haemal arch; HY, hypural; IM, intramuscular bone; NA, neural arch; R, radial element of pterygiophore. PROCEEDINGS CALIFORNIA ACADEMY OF SCIENCES Figure 28. Caudal skeleton of Myrophis vafer, SIO 68-242. Scale represents 1 mm. Ab- breviation CX is for cartilaginous extension of terminal vertebra. 47 48 caudal skeletons is a conspicous foramen be- neath the terminal centrum. It is well-developed in Myrophis, but reduced or absent in Ophich- thus and other ophichthines, resulting from the reduction of the lower HY. The haemal aches (HA) of the posterior caudal vertebrae of Ophich- thus and other ophichthines differs from that of Myrophis and other myrophines. The ophich- thine condition appears to be that of a simple rectangular lateral plate, whereas in the myro- phines a wide gap separates the HA into an an- terior and a posterior lateral flange. The presence of a myrophin-like caudal fin in species of Echelus and Leptenchelys requires further explanation. As stated above, rudiment- ary fin rays are present in the sharp-tailed ophichthines. In the discussion of the evolution of the Ophichthidae it is hypothesized that the elongate bascanichthyins separated early in the evolution of the Ophichthinae, and in general, they possess a blunt rather than extremely sharp pointed tail tip. The tail of Bascanichthys tenuis is surrounded by a weak epidermis, which with- out careful inspection gives the appearance of a rayed caudal fin. The caudal tip of B. tenuis, when viewed with transmitted light, was found to lack caudal fin rays. The type and only known specimen of Leptenchelys vermiformis is similar to B. tenuis in possessing loose epidermis at the caudal tip, although minute fin rays appear to be present. The fin ray development in this juv- enile specimen may be anomalous, or may repre- sent a redevelopment of the rudimentary fin rays characteristic of the ancestral condition. The caudal fin of Echelus myrus, in contrast to the bascanichthyin fins, has well developed fin rays. The caudal skeletons of Ophichthus zophochir and £. myrus do not markedly differ other than in the development of fin rays. The produced rays appear to be a primitive retention of an ancestral condition, whereas the hard-pointed tail tip of other ophichthines was developed early in the evolution of the family. Other primi- tive morphological characters of Echelus that bear similarities to the generalized ophichthines and certain congrids would suggest that Echelus is a primitive ophichthid not far from the basal ophichthine stock. The caudal skeletons of several ophichthids have been illustrated by earlier authors. Included are: Benthenchelys cartieri (Castle, 1972); Echelus myrus, E. pachyrhynchus, and Myrophis plumb- eus (Blache, 1968); Mystriophis rostellatus, M. crosnieri, and Echiophis intertinctus (Blache, SERIES 4, V. 41, #1 McCOSKER — EELS 1971); Myrichthys pardalis, Bascanichthys spp., and Callechelys spp. (Blache and Cadenat, 1971); and Muraenichthys cookei and Cirrhimuraena macgregori (Gosline, 1951a). Visceral Anatomy The digestive tract and gas bladder have been shown by Asano (1962) to be useful taxonomic characters within the Congridae. This study is concerned primarily with osteology and the soft anatomy was therefore not examined in a sys- tematic manner. A cursory examination of a myrophine, Myro- phis vafer (SIO 68-286, 240 mm TL), and an ophichthine, Ophichthus zophochir (SIO 65-166, 335 mm TL), disclosed very similar digestive tracts and gas bladder morphologies (fig. 29). The digestive tract in both species includes a gut diverticulum, or stomach (fide D. Smith, 1971), which branches off the anterior trunk region and extends posteriorly as a blind sac. The intestine, in both species examined, is a straight tube opening directly into the anus, whereas in other ophichthids it appears to extend partially into the caudal region and then to loop forward to the anus. The gas bladder (GB) connects anter- iorly to the intestine through the pneumatic duct (PD) at the mid-trunk level. The GB of both species is thin walled, surrounded by a thin mesentery, and lies alongside the intestine and dorsally within the peritoneal cavity. The GB is white, shiny and flexible; the digestive tract is pale in both species. The gas bladder of the above-mentioned spe- cies occupies little of the peritoneal cavity, as might be expected from their fossorial habits. The GB of the pelagic species Benthenchelys car- tieri however, is considerably longer than that of either of the two fossorial species (Castle, 1972; fig. 25). Its length is approximately one- third of the trunk length, whereas the other two are approximately one-fourth to two-ninths. The GB of congrid eels is longer than the stomach (cf. Asano, 1962; D. Smith, 1971). The GB of the three ophichthids, by contrast, is consider- ably shorter than the stomach. TAXONOMY The following section includes an osteological definition of the family, a dichotomous key for the identification of genera, a diagnosis of the subfamilies and tribes, and an osteological and external morphological description of each genus of the Ophichthidae. PROCEEDINGS CALIFORNIA ACADEMY OF SCIENCES 49 Cc Figure 29. Comparative anatomy of congrid and ophichthid digestive tract and gas bladder. The gas bladder has been separated from the intestine and the mesentery removed to improve clarity. The gas bladderx is stippled for identification, not to indicate pigmentation. Not drawn to scale. Abbreviations are: A, anus; E, esophagus; GB, gas bladder; H, heart; |, intestine; PD, pneumatic duct; S$, stomach. A. Ophichthus zophochir, SIO 65-166 B. Benthenchelys cartieri, after Castle (1972) C. Conger myriaster, after Asano (1962) The generic key is constructed using both osteological and external morphological char- acters. An attempt is made to group genera with- in the key in a natural manner to indicate rela- tionships. In this section, redundancy is avoided wher- ever possible, however in several instances im- portant characters are repeated both in the tribal diagnoses and generic descriptions to facilitate comparisons. The generic descriptions are based, whenever possible, on the type species as well as the most morphologically divergent species within each genus, in an attempt to include the range of variation for each character examined. In a few instances the type species of the genus was unavailable for osteological study. Those genera are identified in the remarks section fol- lowing each description. The included nominal species of each genus are listed under the head- ing “distribution”. Those species known to me only from literature records are indicated by an asterisk (*). Abbreviations of several morphological char- acters and conditions are included for the sake of brevity. The reader is referred to the listing of abbreviations in the Materials and Methods of this paper. Also note that Body = Head + Trunk when used in body and tail length comparisons. The symbol = means ‘‘approximately equal to”. Osteological Definition of the Ophichthidae From the present study the following osteo- logical definition of the Ophichthidae may be developed: (a) branchiostegal rays numerous and broad- ly overlapping along the ventral midline; (b) supraorbital canals united by a trans- verse commissure through the fused frontals; (c) temporal canal present; (d) frontals of adults fused for their entire length and lacking an obvious suture; (e) first epibranchial connected by a con- tinuous cartilaginous strap to the second infra- pharyngobranchial; (f) no more than first basibranchial ossified; (g) third hypobranchial usually cartilaginous; (h) neural spines rudimentary or absent; (i) tongue adnate; (j) palatine absent; (k) pterygoid well separated from vomer and generally free from hyomandibular. Analytical Key to the Genera of Ophichthidae la. Accessory branchiostegal rays originate be- hind ends of epihyal (EH), free rays more numerous than attached; caudal fin rays conspicuous, confluent with dorsal and anal, tail tip flexible; gill openings (GO) mid-lateral, a constricted opening .............. pe eee ey tee ae IMMWAROVOUIMUIMAG sacensee teat een 1b. All branchiostegal rays originate either in association with hyoid or before level of EH tips; free rays, when present, fewer than attached; tail tip a hard or fleshy fin- less point; GO mid-lateral to entirely ven- tral, un-constricted _...... Ophichthinae........ 9 2a. 2b. 3a. 3b. 4a. 4b. SERIES 4, V.41,#1 McCOSKER — EELS Neurocranium short, pointed anteriorly, broad posteriorly, length/depth = 3; eye large, ca. 6 times in head, orbital foramen large, its depth ca. 0.5 skull depth; an- terior nostril non-tubular; (posterior nostril before eye; pectoral fin moderately de- veloped) ........ Benthenchelyini .................... SRR er ee Oe, NAP ROR Benthenchelys Neurocranium more elongate, length/depth = 4; eye smaller, 10 or more in head, orbital foramen smaller, its depth much less than 0.5 skull depth; anterior nostril TUMOUR cece cccssnccee My roonitiiip eens 3 Pleural ribs absent behind 15th-20th trunk vertebra; pectoral fin well developed ...... 4 Pleural ribs present on all trunk vertebrae; pectoral fin either absent or moderately developed)... eee 5 Vomerine teeth absent; dorsal fin origin (DFO) above or behind anus; maxilla stout, not tapering posteriorly, and abutting ptery- goid (fig. 30A); hypohyals (HH) absent, glossohyal (GH) rudimentary ............ Ahlia Vomerine teeth present; DFO anterior to mid-trunk region; maxilla thin and tapering posteriorly, not closely associated with pterygoid (fig. 30B); HH separated from Figure 30. Vomer, maxillae, and pterygoid of Ahlia egmontis (A) and Myrophis vafer (B). Scale represents 1 mm. Abbreviations are: MX, maxillae; PG, pterygoid. PROCEEDINGS CALIFORNIA ACADEMY OF SCIENCES 5a. 5b. 6a. 6b. ae 7b. 8a. 8b. 51 Figure 31. ceratohyal by a gap, GH normally devel- oped Pectoral fin present, coracoid (Co) and sometimes scapula (Sc) present; posterior nostril lateral; transverse processes of Cau- Galieventebraen (GiiR)malbSemts see 6 Pectoral fin absent, girdle reduced to clei- thrum (Cl) and supracleithrum (SCI); pos- terior nostril labial; CTP present —.......... 7 Pectoral fin minute, Sc and actinosts ab- Sant 255s ee eeeaee Pseudomyrophis Pectoral fin well-developed, Sc, Co, and AGCMIMOSUS PORESEMN sce Neenchelys Teeth absent on vomer, absent or em- bedded on intermaxillary, those on max- illa and dentary minute or villiform; supra- occipital (SO) extends anteriorly to frontals, completely separating parietals; SO crest Al SC Il er ie NP RCE tc Schultzidia Teeth present on intermaxillary, maxilla, dentary, and vomer; SO not extending be- VONGmpanletalSaSOmcrest pnesenity asses 8 A prominent toothed groove on underside of snout, bordered by dermal folds, extend- ing anteriorly to anterior nostrils; hypohyals (HH) fused to ceratohyals (CH); suspen- sorium forwardly inclined ....................-------- Underside of snout without a prominent median toothed groove bordered by der- mal folds; HH broadly separated from CH by a suture; suspensorium nearly vertical Muraenichthys 9a. 9b. 10a. 10b. Wile 11b. 12a. 2b: Diagrammatic representation of a species with well developed head pores. Neurocranium short, terete, length/depth ca. 3 or less; dorsal fin origin (DFO) on nape, above supraoccipital (SO); head pores reduced pop’, pop‘, and tp? absent (fig. 31); hyoid stout, thickened; (gill open- ings (GO) inferior, parallel or converging forward, isthmus narrower than GO length; pectoral fin absent) ...... Callechelyini -..... 10 Neurocranium longer, length/depth 4 or more; DFO, if present, behind nape; head pores generally not reduced, may include pop’, pop*, tp?; hyoid more slender ...... 14 Intermaxillary teeth absent; hypohyals (HH) absent; third hypobranchial (Hs3) ossified -. Aprognathodon Intermaxillary teeth present; HH_ broadly separated from ceratohyal by a suture; Hs Cantilaeiim@ USa tere en ee oe eee lit Anterior nostril rim not raised; dorsal fin origin (DFO) above epiotics; neurocranium slightly depressed, not convex across pari- etal-frontal region; four supraorbital pores; (anal fin absent; snout not grooved) ........ Letharchus Anterior nostril tubular; DFO above supra- occipital; neurocranium rounded across parietals and frontals; three supraorbital ONES) 7 at es veceen se tees eee eee ath one eae carareee 72 Anal fin absent; gill openings (GO) ex- panded ventrolaterally, forming broad DOGKEtS iis settee ee eee Paraletharchus Anal fin present; GO only slightly ex- panded, not forming broad pockets ...... 433 ho SERIES 4, V.41, #1 McCOSKER — EELS Figure 32. Representation of underside of callechelyin snouts. A. Median groove pre- sent (13a in key), as in Callechelys. B. Median groove absent, as in Letharchus. Figure 33. Diagrammatic representation of posterior trunk vertebrae (37th), ventral view. A. Parapophyses with anterior projection (16a in key), as in Apterichtus. B. Anterior projection absent (16b in key), as in Stictorhinus. PROCEEDINGS CALIFORNIA ACADEMY OF SCIENCES 13a. 13b. 14a. 14b. 15a. 15b: 16a. 16b. lias 17b. 18a. 18b. 19a. Vomer toothed; median groove on under- side of snout (fig. 32A); snout and nape not heavily papillate -...............-- Callechelys Vomerine shaft toothless; no median groove on underside of snout; snout and nape with numerous papillae ..Leuropharus Pectoral fin absent or vestigial; pectoral girdle reduced, consisting of a cleithrum (Cl), and may include a_ supracleithrum (SCI) and reduced scapula (Sc) and cora- coid (Co); median fins reduced or absent ok 8 2 ae ee ere eo 15 Pectoral fin present, generally well devel- oped; pectoral girdle consists of Cl, SCI (except in Scytalichthys), and generally Sc, Co, and actinosts; median fins generally elevated .............. @phirchithinieeee 29 Neurocranium depressed and_ elongate, length/depth = 4; head pores developed, tp? and pop® generally present; gill open- ings (GO) entirely ventral (except in Yir- nkalal\ ee Sphagebranchini Neurocranium deeper and shorter, length/ depth = 4; head pores reduced, tp*® and pop? absent; GO low lateral, crescentic -... Bascanichthyini All fins absent; parapophyses of posterior trunk vertebrae with an anterior marginal projection (fig. 33a); pectoral girdle re- duced to a cleithrum and a reduced or ab- sent supracleithrum; branchiostegal rays few, generally fewer than 20 pairs; second basiloranchiall(B>)) absent 22222 all Median fins present; anterior margin of posterior trunk vertebral parapophyses en- tire (fig. 33b); pectoral girdle includes cleithrum, supracleithrum, and reduced scapula and coracoid; branchiostegal rays more numerous, more than 20 pairs; Be CaltilalimOUse ee eet * 19 Upper pharyngeal tooth plates (UP; and UP.) fused; cirri present on upper lip ........ Cirricaecula UP; and UPs separate; upper lip smooth.... oe eet irae ee OR a 18 Posterior nostril opening outside mouth, with a flap; anterior nostril tubular; eye moderately developed .............. Apterichtus Posterior nostril opening inside mouth, with or without a flap; anterior nostril flush with snout; eye minute .............. Ichthyapus Lateral head profile, from above, narrows sharply from epiotics to interorbital, then extends evenly to a pointed snout; body 19b. 20a. 20b. 21a. 21b. Daas 22b. Disa 23b. 24a. 24b. DE), 25b. 53 stout, its depth less than 30 in its length; vomerine teeth enlarged, pointed and re- curved Lamnostoma Lateral head profile narrows evenly from epiotics to snout; body moderately elong- ate, its depth more than 40 in its length; vomerine teeth conical, not enlarged ....20 Eye minute, = 5 in snout; anterior nostril flush along snout; interopercle (IOP) absent 4 ens tns 5 Men, SC Re Sa er eae aE Stictorhinus Eye larger, = 3 in snout; anterior nostril tubular, or with a short but noticeable higmcal@© PRE TONES Citi ete oe ee 21 Neurocranium nearly flat across parietals and epiotics; gill openings (GO) with an anterolateral duplication forming a pouch; accessory branchiostegals loosely attached to hyoid, fewer than half associated with epihyal (EH); interopercle (IOP) subrec- tangular, margin entire -................. Caecula Neurocranium raised along dorsal midline, not broad and flat across parietals and epiotics; gill membrane without a duplica- tion; accessory branchiostegals closely as- sociated with hyoid, more than half as- sociated with EH; IOP rounder, serrated along margin Yirrkala Tail short, .300-.360 of total length (TL); body extremely elongate, its depth ca. 75- NGO tina Syw linia eget eee ee eee eee 23 Tail longer, .395-.530 of TL; body not ex- tremely elongate, its depth usually less than. OWI ieee ee eee eee 24 Dorsal fin originating just behind occiput and ending less than 2 head lengths behind gillvopeningsssanall finmlackinigy: eee Bee URE Seo RI shad Lean Aaa Phaenomonas Vertical fins low, but extending nearly to et atid pp ee oe oe « Hanet ke ge Gordiichthys Pectoral fin absent; supraoccipital crest (SOC) extends from a parietal ridge, be- coming a raised point posteriorly _.......... 25 Pectoral fin a minute flap in upper gill opening corner; SOC nearly rounded, little OFMNOMDOStCHOR DO tee ees 28 Dorsal fin origin (DFO) behind gill open- ings (GO); tail longer than body; gill arches stout, third hypobranchial (Hs3) ossi- fied, fifth ceratobranchial (C;) a slender os- SifiedexhOch es ees eee ae et le Dalophis DFO above or behind GO; body = tail; gill arches reduced, Hs cartilaginous, C; ab- sent 54 26a. 26b. 30b. 31a. 31b. Bae 32b. 33a. 33b. 34a. 34b. B5ar Anterior nostril not tubular, its rim not raised, developed as an opening with lat- eral projections into it; underside of snout not grooved; intermaxillary teeth incon- SDLCUOQUSS ae. pee Fen ee Caralophia Anterior nostril tubular; underside of snout grooved; intermaxillary teeth conspicuous Pt ted Eh ee re Ae 27 Median fins continuous around caudal, caudal fin rays evident ............ Leptenchelys Caudal tip blunt, finless -........... Ethadophis Dorsal fin origin on head ....Bascanichthys Dorsal fin origin more than a head length behindeuhiecadi asta aes Pee eA ey Allips Fifth ceratobranchial (Cs) absent ............ 30 Cs present as a slender rod, either ossified OT CantilasinOUs essa ee ee eee 34 Third preopercular pore (pop*) present; pectoral fin rudimentary, smaller than eye; pectoral girdle reduced to cleithrum and SUpraclenthmnUinne ee ee Quassiremus pop® absent; pectoral fin well developed, longer than eye; scapula and coracoid of pectorallcirdlem present, —— 31 Hypohyals (HH) absent; maxilla with a for- ward projection, articulated ca. mid-vomer; supraoccipital (SO) rounded, lacking a pos- terior projection; urohyal (UH) deeply notched anteriorly; anterior nostrils with conspicuous leaflike appendages ................ Fie Bitte ad OG Soe Se Cree Phyllophichthus HH present, separated from ceratohyal by a suture; maxilla without anterior projec- tions, articulated before mid-vomer; SO with a posterior projection; UH not notch- ed beyond midpoint of basal plate; an- terior nostrils without leaflike appendages Pnnenn ann es. Sones $F ONS ol | a AA ey 32 Jaws subequal; upper pharyngeal tooth plates (UP3-UP4) separate _........ Pogonophis Lower jaw inferior; UP3-UP4 fused ........ 33 Third hypobranchial (Hs) ossified; actinosts present; vomerine teeth present ..Flapsopis Hs cartilaginous; actinosts absent; vomer- ines absent, or 1-3 small teeth ....Leiuranus Teeth molariform or granular; pectoral fin broad-based (fig. 34A) Teeth pointed; pectoral fin base restricted, opposite upper half of gill openings (fig. AID) Wes: be 2 oes amd lee ee 36 Dorsal fin origin above or behind gill open- ings (GO); third preopercular pore (pop?) usually present; hypohyals (HH) narrowly separated from ceratohyal (CH); supraocci- 35b. 36a. 36b. 37a. 37b: 38a. 38b. 39a. 39b. 40a. 40b. 41a. 41b. 42a. SERIES 4, V.41,#1 McCOSKER — EELS pital (SO) with a posterior projection ........ Soa ae see et Re ER Pisodonophis DFO well in advance of GO; pop® absent; HH broadly separated from CH; SO rounded, without a posterior projection... Ce ne el a atresia A cman eee Myrichthys Eye before middle of upper jaw, preorbital portion of neurocranium not extending be- yond posterior 2/3, rostral portion of eth- moid shorter than orbit; some teeth long and_ fanglike Eye over middle of upper jaw, preorbital portion of neurocranium extends nearly to or beyond middle of skull; rostral portion of ethmoid about equal in length to orbit; teeth inot aan elie eee ne 42 Lower jaw projects considerably; anterior teeth of both jaws long fanglike canines ex- tending far outside mouth; frontal crest a conspicuous sharp ridge ~.......... Aplatophis Lower jaw inferior or jaws nearly subequal; anterior teeth in jaws not fangs extending beyond snout tip; frontals not forming a sharp ridge, neurocranium rounded or flat dorsally Tail longer than body, compressed posteri- orly; pectoral developed, 5 or less in head length; third preopercular pore (pop’) pres- QNiti «(fee ee 39 Tail shorter than or nearly equal to body; pectoral reduced, 7 or more in head leneth a pOp> absentee 40 Snout short, 7-12 in head length; second- ary cephalic papillae absent ........ Echiophis Snout longer, 6 or less in head length; secondary cephalic papillae well developed sh oe gee LY i a Mystriophis Postorbitals strongly developed, forming a postorbital strut; branchiostegals fewer than 20; postorbital region with a conspicuous transverse depression; lips fringed; colora- CON PAU ieee Brachysomophis Postorbitals moderately developed, not forming a_ strut; branchiostegals 20 or more; dorsolateral profile of head even; lips entire; body spotted —...................... 41 Pectoral fin minute, more than 8 in head length; body much longer than tail ............ wine Es 9 reise Trad Freee te tS ae See Scytalichthys Pectoral fin better developed, ca. 7 in head length; body and tail nearly subequal -....... sik ct eRe ote fs Se Se EE in Xyrias Pectoral fin rudimentary, =~ eye; body longer than tail; (third preopercular pore PROCEEDINGS CALIFORNIA ACADEMY OF SCIENCES Figure 34. Diagrammatic representation of head and pectoral fin of two ophichthins. A. Pectorals broad-based (34a in key), as in Myrichthys. B. Pectoral base restricted (34b in key), as in Ophichthus. 55 56 42b. 43a. 43b. 44a. 44b. 45a. 45b. 46a. 46b. SERIES 4, V. 41, #1 McCOSKER — EELS Figure 35. in key). Fl OSCE LG hehe aoe ee aw eat oe act Selene oe A IR a Evips Pectoral fin developed, noticeably longer than eye; tail longer than body ............ 43 Caudal fin present, confluent with dorsal and anal; temporal, postorbital, and inter- Ofbitallporeswalbsentsse ee Echelus Tip of tail a finless point; temporal, post- orbital, and interorbital pores present ....44 Dorsal fin origin (DFO) before gill open- ings (GO); third preopercular pore (pop®) absent; upper pharyngeal tooth plates (UP3;- UP.) fused; pectoral girdle reduced, scap- ula (Sc), coracoid (Co), and actinosts ab- SC Mitac ete es! sad seks he Malvoliophis DFO behind GO, or if before, the upper lip is fringed; pop’ usually present; UP3s- UP, separate; Sc, Co, and actinosts present So UTE ee eI OS ee en ea PE ee Eee 45 Snout very long, attenuate, ethmoid/neuro- cranium = .500; jaws slender and elongate, incapable of closing completely in adults .. Ophisurus Snout moderate or short, ethmoid/neuro- cranium < .500 jaws not slender and elongate, capable of closing completely..46 Upper lip not fringed, although a barbel may be present; dorsal fin origin (DFO) behind gill openings (GO); opercular series stout, not weak and serrated along margin; aeons wistrlly S4) Ophichthus Upper lip fringed (fig. 35); DFO generally on head, or above GO; opercular series weak, subopercle reduced; actinosts 1-2... Cirrhimuraena Diagrammatic representation of an ophichthin with a fringed upper lip (46b Kaup’s Genera Kaup published his generic and specific de- scriptions twice in 1856. The earlier treatment, “Ubersicht der Aale’’ (1856a), is mentioned by John Edward Gray in the preface of the second, the Catalogue of Apodal fish in the Collection of the British Museum (1856b), published on 30 December 1856. Gray stated that ‘During the printing of the work and the engraving of the plates . . . a synopsis of the genera and new species has been published by Dr. Kaup, in Ger- man, in the Archiv. fur Naturgeschichte, xxii. 41, 1856." Kaup altered several generic names in the latter work (Echiophis became Echiopsis, Ophisurapus - Ophisuraphis, and Pisodonophis - Pisoodonophis), which has resulted in variant spellings of these generic names. Bleeker (1865) emended the spelling of several of Kaup’s gen- era, but Jordan (1919b) returned to Kaup’s earlier work. In the following listings of generic synony- mies, the pagination of Kaup’s later work (1856b) will follow that of the earlier, set off in paren- theses. Type species were not designated by Kaup for his numerous and short-lived genera. Bleeker (1865), as first reviser, synonymized most of Kaup’s genera before types were designated. Jordan seems to have been the first to designate types for Kaup’s invalid genera, and it appears that in each case the first species listed by Kaup was regarded as the type species (cf. Jordan, 119 22))r Subfamilial and Tribal Diagnoses and Generic Descriptions PROCEEDINGS CALIFORNIA ACADEMY OF SCIENCES Subfamily Myrophinae DIAGNOSIS: GO mid-lateral, a constricted open- ing; DFO behind mid-trunk; caudal fin rays not reduced, externally visible, confluent with dor- sal and anal, tail tip flexible; nasals cartilaginous or absent; ceratohyal not divided into a short median and a long distal portion (fig. 17B); only basal plate of urohyal ossified, posterior exten- sion cartilaginous; accessory branchiostegal rays originate behind tips of epihyal, free rays more numerous than attached; branchial skeleton re- duced, basibranchials generally limited to first, fifth ceratobranchial absent; coloration uniform or darkened dorsally. Tribe Benthenchelyini TYPE GENUS: Benthenchelys Fowler, 1934 DIAGNOSIS: Body moderately elongate, laterally compressed behind head; tail much longer than body; anterior nostril not tubular, posterior nos- tril lateral, before center of orbit; GO lateral, a horizontal ellipse; median fins elevated; pectoral fin moderately developed; head pores enlarged, a single preopercular pore, frontal commissure weakly developed; LL ossicles fragmentary, near- ly absent; neurocranium short, rounded (fig. 4); otic bulla, nasals, and SOC absent; maxilla broad, not produced posteriorly (fig. 16), articulating beneath anterior margin of orbit; gill arches weakly ossified, B: cartilaginous, Hs ossified, UP3;-UP, fused; pectoral girdle moderately de- veloped, SCI, Cl, Sc, and Co present; IM bones and ribs weakly developed, CTP absent; epi- pleurals limited to anterior 14-16 vertebrae; ver- tebrae distinctive (fig. 25), neural arches promi- nent; caudal more numerous than precaudal ver- tebrae; coloration uniform, slightly darker dor- sally. Other characters those of the single genus. REMARKS: Benthenchelys cartieri was described by Fowler (1934) and referred to the Derichthy- idae (also Beebe, 1935). Gosline (1952) referred it to the Congridae. Subsequently, Castle (1972) has recognized it as an ophichthid after a thor- ough osteological study. The distinctness of this monotypic genus is herein felt to merit tribal recognition. The Benthenchelyini appear to be a distinctive offshoot from the generalized Myro- phini, specialized for a pelagic mode of life. Spe- cializations include the large eye, compressed body, enlarged median fins, enlarged head pores, and slender dentition. These characteristics con- verge with those of other pelagic eels, especially 57 the genus Derichthys. A myrophin relationship, particularly to the generalized Myrophis, is evi- denced in the hyoid and branchial arches, gill opening, frontal commissure, and disappearance of the epipleural ribs. The pelagic life style of Benthenchelys (and the associated eye enlarge- ment), unique to the Ophichthidae, could have evolved from the epipelagic breeding migrations of certain myrophines (see Cohen and Dean, 1970). Benthenchelys Fowler Benthenchelys Fowler 1934: 267. (Type species; B. cartieri Fowler 1934, by original designa- tion.) DESCRIPTION (supplementing tribal diagnosis): snout blunt; jaws nearly subequal; eye large; an- terior nostril not tubular, a large anteriad open- ing; DFO slightly before vent; jaw and vomer- ine teeth conical, recurved, and uniserial, inter- maxillary teeth flattened and directed anteriorly, separated from those of vomer by a gap; nasal cartilage weakly developed; suspensorium an- teriorly inclined, jaw angle ca. 95°; maxilla broad, not produced posteriorly, articulating be- neath anterior margin of orbit; hyoid weak, GH elongate, HH separated from CH by a gap, UH a subrectangular plate anteriorly, a cartilaginous filament posteriorly; branchiostegal rays numer- ous, 8 along EH, the last 2 joined basally. ETYMOLOGY: From the Greek 3 €2/90f(ben- thos), deep, and ‘Ev LEaus (enchelys; either masculine or feminine, here to be treated as masculine), eel. DISTRIBUTION: A single pelagic species (100- 250 meters) over deep water in the central Indo- Pacific. Tribe Myrophini TYPE GENUS: Myrophis Lutken, 1851 DIAGNOSIS: Body short to extremely elongate, laterally compressed behind head; tail generally longer than body; lower jaw included; anterior nostril tubular; posterior nostril either lateral or labial; GO lateral, a constricted opening; median fins low or elevated, DFO behind mid-trunk; pectoral fin present or absent; head pores vari- ably developed; LL canal weakly ossified; inter- maxillary dentition and vomerine, when present, continuous; neurocranium not raised along fron- tal or parietal midline, SO crest developed in 58 some genera; orbit moderately developed; gill arches reduced, weakly ossified, Bi often absent, Bs.4 absent or rudimentary; pectoral girdle de- velopment variable; IM bones and ribs moder- ately to weakly developed, transverse processes of caudal vertebrae present in some genera; epi- pleurals limited to anterior trunk vertebrae in some genera; caudal vertebrae more numerous than precaudal; coloration uniform, often darker dorsally. Ahlia Jordan and Davis Ahlia Jordan and Davis 1891: 639. (Type species; Myrophis egmontis Jordan 1889, by original designation.) DESCRIPTION: General characters those of My- rophis. Differences include: snout sub-conical, broad; DFO above or behind anus; vomerine teeth absent; maxilla broad, not tapering poster- iorly, closely abuts the short and broad pterygoid (fig. 30); HH absent, GH rudimentary; gill arches reduced, Hs and lz absent. ETYMOLOGY: Named for Jonas Nicolas Ahl, au- thor of “De Muraena et Ophichtho”, with the noun suffix -ia (neuter). DISTRIBUTION: A single Caribbean species. REMARKS: The controversy regarding generic synonymy of Ahlia has never involved a detailed osteological study. Those considering it synony- mous with Myrophis (Parr, 1930: 8; Hildebrand, in Longley and Hildebrand, 1941: 17; Schultz and Woods, 1949: 171) did not consider the ab- sence of vomerine teeth to represent a generic character, but it was assumed by Jordan and Davis (1891: 639), Myers and Storey (1939: 158), and Wade (1946: 199) that this warranted sep- aration. Nelson (1966a: 398) considered Ahlia to be distinct on the basis of gill arch characters. The generic differences that | have identified are clearly related to feeding specialization in A. eg- montis, viz., tooth loss, maxillary-pterygoid brac- ing, and gill arch reduction, yet the universality of these characters among the species of Myro- phis suggests that the species of Myrophis form a natural group from which Ahlia is a specialized offshoot. Cohen and Dean (1970) have recorded an in- teresting observation of offshore movements and a change in eye size accompanying the onset of sexual maturity in this species. Their observa- tions were made off Honduras, approximately 145 km from shore. | have made similar observa- SERIES 4, V.41, #1 McCOSKER — EELS tions within | km from land in the San Blas Archipelago, off the Atlantic coast of Panama. Muraenichthys Bleeker Muraenichthys Bleeker 1853b: 505. (Type spe- cies; M. gymnopterus Bleeker 1853, by original designation.) Scolecenchelys Ogilby 1897: 246. Spelt Scolen- chelys by other authors. (Type species; Mur- aenichthys australis Macleay 1881, by original designation.) Myropterura Ogilby 1897: 247. (Type species; Myropterura laticaudata Ogilby 1897, by origi- nal designation.) ?Aotea Phillipps 1926: 533. (Type species; Aotea acus Phillipps 1926, by monotypy.) DESCRIPTION: Body short to moderately elong- ate, tail generally longer than body, laterally com- pressed posteriorly; snout sub-conical to blunt, not deeply grooved on underside; posterior nos- tril either along edge of lip beneath a flap or opening into mouth; DFO from mid-trunk to well behind anus; pectoral fin absent; pop® pres- ent; LL ossicles continuous, well developed for a myrophin; dentition variable, teeth often multi- serial, either conical or blunt, dentition of vomer continuous with that of intermaxillary; skull sub- truncate posteriorly; SOC present; maxilla elong- ate, slender posteriorly (fig. 16); suspensorium nearly vertical; opercular series weakly ossified, subopercle generally rudimentary, produced pos- teriorly in some species (as in Myrophis, fig. 36); otic bulla weakly developed; PG short, not brac- ing maxillae, reduced and slender in one sub- genus; HH separated from CH by a narrow gap; gill arches reduced, basibranchials absent, ls os- sified and UP;-UPs fusion variable; pectoral girdle reduced to a slender Cl and SCI; epipleural ribs on all precaudal vertebrae; CTP moderately developed. ETYMOLOGY: From the Greek se UPA LU a an eel, and trots (ichthys; masculine), fish. DISTRIBUTION: Nineteen recognized species from the tropical, subtropical, and warm temper- ate Indo-Pacific Ocean, including a single species from the eastern south Pacific. Material exists of undescribed Red Sea and western Pacific species. REMARKS: Subgeneric lines within Muraenichthys were indicated by McCosker (1970) but were not designated pending a thorough osteological study. My examination and comparison of M. PROCEEDINGS CALIFORNIA ACADEMY OF SCIENCES gymnopterus, M. chilensis, and M. macropterus did not uncover osteological differences compar- able to those used to separate other closely re- lated ophichthid genera. This result was unex- pected in that the external morphology differs considerably within the genus, including differ- ences in snout form (from blunt to acute), in body depth (15-50 times in total length), in the number and position of head pores, in dentition, and in the character of the posterior nostril. Sub- generic lines may be identified in the following manner: Posterior nostril opens on the outer lip as an elongate slit with an anterior flap; a single pore between the anterior and_ posterior nostrils; jaw teeth usually in bands, inter- maxillary teeth in a patch; UP3-UPs fused in species examined; snout usually blunt; body usually stout, its depth ca. 15-25 in TL ........ ee Subgenus Muraenichthys Bleeker Posterior nostril opens into mouth, covered by an exterior valvular flap; two pores be- tween nostrils; jaw teeth uniserial or bi- serial, intermaxillary teeth not in a broad patch; UP3-UPs separate in species exam- ined; snout usually acute; body moderately elongate, its depth usually more than 25 in see ene. Subgenus Scolecenchelys Ogilby The subgenus Muraenichthys includes M. gym- nopterus Bleeker (the type species), M. hattae Jordan and Snyder, M. schultzei Bleeker, and possibly M. macrostomus Bleeker, M. philippin- ensis Schultz and Woods, M. sibogae Weber and de Beaufort, and M. thompsoni Jordan and Rich- ardson. The subgenus Scolecenchelys includes M. australis Macleay* (the type species), M. chil- ensis McCosker, M. acutirostris Weber and de Beaufort, M. cookei Fowler, M. gymnotus Bleeker, M. macropterus Bleeker, M. breviceps Gunther, and possibly M. iredalei Whitley. Also included in Muraenichthys, but not here allocated to a subgenus, are Myropterura laticaudata Ogilby, Chilorhinus vermiformis Peters, Muraenichthys devisi Fowler, M. xorae Smith*, and M. godeffroyi Regan. My specimens of M. macropterus dis- agreed with Nelson’s (1966a) description in hav- ing an unfused UP3-UP4. The genera Muraenichthys, Schultzidia, and Schismorhynchus display obvious similarities at- tributable to a common ancestry. Primitive and advanced conditions of certain characters may be identified among the approximately 24 species 59 involved (many of the species included are known to me only from the literature and not from specimens). For example, postulated primi- tive conditions include the moderately elongate body, sub-conical snout, posterior nostril open- ing into the mouth, numerous head pores, uni- serial or biserial conical teeth, presence of the SO crest, posterior development of the sub- opercle, separate hypohyals, ossified second in- frapharyngobranchial, separate UP3-UPs, and con- spicuous cleithrum and supracleithrum. Species of the subgenus Scolecenchelys are clearly the most primitive, with the species of the subgenus Muraenichthys, and Schultzidia and Schismor- hynchus as specialized offshoots. The develop- ment of transverse processes on the caudal verte- brae, shared by these genera, is unique among the Myrophinae and without apparent antece- dents in more primitive genera such as Myrophis. Aotea, type species A. acus, was described by Phillipps (1926) on the basis of a partially di- gested specimen from New Zealand waters, and placed in Muraenichthys by Castle (1967). Whit- ley (1968) placed A. acus in the synonymy of Muraenichthys breviceps Gunther, yet Phillipps (1926: 533-534) characterized A. acus as having “fins absent’ and ‘a hard folded portion be- neath body posterior to head apparently indi- cat(ing) gill-openings...”, both of which would exclude Aotea from the subfamily Myrophinae. Phillipps’ sketchy description of A. acus does not obviously agree with any known _ ophichthine genus, but best fits Apterichtus, Ichthyapus, and Cirricaecula. Further examination may discover that Aotea acus is a species of Apterichtus in that the species of the latter two genera are not known from even as far south as Australian waters. Myrophis Lutken Myrophis Lutken 1851: 14. (Type species; M. punctatus Lutken 1851, by monotypy.) Paramyrus Gunther 1870: 51. (Type species; Conger cylindroideus Ranzani 1838, by Jordan and Davis (1891) as first revisers.) Holopterura Cope 1871: 482. (Type species; H. plumbea Cope, 1871, by monotypy.) Hesperomyrus Myers and Storey 1939: 157. (Type species: H. fryi Myers and Storey 1939 = My- rophis vafer Jordan and Gilbert, by original designation.) DESCRIPTION: Body stout to moderately elong- ate, laterally compressed throughout; snout sub- 60 conical to conical and moderately elongate; eye moderate; posterior nostril along edge of lip be- neath a flap or opening into mouth; DFO before mid-trunk region; pectoral fin moderately devel- oped, longer than eye; pop® present; teeth conical, uniserial or biserial in jaws and vomer; skull sub- truncate posteriorly (fig. 5); SOC present; maxilla elongate, slender posteriorly (fig. 30B); sub- opercle produced posteriorly as a posteroventral border to the opercle (fig. 36); otic bulla weakly developed; PG short, not bracing maxilla; Hs cartilaginous, UP3-UP, fused in one species; Cl and SCI slender, Sc, Co, and an actinost (2) well developed; epipleural ribs limited to anterior- most 15-20 vertebrae; CTP absent. ETYMOLOGY: From the Greek Bu Pos, Myrus, and OPCS (ophis; masculine), snake. DISTRIBUTION: A circumtropical genus of nine nominal species. Included are: Myrophis punc- tatus Lutken (WA), M. australis Castelnau (IP)*, M. cheni Weng (IP)*, M. lepturus Kotthaus (IP)*, M. platyrhynchus Breder (WA)*, M. vafer Jordan and Gilbert (EP), Conger uropterus Temminck and Schlegel (IP), C. cylindroideus Ranzani (EA)*, Holopterura plumbea Cope (EA). Incertae sedis: Myrophis frio Jordan and Davis (WA)*. REMARKS: Schultz, et al. (1953: 68) erroneously included Parabathymyrus Kamohara in the syn- onymy of Myrophis. D. Smith (1971) recognized it as a valid congrid genus of the subfamily Ba- thymyrinae. Figure 36. Opercular series of Myrophis vafer, SIO 68-242. Right side, distal view. Scale repre- sents 1 mm. Abbreviations are: IO, interopercle; OP, opercle; PO, preopercle; SOP, subopercle. SERIES 4, V. 41, #1 McCOSKER — EELS Castle (1963: 16) has discussed the identity of the congrid Gnathophis heterognathus (Bleeker) Which has been erroneously included in Myro- phis by recent authors. Neenchelys Bamber Neenchelys Bamber 1915: 479. (Type species; N. microtretus Bamber 1915, by monotypy.) DESCRIPTION: Body moderately elongate, com- pressed posteriorly; body shorter than tail; snout sub-conical; eye moderate; posterior nostril an elongate slit before lower margin of orbit; DFO before mid-trunk; pectoral fin moderately de- veloped, longer than eye; pop*® absent; teeth conical, uniserial except at vomerines and inter- maxillary; skull rounded posteriorly; nasal condi- tion unknown; SOC absent; maxilla elongate and slender posteriorly; subopercle not developed posteriorly (fide Nelson 1966b, fig. le); otic bulla weakly developed; Bi rudimentary, lz ossified, UP;-UP, separate; pectoral girdle developed, SCI, Cl, Sc, Co, and an actinost (?) present; epipleu- ral rib condition unknown. ETYMOLOGY: Presumably from the Greek W€z , new, and é UZLEDLS (enchelys; feminine or masculine, treated as masculine by Bamber), eel. DISTRIBUTION: Two. species, N. microtretus Bamber* from the Red Sea, and N. buitendijki Weber and de Beaufort* from the Indian Ocean. REMARKS: Specimens of Neenchelys were un- available for this study. The description is pre- pared from Nelson’s (1966b) osteological and Mohamed’s (1958) morphological description of N. buitendijki. Nelson (1967) noted the presence of overlapping branchiostegals in the holotype of N. microtretus (apparently the only known specimen) but did not compare it with N. buiten- dijki. Nelson (1966b: 323), in commenting on Wade’s (1946) description of Pseudomyrophis micro- pinna, stated that ‘‘there is no character signifi- cant enough to maintain Pseudomyrophis as a genus distinct from Neenchelys. P. nimius, on the other hand, seems distinctive enough to be placed in a genus of its own.” My osteological examination of the species of Pseudomyrophis, which | have found to be congeneric, allows further comment on this relationship. The two genera show certain similarities in morphology and habitat (living in mud bottoms in moderately deep water) and are more closely related to each other than to other genera. Osteological simili- larities include the shapes of the neurocrania, PROCEEDINGS CALIFORNIA ACADEMY OF SCIENCES pterygoids, maxillae, gill arches, hyoid arches, and caudal vertebrae. The characters used in the generic key to separate these genera may be ex- panded in the following manner: DFO in anterior trunk region; snout conical; pectoral fin well developed, = snout; pop? absent; subopercle small, sub-rectangular; (pop? and subopercle condition of N. micro- ARERUISMUIMNIKTMOWWIMN)) 22---2020e-002s-c2e 2222-2 Neenchelys DFO in posterior trunk region; snout broad, tumid; pectoral fin minute, = eye; pop? present; subopercle produced posteriorly along ventral and posterior margins of op- BCS SR ee Pseudomyrophis Pseudomyrophis Wade Pseudomyrophis Wade 1946: 199. (Type species; P. micropinna Wade 1946, by original designa- tion.) DESCRIPTION: Body moderately to extremely elongate, laterally compressed throughout; snout broad, tumid; eye small to moderate; posterior nostril an elongate slit before lower margin of orbit; DFO behind mid-trunk; pectoral fin min- ute, smaller than eye; pop’ present; teeth coni- cal, uniserial throughout, except at anterior vo- mer and intermaxillary; skull rounded _ posteri- orly; nasals ossified along canal only, nasal car- tilage weakly developed; SOC absent; maxilla elongate, slender posteriorly; subopercle similar to Myrophis (Fig. 36), produced posteriorly as a slender posteroventral border to opercle; otic bulla weakly developed; PG short, not produced anteriorly; Bi and ls ossified, UP3-UP4 separate; pectoral girdle reduced to Cl and SCI (and frac- tional Co? in P. micropinna); epipleural ribs on all precaudal vertebrae. Cha ETYMOLOGY: From the Greek W€EUPAYS (pseudes), false, -o-, and Myrophis (masculine), a genus of ophichthids. DISTRIBUTION: Two New World species, P. ni- mius Bohlke (Caribbean) and P. micropinna Wade (eastern Pacific). REMARKS: The species of Pseudomyrophis are strikingly different in body depth and head length, yet an osteological comparison did not uncover differences that are clearly generic. Other proportional differences are also related to the extreme elongation of P. nimius. The mean difference in vertebral number between species (ca. 50) is less than that between species 61 of Phaenomonas (ca. 70). Both species are ap- parently adapted to soft mud bottoms in water relatively deep for ophichthids (P. micropinna from depths of 45-60 fms, P. nimius to 400 fms). Schismorhynchus McCosker Schismorhynchus McCosker 1970: 509. (Type species; Muraenichthys labialis Seale 1917, by original designation.) DESCRIPTION: General characteristics those of Muraenichthys. Differences include: body mod- erately elongate; body shorter than tail; snout conical, elongate, with a prominent toothed groove on underside; anterior nostril an elong- ated tube as long as eye; posterior nostril opens into mouth; DFO in posterior trunk region; pop? and median interorbital pore absent; teeth coni- cal, uniserial; maxilla broad, not becoming slender posteriorly; subopercle produced posteri- orly as in Myrophis (fig. 36); suspensorium an- teriorly inclined; HH fused to CH (or absent?), GH rudimentary; gill arches extremely reduced, Bi, Hs, and lI» absent, UPs3-UPs fused, lower tooth plate elongate (see Nelson, 1966a: figs. 14-15); pectoral girdle reduced to a slender Cl. ETYMOLOGY: From the Greek CLIT UN : : O 4 : (schisme), cleft, and 7 V zx Of — (latin- ized as rhynchus, masculine in accordance with item 30(a)(3) of the International Code of Zoo- logical Nomenclature), nose. DISTRIBUTION: A single species, widespread in the central and western Pacific ocean. Schultzidia Gosline Schultzidia Gosline 1951a: 309. Described as a subgenus of Muraenichthys Bleeker. (Type spe- cies; Muraenichthys johnstonensis Schultz and Woods 1949, by original designation.) DESCRIPTION: General characteristics those of Muraenichthys. Differences include: body stout to moderately elongate, body shorter than tail; posterior nostril opens into mouth; DFO behind anus; pop® absent; teeth absent on vomer, ab- sent or imbedded on intermaxillary, those on maxilla and dentary minute or villiform; SOC absent, SO extends anteriorly to frontals, com- pletely separating parietals; opercular series rudi- mentary, subopercle not produced posteriorly; PG slender; hyoid rudimentary, HH separated from CH by a gap, UH fragmentary, separated medially; gill arches extremely reduced, B: and I» 62 absent, UPs-UP4 fused, tooth plates elongate and vermiculated; CTP weakly developed. ETYMOLOGY: Named for Leonard P. Schultz, ichthyologist, with the assumed noun suffix -idia (feminine). DISTRIBUTION: Two central and western Pacific species, described as Muraenichthys johnstonen- sis Schultz and Woods, and M. retropinnis Seale. Subfamily Ophichthinae DIAGNOSIS: Gill opening variable, mid-lateral to entirely ventral and longitudinal, generally elong- ate and unconstricted; DFO variable, from nape to behind anus; tail tip generally a hard finless point, rudimentary rays visible in certain genera; nasals ossified, generally well developed, but re- duced or absent in certain genera; CH divided into a short median and longer distal portion, the median section connecting to the CH by a car- tilaginous splint (see fig. 17A); UH generally continues posteriorly from basal plate as a slender ossified spike; all branchiostegal rays originate either in association with hyoid or be- fore level of EH tips; gill arch skeleton variably developed, first basibranchial always ossified, second through fourth generally present in either a cartilaginous or rudimentary condition, Cs pres- ent in several genera; coloration variable, band- ed, barred, striped, spotted, or uniform patterns. Tribe Callechelyini TYPE GENUS: Callechelys Kaup, 1856 DIAGNOSIS: Body (head and trunk) and tail moderately elongate, laterally compressed; body longer than tail; snout acute, rounded at tip; lower jaw included; eye small; posterior nostril opens into mouth; GO low lateral to entirely ventral, converging forward, length much greater than isthmus width; dorsal fin originating on nape; pectoral fin absent; tail tip a hard finless point; head pores reduced, pop*® and tp? absent; LL ossicles nearly continuous; teeth conical, jaw teeth uniserial, those of vomer separated from those of intermaxillary by a gap; skull short, sloping posteriorly, its height ca. 3 or less in its length (fig. 7); orbit moderately developed; SO rounded, without a posterior projection; PG slender, elongate, free and tapering posteriorly; margin of opercular series irregularly ossified, with cartilaginous gaps; suspensorium nearly vertical; otic bulla well developed; hyoid stout; branchiostegal rays numerous; gill arches re- SERIES 4, V.41,#1 McCOSKER — EELS duced, C; absent, UP;-UPs separate; pectoral girdle reduced to Cl, SCI, and 1 or 2 rod-shaped elements; IM bones, ribs, and CTP developed; precaudal vertebrae more numerous than caudal; coloration variable, either striped, barred, spot- ted, mottled, or uniform. REMARKS: The Callechelyini constitute the most distinct and compact of ophichthid tribes. Avail- able material or radiographs of 19 of the 21 spe- cies of Callechelyini has allowed an in depth study of this tribe. The results of two computer- programmed numerical taxonomic evaluations of this tribe are presented in the discussion sec- tion. Several important morphological and osteo- logical characters of the species are listed in Table 8. Aprognathodon Bohlke Aprognathodon Bohlke 1966: 99. (Type species; A. platyventris Bohlke 1966, by original desig- nation.) DESCRIPTION: Anterior nostril tubular; snout moderate, rounded at tip; median groove absent on underside of snout; intermaxillary teeth ab- sent, vomerine teeth present; DFO above SO; anal fin present; 3 supraorbital pores; neuro- cranium well rounded, highest anterior to front- al-parietal suture; hyoid arch very stout, inflex- ible along CH-EH suture; HH absent; branchio- stegal rays numerous, along arch, often joined basally, distal 4-6 rays along EH broadened bas- ally; UH a simple cartilaginous projection pos- teriorly, basal plate ossified; gill arches stout, Hs partially or completely ossified; pectoral girdle contains Cl, SCI, and 2 rod-shaped elements; body coloration strongly banded longitudinally. ETYMOLOGY: From the Greek XH , without, rf (pro), forward, VVA eos site jaw, and OfavV (odon; masculine), tooth, reference to the lack of intermaxillary pe. DISTRIBUTION: A single western Atlantic species, known from the Bahamas through the Lesser An- tilles to Venezuela. Callechelys Kaup Callechelys Kaup 1856: 51 (28). (Type species; C. guichenoti Kaup 1856 = Dalophis marmor- ata Bleeker 1853, by monotypy.) Cryptopterygium Ginsburg 1951: 482. (Type spe- cies; Cryptopterygium holochroma Ginsburg 1951, by original designation.) PROCEEDINGS CALIFORNIA ACADEMY OF SCIENCES DESCRIPTION: Anterior nostril tubular; snout short, rounded at tip; a median groove on un- derside of snout (Fig. 32A); intermaxillary and vomerine teeth present; DFO above SO; anal fin present; 3 supraorbital pores; neurocranium well rounded, highest anterior to level of frontal- parietal suture (fig. 7); hyoid arch stout, moder- ately flexible along CH-EH suture; HH separated from CH by a narrow gap; branchiostegal rays numerous, along arch; distal rays along EH broadened basally in some species; UH either a simple slender filament posteriorly, or split into two divergent rays; Hs cartilaginous; pectoral girdle contains Cl, SCI, and either one or two rod-shaped elements; coloration variable, either uniform, spotted, mottled, or banded. ETYMOLOGY: Kaup (1856a, b) did not give the derivation of the generic name nor did he desig- nate its gender. From his description (1856b: 28), “this handsome eel .. .’”, one must assume that he intended the generic name to be derived from KANNOS (beauty) and € Y LEAVS (enchelys, eel) which is feminine, but according to Liddell and Scott (1801), was later also mascu- line. Bleeker (1865), as first reviser, further con- fused matters by recognizing Dalophis marmor- ata Bleeker as Callechelys marmoratus and also describing Callechelys melanotaenia. To date, the gender of Callechelys has not been estab- lished, although the most recent revisers (McCos- ker and Rosenblatt, 1972) have regarded Calle- chelys as masculine. DISTRIBUTION: A cosmopolitan genus with 15 tropical and subtropical species. Nominal species include: Callechelys bilinearis Kanazawa (WA), C. cliffi Bohlke and Briggs (EP), C. eristigmus McCosker and Rosenblatt (EP), C. galapagensis McCosker and Rosenblatt (EP), C. /uteus Snyder (IP), C. melanotaenia Bleeker (IP), C. muraena Jordan and Evermann (WA), C. nebulosus Smith (IP), C. perryae Storey (WA, EA), C. striatus Smith (IP), Ophichthys bitaeniatus Peters (IP)*, Crypto- pterygium holochroma Ginsburg (WA), Caecula leucoptera Cadenat (EA)*, Dalophis marmorata Bleeker (IP), Gordiichthys springeri Ginsburg (WA). REMARKS: Numerous authors (Gunther, 1910: 404; Pellegrin, 1912; Storey, 1939: 63; Smith, 1957: 83; McCosker and Rosenblatt, 1972: 22) have discussed the validity of C. guichenoti, the generic type, with the majority supporting its synonymy with C. marmoratus. Marie-Louise 63 Bauchot of the Paris Museum has kindly furn- ished measurements and a radiograph of the type specimen (MNHN 2126) of C. guichenoti. Its morphometry and osteology (183 vertebrae and a single pectoral girdle horizontal element) are further evidence of its synonymy with C. marmoratus. Subgeneric lines within Callechelys were sug- gested by McCosker and Rosenblatt (1972). They recognized two major groups, one containing species with a simple urohyal and a single rod- shaped pectoral element (fig. 19L) and another with species having the urohyal split posteriorly into two slender divergent rays and two rod- shaped pectoral elements (as in Aprognathodon, fig. 19M). A third can be recognized, which pos- sesses a mosaic of characters, including slightly broadened branchiostegal rays along the epihyal, and urohyal and pectoral girdle conditions that do not conform to either of the above groups. Programs REGROUP and WVGM showed little af- finity between C. nebulosus of this last group and the remainder of the genus. It appears that the simple urohyal, broadened rays, and paired girdle elements are primitive conditions within the Callechelyini, characters shared by C. nebu- losus and C. springeri. Letharchus Goode and Bean Letharchus Goode and Bean 1882: 437. (Type species L. velifer Goode and Bean 1882, by original designation.) DESCRIPTION: Anterior nostril a hole, its rim not raised; snout moderate, acute, not rounded at tip; median groove on underside of snout absent (fig. 32b); intermaxillary and vomerine teeth present; DFO above epiotics; anal fin ab- sent; four supraorbital pores; neurocranium de- pressed, not rounded across parietal-frontal re- gion, highest at frontal-parietal suture; hyoid arch stout, flexible along CH-EH suture, HH sep- arated from CH by a gap; branchiostegal rays numerous, slender, all along arch; UH a slender filament posteriorly; Hs cartilaginous; pectoral girdle contains Cl, SCI, and 2 rod-shaped ele- ments; body coloration uniformly dark, con- trasting strongly with the white dorsal fin. ETYMOLOGY: From the Greek NMNOEOOMAKC to forget, and ie or (archos; mascu- line), anus, in reference to the lack of an anal fin. DISTRIBUTION: Known from three New World species: L. velifer from the western Atlantic 64 (North Carolina to the northern Gulf of Mexico), L. aliculatus McCosker from off Brazil, and L. rosenblatti McCosker from the eastern Pacific. REMARKS: The genera Letharchus and Paraleth- archus were recently revised by McCosker (1974). The species of Letharchus form a unique and dis- tinctive offshoot from the generalized callech- elyin condition in their combination of anal fin absence, non-tubular nostrils, an additional supra- orbital pore, and the acute snout, differing mark- edly from the characters of other genera within the tribe. Letharchus pacificus Osborne and Nichols and L. opercularis Myers and Wade are obviously similar to the above species in lacking an anal fin, and probably for that reason were assumed by their describers to be congeneric with L. velifer. After examining considerable material and the types of these five species | have concluded that L. opercularis and L. pacifi- cus represent a separate generic line within the Callechelyini. The osteological description of Letharchus is based on the eastern Pacific species in that mate- rial of L. velifer was unavailable for dissection. Leuropharus Rosenblatt and McCosker Leuropharus Rosenblatt and McCosker 1970: 502. (Type species; L. lasiops Rosenblatt and Mc- Cosker 1970, by original designation.) DESCRIPTION: Anterior nostril tubular; snout short, rounded at tip; median groove on under- side of snout absent; snout, nape, and much of surface of jaws papillate; intermaxillary teeth present, vomerine teeth absent; DFO above SO; anal fin present; three supraorbital pores; neuro- cranium well rounded, highest anterior to front- al-parietal suture; HH separated from CH by a gap; branchiostegal rays numerous, slender, and along arch; Hs cartilaginous; pectoral girdle con- tains Cl, SCl, and 2 rod-shaped elements; body coloration nearly uniform, median fins white. La ETYMOLOGY: From the Greek NE VUPOS Ae (leuros), smooth, and ACOs (pharos; neuter), plow, in reference to the toothless vo- mer. DISTRIBUTION: A_ single species, known only from the type specimen from Manzanillo Bay, western Mexico. Paraletharchus McCosker Paraletharchus McCosker 1974: 620. (Type spe- cies; Letharchus pacificus Osburn and Nichols, SERIES 4, V.41,#1 McCOSKER — EELS 1916, by original designation.) DESCRIPTION: Anterior nostril tubular; snout short, rounded at tip; median groove on under- side of snout absent; LL ossicles block-like cyl- inders, not heavily fractionated (compare figs. 22H and 221); intermaxillary and vomerine teeth present; GO with a deep anterolateral pocket; DFO above SO; dorsal fin elevated; anal fin ab- sent; three supraorbital pores; neurocranium well rounded, highest anterior to frontal-parietal su- ture; hyoid arch stout, only slightly flexible along EH-CH suture; HH separated from CH by a nar- row gap; branchiostegal rays numerous, along arch, distal rays along epihyal broadened basally; UH a simple slender filament posteriorly; Hs cartilaginous; pectoral girdle contains Cl, SCI, and 2 rod-shaped elements; coloration nearly uniform to mottled. 4 ETYMOLOGY: From the Greek ITAL ERK? (para), near, and Letharchus (masculine), a re- lated genus. DISTRIBUTION: Known from two eastern Pacific species, P. opercularis (Myers and Wade), a Galapagos endemic, and P. pacificus (Osburn and Nichols), ranging from Baja California to Costa Rica. REMARKS: As mentioned in the remarks on Letharchus, the above-mentioned species are not congeneric with L. velifer. Schultz and Barton (1960) placed L. opercularis in the synonymy of L. pacificus, however McCosker (1974) provision- ally recognized the Galapagos population as dis- tinct on the basis of the difference in the mean vertebral number (95% confidence limits: P. opercularis 172.7 - 176.6, P. pacificus 158.5 - 162.3). Tribe Sphagebranchini TYPE GENUS: Caecula Vahl 1794, the senior ob- jective svnonym of Sphagebranchus Bloch, 1795 (see following remarks). DIAGNOSIS: Body (head and trunk) and tail moderately elongate, cylindrical, often compres- sed posteriorly; body either nearly equal to or shorter than tail; snout pointed, often broad and depressed; lower jaw included; posterior nostril opens into mouth in most genera; GO entirely ventral; median fins either very low or absent; pectoral fin absent; tail tip sharply pointed; head pores well developed, tp? and pop* gener- ally, and pop* sometimes present; LL ossicles PROCEEDINGS CALIFORNIA ACADEMY OF SCIENCES continuous; teeth conical, not caniniform, and generally uniserial; intermaxillary teeth sepa- rated by a gap from those of vomer; neurocran- ium elongate, generally depressed and truncate posteriorly; orbit reduced; otic bulla moderately to well developed; hyoid arch generally slender; branchiostegal rays few to numerous, generally closely associated with hyoid; gill arches re- duced, C; absent in most genera; intramuscular bones, ribs, and caudal transverse processes well developed; precaudal either nearly equal to or more numerous than caudal vertebrae; colora- tion generally uniform, or darker dorsally. REMARKS: Included among the genera of the Sphagebranchini are the most frustrating and problematical of ophichthid taxa. Their nearly complete lack of superficial characters has re- sulted in a history of repeated lumping and split- ting. Further confusion stems from the inade- quate and misleading descriptions of the two oldest generic names, Caecula Vahl (1794) and Sphagebranchus Bloch (1795). Gosline (1951: 311) summarized the problem in stating that ‘‘the definition, and consequently the limits, of the genus Caecula are agreed upon by no two au- thors as far as | know.” The tribal name Sphagebranchini is derived from Swainson’s (1838) family name Sphage- branchidae. Swainson’s family name was subse- quently rejected (see page 10) long before Sphagebranchus Bloch (1795) was shown to be a junior objective synonym of Caecula Vahl (1794) (see Bohlke and McCosker, 1975). The family-group name Caeculidae has not appeared in the ichthyological literature. In accordance with Article 40 of the International Code of Zoo- logical Nomenclature (1964), the family-group name Sphagebranchini therefore has priority as the tribal name, with Caecula as the type genus of the tribe. Achirophichthys Bleeker Achirophichthys Bleeker 1865: 41. (Type species; A. typus Bleeker 1865, by original designation.) DESCRIPTION (based on Bleeker, 1865, and Weber and de Beaufort, 1916): Body stout, slightly longer than tail; snout pointed; eye small; anterior nostril not tubular, posterior nos- tril opens into mouth; lips with one row of min- ute tubercular papillae; DFO slightly behind GO; GO low lateral to inferior; teeth conical, long and nearly caniniform anteriorly and along vo- mer, uniserial on vomer, biserial on maxilla. 65 W ETYMOLOGY: From the Greek E¢ as (achir), without hands, and Ophichthys, (mascu- line), the amended spelling of Ophichthus. DISTRIBUTION: A single western Pacific species. REMARKS: An osteological diagnosis of Achiro- phichthys is not included in this study in that material of A. typus, the generic type, was un- available. A. kampeni (Weber and de Beaufort), its sole described congener, is herein referred to Lamnostoma. Jordan and Davis (1891: 636) sug- gested that A. typus might be the young of Brachysomophis crocodilinus, but subsequent authors have neither accepted nor commented upon this action. Weber and de Beaufort (1916) considered Achirophichthys to be a subgenus of Brachysomophis. This too was ignored by most subsequent authors. Their description of A. typus, based on the type specimen, strongly in- dicates that it is congeneric with A. kampeni, which if true, would place Achirophichthys in the synonymy of Lamnostoma. Apterichtus Dumeéril Caecilia Lacépede 1800: 134. Preoccupied by Caecilia Linnaeus, a genus of Amphibia. (Type species; C. branderiana Lacépede 1800, by monotypy.) Apterichtus Duméril 1806: 331. Also spelled Apterichthys, Apterichthus, Apterichthe, and Apterichtes by other authors. (Type species; Muraena caeca Linnaeus 1758, by monotypy.) Typhlotes Fischer 1813: 81. A replacement name for Caecilia Lacépede, preoccupied. Branderius Rafinesque 1815: 93. A replacement name for Caecilia Lacépede, preoccupied. Ophisurapus Kaup 1856a: 52. (Type species; O. gracilis Kaup 1856, by monotypy.) Ophisuraphis Kaup 1856b: 29. Emend. pro Oph- isurapus Kaup 1856a. Verma Jordan and Evermann 1896: 374. (Type species; Sphagebranchus kendalli Gilbert 1889, by original designation.) ?Microrhynchus Blache and Bauchot 1972: 728. Preoccupied by Microrhynchus Dejean 1821, a genus of lepidoptera, as well as mammalia (jourdan 1834), crustacea (Bell 1835), aves (Les- son 1843) and vermes (Kepner 1935). (Type species; Sphagebranchus foresti Cadenat and Roux 1964, by original designation.) DESCRIPTION: Body very elongate, cylindrical, and pointed at both ends; body and tail nearly subequal; snout pointed, sub-conical, grooved 66 and flattened on underside; lips without barbels; eye moderately developed; anterior nostril tubu- lar, posterior nostril a horizontally ovate slit out- side of mouth; GO ventral, converging forward; isthmus short; all fins absent; tp” and pop‘ pres- ence variable among species; teeth pointed, uni- serial in jaws, and largest at intermaxillary which is separated from those of vomer by a short gap; skull slightly depressed, sub-truncate to rounded posteriorly; orbit reduced; nasals and nasal car- tilage developed; SOC short and blunt posteri- orly; maxilla pointed posteriorly; opercular mar- gins entire; suspensorium anteriorly inclined, jaw angle ca. 100°; PG slender, pointed and very reduced; hyoid arch slender, HH separated from CH by a gap; branchiostegal rays closely asso- ciated with hyoid; UH cartilaginous posteriorly; C; absent, UP;-UPs separate; Cl broad, SCI re- duced, Co and Sc absent; posterior trunk parapo- physes with an anterior marginal projection (fig. 33). ETYMOLOGY: From the Greek AMT fov (apteron), without fins, and (ichtus, more cor- rectly written ichthys; masculine), fish. DISTRIBUTION: From 10-12 described and valid species, represented in all tropical oceans. Bohlke (1968) provisionally reviewed the species of Verma (=Apterichtus). The genus Apterichtus can be expanded to include: Muraena caeca Lin- naeus (M), Caecula gymnocelus Bohlke (EP), C. monodi Roux (EA)*, C. equatorialis Myers and Wade (EP), Sphagebranchus klanzingai Weber (IP), S. flavicaudus Snyder (IP), S. kendalli Gilbert (WA), Verma ansp Bohlke (WA), Ophisurapus gracilis Kaup (EA)*, Ophichthys anguiformis Peters (EA)*, and possibly Sphagebranchus for- esti Cadenat and Roux (EA)* and Microrhynchus epinepheli Blache and Bauchot (EA)*. REMARKS: Blache and Bauchot (1972) recog- nized Verma as distinct from Apterichtus on the basis of a minor difference in posterior nostril location. Through the kindness of Enrico Tor- tonese | have examined a specimen of the Medi- terranean Apterichtus caecus, and have con- cluded that it is clearly congeneric with the At- lantic and Pacific species previously referred to Verma. Microrhynchus Blache and_ Bauchot (1972) is based upon two species known only from the holotypes. Both were unavailable for study. Their sketchy description of the external morphology of the species, upon which the pres- ent study is based, does not provide characters which would allow their generic separation from SERIES 4, V. 41, #1 McCOSKER — EELS Apterichtus, or possibly Ichthyapus. Should Mic- rorhynchus prove to be a valid genus a substitute name will be required. Caecula Vahl Caecula Vahl 1794: 149. (Type species; C. ptery- gera Vahl 1794, by original designation.) Sphagebranchus Bloch 1795: 88. (Type species; S. rostratus Bloch 1795 = Caecula pterygera Vahl 1794, by monotypy.) DESCRIPTION: Body moderately elongate, cylin- drical, compressed posteriorly; body and _ tail nearly subequal; snout pointed, depressed, and broad dorsally, grooved and flattened on under- side; eye moderate; anterior nostril flush with snout anteriorly, produced as a tube posteriorly, posterior nostril associated with a barbel; GO entirely ventral, converging anteriorly, much longer than isthmus, and with an anterolateral duplication forming a deep pouch; DFO slightly behind GO; pop* and tp? present, pop* absent; teeth conical and uniserial, largest at intermaxil- lary which is widely separated from those of vomer; skull depressed, broad, and truncate pos- teriorly; orbit extremely reduced; nasals and nasal cartilage well developed; SOC a short broad point posteriorly; maxilla elongate, pointed posteriorly; coranoid process of articular moder- ately enlarged; operculum well developed, pre- operculum narrow and thin, their margins entire; suspensorium nearly vertical; hyomandibular broad, expanded posterodorsally and anteriorly, strongly ridged; otic bulla weakly developed; PG slender, pointed at each end, and braced against hyomandibular by a posteromedial extension; hyoid arch slender; HH separated from CH by a gap; branchiostegal rays numerous, flat and unbranched, loosely associated with hyoid; out- ermost rays along epihyal slightly broadened; UH a slender filament posteriorly; C; absent, UP3-UPs separated, anterior half of Bi ossified; Cl, SCI, Sc and Co present. ETYMOLOGY: A diminutive of the Latin caecus, blind, regarded as feminine. DISTRIBUTION: Two western Pacific species. REMARKS: Various species have been haphazard- ly assigned to Caecula to such an extent that it has become a catch-all for most finned and fin- less ophichthids lacking pectoral fins. Smith (1964) redescribed the type of C. pterygera and began the dissection of this confusing assemb- lage. Smith erred in presuming C. pterygera and PROCEEDINGS CALIFORNIA ACADEMY OF SCIENCES Lamnostoma pictum to be synonymous, and in fact, congeneric. Osteologically these genera are similar in their coronoid processes, hyoid, gill arch, and nostril conditions, but differ tren- chently in their neurocrania. Blache and Bauchot (1972) placed the type species of Sphagebranchus, S. rostratus Bloch, in Caecula. BOhlke and McCosker (1975) considered the type species of those genera to be conspe- cific, and suggested that the collection locality of S. rostratus (“river in Surinam’’) was erroneous. Cirricaecula Schultz Cirricaecula Schultz 1953: 49. Type species; C. johnsoni Schultz 1953, by original designa- tion.) DESCRIPTION: General characters those of /Ich- thyapus. Differences include: body and tail near- ly subequal; numerous cirri along edge of upper lip, a prominent barbel between the nostrils; an- terior nostril nearly flush with snout anteriorly, slightly produced as a tube posteriorly; GO en- tirely ventral, consisting of 2 parallel longitudinal slits with thin medial membranes; isthmus min- ute; pop’ and tp? present; otic bulla weakly de- veloped; PG slender, pointed at each end; hyoid arch slender, HH separated from CH by a gap, UH a slender filament posteriorly; Cs; ossified, UP;-UP, fused; pectoral girdle reduced to a ven- trally located Cl pair which are broad and anteri- orly expanded (fig. 19); CTP strongly developed. ETYMOLOGY: From the Latin cirrus, tendril, and Caecula (feminine), a related genus. DISTRIBUTION: A single central Pacific species, known only from the type series collected at Eniwetok, Marshall Islands. Hemerorhinus Weber and de Beaufort, incertae sedis Hemerorhinus Weber and de Beaufort 1916: 280. (Type species; Sphagebranchus heyningi Web- er 1913, by original designation.) DESCRIPTION: Body moderately elongate, cylin- drical, pointed at each end; body longer than tail; snout pointed, grooved on underside; eye small; anterior nostril flush with snout; posterior nostril below eye, a long slit in upper lip; GO inferior, vertical; vertical fins low; tp? present(?), pop® absent(?). / ETYMOLOGY: Presumably from the Greek 72 ME efi (hemeros), cultivated, and eC C EI 6 67 (rhinos; masculine in accordance with item 30(a)(3) of the International Code of Zoological Nomenclature), nose. DISTRIBUTION: Known from the type specimen, collected in deep water (69-91 meters) from Flores, Indonesia, and H. opici Blache and Bau- chot (EA)*. REMARKS: Species of Hemerorhinus were un- available for study. Blache and Bauchot’s (1972) redescription of the type species clarified several confusing aspects of the original description, par- ticularly in correcting the mistaken intepretation of the nostrils and fin position. From their dis- cussion however, | am unable to confidently find its placement within this tribe. On the basis of Blache and Bauchot'’s illustrations, the cephalic pore condition would indicate a similarity to the species of Yirrkala. Ichthyapus de Barneville Ichthyapus de Barneville 1847: 219. (Type spe- cies; /. acutirostris de Barneville 1847, by monotypy.) Rhinenchelys Blache and Bauchot 1972: 718. (Type species; Sphagebranchus ophioneus Evermann and Marsh 1902, by original desig- nation.) DESCRIPTION: Body elongate, cylindrical, point- ed at both ends; tail longer than body; snout pointed, depressed, and broad dorsally, grooved and flattened on underside; lips without barbels; eye small; anterior nostril flush with snout, pos- terior opens into mouth; GO entirely ventral, with a thin medial membrane, converging for- ward, isthmus small; all fins absent; tp? always and pop? usually present; teeth pointed, uni- serial, and largest at intermaxillary which are separated from those of vomer by a gap; skull depressed, broad, and truncate posteriorly (fig. 8); orbit extremely reduced; nasals moderately, and nasal cartilage well developed; SOC moder- ately projecting posteriorly; maxilla elongate and pointed posteriorly; suspensorium nearly verti- cal; opercular margins entire, preopercle re- duced; hyomandibular broad, expanded anteri- orly and posterodorsally; otic bulla moderately developed; PG elongate and rectangular posteri- orly, with a slender projection from the antero- dorsal corner; hyoid thickened (not as slender as in related genera); HH separated from CH by a narrow gap; branchiostegal rays not numerous, slender and generally unbranched, closely asso- 68 ciated with hyoid, outermost rays along EH broadened basally; UH cartilaginous for pos- terior two-thirds; C; reduced (see Remarks), UP3- UP, separate; Cl broad, SCI usualy absent, Sc and Co absent; posterior trunk parapophyses with an anterior marginal projection as in Apter- ichtus (fig. 33). ETYMOLOGY: From the Greek ¢Y% Obs (ichthys), fish, and Gx ae) Sp (apous; masculine), without foot, presumably in _refer- ence to the lack of pectoral fins. DISTRIBUTION: Circumtropical, with 6-7 valid species. Included are: /. acutirostris de Barneville (locality unknown)*, Sphagebranchus — vulturis Weber and de Beaufort (=Caecula platyrhyncha Gosline) (IP), S. ophioneus Evermann and Marsh (WA), Apterichthys selachops Jordan and Gilbert (EP), and three undescribed eastern and western Pacific forms. Incertae sedis: Sphagebranchus omanensis Norman (IP). REMARKS: The sharp-snouted finless species pre- viously placed in Sphagebranchus, excluding the species of Cirricaecula and Apterichtus as herein defined, are referred to Ichthyapus. The generic type, /. acutirostris, is obviously congeneric with those species according to descriptions of the type made by de Barneville (1847, fide Fowler, 1936: 293), Kaup (1856b: 29) and Blache and Bauchot (1972: 718-728). Blache and Bauchot (1972) differentiated Rhinenchelys from Ichthyapus on the basis of minor differences in nostril condition and inter- maxillary tooth location. My examination of the osteology of ophioneus, the type of Rhinench- elys, indicates that it is congeneric with sela- chops, vulturis, and presumably acutirostris. Nelson (1966a: table 1, figure 19) has de- scribed and illustrated the gill arch condition of I. vulturis (as Caecuia platyrhyncha). The fifth ceratobranchial (Cs) is reduced and fused to the lower pharyngeal dermal tooth plate. | have found the gill arches of J. ophioneus, |. vulturis, and /. selachops to be similar in this condition. Cirricaecula, with a prominent Cs, appears inter- mediate between /Ichthyapus and most Ophich- thyini in this condition. The C; condition of the related genus Apterichtus, based on my examina- tion of A. flavicaudus, is the most reduced in the group. Lamnostoma Kaup Lamnostoma Kaup 1856: 49 (23). (Type species; L. pictum Kaup 1856 = Dalophis orientalis SERIES 4, V.41,#1 McCOSKER — EELS McClelland 1844, by Jordan 1919b as first re- viser.) Anguisurus Kaup 1856: 50 (24). (Type species; A. punctulatus Kaup 1856 = Dalophis orien- talis McClelland 1844, by monotypy.) DESCRIPTION: Body stout, cylindrical, pointed at each end; body slightly longer than tail; snout pointed, its underside grooved; eye small to moderate; anterior nostril flush along snout, its posterior rim produced, posterior nostril usually associated with a pendulous flap; GO inferior, ca. equal to isthmus; DFO above or behind GO; tp? and pop? absent; teeth slender, pointed, and recurved, uniserial or biserial in jaws, those of intermaxillary and vomer largest and _ widely spaced; neurocranium truncate posteriorly, elongate and narrow, particularly along ethmoid and interorbital region; orbit depressed; nasals and nasal cartilage moderately developed; SOC present; maxillae moderately elongate, slender but not pointed posteriorly; coronoid process of articular greatly enlarged; opercular series mod- erately developed, their margins entire; suspen- sorium anteriorly inclined, jaw angle ca. 100°; hyomandibular broad, expanded posterodorsally and strongly ridged; otic bulla well developed; PG slender, very reduced; hyoid slender, equal to branchiostegal rays in thickness, HH separated from CH by a broad gap, rays numerous, flat and unbranched, only the distal-most associated with hyoid, others terminate anteriorly behind tip of slender UH, outermost rays along EH slightly broadened; gill arches reduced, C; absent, UPs- UP, separate, B; cartilaginous except at anterior tip which is ossified; Cl, SCI, and reduced Co and Sc present; coloration generally darker dor- sally, a series of white spots across nape. ETYMOLOGY: From the Greek AMX AL CX (lamna), a horrible anthropophagous monster, a bugbear used by the Greeks to frighten refract- ory children (Jordan and Evermann, 1896: 49), and g TO (stoma; neuter), mouth. DISTRIBUTION: From the western Pacific, pro- visionally including four species. REMARKS: Lamnostoma has been placed by re- cent authors in the synonymy of Caecula. It is herein found to differ markedly in numerous osteological and morphological characters, and is COnsequently resurrected. The species of this genus are generally collected in freshwater, and are easily recognized by their slender jaws and conspicuous white spotting on the head and PROCEEDINGS CALIFORNIA ACADEMY OF SCIENCES lateral line. Several nominal species are included in Lamnostoma, most of which fall into the syn- onymy of L. orientalis (McClelland). Caecula min- dora Jordan and Richardson and C. tay/ori Herre are also referable to this genus. Achirophichthys kampeni (Weber and de Beaufort) is clearly a Lamnostoma. Its description, and recent refer- ences containing general morphological descrip- tions (Herre, 1924; Nichols, 1955; La Monte, 1961; Tortonese, 1964), concern large adults which have been collected in freshwater, each displaying the slender jaws and the head and body spotting of Lamnostoma. Stictorhinus Bohlke and McCosker Stictorhinus Bohlke and McCosker 1975: 5. (Type species; S$. potamius Bohlke and McCos- ker 1975, by original designation.) DESCRIPTION: General characteristics those of Ichthyapus. Differences include: Median fins present, low but distinct, DFO ca. mid-trunk; tp? present, pop* absent; LL ossicles with a short gap at pores; basisphenoid not elongate as in Ichthyapus (compare Figs. 8 and 9); SOC re- duced; opercular series reduced, interopercle absent, preopercle reduced; branchiostegal rays numerous, branched basally in some individuals, closely associated with hyoid, outermost rays along EH broadened basally; UH a slender ossi- fied filament posteriorly; C; absent; Cl broad and expanded anteriorly, SCI absent, Sc and Co reduced; trunk parapophyses lack an anterior projection (fig. 33). ETYMOLOGY: From the Greek TTLKTOS (stiktos), pricked or punctured, and eC Llp (rhin, latinized to rhinus, nominative case), nose, in reference to the nature of the anterior nostrils. DISTRIBUTION: A single western Atlantic species, known from tidal rivers in Brazil. REMARKS: In 1971 Naercio Menezes sent the author a series of eels from freshwater in Brazil. These were identified as Sphagebranchus rostra- tus Bloch, a species known only from the type specimen said to have been from a “river in Suri- nam.”” Subsequent examination of the type speci- mens of Caecula pterygera and Sphagebranchus rostratus by J. E. BOhlke found them to be based on the same species, and the Brazilian specimens to be an undescribed genus and species (Bohlke and McCosker, 1975). My treatment (McCosker, 1973) of Sphagebranchus was therefore based on Stictorhinus. 69 Yirrkala Whitley Yirrkala Whitley 1940: 410. (Type species; Y. chaselingi Whitley 1940 = Sphagebranchus lumbricoides Bleeker 1865, by original desig- nation.) Pantonora Smith 1964: 719. (Type species; Oph- ichthys tenuis Gunther 1870, by original desig- nation.) DESCRIPTION: Body elongate, cylindrical, shorter than tail; snout conical, moderately developed, its underside grooved; eye moderate; anterior nostril in a short tube, posterior opens into mouth; GO ventral, longer than isthmus; DFO above or slightly behind GO; tp” usually present, pop® absent; teeth conical, pointed, nearly sub- equal, uniserial in jaws, those on vomer biserial anteriorly, separated from those of intermaxillary by a gap; skull not depressed, rounded posteri- orly; orbit not strongly depressed; nasals and nasal cartilage moderately developed; SOC ab- sent; maxilla pointed posteriorly; opercular ser- ies well developed, their margins entire; sus- pensorium anteriorly inclined, jaw angle ca. 100°; PG slender, pointed anteriorly; HH sepa- rated from CH by a short gap (HH absent in Y. misolensis); branchiostegal rays closely associ- ated with hyoid; UH with a short projection pos- teriorly; C; absent, UP3-UPs separate; pectoral girdle contains SCI, Cl, and reduced Co and Sc; posterior trunk parapophyses lack anterior mar- ginal projections. ETYMOLOGY: Named for Yirrkala, northern Aus- tralia, the type locality of the type species, mas- culine, in accordance with item 30(b)(ii) of the International Code of Zoological Nomenclature. DISTRIBUTION: Contains approximately 12 spe- cies, from the eastern Atlantic, the Red Sea, and Indian and western Pacific Oceans. Included are: Sphagebranchus lumbricoides Bleeker, Ophich- thys tenuis Gunther, Caecula maculata Klause- witz, and an undescribed species from the Mar- quesas. Incertae sedis: Sphagebranchus macro- don Bleeker*, S. gjellerupi Weber and de Beau- fort*, S. kaupi Bleeker, Dalophis moluccensis Bleeker*, Ophichthys misolensis Gunther, Mur- aena fusca Zuiew*, and Caecula natalensis Fow- ler*. REMARKS: The description of Yirrkala is so lack- ing in diagnostic characters as to preclude its proper placement, and consequently has been ignored by most recent authors. The only sub- sequent reference to Y. chaselingi, the generic 70 type, which | have encountered is that of Munro (1957) in which this species is placed in Sphage- branchus. | have examined and x-rayed the badly damaged and dessicated paratype of Y. chase- lingi. The anterior nostrils, head pores, and gill openings are too badly damaged to allow accur- ate examination, however the dentition is well preserved and agrees well with Whitley’s illus- tration. In his description, Whitley stated that Y. chaselingi and Sphagebranchus lumbricoides are congeneric, differing in the ‘position of dorsal origin, proportion of head to trunk, in having vomerine teeth largest, and other details of porportions and dentition.” My comparison of the paratype of Y. chaselingi and specimens of S. lumbricoides from Vietnam (CAS 13969) and the Philippines (CAS reg. 1607) indicates that they are conspecific. The proportions, dorsal fin origins, dentition, and vertebral numbers (Y. chaselingi = 153, S. lumbricoides = 151,154) are not different. Pantonora Smith (1964) is herein considered a synonym of Yirrkala. Sphagebranchus kaupi and Ophichthys misol- ensis are provisionally referred to Yirrkala. Y. kaupi differs in having the major axis of its GO vertical. Y. misolensis differs in having basihyals fused to the ceratohyals, lateral line ossicles con- sisting of numerous short coils rather than small block-like segments, and lacking tp. The external morphology of the species of Hemerorhinus, as described by Blache and Bau- chot (1972), indicates that those species might be conspecific with the species within this com- plex. A comparative study of the type species of Yirrkala, Hemerorhinus, and Pantonora is clearly needed. Tribe Bascanichthyini TYPE GENUS: Bascanichthys Jordan and Davis, 1892. DIAGNOSIS: Body (head and trunk) and tail moderately to extremely elongate, generally cyl- indrical, and compressed posteriorly in some genera; body either equal to or longer than tail; lower jaw included; posterior nostril opens into mouth; GO low lateral, crescentric, never en- tirely ventral; median fins generally low, DFO on head in most genera; pectoral fin absent or present as a minute flap in upper GO corner; head pores reduced, pop* and tp? absent; LL ossicles nearly continuous or separated by a short gap at pores; teeth conical, not caniniform; SERIES 4, V.41,#1 McCOSKER — EELS neurocranium variable in proportions and pos- terior shape; orbit generally reduced; otic bulla moderately to well developed; hyoid condition variable; branchiostegal rays numerous, associ- ated with the hyoid; gill arches reduced, Cs; ab- sent in most genera; IM bones, ribs, and CTP developed; precaudal vertebrae generally more numerous than caudal; coloration nearly uni- form, or darker posteriorly. Allips McCosker Allips McCosker 1972: 116. (Type species; A. concolor McCosker 1972, by original designa- tion.) DESCRIPTION: Body elongate, cylindrical for most of its length; body much longer than tail; head markedly rugose; snout blunt, its under- side grooved; eye small; anterior nostril tubular; GO oblique, lateral, and less than isthmus in length; median fins low, DFO in anterior trunk region; pectoral minute; caudal blunt; tip of snout with numerous sensory hairs; LL ossicles nearly continuous; teeth small, conical, uniserial, intermaxillary teeth largest, separated from those of vomer; skull sloping posteriorly, orbit a narrow slit; SOC weakly developed; maxillae taper posteriorly; HH separated from CH by a gap; branchiostegal rays numerous, slender; pec- toral girdle reduced, only Cl, SCI, and a thin Co (2). yA ETYMOLOGY: From the Greek & VAOS (allos), another, and oe yy (ips; mascu- line), a worm. DISTRIBUTION: A_ single species, known only from the type specimen from Thailand. Bascanichthys Jordan and Davis Bascanichthys Jordan and Davis 1891: 621. (Type species; Caecula bascanium Jordan 1885 = Sphagebranchus teres Goode and Bean 1882, by original designation.) DESCRIPTION: Body elongate, cylindrical, and compressed posteriorly; body longer than tail; head markedly rugose; snout short and blunt, its underside grooved; mouth small; eye small; anterior nostril tubular; GO nearly horizontal, low lateral, = isthmus in length; median fins low, DFO on head; pectoral minute; caudal blunt; LL ossicles nearly continuous; teeth small, conical, uniserial in jaws, intermaxillary teeth separated from those of vomer by a short gap; neurocranium sloping posteriorly (fig. 10), orbit PROCEEDINGS CALIFORNIA ACADEMY OF SCIENCES reduced; nasals and nasal cartilage moderately developed, the nasals closely associated with the ethmoid for their entire length; SOC absent; maxilla weak, slender, and pointed posteriorly, attached before mid-vomer (fig. 16); opercular series moderately developed, their margins en- tire; suspensorium nearly vertical; otic bulla moderately developed; PG pointed anteriorly; HH separated from CH by a narrow suture; branchiostegal rays numerous, often branched basally, all along arch, the distal rays not broad- ened basally; UH either ossified or cartilaginous posteriorly; C; reduced or absent, UPs3-UPs sep- arate; pectoral girdle consists of reduced Cl, SCI, Sc and Co. ETYMOLOGY: From Bascanion (diminutive, from the Greek 32a 7K X vos , malignant), the generic name for the black snake, and oe e&f (ichthys; masculine), fish. DISTRIBUTION: All tropical oceans, with ap- proximately 16 presently recognized species, in- cluding an undescribed eastern Pacific species from Cocos Island. Included are: B. bascanoides Osburn and Nichols (EP), B. cylindricus Meek and Hildebrand (EP), B. panamensis Meek and Hilde- brand (EP), B. ceciliae Blache and Cadenat (EA)*, B. congoensis Blache and Cadenat (EA)*, B. paul- ensis Storey (WA), B. pusillus Seale (IP)*, Sphage- branchus teres Goode and Bean (WA), S. longi- pinnis Kner and Steindachner (IP)*, S. scuticaris Goode and Bean (WA)*, Callechelys myersi Herre (IP), Ophichthys filaria Gunther (IP)*. I/Incertae sedis: Callechelys longissimus Cadenat and Mar- chal (EA)*, and Ophichthys kirkii Gunther (IP)*. Leptenchelys tenuis Tortonese, from New Guinea, is probably a junior synonym of B. longipinnis. REMARKS: The species of Bascanichthys have been reviewed by Storey (1939), Ginsburg (1951) (western Atlantic), and Blache and Cadenat (1971) (eastern Atlantic). The status of Bascanichthys pusillus, considered a junior synonym of B. filaria by Fowler (1931: 316), is uncertain. Caralophia Bohlke Caralophia Bohlke 1955: 1. (Type species; C. loxochila Bohlke 1955, by original designa- tion.) DESCRIPTION: Body elongate, cylindrical, mod- erately compressed posteriorly; body longer than tail; snout subconical, blunt from above, lacking a groove on its underside; eye moderate; anter- ior nostril non-tubular, a hole with lateral projec- ail tions into it; GO ventral, converging forward, longer than isthmus; median fins low, DFO on head; tail tip pointed; LL narrowly separted at pores; teeth bluntly conical, uniserial in jaws and on vomer, intermaxillary teeth slightly larger, separated from those of vomer by a short gap; skull sloping posteriorly, depressed dorsally, or- bit reduced; nasals stout and well developed, their margins entire, tightly joined to ethmoid; nasal cartilage well developed; frontals extend posteriorly to mid-parietals; SOC moderately de- veloped, SO extends from mid-frontals posteri- orly to a strong point; maxilla weak, slender, and pointed posteriorly; opercular series developed, their margins entire; suspensorium posteriorly inclined, jaw angle ca. 100°; otic bulla moder- ately developed; PG broad centrally, closely as- sociated with maxilla anteriorly; hyoid stout, HH fused to CH; branchiostegal rays numerous, along arch, and often branched basally; distal rays along EH broadened basally; UH reduced to a well ossified basal plate with a posterior car- tilaginous filament; tooth plates reduced, UP3-UP4 fused; pectoral girdle reduced to a Cl and frag- ments of a SCI, Sc, and Co. ETYMOLOGY: From the Greek KX PK , head, and NAO Pf cot (lophia; feminine), a mane, in reference to the anterior dorsal fin origin. DISTRIBUTION: A single western Atlantic spe- cies, extending from the Bahamas to the lesser Antilles and Panama. Dalophis Rafinesque Dalophis Rafinesque 1810a: 68. (Type species; D. serpa Rafinesque 1810a = Sphagebranchus imberbis De la Roche 1809.) Pterurus Rafinesque 1810b: 59. (Type species; P. flexosus Rafinesque 1810b = Sphagebranch- us imberbis De la Roche 1809, by monotypy.) Scytallurus Duméril 1856: 199. (Type species; Sphagebranchus imberbis De la Roche 1809, by monotypy.) Pelia Bleeker 1863: 128. (Type species; P. cepha- lopeltis Bleeker 1863.) DESCRIPTION: Body moderately elongate, cyl- indrical, slightly compressed posteriorly; _ tail longer than body; snout sub-conical, grooved on underside; eye small; anterior nostril tubular; GO low on body, extending onto venter, its length ~ isthmus; vertical fins low, DFO well behind GO; pectoral fin rudimentary if present; tail tip blunt; LL ossicles widely separated at 72 pores; teeth conical, uniserial, intermaxillary teeth separated from those of vomer by a gap; skull subtruncate posteriorly, orbit depressed; nasals well developed; SOC weakly developed; maxilla tapers posteriorly; suspensorium nearly vertical; HH separated from CH by a gap; UH ossified posteriorly; gill arches stout, Hs ossified, C; a slender ossified rod, UP3-UPs separate; pec- toral girdle reduced, only a Cl and SCI present. ETYMOLOGY: The significance of the prefix is not obvious, but possibly derived from Dalmatia, then a country on the east side of the Adriatic Sea, and ro) na (ophis; masculine), a snake. DISTRIBUTION: Five eastern Atlantic and Medi- terranean species are presently recognized, in- cluding: D. boulengeri Blache and Bauchot (EA)*, D. multidentatus Blache and Bauchot (EA)*, D. obtusirostris Blache and Bauchot (EA)*, Pelia cephalopeltis Bleeker (EA)*, and Sphagebranchus imberbis De la Roche (M). REMARKS: Blache and Bauchot (1972) have re- cently expanded Dalophis to include Pelia and five nominal species. Their finding (p. 746) that the pectoral fin, though generally absent, may be represented by a ‘miniscule filament’ is in agreement with Lozano Rey’s (1947, p. 546). Through the kindness of Enrico Tortonese | have been able to examine and partially dissect a specimen of Dalophis imberbis. A complete os- teological preparation, however, was impossible. Dalophis has been summarily synonymized with Caecula by previous authors. Its general facies (physiognomy, low median fins, and body depth and taper) and certain osteological fea- tures (dentition, pectoral girdle, and anterior trunk vertebrae) suggest a relationship with Ethadophis and Leptenchelys, genera restricted to the east- ern Pacific. The gill arch skeleton and body/tail proportions are typically ophichthin, and for that reason, Dalophis is placed in the Bascanichthyini with reservations. Ethadophis Rosenblatt and McCosker Ethadophis Rosenblatt and McCosker 1970: 498. (Type species; EF. byrnei Rosenblatt and Mc- Cosker 1970, by original designation.) DESCRIPTION: Body moderately elongate, cyl- indrical anteriorly, becoming compressed pos- teriorly; body equal to or slightly longer than tail; snout rounded, conical from above; eye small; anterior nostril tubular; GO low on body, SERIES 4, V. 41, #1 McCOSKER — EELS extending onto venter, their length = to isthmus; vertical fins low, DFO before GO; tail tip blunt, fleshy in one species; LL ossicles separated at pores; teeth conical, uniserial, intermaxillary teeth separated from those of vomer by a gap; skull subtruncate posteriorly, orbit depressed; nasals well developed; SOC moderately devel- oped, pointed as in Bascanichthys; maxilla tap- ers posteriorly; suspensorium nearly vertical; HH separated from CH by a gap; branchiostegal rays numerous, slender and along hyoid; UH ossified posteriorly; UP3-UP,s separate; pectoral girdle re- duced, only a Cl, SCI, and Co (?) remnant visible by radiograph. Zz ETYMOLOGY: From the Greek cen (ethas), customary or ordinary, and O PoG (ophis; masculine), serpent. DISTRIBUTION: Two species, £. byrnei Rosen- blatt and McCosker and £. merenda Rosenblatt and McCosker, known only from the type speci- mens from Baja California, Mexico. Gordiichthys Jordan and Davis Gordiichthys Jordan and Davis 1891: 644 (Type species; G. irretitus Jordan and Davis 1891, by original designation.) DESCRIPTION: Body extremely elongate, cylin- drical, much longer than tail; snout tapering; eye moderate; GO low lateral; DFO behind nape: teeth conical, recurved, and uniserial. ETYMOLOGY: From Gordius, the horsehair worm, named after I-76 P Alos, the king whose complicated knot was cut by Alexander, and ox evr (ichthys; masculine), fish. DISTRIBUTION: Known from a single deep-water western Atlantic species. REMARKS: The above description is based on Jordan and Davis (1891) and Ginsburg (1951) in that material of G. irretitus, the type species, was unavailable for study. Gordiichthys will be rede- scribed by J. E. Bohlke (personal communication). G. springeri Ginsburg, its sole described con- gener, is a species of Callechelys (fide Rosenblatt and McCosker, 1970, and McCosker and Rosen- blatt, 1972). Leptenchelys Myers and Wade Leptenchelys Myers and Wade 1941: 72. (Type species; L. vermiformis Myers and Wade 1941, by original designation.) PROCEEDINGS CALIFORNIA ACADEMY OF SCIENCES DESCRIPTION: Body elongate, cylindrical, slightly compressed posteriorly; body longer than tail; snout short, subconical; top of head and snout covered with numerous papillae; eye small; an- terior nostril tubular; GO low lateral, crescentic, = isthmus; DFO on head; median fins conflu- ent around tail tip; teeth weak, pointed, uni- serial; skull sloping posteriorly, orbit reduced; SO rounded, not projecting posteriorly; suspen- sorium anteriorly inclined, jaw angle ca. 100°; hyoid arch slender, HH separated from CH by a narrow gap: branchiostegal rays slender, appear from radiograph to lie along hyoid; pectoral girdle reduced, only SCI and Cl visible in radio- graph. ETYMOLOGY: From the Greek NEMTTOS (leptos), thin, and é YK ENS (enchelys; either feminine or masculine, treated as mascu- line), eel. DISTRIBUTION: A _ single species, known only from the type specimen from Playa Blanca, Pacific Costa Rica. REMARKS: This poorly known genus is provision- ally referred to the Bascanichthyini. The generic type, apparently uncollected since the capture of the 115 mm type specimen, was examined and radiographed for the purpose of this study. Its inclusion as a bascanichthyin is based on the numerous branchiostegal rays which appear to contact the hyoid, the posteriorly sloping neuro- Cranium, the low unconstricted gill openings, and the head pore configuration. Caudal fin rays are present on the type specimen, but are more poorly developed than any myrophine’s. Leptenchelys has had an erratic history. Origi- nally placed in the Echelidae on the basis of its having a caudal fin (Myers and Wade, 1941), it was then synonymized with Muraenichthys by Schultz and Woods (1949), re-erected and ex- panded by Schultz, et al. (1953), and finally, re- stricted to a single species (McCosker, 1970) differing markedly from Muraenichthys and re- lated genera. Leptenchelys tenuis Tortonese (1964) from New Guinea is a species of Bascan- ichthys, which is possibly synonymous with B. longipinnis described from Samoa. My exami- nation of the holotype of L. tenuis indicates that Tortonese overlooked the rudimentary pectoral fin, similarly developed in B. longipinnis accord- ing to Storey (1939). The tail tip of L. tenuis is soft and fleshy, but not unlike that of other species of Bascanichthys. 7k) Phaenomonas Myers and Wade Phaenomonas Myers and Wade 1941: 77. (Type species; P. pinnata Myers and Wade 1941, by original designation.) DESCRIPTION: Body elongate, cylindrical, much longer than tail; head markedly rugose; snout blunt, grooved ventrally; eye minute; anterior nostril tubular; GO low lateral, oblique and elongate, = isthmus; DFO mid-head, low, end- ing in anterior trunk region; anal fin absent; teeth small, conical, uniserial, intermaxillary teeth largest, separated from those of vomer; skull sloping posteriorly, orbital foramen a_nar- row slit; nasal cartilage weak; SOC weakly de- veloped; maxilla tapers potseriorly; opercular series reduced, all but opercle serrated at mar- gin; suspensorium anteriorly inclined, jaw angle ca. 100°; otic bulla well developed; PG free and tapering anteriorly; HH separated from CH by a gap; branchiostegal rays numerous, slender and along hyoid; UH club-shaped, cartilaginous pos- teriorly; pectoral girdle reduced, only SCI, Cl, and a thin Co (?); IM bones, ribs, and CTP weakly developed. ETYMOLOGY: Presumably from the Greek PAC Ya>(phainos), to show, and Lo va (monas; feminine), single or alone, in reference to the unique dorsal fin condition. DISTRIBUTION: Two described species, P. pin- nata Myers and Wade ranging from the Gulf of California to Colombia, and P. cooperae Palmer, ranging from Hawaii to the western Indian Ocean (McCosker, In Press). Tribe Ophichthini TYPE GENUS: Ophichthus Ahl, 1789 DIAGNOSIS: Body (head and trunk) and _ tail moderately elongate, cylindrical anteriorly, gen- erally compressed posteriorly; body usually shorter than tail; snout conical or subconical; lower jaw usually included; anterior nostril gen- erally tubular; posterior nostril opens into mouth in most genera; gill openings lateral, their major axis vertical, crescentic, about equal to, less than, or longer than isthmus; median and pectoral fins present; tail tip sharply to bluntly pointed; tp? and pop‘ absent; dentition variable, including caniniform and molariform conditions; neuro- cranium moderately elongate; orbit well devel- oped; otic bulla moderately to well developed; hyomandibular moderately to strongly ridged; 74 opercular series well developed, margins entire; gill arches well developed, C; present in many genera; pectoral girdle generally well developed; intramuscular bones, ribs, and caudal transverse processes well developed; coloration variable. Aplatophis Bohlke Aplatophis Bohlke 1956b: 1. (Type species; A. chauliodus Bohlke 1956b, by original designa- tion.) DESCRIPTION: Body stout, not elongate, later- ally compressed posteriorly; body slightly longer than tail; jaws elongate, the lower projecting be- yond the upper; eye small; posterior nostril in a short tube before and beneath eye; GO verti- cal, lateral and elongate, less than isthmus; DFO behind pectoral tips; tail tip blunt; pop* absent, median head pores absent although frontal com- missure and supratemporal canal are present; LL ossicles widely separated at pores; anterior teeth of both jaws excessively developed as long fang- like canines extending far outside mouth when closed, jaw teeth biserial, those of vomer few and uniserial and continuous with those of inter- maxillary; skull subtruncate posteriorly, orbit large, eyeball displaced anteriorly; PO ossicles weak, not fused as a strut to maxilla; nasals ex- tremely reduced to a short and slender canal, nasal cartilage absent; frontal midline elevated anteriorly forming a sharp ridge extending to ‘ $O; SOC absent, SO anteriorly contacts frontals and separates parietals; maxilla elongate, round- ed posteriorly; opercular series developed, sub- opercle scythe-shaped, preopercle enlarged; sus- pensorium posteriorly inclined, jaw angle ca. 80°; hyomandibular ridged and elongate pos- terodorsally for muscle attachment; otic bulla well developed; PG broad posteriorly, free and terete anteriorly, HH separated from CH by a gap; branchiostegal rays slender, rays of EH joined basally; UH a spike posteriorly; Cs ossi- fied, UPs-UP,s separate, Bo-, absent; pectoral fin moderately developed, girdle well developed, Sc and Co large, actinosts absent (fig. 19F); cau- dal vertebrae fewer than precaudal; coloration nearly uniform, slightly darker dorsally. Ya ETYMOLOGY: From the Greek XT AAT OS (aplatos), unapproachable, and S PSS (ophis; masculine), serpent. DISTRIBUTION: Known from a single trans-At- lantic species. SERIES 4, V.41,#1 McCOSKER — EELS Brachysomophis Kaup Brachysomophis Kaup 1856: 45 (9). (Type spe- cies; B. horridus Kaup 1856 = Ophisurus croc- odilinus Bennett 1833, by monotypy.) DESCRIPTION: Body cylindrical, moderately elongate, longer than tail; snout very short, flat, blunt; jaws elongate, with lower jaw slightly pro- truding; eye moderate; a conspicuous transverse depression exists in the postorbital region of some species; lips with numerous small cirri; an- terior nostril in a very short tube, posterior opens into mouth; GO low lateral, crescentic, longer than isthmus; DFO well behind pectoral tips; tail tip sharply pointed; pop* absent; LL ossicles separated at pores; teeth strong, conical, maxil- lary biserial, dentary uniserial, those of vomer uniserial, large, and widely separated anteriorly but continuous with those of intermaxillary; skull subtruncate posteriorly, orbit reduced; dorsal- most postorbital with a broad anterior flange forming posterior margin of orbit, tightly sutured to smaller postorbitals to form a maxillary strut (fig. 15); nasals and nasal cartilage much re- duced; SOC absent, SO broad posteriorly; max- illa elongate, tapering posteriorly; opercular series well developed, margins slightly serrate; suspensorium nearly vertical; hyomandibular ridged; otic bulla well developed; PG broad posteriorly, tapering to a fine point anteriorly; hyoid arch slender, HH separated from CH by a gap; branchiostegals slender and unbranched, closely associated with hyoid; UH a spike pos- teriorly; Cs ossified, UPs3-UPs separate; pectoral fin and girdle reduced, SCI and actinosts absent, SC reduced; caudal vertebrae fewer than pre- caudal; coloration uniform or darkened dorsally. a ETYMOLOGY: From the Greek 8 PX US (brachys), short, TW MA (soma), body, and op cf (ophis; masculine), serpent. DISTRIBUTION: Known from four central and western Pacific and a single eastern Atlantic spe- cies. Gunther’s (1870) report of B. crocodilinus from the Galapagos Islands was probably er- roneus (Rosenblatt et al., 1972). REMARKS: The species of Brachysomophis are clearly in need of revision (Schultz, 1943; Smith, 1962) but | am lacking the material to make the required comparisons. Brachysomophis atlanticus Blache and Saldana (1972) was recently described from Senegal. The Pacific species of Brachysomo- phis include: Ophisurus cirrhocheilos Bleeker*, PROCEEDINGS CALIFORNIA ACADEMY OF SCIENCES O. corcodilinus Bennett*, Brachysomophis hen- shawi Jordan and Snyder, and B. sauropsis Schultz. The last three nominal species are quite similar and perhaps conspecific. B. cirrhocheilos differs from those species in apparently lacking the postorbital depression, in its longer pectoral fin, and faintly banded coloration (fide Deraniya- gala, 1929). The neurocranium and suspensorium of B. henshawi were illustrated by Gosline (1951a: fig. 2). The formation of the fused postorbital strut in species of Brachysomophis is functionally con- vergent with that of Ophichthus zophochir, how- ever the third, dorsal-most postorbital of B. sauropsis is longer than the second, and enlarged anteriorly to form a cup that braces the eye. Cirrhimuraena Kaup Cirrhimuraena Kaup 1856; 51 (27). (Type species; C. chinensis Kaup 1856, by monotypy.) Jenkinsiella Jordan and Evermann 1905: 83. (Type species; Microdonophis macgregori Jenkins 1903, by original designation.) Calamuraena Whitley 1944: 261. (Type species; Ophichthys calamus Gunther 1870, by original designation.) Calamuraena Gunther 1870, sic Rosenblatt and McCosker 1970: 496, lapsus pro Calamuraena Whitley 1944. DESCRIPTION: General characteristics those of Ophichthus. Differences include: DFO generally on head or above GO; pectoral fin moderately to well developed; upper lip with numerous cirri; pop® generally present; teeth conical, pointed but not enlarged, often multiserial in jaws and on vomer, those of intermaxillary and vomer discontinuous; skull subtruncate posteri- orly (more so than in Ophichthus); PO strut ab- sent; SO and frontal crests weakly developed; maxilla elongate posteriorly; otic bulla not strongly developed; actinosts absent, or 1-2; col- oration uniform, often darkened dorsally. ETYMOLOGY: From the Latin cirrus (cirrh is a wrong but very common form of cirr, mistakenly considered to be from the Greek 4e PEOS [kirrhos] meaning yellow or tawny), tendril, and Muraena, a genus of muraenid eels. DISTRIBUTION: An Indo-Pacific and Red Sea genus, with ca. 9 valid species. REMARKS: Cirrhimuraena is broadly defined to include species that have been referred to Cala- muraena and Jenkinsiella. Subgeneric lines within 75 Cirrhimuraena may be separated on an external morphological basis in the following manner: Mandibular and vomerine teeth uniserial; pop® absent (not determined for all species) Baca Be sci it 3 ee ee subgenus Jenkinsiella Mandibular and vomerine teeth multiserial; pop® present (not determined for all species) Re iS Rie ea eee S83 subgenus Cirrhimuraena The subgenus Jenkinsiella Jordan and Evermann includes Microdonophis macgregori Jenkins, Ophichthys playfairii Gunther*, and may include Jenkinsiella oliveri Seale* and J. inhacae Smith*. The remaining species belong in the subgenus Cirrhimuraena Kaup (which includes Calamur- aena Whitley), including: Cirrhimuraena chinen- sis Kaup, C. taeniopterus Bleeker, C. paucidens Herre and Myers (=C. chinensis?), Ophichthys calamus Gunther, and Ophisurus cheilopogon Bleeker*. Cirrhimuraena may merit further gen- eric division, particularly if confirmed by an osteological study of all the included species. Nelson’s (1966a: 395) description of the fifth ceratobranchial of C. macgregori as having ‘‘the proximal portion of Cs cartilaginous... (and) the distal portion extending posteriorly as a thin filament of cartilage’’ does not agree with my findings on available specimens. In them the distal portion is well ossified. Gosline (1951a: fig. 6) has illustrated the neurocranium and sus- pensorium of C. macgregori. Echelus Rafinesque Echelus Rafinesque 1810a: 63. (Type species; E. punctatus Rafinesque 1810a = Muraena myrus Linnaeus 1758, by Bleeker 1864 as first reviser.) Myrus Kaup 1856: 53 (31). (Type species; M. vul- garis Kaup 1856 = Muraena myrus Linnaeus 1758, by monotypy.) DESCRIPTION: General characteristics those of Ophichthus. Differences include: DFO before pectoral tips; median fins continuous around tail; head pores reduced, pop® present, temporal, postorbital, and interorbital pores absent; ceph- alic surface sensory pores well developed; teeth multiserial, small, nearly granular, intermaxillary teeth continuous with those of vomer; PO strut absent, but a cartilaginous antorbital strut present in E. myrus; SO and frontal crests moderately developed; maxilla slender posteriorly; branch- iostegal rays few; coloration nearly uniform, darker dorsally. 76 2, ETYMOLOGY: From the Greek éypenue (enchelys, considered either feminine or mascu- line, regarded as masculine by Rafinesque, Kaup, and Bleeker), eel. DISTRIBUTION: Known from a shallow and a deep-water species from the Mediterranean and eastern Atlantic, respectively, E. myrus (Linnaeus) and Myrus pachyrhynchus Vaillant. REMARKS: The osteology of the species of Echelus indicates that they are closely related to species of Ophichthus, with particular similarities evidenced in the hyoid arches, gill arches, and neurocrania. The conspicuous presence of a caudal fin led previous workers to consider Echelus to belong to a family or subfamily sep- arate from Ophichthus. The tail fin is clearly a superficial condition that does not merit such suprageneric separation, and thus the family name Echelidae is referred to the synonymy of the Ophichthidae, and the subfamily Echelinae to that of the Ophichthinae. The species of Echelus are easily separable from other ophichthins on the basis of two char- acters, the presence of extremely visible caudal fin rays, and the head pore reduction. The pro- duced caudal fin appears to be a primitive re- tention of an ancestral conditon, although the hard-pointed tail tip characteristic of other oph- ichthines was probably achieved early in the evolution of the subfamily. The absence of the temporal, postorbital, and interorbital pores (as noted by Gosline, 1952) probably reflects the loss of these openings, in that the cephalic later- alis canals and frontal commissures of E£. myrus (Allis, 1903: figs. 5-8, and personal observation) and £. pachyrhynchus (personal observation) have been retained. The antorbital cartilaginous strut of E. myrus, as identifed by Gosline (1952), is not present in E. pachyrhynchus and does not appear ot be an important indicator of phylo- geny. Gosline (1952: 133) suggested that “Ech- elus myrus appears, despite specializations, to be by far the most primitive (or generalized) ophich- thid known. In fact, to a considerable degree it fills in the gap between the Ophichthidae and the Congridae.” | concur with Gosline in consid- ering Echelus to be a very generalized (primitive) ophichthid, however its gill arch reduction, labial posterior nostril, and hyoid apparatus suggest that it is well separated from any congrid-like ancestor. The neurocranium of Echelus myrus has been discussed and illustrated by Storms (1896, as SERIES 4, V. 41, #1 McCOSKER — EELS Myrus vulgaris) and by Gosline (1952), and the external morphology of £. myrus and E. pachy- rhynchus by Blache (1968). Echiophis Kaup Echiophis Kaup 1856a: 46. (Type species; Ophis- urus intertinctus Richardson 1844b, by mono- typy.) Echiopsis Kaup 1856b: 13. Emend. pro Echiophis Kaup 1856a. Crotalopsis Kaup 1860: 12. Also spelled Crotalo- pis, Crotalophis, by other authors. (Type spe- cies; C. punctifer Kaup 1860, by monotypy.) Macrodonophis Poey 1867: 251. (Type species; Conger mordax Poey 1861, by monotypy.) DESCRIPTION: Body cylindrical, moderately elongate, laterally compressed posteriorly; body shorter than tail; snout short, subconical, slightly constricted near tip; jaws elongate and sub- equal; eye moderate; posterior nostril in a short tube before and beneath eye; GO vertical, lateral and elongate, = ishtmus; DFO behind pectoral tips; tail tip blunt; pop*® present; LL ossicles sep- arated at pores; teeth strong, pointed and bi- serial, largest anteriorly in jaws and on ethmoid, those of intermaxillary and vomer biserial and nearly continuous; skull subtruncate posteriorly, preorbital region reduced, orbit large; PO fused to skull and maxilla, forming a strut; nasals slender, nasal cartilage reduced or absent; SOC absent; maxilla elongate, rounded posteriorly, toothed along most of its length; opercular mar- gins entire, subopercle scythe-shaped; suspen- sorium posteriorly inclined, jaw angle ca. 85°; hyomandibular ridged and elongate posterodor- sally for muscle attachment; otic bulla well de- veloped; PG broad posteriorly, free and terete anteriorly; hyoid arch slender, HH separated from CH by a narrow gap; branchiostegal rays numerous, all slender, unbranched, and along hyoid; UH a spike posteriorly; C; ossified, UPs- UP, fused; pectoral girdle and fin developed, Sc and Co large, actinosts present; caudal vertebrae slightly more than precaudal; coloration gener- ally strongly spotted. aif ETYMOLOGY: From the Greek ees (echis), viper, and O PL (ophis; mascu- line), serpent. Kaup emended the suffix to ¥ tf (Opsis), appearance. DISTRIBUTION: A largely New World genus con- taining 3-5 Atlantic and Pacific species. Included are: Ophisurus intertinctus Richardson (WA, and _Cyclophichthys Whitley 1951: — - apparently PROCEEDINGS CALIFORNIA ACADEMY OF SCIENCES EA as Mystriophis cruetzbergi Cadenat), Conger mordax Poey (WA), Crotalopsis punctifer Kaup (WA), and an undescribed eastern Pacific species ranging from the Northern Gulf of California to Panama. Incertae sedis: Mystriophis blastorhinos Kanazawa (WA). REMARKS: The species of Echiophis form a dis- tinctive complex within the Ophichthini, and are difficult to separate at the specific level. The At- lantic species are separable from each other on the basis of spot size, yet a continuous grade is clearly present. Opinions as to the distinctiveness of the Atlantic species have been presented (Jor- dan and Davis, 1891; Springer and Allen, 1932; Ginsburg, 1951) yet a conclusive study is still lacking. The morphology of eastern Atlantic specimens of EF. intertinctus is described by Blache (1971). Elapsopis Kaup Elapsopis Kaup 1856: 45 (9). Emended to Elaps- opsis by other authors. (Type species; Ophi- surus versicolor Richardson 1844, by mono- typy.) 392. Described as a subgenus of Malvoliophis Whitley 1934. (Type species; Ophichthus cyclorhinus Fraser- Brunner 1934, by original designation.) DESCRIPTION: General characteristics those of Leiuranus. Differences include: anterior nostril in a short tube, its rim flared distally; DFO above or slightly before GO; pectoral fin reduced, ~ GO length; vomerine teeth present; opercular margin fringed; C; ossified; actinosts of pectoral girdle present. ETYMOLOGY: From the Latin e/aps (masculine), a snake, and opis, an erroneous spelling of the Greek © roy (opsis), Meaning appearance. DISTRIBUTION: Included are two, probably con- specific, species, from Australia, Lord Howe Is- land, and the southern Caroline Archipelago. REMARKS: This genus is provisionally recognized as distinct from Leiuranus on the basis of the secondary ossification of the fifth ceratobranchial and the presence of actinosts. _ The occasional presence of 1-2 vomerine teeth in specimens of L. semicinctus approaches the | condition of certain specimens of EF. versicolor and E. cyclorhinus. Further investigation may re- sult in the recognition of Elapsopis as a subgenus of Leiuranus. Te, Evips McCosker Evips McCosker 1972: 113. (Type species; E. per- cinctus McCosker 1972, by original designa- tion.) DESCRIPTION: Body moderately elongate, cyl- indrical, laterally compressed posteriorly; body slightly longer than tail; snout blunt, conspicu- ously papillate; lower jaw included; eye large; DFO behind GO; pectoral fin minute; tail tip pointed; pop® absent; LL ossicles continuous, but heavily fractionated; teeth pointed, uniserial, ex- cept those of maxilla which are biserial, largest at intermaxillary which are continuous with those of vomer; skull subtruncate posteriorly, orbit large; SOC pointed posteriorly; maxilla moder- ately produced, slender posteriorly; suspensor- ium nearly vertical, jaw angle ca. 95°; hyoid stout, HH separated from CH by a gap; branchi- ostegal rays numerous, all slender and associated with hyoid; UH a spike posteriorly; C; present, UP3-UP4 separate; pectoral girdle reduced to a slender Cl (SCI not visible in radiograph), actin- osts absent; caudal = precaudal vertebrae; col- oration strong banded. ETYMOLOGY: From the Greek € % (eu, latinized to ev for euphony before a vowel), good, and Ga af (ips; masculine), a worm. DISTRIBUTION: A_ single species known only from the type specimen from the Southern Car- oline Archipelago. Leiuranus Bleeker Leiuranus Bleeker 1853a: 24. (Type species; L. lacepedii Bleeker 1853a = Ophisurus semi- cinctus Lay and Bennett, by monotypy.) Stethopterus Bleeker 1853a: 24. (Type species; Ophisurus (Sphagebranchus) vimineus Rich- ardson 184442 = Ophisurus semicinctus Lay and Bennett 1839, by monotypy.) Machaerenchelys Fowler 1937: 85. (Type species; M. vanderbilti Fowler 1937 = Ophisurus semi- cinctus Lay and Bennett 1839, by original des- ignation.) DESCRIPTION: Body moderately elongate, cy- lindrical, laterally compressed posteriorly; body and tail subequal; snout conical, flattened and grooved on underside; lower jaw included; eye moderate; anterior nostril in a short tube, pos- terior along lip margin; GO vertical, crescentic, shorter than isthmus; median fins low, expanded before tail tip; DFO above GO; pectoral fin moderately developed; tail tip sharply pointed; 78 pop® absent; LL ossicles narrowly separated at pores; teeth small, recurved, uniserial, absent from vomer or 1-2 small teeth; neurocranium subtruncate posteriorly, orbit large; PO strut ab- sent; nasals and nasal cartilage moderately de- veloped; frontal-parietal ridge and SOC devel- oped; maxilla short, articulated ca. mid-vomer; maxillae closely apposed anteromedially (but not abutting as per Gosline, 1951a: 301); suspensor- ium anteriorly inclined, jaw angle ca. 100°; otic bulla well developed; PG slender, braced pos- teriorly by hyomandibular, pointed anteriorly; HH separated from CH by a short gap; branchio- stegal rays numerous, slender (distal-most broad- ened slightly), unbranched, and associated with hyoid (except for innermost rays); UH split an- teriorly (less so than in Phyllophichthus), pro- duced posteriorly as a spike; gill arches weak, Cs absent, anterior half of H: ossified; pectoral girdle reduced, actinosts absent; precaudal ver- tebrae more numeyous than caudal; coloration strongly banded. rs ETYMOLOGY: From the Greek AECL (leios), smooth, and 0 UP KUS, (ouranos; masculine), sky, in reference to the toothless vomer. DISTRIBUTION: A single species, widespread in the central and western Pacific and Indian oceans. REMARKS: Gosline (1951a: fig. 4) has illustrated the neurocranium and suspensorium of L. semi- cinctus. Malvoliophis Whitley Malvoliophis Whitley 1934: 154. (Type species; Bascanichthys hemizona Ogilby = Ophichthys pinguis Gunther 1872, by original designation.) DESCRIPTION: Body elongate, subcylindrical, laterally compressed; body shorter than tail; snout developed, subconical, a short groove on underside; lower jaw included; eye large; an- terior nostril tubular; DFO before GO; pectoral fin moderately developed; pop* absent; LL os- sicles widely separated at pores; teeth conical, not sharply pointed, unserial in jaws, largest at intermaxillary which are widely separated from those of the vomer; skull subtruncate posteriorly, orbit large; PO strut absent; maxilla elongate, slender posteriorly; opercular series weakly de- veloped; suspensorium nearly vertical, jaw angle ca. 95°; otic bulla moderately developed; PG broad _ posteriorly, slender anteriorly; hyoid slender, HH separated from CH by a suture; SERIES 4, V.41,#1 McCOSKER — EELS branchiostegal rays numerous, slender and slightly flattened, the anteriormost ray slightly enlarged; UH broad anteriorly, a short spike posteriorly; Cs; weakly ossified (cartilaginous?), UP;-UPs fused; pectoral girdle reduced, lacking Sc, Co, and actinosts; caudal vertebrae more numerous than precaudal; coloration weakly banded and spotted. ETYMOLOGY: From Malvolio, Lady Olivia’s steward in Shakespeare’s Twelfth Night, and opts (ophis; masculine), serpent. The banded coloration of M. pinguis, the generic type, suggests the cross-gartered legs and yellow socks worn by Malvolio (G. P. Whitley, personal communication). DISTRIBUTION: A single species, reported from southern Queensland, New South Wales, and Lord Howe Island. Myrichthys Girard Myrichthys Girard 1859: 58. (Type species; M. tigrinus Girard 1859, by monotypy.) Chlevastes Jordan and Snyder 1901: 867. (Type species; Muraena colubrina Boddaert 1781, by original designation.) DESCRIPTION: General characteristics those of Pisodonophis and Ophichthus. Differences in- clude: snout short (for an ophichthin), broad from above; DFO well before GO; pectoral fin short, broad-based (fig 34); pop* absent; teeth molariform or granular, multiserial on jaws and vomer, largest at intermaxillary, which are nar- rowly separated from those of vomer; PO strut absent; maxilla elongate, slender posteriorly; teeth of pharyngeal plates generally small and pavement like; pectoral girdle reduced, Co re- duced, Sc and actinosts absent; coloration either spotted, banded, or both. ETYMOLOGY: From the Greek paupPos (myrus), eel, and Ge (2) ba 6 (ichthys; masculine), fish. DISTRIBUTION: A circumtropical genus currently including seven valid species. The genus may be expanded to 12 should certain insular popula- tions of M. maculosus and M. xystrurus prove to be distinct. Included in Myrichthys are: M. bleekeri Gosline (a substitute name for Ophi- surus fasciatus var. semicinctus Bleeker, preoc- cupied) (IP), M. sp. (EP), Muraena acuminata Gronow (WA), M. colubrina Boddaert (IP), M. maculosa Cuvier (IP), Pisodonophis oculatus Kaup PROCEEDINGS CALIFORNIA ACADEMY OF SCIENCES (WA), Ophisurus pardalis Valenciennes (EA)*, and Ophichthys xystrurus Jordan and Gilbert (EP) (see Remarks). REMARKS: Muraena tigrina Ruppell (1826), de- scribed from the Red Sea, is currently placed in the synonymy of Myrichthys maculosus (Cuvier). This creates a homonymy with the eastern Pacific Myrichthys tigrinus Girard (1859), the type spe- cies of the genus. Myrichthys xystrurus (Jordan and Gilbert, 1882) is the next available name for the eastern Pacific species. The genera Pisodonophis and Myrichthys are offshoots from a generalized Ophichthus-like an- cestor. Within the examined species of Pisodon- ophis a trend in characters is evident, with P. cancrivorus the most Ophichthus-like and P. daspilotus the most Myrichthys-like (the char- acters however, are discontinuous enough to al- low generic separation). These trends include: the transformation from pointed to molariform dentition; the advancement of the DFO; the re- duction in length and basal broadening of the pectoral fin; and the reduction of the coracoid and loss of the scapula. Harry (1949) has pointed out that the type locality of M. tigrinus is properly Adair Bay, Mex- ico, not Oregon. Gosline (1951a) has illustrated the neurocran- ium, vertebrae, gill arches, and pectoral girdle of M. maculosus. Mystriophis Kaup Mystriophis Kaup 1856: 45 (10). (Type species; Ophisurus_ rostellatus Richardson 1844, by monotypy.) ; DESCRIPTION: Body cylindrical, moderately elongate, laterally compressed posteriorly; body shorter than tail; snout short, subconical, con- stricted near tip; jaws elongate, the lower slightly inferior; eye moderate; posterior nostril at edge of upper lip, covered by a flap; GO vertical, lat- eral and elongate, = isthmus; DFO above or be- hind pectoral tips; tail tip blunt; pop*® absent, secondary cephalic pores well developed; LL ossicles separated at pores; teeth strong, pointed, biserial, largest anteriorly in jaws and on eth- moid, intermaxillary and uniserial vomerine teeth nearly continuous; coloration nearly uniform, darkened dorsally. ETYMOLOGY: From the Greek pau oT Ploy (mystrion), a spoon, and Spe (ophis; masculine), serpent, in relation to the snout shape of M. rostellatus, the generic type. 7/2) DISTRIBUTION: Two eastern Atlantic and one Japanese species are provisionally placed in My- striophis (see following Remarks). REMARKS: Specimens of Mystriophis were un- available for osteological examination, therefore the above description was prepared from Blache (1971). The genera Mystriophis and Echiophis have been separated by recent authors (Gins- burg, 1951; Kanazawa, 1963; Rosenblatt and McCosker, 1970; Blache, 1971) on the basis of the snout shape and vomerine dentition. M. ros- tellatus Richardson and M. crosnieri Blache, from the eastern Atlantic, are undoubtedly congeneric and appear to be recently derived from an Ophichthus-like ancestor (evidenced in the pop? absence, secondary cephalic pore development, moderate snout length, and body coloration). The species of Echiophis however, appear to have been derived separately from a rather dif- ferent Ophichthus-like ancestor (pop* present, secondary cephalic pores absent, shorter snout, and spotted coloration). Ophisurus porphyreus Temminck and Schlegel, from Japan, tentatively placed in Brachysomophis by Kanazawa (1963), probably belongs in Mystriophis. M. blastorhinos Kanazawa, from French Guiana, is either an Echiophis or belongs in a separate genus. Kana- zawa’s illustration of the pore pattern of M. blas- torhinos is somewhat inaccurate; pop*® and a single, rather than a double, interorbital pore are present (Kanazawa, in litt.). Ophichthus Ahl Ophichthus Ahl 1789: 5. Emended to Ophichthys by other authors. (Type species; Muraena ophis Linnaeus 1758, by original designation.) Innominado Parra 1781: 96. A junior synonym of Muraena ophis Linnaeus 1758, non-binomial. Ophis Turton 1807: 87. (Type species; ““O. macu- lata . . . Spotted Serpent. Shaw Zool., iv. p. 22 . . . Bloch t. 154,” presumably based on Mur- aena ophis Linnaeus 1758, by monotypy.) Cogrus Rafinesque 1810a: 62. (Type species; C. maculatus Rafinesque 1810a, by monotypy.) Ophithorax McClelland 1844: 212. (Type spe- cies; Ophisurus ophis Lacépede 1800, presum- ably based on Muraena ophis Linnaeus 1758, by Jordan, 1919b, as first reviser.) Centrurophis Kaup 1856: 42 (2). (Type species; Ophisurus spadiceus Richardson 1844, as a misidentification of Ophichthys cephalazona Bleeker 1864, by Jordan, 1919b, as first re- viser.) Poecilocephalus Kaup 1856. 43 (5): (Type spe- 80 cies; P. bonaparti Kaup 1856, by monotypy.) Microdonophis Kaup 1856: 43 (6). (Type species; M. altipinnis Kaup 1856, by monotypy.) Coecilophis Kaup 1856: 44 (6) (Type species; Ophisurus compar Richardson 1844 = Ophi- surus apicalis Bennett 1830, by monotypy.) Herpetoichthys Kaup 1856: 44 (7). (Type species; H. ornatissimus Kaup 1856, by Jordan, 1919b, as first reviser.) . Muraenopsis LeSueur, sic Kaup 1856: 46 (11), lapsus pro Muraenophis Lacépéde 1825. Scytalophis Kaup 1856: 46 (13). (Type species; S. magnioculis Kaup 1856, by Jordan, 1919b, as first reviser.) Leptorhinophis Kaup 1856: 46 (14). (Type spe- cies; Ophisurus gomesi Castelnau 1855, by Jordan, 1919b, as first reviser.) Cryptopterus Kaup 1860: 11. (Type species; C. puncticeps Kaup 1860, by monotypy.) Uranichthys Poey 1867: 256. (Type species; Mur- aena hauannensis Bloch and Schneider 1801 = Muraena ophis Linnaeus 1758, by Jordan and Davis, 1891, as first revisers.) Oxydontichthys Poey 1880: 254. (Type species; Ophichthys macrurus Poey 1880 = Ophisurus gomesi Castelnau 1855, by original designa- tion.) Omochelys Fowler 1918: 3. Described as a sub- genus of Pisodonophis Kaup. (Type species; Pisodonophis cruentifer Goode and Bean 1895, by original designation.) Syletor Jordan 1919a: 343. (Type species; Piso- odonophis cruentifer Goode and Bean 1895, by original designation.) Styletor Jordan 1919a, sic Jordan, Evermann, and Clark 1930: 86, lapsus pro Syletor Jordan 1919a. Acanthenchelys Norman 1922: 296. (Type spe- cies; A. spinicauda Norman 1922, by original designation.) Cryptopterenchelys Fowler 1925: 1. Described as a subgenus of Ophichthus Ahl. (Type spe- cies; Cryptopterus puncticeps Kaup, as a sub- stitute name for Cryptopterus Kaup 1860, pre- occupied.) Zonophichthus Whitley 1930: 250. (Type spe- cies; Ophichthys cephalazona Bleeker 1864, by original designation.) Gisenchelys Fowler 1944: 188. Described as a subgenus of Ophichthus Ahl. (Type species; Ophichthys zophochir Jordan and_ Gilbert 1881, by original designation.) Syletophis Whitley 1950: 44. Substitute name for Syletor Jordan 1919a, preoccupied. SERIES 4, V. 41,#1 McCOSKER — EELS Antobrantia Ypiranga Pinto 1970: 13. (Type spe- cies; A. ribeiroi Ypiranga Pinto 1970 = Mur- aena ophis Linnaeus 1758, by original designa- tion.) DESCRIPTION: Body moderately to very elongate, laterally compressed posteriorly; body shorter than tail; snout moderately developed, subconi- cal; lower jaw included; eye moderate to large; anterior nostril tubular, posterior along lower edge of lip or opening into mouth; DFO above or behind GO, but generally before pectoral tips; pectoral fin well developed; LL ossicles sep- arated at pores; dentition variable, from num- erous, small, and multiserial to few, large, and uniserial or biserial, never caniniform; teeth larg- est at intermaxillary and on vomer and some- times separated by a short gap; skull subtruncate posteriorly (fig. 2), orbit large; PO strut devel- oped in some species (fig. 14); nasals and nasal cartilage moderately developed; frontal and SO crests moderately to well developed; maxilla elongate, slender or rounded posteriorly (figs. 14, 16); opercular series well developed, their mar- gins entire (fig. 14); suspensorium nearly vertical to posteriorly inclined; otic bulla well devel- oped; PG broad posteriorly, free and terete an- teriorly; hyoid slender, HH separated from CH by a narrow gap (fig. 17); branchiostegal rays slender, often numerous, unbranched and along arch; UH a spike posteriorly; gill arches com- plete, C; present, UP3-UP.s separate (fig. 18); pec- toral girdle well developed, actinosts present (fig. 19A); caudal vertebrae more numerous than precaudal; coloration generally uniform, although banded or spotted species may be included. Xf ETYMOLOGY: From the Greek O MLS (ophis), snake, and ix evs (ichthus, more correctly written ichthys; masculine), fish. DISTRIBUTION: A circumtropical genus with ap- proximately 55 species. REMARKS: Ophichthus contains approximately 55 valid species, for which no fewer than 25 nominal genera have been erected. The genera Gisenchelys Fowler and Zonophichthus Whitley are here included in the synonymy of Ophich- thus. Hubbs’ (1932) inclusion of Ophis Turton in the synonymy of Ophichthus was neither in- cluded nor commented upon by Bohlke and Robins (1959) in their synonymy of Ophichthus ophis. Bohlke and Menezes (in litt.) have found Antobrantia, type species A. ribeiroi Ypiranga Pinto, to be an exact synonym of Ophichthus ophis. PROCEEDINGS CALIFORNIA ACADEMY OF SCIENCES The generic concepts of earlier authors, par- ticularly Kaup, are nearly all based on differ- ences in coloration or dentition. | have at- tempted to examine as many of these subgenera as possible. As broadly defined herein, Ophich- thus exhibits a wide variety of morphological conditions, yet among the diverse types ex- amined, | have found no accompanying differ- ences that are clearly generic. It is particularly perplexing however, that the few osteological variations within the genus that might merit gen- eric recognition are present in O. triserialis, O. zophochir, and the type species, O. ophis. These concern the development of the postorbital strut and the posterior shortening of the maxilla. | have attempted to identify subgeneric line- ages within Ophichthus, using available generic synonyms as subgeneric names. The following analysis however, must be regarded as strictly provisional: Body = tail; DFO above or before GO; teeth minute and uniserial throughout; col- oration of several species strongly spotted)... nonseee subgenus Microdonophis Kaup Body shorter than tail; DFO behind GO, generally above pectoral tips; teeth larger, ier GANS) ee eee Anterior nostril a broad tube, flared at tip; intermaxillary block hooked down- ward, and intermaxillary teeth directed horizontally backward; tail tip hard and sharply-pointed; median fins conspicu- ously expanded before tail tip -................... a subgenus Centrurophis Kaup Anterior nostril tubular, not flared distally; intermaxillary teeth erect, with tips hook- ed back; tail tip hard and blunt, not sharply-pointed; median fins conspicu- spicuously expanded before tail tip -....... ah ee ee subgenus Centrurophis Kaup Postorbital bones tightly sutured and bracing maxilla, forming a strut; maxilla rounded posteriorly; jaw and vomerine dentition bi- or triserial -....................... ie eee ee A subgenus Ophichthus Ahl Postorbital bones not tightly sutured and not forming a strut with maxilla; maxilla elongate and slender posteri- orly; dentition variable, either uniserial OTN UNI CUS tile lee Se eee es oe een subgenus Coecilophis Kaup Included in the subgenus Microdonophis are: M. altipinnis Kaup (IP), M. erabo Jordan and 81 Snyder (IP), Ophichthys polyophthalmus Bleeker (IP)*, and O. melanochir Bleeker (IP). Included in the subgenus Centrurophis are the generic types of Zonophichthus and_ probably Poecilocephalus. \Included species: Ophichthys cephalazona Bleeker (IP), and Poecilocephalus bonaparti Kaup (IP)*. Included in the subgenus Ophichthus are the generic types of Innominado, Ophis, Cogrus, Uranichthys, Ophithorax, and Gisenchelys. In- cluded species: Muraena ophis Linnaeus (WA, EA), Muraenopsis triserialis Kaup (EP), and Oph- ichthys zophochir Jordan and Gilbert (EP). A complete synonymy of O. ophis is provided by Bohlke and Robins (1959). Included in the subgenus Coecilophis is Ophi- surus apicalis Bennett (IP), and presumably many other species listed below as incertae sedis. Incertae sedis: Acanthenchelys spinicauda Re- gan (WA)*, Cogrus maculatus Rafinesque (M)*, Conger urolophus Schlegel (IP), Cryptopterus puncticeps Kaup (WA), Echelus rufus Rafinesque (M)*, Omochelys marginatus Fowler (WA) (= Ophichthus cruentifer 2), Ophichthus asakusae Jordan and Snyder (IP), O. manilensis Herre (IP)*, O. melanoporus Kanazawa (WA), O. multiserialis Norman (IP)*, O. retifer Fowler (IP)*, O. roseus Tanaka (IP)*, O. stenopterus Cope (IP)*, Ophich- thys ascensionis Studer (WA)*, O. ater Peters (EP)*, O. callensis Gunther (EP), O. derbeyensis Whitley (IP)*, O. episcopus Macleay (IP)*, O. evermanni Jordan and Snyder (IP), O. frontalis Garman (EP), O. garretti Gunther (IP)*, O. limkouensis Chen (IP)*, O. macrops Gunther (IP)*, O. madagascari- ensis Fourmanoir (IP)*, O. melanochir Bleeker (IP), O. pacifici Gunther (EP), O. retropinnis Eigen- mann (WA)*, O. unicolor Regan (IP)*, O. woo- suitingi Chen (IP)*, Ophisurus celebicus Bleeker (IP)*, O. gomesii Castelnau (WA), O. grandoculis Cantor (IP)*, O. macrochir Bleeker (IP), O. mar- ginatus Peters (IP)*, O. parilis Richardson (WA)*, O. reguis Richardson (EA)*, O. remiger Valen- ciennes (probably EP, no type locality given)*, O. rutidoderma Bleeker (emended by most au- thors to rhytioderma), (IP)*, O. rutidodermatoides Bleeker (emended by most authors to rhytioder- matoides), (IP), Pisodonophis cruentifer Goode and Bean (WA), Scytalophis magnioculis Kaup (WA)*, and two undescribed Eastern Pacific spe- cies. Ophisurus Lacépede Ophisurus Lacépéde 1800: 195. (Type species; Muraena serpens Linnaeus 1758, as restricted 82 by Risso 1826.) Leptognathus Swainson 1838: 334. (Type species; L. oxyrhynchus Swainson 1838 = Muraena ser- pens Linnaeus 1758, by original designation.) Leptorhynchus Smith 1847: 244. (Type species; L. capensis Smith 1847 = Muraena serpens Linnaeus 1758.) Anepistomon Gistel 1848: ix. A substitute name for Leptorhynchus Smith 1847, preoccupied. DESCRIPTION: General characteristics those of Ophichthus. Differences include: body very elongate; snout attenuate, jaws elongate, slender, incapable of closing completely in adults; lower jaw included; anterior nostril lateral, at mid- snout (fig. 24B), posterior on outer lip and cov- ered by a flap; DFO beyond tips of pectoral fins; cephalic surface sensory papillae well developed (figs. 24B, C); teeth conical, nearly uniserial, en- larged along vomer; maxilla elongate, slender and produced posteriorly; coloration uniform, darker dorsally. ETYMOLOGY: From the Greek & pey (ophis), serpent, and OY PX (oura), tail, and -of (-us), masculine suffix. DISTRIBUTION: Presumably a single species, known from the Mediterranean, eastern Atlantic, Cape of Good Hope, the western and central Indian Ocean, eastern Australia, New Zealand, and Japan. Ophisurus macrorhynchus Bleeker 1852, from Japan, differs slightly in morphometry from the Mediterranean Muraena serpens, yet a conclusive comparison has not been made. REMARKS: The Atlantic and Pacific populations of O. serpens, if indeed conspecific, represent the most disjunct of known ophichthid distribu- tions. Historically, the Mediterranean population could be explained as a relict of a Tethyan spe- cies. Parallels may be found in the disjunct pop- ulations of other eel species (see D. Smith, 1969), such as Kaupichthys diodontus, which presum- ably ranges from the western Atlantic, and the central and western Pacific, and Indian oceans J. L. B. Smith, 1965), and Dysomma anguillare, presumably ranging from the tropical western Atlantic, Indian Ocean, and off Japan and China (Robins and Robins, 1970). The identity of Ophichthus multiserialis Nor- man 1939 from the Gulf of Aden, included by J. L. B. Smith (1962: 455) in Ophisurus, has not been determined. Oxystomus Rafinesque (1810b: 62) has been considered by most authors to be synonymous with Ophisurus. Castle (1969) however, has con- SERIES 4, V.41, #1 McCOSKER — EELS firmed Bertin’s (1935) suggestion that the type species, Oxystomus hyalinus, is based on a post- metamorphic Serrivomer rather than a_ larval Ophisurus serpens. The external morphology and osteology of Ophisurus serpens (as O. novaezelandiae Hector) was described by Knox (1870). Phyllophichthus Gosline Phyllophichthus Gosline 1951a: 316. (Type spe- cies; P. xenodontus Gosline 1951a, by original designation.) DESCRIPTION: Body moderately elongate, cylin- drical, laterally compressed posteriorly; body = tail; snout produced, subconical, grooved on underside; lower jaw included; eye large; an- terior nostril subtubular, its posterior borders ex- tending downward into leaflike appendages, pos- terior nostril opens into mouth; GO vertical, crescentic, shorter than isthmus; dorsal and anal fins low, expanded before the sharply pointed tail tip; DFO above GO; pectoral fin well de- veloped; pop® absent; LL canal ossicles are short coils separated at pores; teeth conical, minute, and uniserial in jaws, recurved on mandible, ab- sent from vomer; neurocranium elongate, sub- truncate posteriorly, orbit large; PO strut absent; nasals and nasal cartilage well developed; SOC absent; maxilla fragile and not produced, articu- lated ca. mid-vomer (fig. 16); preopercle re- duced; suspensorium anteriorly inclined, jaw angle ca. 100°; otic bulla moderately developed; PG slender, short, and pointed anteriorly; hyoid slender, HH absent, CH with a minute anterior condyle (possibly a remnant of a fused HH2); branchiostegal rays numerous, filamentous, free from hyoid; UH split anteriorly, produced pos- teriorly as a slender spike; gill arches weak, pharyngeal teeth minute, C; absent, Hi ossified only at tip; pectoral girdle well developed, acti- nosts present; precaudal vertebrae more num- erous than caudal; coloration uniform. ETYMOLOGY: From the Greek P vAroY (phyl- lon), leaf, and Ophichthus (masculine), a genus of ophichthids, in reference to the leaf-like an- terior nostrils. DISTRIBUTION: Two described Pacific species, P. xenodontus Gosline from the central and west- ern Pacific and Indian oceans, and P. macrurus McKay*, described from Australia. Pisodonophis Kaup Pisodonophis Kaup 1856a: 47. Also spelt Piso- PROCEEDINGS CALIFORNIA ACADEMY OF SCIENCES dontophis by other authors. (Type species; Ophisurus cancrivorus Richardson 1844, as restricted by Bleeker 1865.) Pisoodonophis Kaup 1856b: 17. Emend. pro Pisodonophis Kaup 1856a. Brachycheirophis Fowler 1944: 190. (Type spe- cies; Pisodonophis daspilotus Gilbert, in Jor- dan and Evermann 1898, by original designa- tion.) DESCRIPTION: General characteristics those of Ophichthus. Differences include: pectoral fin broad-based; pop® present or absent; teeth mol- ariform or granular, multiserial on jaws and vo- mer, largest at intermaxillary with a short gap separating those of the vomerine block; PO strut absent; maxilla elongate, slender posteriorly; innermost branchiostegal rays free from hyoid; teeth of pharyngeal plates generally small, pave- ment-like; pectoral girdle reduced, actinosts ab- sent, Sc absent in one species; coloration gen- erally uniform, spotted in one species. ETYMOLOGY: From the Greek 77 © oof (pisos), pea, °o gwovys (odons), tooth, and Oo Pry (ophis; masculine), serpent. DISTRIBUTION: Known from 8-10 species, a circumtropical genus. Included are: Ophisurus cancrivorus Richardson (IP), O. boro Hamilton- Buchanan (IP), O. hijala Hamilton-Buchanan (IP)*, O. hoevenii Bleeker (IP), O. hypselopterus Bleeker (IP)*, O. semicinctus Richardson (EA)*, Pisodonophis copelandi Herre (IP), P. daspilotus Gilbert (EP), Pisoodonophis zophistus Jordan and Snyder (IP) (=P. cancrivorus?), Ophichthys dro- micus Gunther (EA)*. REMARKS: The relationships of Pisodonophis, Myrichthys, and their Ophichthus-like ancestors are discussed in the remarks on Myrichthys. Omochelys Fowler, considered by recent au- thors to be a junior synonym of Pisodonophis, is here referred to Ophichthus. Omochelys cru- entifer is unlike the species of Pisodonophis in having few branchiostegal rays, all of which are attached to the hyoid. Ganguly and Nag (1964) described the func- tional morphology of the pectoral girdle and the acranial myomeric musculature of a fish they identified as Ophichthus boro. The pectoral girdle they have crudely illustrated is clearly not that of Pisodonophis boro, and is probably from a species of Ophichthus. Tilak and Kanji (1967) noted their error and properly described the morphology of the pectoral girdle of Pisodono- 83 phis boro in relation to its habit. Pogonophis Myers and Wade Pogonophis Myers and Wade 1941: 78. (Type species; P. fossatus Myers and Wade 1941, by original designation.) DESCRIPTION: Body moderately elongate, cylin- drical, laterally compressed posteriorly; body slightly longer than tail; snout blunt; jaws sub- equal, a single pendulous barbel on upper lip; eye large; anterior nostril tubular, with a pos- terior tag; DFO behind GO; pectoral fin minute; tail tip pointed; preopercular pores vertically, not obliquely aligned, pop*® absent; LL ossicles nearly continuous; teeth pointed, uniserial ex- cept on maxillae where biserial, largest at inter- maxillary which is continuous with those of vo- mer; skull subtruncate posteriorly, orbit large; PO strut absent; nasals and nasal cartilage re- duced; SOC weakly pointed; maxilla elongate, slender posteriorly; suspensorium posteriorly in- clined, jaw angle ca. 80°; otic bulla moderately developed; PG broad posteriorly, free and terete anteriorly; hyoid slender; HH separated from CH by a broad gap; branchiostegal rays numer- ous, all slender and associated with hyoid, the anteriormost 3-4 rays fused basally; UH broad anteriorly, short and slender posteriorly; _ gill arches reduced, C; absent, UP3-UPs separate; pectoral girdle weakly developed, actinosts ab- sent; caudal = precaudal vertebrae; coloration strongly spotted. ETYMOLOGY: From the Greek Tf WyYwWwyv (pogon), beard, and 6 ees (ophis; masculine), serpent, in reference to the consipcu- ous barbel on the snout. DISTRIBUTION: A single eastern Pacific species, ranging from the Gulf of California to Peru (as Ophichthus afuerae Hildebrand). Quassiremus Jordan and Davis Quassiremus Jordan and Davis 1891: 622. (Type species; Ophichthus evionthas Jordan and Bollman 1889, by original designation.) DESCRIPTION: Body moderately elongate, cylin- drical, laterally compressed posteriorly; body slightly longer than tail; snout developed, coni- cal, a short groove on underside; lower jaw in- cluded; eye large; DFO behind GO; pectoral fin minute; tail tip pointed; pop*® present; LL os- sicles continuous; teeth pointed, uniserial, larg- est at intermaxillary and continuous with those 84 of vomer; skull truncate posteriorly, orbit large; PO strut absent; nasals well developed, nasal cartilage weakly developed; SOC strongly point- ed; maxilla elongate, slender posteriorly; oper- cular margins entire; suspensorium anteriorly in- clined, jaw angle ca. 100°; otic bulla well de- veloped; PG broad posteriorly, free and terete anteriorly; hyoid slender, HH separated from CH by a broad gap: branchiostegal rays numerous, all slender and associated with hyoid, the an- teriormost 3-4 rays fused basally; UH broad an- teriorly, short and slender posteriorly; gill arches reduced, Cs absent, UP3;-UPs separate; pectoral girdle reduced to a slender Cl and SCI; caudal = precaudal vertebrae; coloration strongly spot- ted. ETYMOLOGY: From the Latin quassus, obliter- ated, and ramus (masculine), oar, in reference to the minute pectoral fins. DISTRIBUTION: Known from 3 New World spe- cies. Included are: Ophichthus evionthas Jordan and Bollman (EP), Ophichthys nothochir Gilbert (EP) (including Q. parvipinna Seale), and Quas- siremus productus Seale (WA) (including Q. gos- lingi Beebe and Tee-Van). REMARKS: The osteological description of Quas- siremus is based on Q. nothochir. Q. evionthas, the generic type, was examined from radiographs and gill arch dissection. Scytalichthys Jordan and Davis Scytalichthys Jordan and Davis 1891: 635. De- scribed as a subgenus of Mystriophis Kaup. (Type species; Ophichthys miurus Jordan and Gilbert 1882, by original designation.) DESCRIPTION: Body cylindrical, moderately elongate, longer than tail; snout very short, flat, subconical; jaws elongate, lower jaw inferior; eye moderate; anterior nostril in a short tube, posterior opens into mouth; GO _ low lateral, crescentic, longer than isthmus; DFO well be- hind pectoral tips; pectoral fin reduced; tail tip bare, sharply pointed; pop® absent; LL ossicles continuous, not separated at pores; teeth strong, conical, those on maxilla in two widely separated rows, uniserial on dentary; those on vomer large, uniserial, widely separated, and continuous with those on intermaxillary; skull subtruncate pos- teriorly, orbit reduced; PO not bracing maxilla; nasals very reduced, nasal cartilage reduced; SOC absent, a short broad posterior projection; maxilla elongate, bifurcate posteriorly, the distal SERIES 4, V. 41, #1 McCOSKER — EELS section elongate and terete, the medial shorter and pointed; opercular series well developed, their margins entire; suspensorium nearly verti- cal, hyomandibular strongly ridged; otic bulla well developed; PG slender, a fine point anteriorly; hyoid arch slender, HH widely separated from CH; branchiostegal rays numerous, all slender, unbranched, and along arch; UH a short spike posteriorly; Cs short, UP3-UPs nearly fused, lower pharyngeal plate elongate; pectoral girdle re- duced, SCI and actinosts absent; IM bones, ribs, and CTP well developed; caudal vertebrae fewer than precaudal; coloration strongly spotted. ETYMOLOGY: From the Greek GZKUTAKANR (skytale), viper, and cyous (ichthys; mas- culine), fish. DISTRIBUTION: A single eastern Pacific species, ranging from the Galapagos Islands to the Gulf of California and Guadalupe Island. REMARKS: Schultz (1942) has commented on the generic status of Scytalichthys with regard to it dentition. Xyrias Jordan and Snyder Xyrias Jordan and Snyder 1901: 864. (Type spe- cies; X. revulsus Jordan and Snyder 1901, by original designation.) DESCRIPTION: Body cylindrical, moderately elongate, longer than tail; snout short, subconi- cal; jaws elongate, the lower inferior; eye mod- erate; anterior nostril non-tubular, laterally ovate, with a minute ventral flap; posterior nostril out- side mouth and covered with a flap; GO low lat- eral, crescentic, longer than isthmus; DFO be- hind pectoral tips; pectoral fin moderate; tail tip bare, pointed; pop* absent; teeth strong, conical, not extremely enlarged; those of vomer largest, uniserial, and nearly continuous with those of intermaxillary, jaw teeth multiserial; orbit depressed; PO not bracing maxilla; nasals and nasal cartilage reduced; SOC with a short posterior projection; maxilla elongate, bifurcate posteriorly; opercular series well developed, their margins entire; suspensorium nearly verti- cal; PG slender, tapering to a fine point an- teriorly; hyoid arch moderately slender, HH sep- arated from CH by a gap; branchiostegals num- erous, slender, unbranched and along arch; C; ossified, UPs-UPs united by a suture; pectoral girdle complete, SCI, Cl, Sc, Co, and actinosts visible in radiograph; IM bones, ribs, and CTP well developed; caudal vertebrae fewer than pre- caudal; coloration of numerous small spots. PROCEEDINGS CALIFORNIA ACADEMY OF SCIENCES ETYMOLOGY: From the Greek JuPioy (masculine), a shaveling, in reference to the ab- sence of cirri on the lips. DISTRIBUTION: A single species from Japan. Glover (1973) reported that it is caught occasion- ally by shrimp trawlers in the Misaki, Kochi, and Kumano prefectures. REMARKS: This description was prepared from radiographs and a gill arch dissection of the holotype. Comparison with Previous Classifications The results of the present study offer certain insights into the reliability of superficial morpho- logical characters as a basis for classification. The most recent attempt at a compilation of genera within the Ophichthidae was that of Rosenblatt and McCosker (1970). That study was based on an analysis of previous literature and an exami- nation of all recognized genera except Neen- chelys, Malvoliophis, and Benthenchelys (Castle, 1972, was the first to suggest that Benthenchelys is an ophichthid). The following discrepancies ex- ist between the results of that study and the present: (a) Echelus was included in the Myrophinae (as Echelinae), following Gosline (1952); in the present study Echelus is shown to be an ophich- thine, and the presence of a caudal fin is not regarded as a definitive character separating the ophichthid subfamilies; (b) Leptenchelys was included in the Myro- phinae, again based on the caudal condition; in the present study it is suggested that its rela- tionships lie with the Bascanichthyini; (c) Lamnostoma was considered to be syn- Oonymous with Caecula; in the present study osteological characters have been identified that allow its generic separation; (d) Omochelys was considered to be synony- mous with Pisodonophis on the basis of tooth characters; in the present study it is indicated that its affinities lie closer to species of Ophich- thus than of Pisodonophis, and Omochelys is tentatively placed in the synonymy of Ophich- thus awaiting a revision of that large and cum- bersome genus; (e) Jenkinsiella, and Microdonophis and Zono- phichthus, were recognized at the generic level; in the present study no osteological characters have been found to support their generic recog- nition, and they are reduced to subgeneric rank under Cirrhimuraena and Ophichthus, respec- 85 tively. The results of this comparison however, have shown that the previous classification, based for the most part on external morphological char- acters, was not seriously upset by a classification based primarily on osteology. The finding of greatest Consequence was that the affinities of Echelus lie with the Ophichthinae rather than the Myrophinae; the caudal fin condition was shown to be trivial when compared with the “Ophichthus-like’’ condition of numerous osteo- logical characters. In that the majority of classification schemes within the teleosts are now based on external morphological characters, it is somewhat reas- suring to realize that at least within the Ophich- thidae, the external morphological classification parallels that based on osteology. EVOLUTION OF THE OPHICHTHIDAE Relationship to other Anguilliforms Recent authors have summarily recognized 23 families within the suborder Anguilloidei (Green- wood, et al., 1966; Gosline, 1971). Subsequent studies have changed this listing in the following manner. Robins and Robins (1970) have ex- panded the Dysommidae to include the Dysom- minidae and the Nettodaridae and (1971) pro- visionally united the Nessorhamphidae with the Derichthyidae. Smith (1971) has provisionally re- moved Coloconger from the Congridae and erected the family Colocongridae. The above- mentioned authors have allocated the anguilloid families to five superfamilies, namely: the An- guilloidea (those eel families with paired fron- tals), Synaphobranchoidea (Synaphobranchidae, Dysommidae and Simenchelyidae), Congroidea (Congridae, Colocongridae, Muraenesocidae, Net- tastomidae, and Macrocephenchelyidae), Nem- ichthyoidea, and Ophichthoidea. The superfamiliy Ophichthoidea is restricted to the Ophichthidae. They appear most closely related to the superfamily Congroidea, but differ in the condition of several major characters (sev- eral fundamental characteristics of the Ophich- thidae and related apodal families are indicated in Table 7). As Smith (1971) has shown, the con- groids possess a complete pterygoid arch and a nearly complete gill arch skeleton. Conversely, the ophichthid pterygoid is reduced and. well separated from the vomer, and their gill arches exhibit various stages of reduction. The ophich- thids also differ from the congrids in possessing 86 numerous and overlapping branchiostegal rays, a fused frontal commissure, and a cartilaginous connection between the first epibranchial and second infrapharyngobranchial of the gill arch skeleton. Certain congrids, primarily within the subfamily Heterocongrinae, are similar to oph- ichthids in the development of laminar ribs, re- duced neural spines, an elongate body, and a fleshy tail tip (cf. Bohlke, 1957; Rosenblatt, 1967). Rosenblatt discussed these similarities and sug- gested (p. 95) that ‘the superficial similarties be- tween Gorgasia and the ophichthids are cer- tainly parallel adaptations to a similar mode of life, and the minor osteological similarities may be adaptations as well.” The general condition of the primitive ophich- thid genera, viz., Echelus and Ophichthus, ap- pears to be derived from a congrid ancestor. To my knowledge, no known congrid exhibits defin- itive ophichthid characters in an intermediate state of development, particularly the fused frontal commissure and overlapping branchio- stegal rays. The achievement of the ophichthid condition may have been a quantum jump in apodal evolution, attained by either an advanced congrid or congrid ancestor by means of a rapid integration of the above mentioned characters. Yet this may only be presumed until further evi- dence from extant or fossilized species is ob- tained. SERIES 4, V. 41, #1 McCOSKER — EELS Evolution within the Ophichthidae The purposes of any classification system are to best reflect the phylogenetic history of the taxa involved and to provide predictability, such that a newly discovered taxon might be properly placed without upsetting the system. An evolu- tionary scheme may only be inferred from exist- ing data, and must carefully weigh the prob- abilities related to each hypothetical pathway. In dealing with supra-specific categories, one is faced with the difficulty of delineating groups on a subjective basis. The history of the Ophich- thidae, however, would suggest that the genus, tribe, and subfamily reflect certain biological realities in an evolutionary sense. The present study has attempted to define and identify these groupings. The presence of several fundamental ostelogical characters in both major lines of the Ophichthidae suggests a monophyletic origin of the family. These characters include the frontal commissure, the epibranchial interconnections, and the unique manner of branchiostegal over- lap. The two subfamilies differ trenchantly in the manner in which the branchiostegal rays are at- tached to the hyoid, yet this difference need not preclude a monophyletic origin. The inferred subfamilial and tribal evolution within the Oph- ichthidae is illustrated below in Figure 37. Gen- eric evolution within each tribe is treated in the following discussion. Callechelyini Sphagebranchini Ophichthini Bascanichthyini Bentnenchelyini / ‘ , a Myrophini CONGRID-LIKE ANCESTOR Figure 37. Proposed evolutionary relationship of ophichthid tribes. PROCEEDINGS CALIFORNIA ACADEMY OF SCIENCES Sehtsmorhynehus (1) Sehultatdta (2) snout, nostril, and hyoid dentition and neurocranium modified reduced, pop? absent ee aes (ea 20) pectoral fin lost, girdle reduced, Ahlta (1) gill arches reduced, CTP developed vomerine teeth lost jaws modified Myrophts (9) Pseudomyrophis (2) pectoral fin and girdle reduced posterior epipleurals lost Neenchelys (2) posterior nostril labial _-Benthenchelys (1) anterior nostril non-tubular, neurocranium shortened anterior nostril tubular, posterior nostril before eye, pectoral fin well developed, scapula and coracoid present CONGRID-LIKE ANCESTOR Figure 38. Proposed evolution of the Myrophinae. Number of species in parentheses. 87 88 Myrophini and Benthenchelyini. Two. condi- tions indicate that the Myrophinae probably arose from the first dichotomy of the ophichthid lineage (Figs. 37-38). These are the presence of a well developed caudal fin and the unique my- rophine branchiostegal condition in which the majority of the rays are free from and well be- hind the epihyal. The attached rays, like those of the congirds, are basally broadened. Although the species of Echelus possess a myrophine-like caudal fin, they also possess an ophichthine-like hyoid and other osteological conditions that link them with the Ophichthinae. Primitive conditions of certain characters with- in the Myrophinae include: (a) teeth present on the dentary, intermaxil- lary, maxilla, and vomer; (b) pectoral fin moderately to well developed; (c) coracoid and scapula present; (d) posterior nostril lateral; (e) first basibranchial ossified; (f) seven attached branchiostegals along epi- hyal; (g) anterior nostril tubular; (h) eye not enlarged, its diameter ten or more in head length. — Benthenchelys cartieri, the single species with- in the Benthenchelyini, is highly specialized for its unique adaptation to a pelagic existence (see Remarks on Benthenchelys). It has however, re- tained certain primitive conditions which would suggest that it separated early from the ancestral myrophine stock. Benthenchelys possess all the primitive conditions listed above except that its first basibranchial has been reduced to cartilage. Whether Benthenchelys separated before, after, or along with the Neenchelys-Pseudomyrophis lineage was not discerned from the available data. The posterior nostril of Neenchelys and Pseudomyrophis, like that of Benthenchelys, is lateral. The nostril of Benthenchelys is extremely atypical within the Ophichthidae in that it is markedly more dorsally located than it is in other genera. The partial re-ossification of the third hypobranchial in Benthenchelys probably represents a secondary specialization; the third hypobranchial of all congrids, and presumably that of the basal ophichthids, is cartilaginous. Similarities between Neenchelys and Pseudo- myrophis were described earlier in this study. These genera have retained the first basibranch- ial, the pectoral fin, and pectoral girdle, but an attached branchiostegal has been lost in two of SERIES 4, V. 41, #1 McCOSKER — EELS the three species examined. Pseudomyrophis has become specialized in the reduction of its pec- toral fin and girdle, and in one species the body has become extremely elongate. Neenchelys is somewhat more primitive in that it has retained a developed pectoral fin and its girdle is not reduced. Beyond the level of the Pseudomyrophis-Neen- chelys separation, the posterior nostril has be- come labial in position. In general, the ophich- thid posterior nostril lies within the mouth or along the lip and is covered by a flap. The rever- sion to the more generalized and presumably primitive condition in which the posterior nostril lies along the outer edge of the lip has been achieved numerous times. This is evidenced in the contrasting posterior nostril conditions of closely related genera such as I/Ichthyapus and Apterichtus, Ophichthus and Ophisurus, and par- ticularly between the species of Muraenichthys. A distinction however, should be made between the condition of the earlier-removed myrophine genera (Benthenchelys, Pseudomyrophis, and Neenchelys) and that of the remaining ophich- thids. In no case does it appear that the posterior nostril has secondarily returned to a congrid-like placement. The next major dichotomy is that of the Ahlia- Myrophis lineage. That they arose from a com- mon ancestor is evidenced by their conspicuous specialization in which the pleural ribs are lim- ited to the anterior trunk vertebrae. The species of Myrophis differ considerably in external ap- pearance, primarily in body and snout elonga- tion, even though they differ little osteologically. Ahlia egmontis, by comparison, is specialized in having lost the vomerine dentition, posteriorly shortened maxillae, modified pterygoids, the hypohyals either lost or fused to the ceratohyal, and the dorsal fin origin withdrawn to the level of the anus. The remainder of the Myrophini comprise the species of Muraenichthys, Schultzidia, and Schismorhynchus, commonly called the ‘worm eels’. The reductions and specializations of the latter two genera must preclude them as ances- tors to Muraenichthys or to each other. Within Muraenichthys however, are species sufficiently generalized that either Schismorhynchus or Schultzidia might have been derived from them. As discussed in the remarks on Muraenichthys, subgeneric lineages, although including rather divergent extremes, are bridged by a broad spec- trum of morphological conditions. The more gen- PROCEEDINGS CALIFORNIA ACADEMY OF SCIENCES eralized species of the subgenus Scolecenchelys are likely ancestors to the subgenus Muraenich- thys and the genera Schultzidia and Schismor- hynchus. Species of Scolecenchelys have the pos- terior nostril within the mouth, separate upper pharyngeal tooth plates, an ungrooved snout, uniserial or biserial conical teeth, and a third pre- opercular pore (pop*). Species of the subgenus Muraenichthys have an external posterior nostril, multiserial granular dentition, a deep body, a blunt snout, and fused upper pharyngeal teeth. The species of Schultzidia have lost the pop® and the dentition is extremely reduced, being absent on the vomer, absent or embedded on the inter- maxillary, and minute or villiform in the jaws. Schismorhynchus labialis is specialized in quite a different manner, and seems to have been de- rived from a Muraenichthys-lineage different from that of Schultzidia. It is an elongate worm eel that has undergone specializations apparently related to its mode of feeding. A prominent toothed groove divides the underside of the snout and extends anteriorly to the elongate tubular nostrils. The suspensorium is forwardly inclined, and although the dentition is generalized, the upper pharyngeal tooth plates are fused, and certain gill arch members (Hs and lz) are absent. Ophichthini. The Ophichthini include the most primitive and generalized of ophichthids. Primi- tive conditions of certain characters include: (a) caudal rays developed; (b) pectoral, dorsal, and anal fins well devel- oped; (c) branchiostegal rays fewer than 20; (d) teeth conical, often multiserial, present on maxilla, dentary, vomer, and intermaxillary; (e) fifth ceratobranchial (Cs) ossified as a slender rod, upper pharyngeal tooth plates sepa- rate; (f) third preopercular pore (pop*) present; (g) neurocranium truncate posteriorly; (h) tail approximately 60 percent of the total length. The species of Echelus possess all of the above characters and represent one branch of the oph- ichthine lineage (Fig. 39). All subsequent oph- ichthins are specialized in having the tail tip re- duced to a finless point. Although differences exist in the condition of the tail tip, in no in- stance are caudal rays developed as in Echelus or the myrophines. The genus Ophichthus, as broadly defined herein, is the most generalized of the remaining ophichthine genera, having 89 character states that embrace most of the varia- tion of the other genera. The changes in other genera have to do with the loss or specialization of characters, or rearrangements of the basic Ophichthus condition. Certain lineages within Ophichthus appear ancestral to lineages within the tribe itself. For example, those species with few branchiostegal rays and an average body taper probably gave rise to the long-jawed pi- scivorous genera. The proposed sphagebranchin lineage can also be derived from generalized ophichthin characters. Ophisurus appears to be an offshoot from a moderately elongate Ophichthus-like species with few branchiostegals, nearly uniform denti- tion, and well developed surface sensory papil- lae. Its jaws and neurocranium modifications merit its generic recognition. The species of Quassiremus share several prim- itive characters with Ophichthus, but have be- come specialized through the reduction of the pectoral fin and loss of the C; and pectoral girdle elements. The species of Pisodonophis, Myrichthys, and Cirrhimuraena probably arose from a common lineage, evidenced in their increased number of branchiostegals and generally multiserial denti- tion. Cirrhimuraena is specialized in the devel- opment of labial cirri. Differences between the species of Cirrhimuraena are probably deserving of subgeneric rank, as discussed in the remarks on this genus. Pisodonophis and Myrichthys share several unique simlarities, primarily their multiserial molariform dentition and broad-based pectoral fins. Myrichthys is further specialized in the loss of pop’, the reduction of the pectoral fin and girdle, and the advancement of the DFO. As mentioned in the remarks on Myrichthys, a nearly continuous character series exists from species of Ophichthus —> Pisodonophis > Myrichthys. The species of Mystriophis, Echiophis, Brachy- somophis, Aplatophis, Xyrias, and Scytalichthys form a natural group of predaceous ophichthines specialized for the capture of large struggling prey. These specializations include the develop- ment of a postorbital strut to brace the maxilla, the strengthening of the suspensorium, enlarge- ment of the vomerine and jaw dentition, the ad- vancement and dorsal location of the eyes on the snout, and the attainment of a large size as adults. Certain species of Ophichthus (O. ophis, O. triserialis, and O. zophochir) are similarly specialized. The character states of Echiophis McCOSKER — EELS SERIES 4, V. 41, #1 90 “1UIYYYIIYDO ay} JO uONNJOAS pasodoig “GE |INBI4 YOLSAINV ANIT-AIYSNOD poonpert soiod pesy (2) 9n2ay40g yTeiquaa s8utuedo [{13 ‘passaidap ZSOT UT} Tepned untueld0inau ‘skel [esa sotysueiq Key ysot $9 *ZsOT uty Te10}Ded ‘quasoid gdod pue 29 peonpel utz Te103.5D0ed (g) snuwaazssnne _———— PS INIHONVUGINVHdS 03 INIAHLHOINVOSV 0} ad SOT gdod (0S *e9) snyzyo1rydQ —————(j1) snanszydo x (1) 82y4d0110a7Dy ae SOT $5 UI aSseoLDUT (9) szydouoposzg (1) sd2zag pedoTeaep 3n14s Od (1) snuvanie7 ysoT~cdod ‘piremi0F OAC ENS (1) s2ydouoBbog (8) sfhyzydrahy (é£) s2ydo2zazshy (42) s2zdosdn7q (6) Duavanutysa19 OSPOIDUT YBUeT YyuUNIy é qsot dod v7, (4£) s2ydo2yog é PetFrpow uot IT JUSp pue ‘prody ‘umtueis0inou if petzytpow AT8u01.s wntuei1d01nau (Z) snyzyo2rydo7z7hyqg we (p-2) s2tydowoskyovag poonpat ATtLepuodsas 4n13s Od (1) #fy2402104 hog petztpow Apsuo014s uOT}TJUSp pue wNTUeIIOINDU (1) sD2ahy NS (1) #24dozn7dy PROCEEDINGS CALIFORNIA ACADEMY OF SCIENCES and Mystriophis are merely further specializa- tions of the condition of those species of Oph- ichthus. As mentioned in the remarks on Mystri- ophis, the specializations of EF. blastorhinos, when further investigated, may merit generic recogni- tion. Remaining genera in the lineage are further specialized by having a proportionately longer trunk region, or conversely, a shorter tail, and have lost the pop*. The examined species of Brachysomophis have developed labial cirri and a massive postorbital strut. The translocation of the orbit to the snout tip and development of the surface sensory papillae system in Brachy- somophis and Scytalichthys correlate with the feeding behavior; these species, like those of Echiophis (personal observation, and Hiatt and Strasburg, 1960), lie within their burrows with only their snout tips exposed, awaiting the pas- sage of prey items. The advantages of the eye position and papillae development are obvious. A similar feeding behavior is practiced by other ophichthines. Species of Callechelys, for ex- ample, have been observed (personal observa- tion, and R. H. Rosenblatt, personal communica- tion) with the head protruding from the sand bottom, apparently awaiting the passage of prey items. Callechelys has not undergone the devel- opment of the strengthened suspensorium and elongated jaws and associated orbit translocation typical of the long-jawed ophichthines, and feeds primarily on relatively smaller and weaker prey. The development of labial cirri in Brachysomo- phis and other ophichthine genera probably serves as a screen to prevent fine sediment from entering the mouth. The absence of the post- orbital strut in the remaining genera is assumed to be a secondary reduction. Xyrias and Scytal- ichthys are quite similar in the shape of their maxillae and in the condition of their multiserial dentition. Scytalichthys has a proportionately longer trunk region and has a further modified snout condition. Aplatophis is conditionally placed at the apex of this lineage. Its numerous reductions and specializations make its place- ment within the lineage difficult. The remaining ophichthine genera have lost the pop*. Malvoliophis has diverged from the Ophichthus condition in having an advanced DFO. Evips has retained an Ophichthus-like DFO, but has a considerably reduced pectoral fin and girdle and a proportionately reduced tail length. The ancestral bascanichthyin probably arose from this lineage, and probably possessed an anterior DFO, an ossified C;, and a reduced pectoral fin 91 and girdle. Subsequent ophichthins have lost the rod-like ossified C; possessed by all other oph- ichthins. Pogonophis, although similar in external appearance to species of Ophichthus, is special- ized in its development of labial barbels and by its loss of the C; and pop’. Leiuranus and Elapso- pis are closely-related genera which have de- parted from Ophichthus in snout shape, UP3-UP4 fusion, and reduced pectoral fins and girdles; Leiuranus is further specialized in the loss of vomerine dentition and certain pectoral ele- ments. Phyllophichthus is aligned with this line- age although its exact placement is undeter- mined. Its suspensorium, jaws, dentition, and neurocranium are extremely modified. Sphagebranchini. The Sphagebranchini com- prise a specialized ophichthin offshoot of highly modified species, assembled on the basis of the absence of the pectoral fin, pectoral girdle reduc- tions, and low or entirely ventral gill openings. Primtive conditions of certain characters include: (a) neurocranium depressed; (b) branchiostegal rays fewer than 20; (c) pop* and tp? pores present; (d) Cs; ossified; (e) dorsal and anal fins present; (f) anterior nostrils tubular, posterior nostrils within mouth; (g) body and tail nearly subequal; (h) gill openings low lateral. The interpretation of intergeneric relationships within this tribe is made difficult by the reduc- tion or loss of numerous characters. Several in- terpretations are possible, depending upon the importance applied to certain characters. The following interpretation, in assuming the tribe to have had a monophyletic origin, assumes that the C; has been lost independently in two lineages. This loss seems plausible through an ossified > cartilaginous transformation, and has apparently occurred elsewhere in the family. The preliminary dichotomy separates the three sharp-snouted, entirely finless genera (Fig. 40). Each possesses a peculiar projection from the parapophyses of the anterior trunk vertebrae (Fig. 33A), a specialization not observed else- where in the tribe. Cirricaecula is primitive in having retained an Ophichthus-like C; but spe- cialized in its UPs3-UPs fusion, entirely ventral non-converging gill openings, and labial cirri de- velopment. The species of /Ichthyapus have a small eye, reduced anterior nostril, and cartilagi- nous C;. Species of Apterichtus have entirely lost McCOSKER — EELS SERIES 4, V. 41, #1 BO) D7TDYAAIT smef pue wntueitdsoinou ‘7SO0T 7d3 ae (7) DuwozsouwnT "1ulyUReIgaseYds ay} JO UOINJOAD pasodosg “Op 91N3!4 YOLSAONV AMIT-NIHLHOTHdO poessoidep wntuels0inou ‘patFrpow qJsor 29 soskydodeied ‘}SO[T suTy [Te ‘skel [TeSBoSOTYIUeIG UL 9SeoIDUT petytpow ssutusedo T1ts fuotsnz Van-fan ee en 2SOT odod poonpoelr S5 Te1IQUSA s8utuedo [{[t3 poonper aka ~~ (L-9) sndohyzyer pots tpow ynojis Apoqg ‘patyFtpow prodky pue sutusedo [{t3 (1) vznaang (1) snuzytoz0279 petFrpow S[Ttaqzsow {£7sSoT S5 \ (1) shyzyorydoa2ryoy (OT-6) 82yoraeqdy PROCEEDINGS CALIFORNIA ACADEMY OF SCIENCES Caralophia (1) ‘\ antervor nostril and snout modified, dentition reduced Phaenomonas (2) Ethadophts (2) tins, reduced tO a short. dorsal, trunk extremely elongate Gordtichthys (1) Leptenchelys (1) Cart tip) fieshy pectoral fin absent Agttpe (1) DFO posteriorly displaced to CALLECHELYINI a Bascanichthys (ca. 16) Co lost C. reduced or absent, DFO advanced to head, trunk elongated Dalophis (5)=——=——? OPHICHTHIN-LIKE ANCESTOR Figure 41. Proposed evolution of the Bascanichthyini. 93 94 the Cs; and the posterior nostril is translocated to the outer lip. The remaining genera are characterized by an increased number of branchiostegals and the loss of the C;. They may be separated into two major lineages. Caecula and Stictorhinus are similar to the finless sphagebranchins in having ventral gill openings, pointed snouts, and moderately de- pressed neurocrania; these similarities, in part, may be adaptations to a similar mode of life. The derivation of Yirrkala and the closely-related Achirophichthys and Lamnostoma is somewhat uncertain. They have lost the pop® and have low lateral to ventral gill openings. The neurocrania of these genera are considerably modified from the broad, depressed state of other sphagebran- chins. The neurocranium of Yirrkafa is rounded dorsally. That of Lamnostoma has become. nar- rowed along the ethmoid, in probable correla- tion with its enlarged dentition and related to its feeding mode. Certain species of Yirrkala have retained the tp®. The species of Yirrkala are gen- erally elongate whereas those of Achirophichthys and Lamnostoma are cylindrical and stout. Bascanichthyini. The Bascanichthyini, like the Sphagebranchini, appear to be derived from a moderately specialized ophichthin-like ancestor. Certain specializations in the form of hyoid mod- ifications and fin reductions had already been achieved by their supposed ophichthin-like an- cestor. As adults, the bascanichthyins have spe- cialized toward a burrowing, vermiform existence, feeding on small prey and rarely leaving the substrate. As noted in the discussion of the axial skeleton (p. 45), the trunk elongation of species of Phaenomonas, Allips, certain Bascanichthys and presumably Gordiichthys is a specialization for this mode of life. Primitive conditions of cer- tain characters within the Bascanichthyini in- clude: (a) pectoral fin rudimentary; (b) eye small, but not minute; (c) Cs present, but reduced; (d) body and tail nearly subequal; (e) branchiostegal rays numerous; (f) DFO behind head; (g) snout grooved on underside; (h) gill openings low lateral in position. The inclusion of Dalophis in the Bascanichthy- ini is uncertain. Although possessing Ophichthus- like body/tail proportions and an ossified C;, it appears more similar to the bascanichthyin condition in its fin reductions and general ce- SERIES 4, V. 41, #1 McCOSKER — EELS phalic appearance. As a_ bascanichthyin, it is placed near the primary separation from the ancestral lineage (Fig. 41). The remaining genera appear to be derived from the generalized conditon of Bascanichthys. The rudimentary pectoral fin is retained only by Allips. Allips is similar to species of Bascanich- thys in its trunk elongation and general physi- ognomy, but differs in its presumably secondar- ily-derived posterior DFO. Phaenomonas is fur- ther specialized through reduction from this line- age and has become nearly finless. Gordiichthys, not examined in this study, is provisionally re- ferred to this lineage on the basis of characters included in its terse description. The remaining bascanichthyins are somewhat similar in their general morphologies, and have undergone sev- eral modifications and reductions from the Bas- canichthys condition. The Callechelyini are here- in assumed to be derived from a bascanichthyin- like ancestor, but have further specialized and radiated along a different complex of adaptive characters. Callechelyini. The Callechelyini is the most distinct and compact of ophichthine tribes. The species are among the most specialized of oph- ichthids and are quite removed from the ances- tral ophichthid stock. Their specializations and reductions impart a particular facies to the group that readily separates it from other ophichthids, particularly evidenced in the ventral, convergent gill openings, laterally compressed body, short tail, anterior dorsal fin origin, small eyes ,and reduced pore systems. These outer similarities are borne out by the shortened neurocranium, stout hyoid, and the osteological reductions in the pectoral apparatus and gill arches which further characterize the group. These conditions appear to have been de- rived from a bascanichthyin-like ancestor or from an ophichthin ancestor which gave rise to the Bascanichthyini. Available specimens or radio- graphs of 20 of the 22 species of the Callechely- ini have allowed an in-depth study of this tribe. The meristic and morphological characters listed in Tables 8-9 were used to generate the com- puter-programmed taxonomic evaluations illus- trated in Figures 43-44. The characteristics of computer programs WVGM and REGROUP are described in the taxonomic methods sections of this study. Primitive conditions of certain key characters within the Callechelyini include: PROCEEDINGS CALIFORNIA ACADEMY OF SCIENCES 95 Letharehus (3) aaterior nostrils lost, /m supraorbital pore added, neurocranium elongated Paraletharechus (2) Calleehelys (15) UH forked, scapula present anak tin lost, branchial pouches UH spiked, expanded scapula absent —- =— nS ‘ Aprognathodon (1) slender branchiostegals \ intermaxillary teeth lost, ‘ \ gill arches modified ; snout grooved , / / / Leuropharus (1) vomerine teeth lost EH branchvosteeals broad, rays 20-25), snout not grooved on underside, trunk not jexcessavely ellongate BASCANICHTHYIN-LIKE ANCESTOR Figure 42. Proposed evolution of the Callechelyini. 96 (a) teeth uniserial and present on the dentary, intermaxillary, maxilla, and vomer; (b) underside of snout not grooved; (c) gill openings low and bascanichthyin-like, not specialized as in Letharchus and Paralethar- chus; (d) branchiostegal rays fewer than 25 pairs, those along the epihyal broadened basally; (e) three supraorbital pores; (f) trunk not extremely elongate, tail 40 per- cent or more of total length; (g) urohyal ossified and spike-like, not forked posteriorly; (h) two longitudinal rod-like pectoral elements. No living tribal member possesses all the prim- itive characters listed above. However Aprogna- thodon platyventris and certain species of Cal- lechelys, except for minor specializations, closely approximate the above conditions. The retention of several primitive characters suggests that Aprognathodon platyventris sepa- rated early from the basal stock (Fig. 42). It has become specialized through the loss of inter- maxillary dentition and the incomplete re-ossi- fication of the third hypobranchial. These condi- tions are probably adaptations to a specialized mode of feeding. The next dichotomy in the tribal evolution in- volved two other New World genera, Letharchus and Paraletharchus (compare Figs. 42-44). They, like Aprognathodon, have retained the broad branchiostegals and ungrooved snouts, but are specialized in having broadly flared branchial pouches and in the loss of the anal fin. Species of Paraletharchus appear externally quite similar to species of Callechelys; species of Letharchus are quite distinctive in having lost the tubular anterior nostril condition, added a fourth supra- orbital pore, and having more elongate and de- pressed neurocrania. The remaining callechelyins include Leuro- pharus lasiops and the numerous species of Callechelys. Leuropharus is somewhat general- ized in having few vertebrae, a moderate tail length, and an ungrooved snout, yet it differs from other callechelyins in lacking teeth on the vomer. Whether it separated before the species of Callechelys or from a Callechelys-like ancestor is questionable in that L. Jasiops appears to have numerous, slender branchiostegals (observed from a radiograph of the type specimen). This condition, if observed correctly, is more ad- vanced than that of certain species of Callech- elys (C. nebulosus, C. springeri, and C. holo- SERIES 4, V. 41, #1 McCOSKER — EELS chromus), and would necessitate the parallel evolution of this condition if Leuropharus sepa- rated earlier than Callechelys. The remaining genus, Callechelys with 15 rec- ognized species, has apparently combined the generalized callechelyin condition with certain minor specializations into a very successful ‘“‘body plan’. Evolution within Callechelys, as discussed on page 63,has included two or three early separations from the ancestral stock, which were followed by two major subgeneric lines. Callech- elys nebulosus, C. springeri, and perhaps C. holo- chromus have retained broadened branchiostegal rays, which is probably correlated to their mod- erate number. These two conditions are probably responsible for the separation of these species from other Callechelys species in programs WVGM and REGROUP (Figs. 43-44). These con- ditions should not preclude their inclusion within Callechelys and _ illustrates the weakness of a numerical taxonomic scheme based on too few characters. The remaining species are special- ized in having more numerous and _ slender branchiostegal rays, and a longer trunk region and an increase in the number of trunk verte- brae. These species have evolved along two line- ages, one containing species which have retained a simple urohyal and lost the posterior pectoral girdle element (the scapula?), and another with species in which the urohyal is split posteriorly into two slender divergent rays but with the sec- ond pectoral element retained. Neither of these specializations, when compared with those of other ophichthids, seems to merit generic sepa- ration. Zoogeography and Comments on Ophichthid Speciation In the absence of a suitable fossil record, it is virtually impossible to reconstruct with certainty the past distribution and center of origin of the Ophichthidae. Certain inferences relating to an- cestral distributions however, can be made based on the present species distribution, assumed en- vironmental tolerances of ophichthids in general, and the presumed geological history of tropical land and water masses. Two major geological events have directly af- fected the distribution of tropical marine organ- isms. These were the Miocene (?) closure of the Tethyan Seaway through the convergence of the European and African continental plates (Phillips and Forsyth, 1972), isolating the Mediterranean and Atlantic from the Indo-Pacific, and the late PROCEEDINGS CALIFORNIA ACADEMY OF SCIENCES 97 Fab -1 .8 6 4 .2 0 .2 4 .6 .8 +1 TEE RII eS A. platyventris L. velifer L.rosenbliatti P pacificus P.opercularis C.nebulosus L, lasiops C. bilinearis C. cliffi C.galapagensis Cc. muraena C. luteus Cspringeri Cmarmoratus Ceristigmus Cmelanotaenius Striatus Figure 43. Phenogram of the relationships of the species of the Callechelyini, using program WVGM. The levels of correlation at which species join are represented by the scale Tab- 98 Cis 4/9 C. springeri marmoratus Callechelys (9) SERIES 4, V. 41, #1 McCOSKER — EELS | 8/9 Letharchus (2) Aprognathodon 2/3 2/27 1/9 Paraletharchus (2) ays Leuropharus nebulosus & C. Figure 44. Interrelationships of species groups of the tribe Callechelyini, as defined by program REGROUP. Significance level set at 0.600. Fractions are the ratios of the number of observed between-group species connections to the maximum number of possible connec- tions. Number of species represented per genus are within parentheses. Not included are Callechelys bitaeniatus, C. holochromus, C. leucopterus, and Letharchus aliculatus. Pliocene to Pleistocene closure (Whitmore and Stewart, 1965) of the Middle American Seaway, separating the New World oceans (Rosenblatt, 1963). Assuming that the environmental toler- ances of the Ophichthidae have always limited them to tropical, sub-tropical, or warm temperate waters, it may be stated that the Tethyan and Central American Seaway closures have delimi- ted, in large part, the waters available to the distribution of living genera. (A single exception may be the distribution of Ophisurus serpens in the Mediterranean, and eastern and western Afri- can shores, probably resulting from a transgres- sion of the Cape of Good Hope.) Applying these assumptions to the known distribution of oph- ichthid genera (Table 10), inferences concerning the evolution and generic interrelationships of the family may be drawn. For example, the seven circumtropical genera must have existed prior to the closure of the Tethyan Seaway, or have passed through the Central American Seaway and transgressed both oceans. Recent informa- tion gained from investigations of plate tectonics and paleomagnetism suggests that the Atlantic Ocean during the Early Cretaceous was much narrower than at present (Phillips and Forsyth, 1972). On that basis, the distribution of an ar- chaic species across the Tethyan Seaway and into the eastern Pacific would seem quite plausble. An Ophichthus-like genus probably existed in the Upper Eocene, as evidenced by Storms’ (1896) description of Eomyrus dolloi from the Wemmelian Formation of western Europe. The neurocranium, as illustrated, is much like that of a modern Ophichthus. Those genera restricted to both coasts of the New World (Table 10, Group III) must have existed prior to the Pana- manian uplift, and now include several pairs of closely related species. Echiophis is also a mem- ber of this group but has presumably extended its distribution to the eastern Atlantic. The genus Muraenichthys is presently limited to the Red PROCEEDINGS CALIFORNIA ACADEMY OF SCIENCES Sea, Indian Ocean, western and central Pacific oceans, with a single south eastern Pacific spe- cies known only from the offshore Chilean is- lands of Juan Fernandez and San Felix. Its ab- sence from the tropical Atlantic, as explained by McCosker (1970), perhaps suggests a post-Teth- yan origin of the genus rather than the lack in the Atlantic of a suitable habitat. The single New World species is probably recently derived from the south-western Pacific by means of eastward transport across the South Pacific. The restriction of many genera to the Indo-Pacific and New World oceans (Table 10, Groups II, IV-V) prob- ably represents radiations since the Tethyan and Central American Seaway closures, respectively. The origin of genera with species distributed across major water masses might be _ inferred from an analysis of the species involved. The genus Phaenomonas, for example, contains an eastern Pacific species (P. pinnata) which ranges from the Gulf of California to Colombia, and an Indo-Pacific species (P. cooperae) which ranges from Hawaii to east Africa. The extreme speciali- zation of P. cooperae, in comparison to P. pin- nata, might suggest that Phaenomonas arose in the eastern Pacific and radiated westward. Its presumed absence from the Atlantic Ocean would indicate that Phaenomonas either evolved from the bascanichthyin stock since the closure of the Central American Seaway, or that proper Phaenomonas habitat is presently unavailable in the Atlantic. Another explanation might assume that the present existence of the primitive P. pinnata in the eastern Pacific is the result of an eastward radiation from an Indo-Pacific ori- gin, followed by further specialization of the an- cestral stock through competition with the more complex ophichthid fauna of the Indo-Pacific. Caution however, must be applied in any of the above assumptions concerning the present dis- tribution of ophichthid genera. Modern collect- ing methods have resulted in the discovery of numerous new species, many of which repre- sented new genera, and it is highly likely that many extant ophichthid species remain uncol- lected. An analysis of the distribution of species among ophichthid genera reveals several inter- esting biological phenomena. The classification proposed here is hopefully more than just a con- venient cataloguing system, and assumes that the disproportionate number of species in vari- ous ophichthid genera, ranging from a single species in nearly half of the genera to more than 99 50 in Ophichthus, reflects something real in the natural system. Figure 45 illustrates the distribu- tion of species among the genera of the Oph- ichthidae, Gobiesocidae, and the blenniid tribe Salariini. These examples were chosen because they represent recent monographic revisions, uncomplicated by the efforts of a multiplicity of authors; the gobiesocid data are from Briggs (1955) and subsequent species descriptions and the blenny data are based on Smith-Vaniz and Springer (1971). A pattern exists in the three groups illustrated, and if transformed to logs, the data would indicate an almost straight-line in- verse relation between the log-number of spe- cies/genus and the log number of genera. C. B. Williams (1964) has discussed similar evidence from a number of terrestrial groups and sug- gested that such a log-normal mathematical pat- tern is a recurrent phenomenon in natural sys- tems. He observed that the fit of these data to a calculated log series is moderately good at most levels, but higher than expected for monotypic genera. The following hypotheses are proposed to ex- plain the distribution of species among genera. One hypothesis might assume that such a dis- tribution of species among genera reflects the evolution of a group with many recently-derived genera, and a decreasing number of genera which have existed for increasingly longer time spans. Those archaic genera have had the op- portunity through geologic time and events to segregate and speciate, whereas the more recent taxa have lacked those opportunities. A more intriguing hypothesis however, suggested by Richard H. Rosenblatt (personal communication), might assume that a combination of characters exist in the ancestral lineage from which taxa radiate by means of specializations and reduc- tions; certain resultant taxa would include a combination of characters which would allow further radiation (dependent upon certain bio- logical factors and geological events) resulting in genera with numerous species, whereas other taxa have specialized in a manner which, in re- lation to the available environments, has a low probability of further radiation. The latter cate- gory contains the numerous monotypic genera of the Ophichthidae. These monotypic genera might represent evolutionary “forays’’ into rather unique environments or life styles and are ap- parently unsuccessful beyond their present lim- ited area of distribution or as ancestral bases for further speciation. The former category, in which SERIES 4, V. 41, #1 McCOSKER — EELS 100 25 re) O0=Gobiesocidae 20 A: Salariini % = O=: Ophichthidae ¢ Oo 0 15 6 = 0 5 10 fe) fe] A Oo P| A 25 O oO Oo Oo Opa A M A AAMOO DADA fe) bed 1 2 3 4 5 678910 20 30 40 5060 Species /Genus Clog scale) Figure 45. Distribution of species among the genera of the Ophichthidae, Gobieso- cidae and Salariini. few genera contain many species, contains such genera as Myrichthys, Callechelys, and Oph- ichthus, in which the combination of adaptive characters selected for have, with minor modi- fications, resulted in the numerous species which occupy similar habitats in all tropical oceans. Biological factors such as the leptocephalus lar- val stage and geological events such as seaway closures lend credence to the first mentioned hypothesis and complicate the second. Yet the present distribution and few species of several presumably archaic ophichthid genera would tend to support the latter hypothesis. Further investigations into the ecology and behavior of species of this intriguing and diverse eel family may offer further insight into the evo- lutionary processes which have shaped _ the Ophichthidae. PROCEEDINGS CALIFORNIA ACADEMY OF SCIENCES LITERATURE CITED Ahl, J. N. 1789. Dissertatio de Muraena et Oph- ichtho. Dissertationes Academicae Upsaliae habitae sub praesidio C. P. Thunberg, 3(1): 1- 12. Allis, E. P., Jr. 1903. The lateral sensory system in the Muraenidae. Intern. Monats. Anat. Physiol., 20: 125-170. Asano, H. 1962. Studies of the Congrid eels of Japan. Bull. Misaki Marine Biol. Inst., no. 1. 143 pp. Bamber, R. C. 1915. Reports on the marine biol- ogy of the Sudanese Red Sea, from collections made by Cyril Crossland . .. XII. The fishes. Linn. J. Zool., 31: 477-485. Beebe, W. 1935. Deep-sea fishes of the Bermuda Oceanographic expeditions. No. 1 — Family Derichthyidae. Zoologica, N.Y., 20(1): 1-23. Bertin, L. 1935. Oxystomus de Rafinesque est une forme bién distincte, parmi les larves lepto- céphaliennes. Comptes Rend. des séances de l’‘Acad. des. Sci., 200: 1878. Bertin, L. and C. Arambourg. 1958. Super-ordre des Teleostéens. Traité de Zoologie, 13, fasc. 3: 2204-2500. Blache, J. 1968. Contribution a la connaissance des Poissons anguilliformes de la c6éte occi- dentale d’Afrique. Huiteéme note: la famille des Echelidae. Bull. de |’l. F. A. N., sér. A, 30(4): 1501-1539. Blache, J. 1971. Contribution a la connaissance des Poissons anguilliformes de la c6éte occi- dentale d’Afrique. Onziéme note: les genres Mystriophis et Echiopsis (Fam. des Ophich- thidae). Bull. de I'l. F. A. N., sér. A, 33(1): 202- 226. Blache, J. and M. L. Bauchot. 1972. Contribution a la connaissance des Poissons anguilliformes de la céte occidentale d’Afrique. Troisieme note: les genres Verma, Apterichthus, Ichthya- pus, Hemerorhinus, Caecula, Dalophis avec la description de deux genres nouveaus (Fam. des Ophichthidae). Bull. de I'l. F. A. N., sér. A, 34(3): 692-773. Blache, J. and J. Cadenat. 1971. Contribution a la connaissance des Poissons anguilliformes de la céte occidentale d'Afrique. Dixiéme note: les genres Myrichthys, Bascanichthys et Cal- lechelys (Fam. des Ophichthidae) Bull. de I'l. F.A.N., sér. A, 33(1): 158-201. Blache, J., J. Cadenat., and A. Stauch. 1970. Clés de détermination des Poissons de mer sig- nalés dans l’atlantique oriental entre le 20¢ 101 parallele nord et le 15¢ parallele sud. O. R. S. T. O. M., Faune Tropicale, 13. 479 pp. Bleeker,, P. 1853a. Dijdrage tot de kennis der Muraenoiden en Symbranchoiden van den Indischen Archipel. Verh. Batav. Genootsch., 25: 1-62. Bleeker, P. 1853b. Diagnostische beschrivnigen van nieuwe of wenig bekende wisschsoorten van Batavia. Tintal I-VI, Nat. Tijdschr. Neder.- Indie, 4: 451-516. Bleeker, P. 1863. Mémoire sur les poissons de la cote de Guinée. Nat. Verh. Holl. Maatsch. Wet. Haarlem, 2(18): 1-136. Bleeker, P. 1865. Systema Muraenorum revisum. Neder. Tijdschr. Dierk., I], 1865: 113-122. Bloch, M. E. 1795. Naturgeschichte der Auslandi- schen Fische. IX. (Not seen). Bohlke, J. E. 1955. A new genus and species of ophichthid eels from the Bahamas. Notul. Nat., no. 282: 1-7. Bohlke, J. E. 1956a. A synopsis of the eels of the family Xenocongridae (including the Chlopsidae and Chilorhinidae). Proc. Acad. Nat. Sci. Phila- delphia, 108: 61-95. Bohlke, J. E. 1956b. A small collection of new eels from western Puerto Rico. Notul. Nat., no. PKS) |= 3}. Bohlke, J. E. 1960. A new ophichthid eel of the genus Pseudomyrophis from the Gulf of Mex- ico. Notul. Nat., no. 329: 1-8. Bohlke, J. E. 1967. The descriptions of three new eels from the tropical West Atlantic. Proc. Acad. Nat. Sci. Philadelphia, 118(4): 91-108. Bohlke, J. E. 1968. A new species of the ophich- thid eel genus Verma from the West Atlantic, with comments on related species. Notul. Nat., no. 415: 1-12. Bohlke, J. E. and J. E. McCosker. 1975. The status of the ophichthid eel genera Caecula Vahl and Sphagebranchus Bloch, and the description of a new genus and species from fresh waters in Brazil. Proc. Acad. Nat. Sci. Philadelphia, 127 (ys aeltt Bohlke, J. E. and C. R. Robins. 1959. The char- acters and synonomy of the Western Atlantic snake eel, Ophichthus ophis Linnaeus. Notul. Nat., no. 320: 1-9. Briggs, J. C. 1955. A monograph of the cling- fishes (order Xenopterygii). Stanford Ichthyol. Bull., 6: 1-224. Burton, P. R. 1956. A comparative ostelogical study of the skulls of the moray eels, Gymno- thorax funebris Ranzani and Gymnothorax moringa (Cuvier). J. Florida Acad. Sci., 19(1): 102 35-44. Castle, P. H. J. 1963. The systematics, develop- ment and distribution of two eels of the genus Gnathophis (Congridae) in Australian waters. Zool. publ. Victoria Univ. New Zeal., no. 34: 15-47. Castle, P. H. J. 1965. Ophichthid leptocephali in Australian waters. Trans. roy. Soc. N. Z., 7(6): 97-123. Castle, P. H. J. 1967. Two remarkable eel-larvae from off Southern Africa. Spec. Publ. Inst. Ich- thyol. Rhodes Univ., (1): 1-12. Castle, P. H. J. 1969. An index and bibliography of eel larvae. Spec. Publ. Inst. Ichthyol. Rhodes Onive, (7) 2nen2ae Castle, P. H. J. 1972. The eel genus Benthench- elys (fam. Ophichthidae) in the Indo-Pacific. Dana-Rep., no. 82. 32 pp. Chabanaud, P. 1936. Le neurocrane osseux des Téleostéens dyssymetriques. Ann. Inst. Ocean- ogr., n. s., 16(3): 223-297. Chan, W. L. 1967. A new species of congrid eel from the South China Sea. J. nat. Hist., 1: 97- Ws Cohen, D. M. and D. Dean. 1970. Sexual matur- ity and migratory behaviour of the tropical eel, Ahlia egmontis. Nature, 227(5254): 189-190. Cope, E. D. 1871. Contribution to the ichthyology of the Lesser Antilles. Trans. Amer. Phil. Soc., 14(2): 445-583. Cuvier, G. L. C. F. D. 1817. Le regne animal dis- tribué d’apres son organisation, pour servir de base a l'histoire naturelle des animaux et d’in- troduction a l’anatomie comparée. Les reptiles, les poissons, les mollusques et les annélides. Deterville, Paris, 2: 1-532. de Barneville, B. 1847. Note sur un nouveau gen- re d’anguilliformes. Rev. Zool., 1847: 219-220. Deraniyagala, P. E. P. 1929. Some anguilliform fishes of Ceylon. Spolia Zeylandica, 15(1): 1- 78) Dumeril, A. M. C. 1806. Zoologie analytique, ou methode naturelle de classification des ani- maux. Paris. 344 pp. Dumeril, A. M. C. 1856. Ichthyologie analitique ou Classification des poissons, suivant la méth- ode naturalle, a l’aide de tableaux synoptiques. Mém. Acad. Sci. Paris, 27, pt. 1.511 pp. Ebeling, A. W. and W. H. Weed. 1963. Melam- phaidae III. Systematics and distribution of the species in the bathypelagic fish genus Scopel- ogadus Vaillant. Dana-Rep., no. 60. 58 pp. Eldred, B. 1966. The early development of the spotted worm eel. Myrophis punctatus Lutken SERIES 4, V. 41, #1 McCOSKER — EELS (Ophichthidae). Florida Bd. Conserv. Mar. Lab., leaf. ser., 4(1): 1-13. Fager, E. W. 1957. Determination and analysis of recurrent groups. Ecology, 38(4): 586-595. Fager, E. W. 1969. Recurrent group analysis in the classification of flexibacteria. J. gen. Microbiol., 58: 179-187. Fager, E. W. and A. R. Longhurst. 1968. Recurrent group analysis of species assemblages of de- mersal fish in the Gulf of Guinea. J. Fish. Res. Bd. Canada, 25 (7): 1405-1421. Fager, E. W. and J. A. McGowan. 1963. Zooplank- ton species groups in the North Pacific. Sci- ence, 140(3566): 453-460. Fischer von Waldheim, G. 1813. Zoognosia, Tab- ulus Synopticus Illustrata. Ed. III, vol. 1. (Not seen). Fowler, H. W. 1918. New and little-known fishes from the Philippine Islands. Proc. Acad. Nat. Sci. Philadelphia, 70: 2-71. Fowler, H. W. 1925. New taxonomic names of West African marine fishes. Amer. Mus. Novi- tates, 162: 1-5. Fowler, H. W. 1934. Descriptions of new fishes obtained 1907 to 1910, chiefly in the Philip- pine Islands and adjacent seas. Proc. Acad. Nat. Sci. Philadelphia, 85: 233-267. Fowler, H. W. 1936. The marine fishes of West Africa. Part |. Bull. Amer. Mus. Nat. Hist., 70: 1-605. Fowler, H. W. 1938. The fishes of the George Vanderbilt South Pacific Expedition, 1937. Monogr. Acad. Nat. Sci. Phila. no. 2. 349 pp. Fowler, H. W. 1944. Results of the Fifth George Vanderbilt Expedition (1941). Monogr. Acad. Nat. Sci. Phila. no. 6: 57-529. Frizzell, D. L. and C. K. Lamber. 1962. Distinc- tive ‘“‘Congrid type’ fish otoliths from the Lower Tertiary of the Gulf Coast (Pisces: An- guilliformes). Proc. Cal. Acad. Sci., 4th ser., 32(5): 87-101. Frost, G. A. 1926. A comparative study of the otoliths of the Neopterygian fishes (contin- ued). Ill. Order Apodes. Ann. Mag. Nat. Hist., ser. 8, 10: 99-104. Ganguly, D. N. and A. C. Nag. 1964. On the functional morphology of the pectoral girdle and the acranial myomeric musculature of a benthozoic teleostean fish Ophichthys boro (Ham. Buch.). Anat. Anz., 115: 405-417. Ginsburg, I. 1951. The eels of the northern Gulf Coast of the United States and some related species, Texas J. Sci., 3(3): 431-485. Girard, C. F. 1859. Ichthyological notices. Proc. PROCEEDINGS CALIFORNIA ACADEMY OF SCIENCES Acad. Nat. Sci. Philadelphia, 1859(1860): 56- 58. Gistel, J. 1848. Naturgeschichte des Thierreichs fur hohere Schiler. (Not seen.) Glover, T. A. 1973. Fishes of southern and west- ern Japan, vol. 1 (groups 1-7). Nagasaki Univ. 429 pp. (In Japanese.) Goode, G. B. and T. H. Bean. 1882. Descriptions of twenty-five new species from the Southern United States, and three new genera, Lethar- chus, loglossus, and Chriodorus. Proc. U. S. Nat. Mus., 1882, 5: 412-437. Goodrich, E. S. 1930. Studies on the structure and development of vertebrates. 8 London. 837 pp. Gosline, W. A. 1950. The osteology and rela- tionships of the echelid eel, Kaupichthys dio- dontus. Pacific Sci., 4(4): 309-314. Gosline, W. A. 1951a. The osteology and classi- fication of the ophichthid eels of the Hawaiian Islands. Pacific Sci., 5(4): 298-320. Gosline, W. A. 1951b. Chilorhinus brocki, a new echelid eel from Hawaii, with notes on the classification of the order Anguillida. Copeia, 1951(3): 195-202. _Gosline, W. A. 1952. Notes on the systematic status of four eel families. J. Wash. Acad. Sci., 42(4): 130-135. Gosline, W. A. 1959. Mode of life, functional morphology, and the classification of modern teleostean fishes. Syst. Zool., 8(3): 160-164. Gosline, W. A. 1971. Functional morphology and classification of teleostean fishes. Honolulu, Univ. Hawaii Press. 208 pp. Grassi, G. B. 1913. Metamorfosi dei murenoidi. Ricerche sistematiche ed ecologische. Jena, Fischer. 211 pp. (Not seen). Greenwood, P. H., D. E. Rosen, S. H. Weitzman, and G. S. Myers. 1966. Phyletic studies of tele- ostean fishes, with a provisional classification of living forms. Bull. Amer. Mus. Nat. Hist., 131: 345-455. Giinther, A. 1870. Catalogue of the fishes in the British Museum. Vol. VIII, Catalogue of the Physostomi. London, Taylor and Francis. 549 pp. Giinther, A. 1910. Andrew Garrett’s Fische des Stidsee, . . . Heft IX. J. Mus. Godeffroy, Ham- burg, Heft 17: 389-515. Harry, R. R. 1948. New records for the fish Myr- ichthys tigrinus, a snake eel of the eastern tropical Pacific, with a relocation of the type locality. Copeia, 1948(2): 145-146. Hector, J. 1870. On a species of Ophisurus found 103 on the coast of New Zealand. Trans. Proc. N. Z. Inst., 1869(1870), 2: 34-40. Herre, A. W. C. T. 1924. Some rare Philippine eels. Philipp. J. Sci., 24(1): 107-111. Hiatt, R. W. and D. W. Strasburg. 1960. Ecologi- cal relationships of the fish fauna on coral reefs of the Marshall Islands. Ecol. Monographs, 30: 65-127. Hopkirk, J. D. 1965. Records of yellow and spotted snake-eels (genus Ophichthus) from San Francisco Bay, California. Cal. Fish and Game, 51(3): 183-186. Hubbs, C. L. 1932. The use of the generic name Ophis for an eel, a snake and a mollusc. Copeia, 1932(1): 26-27. Jordan, D. S. 1919a. New genera of fishes. Proc. Acad. Nat. Sci. Philadelphia, 1918(1919), 70: 341-344. Jordan, D. S. 1919b. The genera of fishes, part Il, from Agassiz to Bleeker, 1833-1858, twenty- six years, with the accepted type of each. A contribution to the stability of scientific no- menclature. Leland Stanford Jr. Univ. Pub., Univ. ser.: 163-284. Jordan, D. S. 1932. The “first species’” and the “first reviser’’. Science, 25(638): 467-469. Jordan, D. S. and B. M. Davis. 1891. A prelimi- nary review of the apodal fishes or eels in- habiting the waters of America and Europe. Rep. U.S. Comm. of Fish and Fisheries for 1888, part 16: 581-677. Jordan, D. S. and B. W. Evermann. 1896. The fishes of North and Middle America: A de- scriptive catalogue . . . Panama. Part I. Bull. no. 47, U. S. Nat. Mus. 1240 pp. Jordan, D. S. and B. W. Evermann. 1905. The aquatic resources of the Hawaiian Islands. Part |. The shore fishes. U. S. Bur. Fisheries, Bull. 1903, 23(1): 1-574. Jordan, D. S. and C. H. Gilbert. 1882. Descrip- tions of thirty-three new species of fishes from Mazatlan, Mexico. Proc. U. S. Nat. Mus., 4: 338-365. Jordan, D. S. and C. H. Gilbert. 1883. On the nomenclature of the genus Ophichthys. Proc. U.S. Nat. Mus., 5: 648-651. Jordan, D. S. and J. O. Snyder. 1901. A review of the apodal fishes or eels of Japan, with de- scriptions of 19 new species. Proc. U. S. Nat. Mus., 23(1239): 837-890. Kanazawa, R. H. 1963. Two new species of oph- ichthid eels from the western Atlantic. Proc. Biol. Soc. Wash., 76: 281-288. Kaup, J. 1856a. Ubersicht der Aale. Arch. Na- 104 turges., 22(1): 41-77. Kaup, J. 1856b. Catalogue of apodal fish in the collection of the British Museum. London. 160 pp. Kaup, J. 1860. Neue aalahnliche Fische des Ham- burger Museums. Abh. Natur. Verein Ham- burg, 1859(1860), 4, Abth. 2. 35 pp. Knox, F. J. 1870. Anatomical observations on Ophisurus novaezelandiae. Trans. Proc. N. Z. Inst., 1869(1870), 2: 34-40. Kotthaus, A. 1968. Fische des Indischen ozeans, ergebnisse der ichthyologen, untersuchungen ... lll Ostariophysi und Apodes. ‘Meteor’ Forschungsergeb Ser. D (Biol.), (3): 14-56. Lacepede, (Comte) B. G. E. (here called Citoyen La Cepede). 1800. Histoire Naturelle des Pois- sons, Vol. Il. Paris. 413 pp. La Monte, F. 1961. Achirophichthys kampeni from Papua. Copeia, 1961(1): 115-116. Liddell, H. G. and R. Scott. 1901. A Greek-Eng- lish lexicon. 8th edit. Oxford, Clarendon Press. 1776 pp. Linnaeus, C. 1758. Systema Naturae. Regnum Animalie. Guilielmi Engeleman, Lipsiae. 824 pp. Longley, W. H. and S. F. Hildebrand. 1941. Sys- tematic catalogue of the fishes of Tortugas, Florida. Papers from Tortugas Lab., vol. 34, Carnegie Inst. Wash. Publ. 535. 331 pp. Lozano Rey, D. L. 1947. Peces Ganoideos y Fiso- stomos. Mem. Real Acad. Cien. Exactas, Fisicas y Nat. de Madrid. Ser. de Cien. Nat., 11. 839 pp. Litken, C. F. 1851. Nogle bemaerkinger om naes- eborenes .. . aalefamilien. Vidensk. Meddel. Naturhist. Foren. Kjobenjavn. 21 pp. (Not seen). Mattioli, P. A. 1568. Les commentaires de M. P. A. Matthiole ... sur les six livres de P. Dis- coride ... de la matiére médicinale . . . mis en Francais sur la derniere édition Latine de l’autheur par M. J. des Moulins. Lyon. 819 pp. (Not seen). McAllister, D. E. 1968. Evolution of branchio- stegals and classification of teleostome fishes. Bull. Nat. Mus. Canada, 221: 1-239. McClelland, J. 1844. Apodal fishes of Bengal. J. Nat. Hist. Calcutta, 5: 150-226. McCosker, J. E. 1970. A review of the eel genera Leptenchelys and Muraenichthys, with the de- scription of a new genus, Schismorhynchus, and a new species, Muraenichthys chilensis. Pacific Sci., 24(4): 506-516. McCosker, J. E. 1972. Two new genera and two SERIES 4, V. 41, #1 McCOSKER — EELS new species of western Pacific snake-eels (Apodes: Ophichthidae). Proc. Cal. Acad. Sci., 4th ser., 39(10): 111-120. McCosker, J. E. 1973. The osteology, classifica- tion, and relationships of the eel family Oph- ichthidae (Pisces, Anguilliformes). PhD diss., Univ. Calif. San Diego. 289 pp. McCosker, J. E. 1974. A revision of the oph- ichthid eel genus Letharchus. Copeia, 1974(3): 619-629. McCosker, J. E. In Press. The eel genus Phaeno- monas (Pisces: Ophichthidae). Pacific Sci. McCosker, J. E. and R. H. Rosenblatt. 1972. East- ern Pacific snake-eels of the genus Callechelys (Apodes: Ophichthidae). Trans. San Diego Soc. Nat. Hist., 17(2): 15-24. Miller, R.V. and J. V. Landingham. 1969. Addi- tional procedures for effective enzyme clear- ing and staining of fishes. Copeia, 1969(4): 829-830. Mohamed, K. H. 1958. On the occurrence of the eel Neenchelys buitendijki Weber and de Beaufort in Indian waters. J. Bombay Nat. Hist. NOE, HHe HVS Myers, G. S. and M. H. Storey. 1939. Hesperomy- rus fryi, a new genus and species of echelid eels from California. Stanford Ichthyol. Bull., 1(4): 156-159. Myers, G. S. and C. B. Wade. 1941. Four new genera and ten new species of eels from the Pacific coast of tropical America. Allan Han- cock Pac. Exped., 9(4): 65-111. Nelson, G. J. 1966a. Gill arches of teleostean fishes of the order Anguilliformes. Pacific Sci., 20(4): 391-408. Nelson, G. J. 1966b. Osteology and relationships of the eel, Neenchelys buitendijki. Copeia, 1966(2): 321-324. Nelson, G. J. 1967. Notes on the systematic status of the eels Neenchelys and Myroconger. Pacific Sci., 21(4): 562-563. Nelson, G. J. 1969. Gill arches and the phylogeny of fishes with notes on the classification of vertebrates. Bull. Amer. Mus. Nat. Hist., 141: 479-522. Nelson, G. J. 1972. Cephalic sensory canals, pit- lines, and the classification of esocid fishes, with notes on galaxiids and other teleosts. Amer. Mus. Novitates, no. 2492. 49 pp. Nichols, J. T. 1955. Results of the Archbold Ex- peditions. No. 71. Two new freshwater fishes from New Guinea. Amer. Mus. Novitates, no. 1735. 6 pp. Norman, J. R. 1922. A new eel from Tobago. Ann. PROCEEDINGS CALIFORNIA ACADEMY OF SCIENCES Mag. Nat. Hist., ser. 9, 10: 296-297. Norman, J. R. 1926. The development of the chondrocranium of the eel (Anguilla vulgaris), with observations on the comparative morph- ology and development of the chondrocranium in bony fishes. Phil. Trans. R. Soc., ser. B, 214: 369-464. Nybelin, O. 1963. Zur morphologie und Termi- nologie des Schwanzskelettes der Actinoptery- gien. Arkiv fur Zoologi, ser. 2, 15(35): 485- 516. Ogilby, J. D. 1897. Some new genera and species of fishes. Proc. Linn. Soc. New South Wales, part 2: 245-251. Parr, A. E. 1930. Jugostegalia, an accessory skele- ton in the gill cover of the eels of the genus Myrophis. Copeia, 1930(3): 71-73. Parra, D. A. 1787. Descripcion de differentes piezas de historia natural, las mas del reino maritimo, representadas en setenta y cinco laminas. Havana. (Not seen). Peden, A. E. 1972. Redescription and distribution of the rare deep sea eel Xenomystax atrarius in the eastern Pacific Ocean. J. Fish. Res. Bd. Canada, 29: 1-12. Pellegrin, J. 1912. Sur une collection de poissons des Nouvelles-Hebrides du Dr. Cailliot. Bull. Mus. Hist. Nat., Paris, 18(4): 205-207. Phillipps, W. J. 1926. New or rare fishes of New Zealand. Trans. Proc. N. Z. Inst., 56: 529-537. Phillips, J. D. and D. Forsyth. 1972. Plate tecton- ics, paleomagnetism, and the opening of the Atlantic. Bull. Geol. Soc. Amer., 83(6): 1579- 1600. Poey y Aloy, F. 1867. Monografia de las Morenas Cubanas. Repertorio Fisico-Natural de la Isla de Cuba, II: 245-268. Poey y Aloy, F. 1880. Revisio piscium cubensum. Anal. Soc. Hispanola Hist. Nat., Madrid, 9: 243-261. (Not seen). Popta, C. M. L. 1904. Les arcs branchiaux de quelques Muraenidae. Ann. Sci. Nat., 19: 367- 390. Rafinesque-Schmaltz, C. S. 1810a. Caratteri di Alcuni Nuovi Generi e Nuove Species di Ani- mal e Piante della Sicilia. Pub. April 1, 1810. Rafinesque-Schmaltz, C. S. 1810b. Indice d’Ittio- logia Siciliana. Pub. May, 1810. Rafinesque-Schmaltz, C. S. 1815. Analyse de la Nature, ou Tableau del’Univers et des Corps Organises ‘‘La Nature est mon Guide et Lin- naeus mon maitre.”” Palermo. 224 pp. Randall, J. E. and J. E. McCosker. 1975. The eels of Easter Island with a description of a new 105 moray. Los Angeles County Mus., Contrib. in Sci., no. 264. 32 pp. Randall, J. E. and C. R. Robins. 1966. Acanthen- chelys spinicauda Norman, a valid West Indian species of the snake-eel genus Ophichthus. Copeia, 1966(3): 610-611. Regan, C. T. 1912. The osteology and classifica- tion of the teleostean fishes of the order Apo- des. Ann. Mag. Nat. Hist., 8th ser., (10): 377- 387. Richardson, J. 1844a. The zoology of the voyage of H. M. S. Sulpher under the command of Captain Sir Edward Belcher . . . during the years 1836-42, no. 5, Ichthyology. Part 1: 51- 70. Richardson, J. 1844b. Ichthyology of the voyage of H. M. S. Erebus and Terror, under the Com- mand of Captain Sir James Clark Ross, R. N., F. R. S. Edward, Newman, London. 139 pp. Robins, C. H. 1971. The comparative morphology of the synaphobranchid eels of the straits of Florida. Proc. Acad. Nat. Sci. Philadelphia, 123(7): 153-204. Robins, C. H. and C. R. Robins. 1967. The xeno- congrid eel Chlopsis bicolor in the western North Atlantic. Bull. Mar. Sci., 17(1): 232-248. Robins, C. H. and C. R. Robins. 1970. The eel family Dysommidae (including the Dysom- minidae and Nettodaridae), including a new genus and species. Proc. Acad. Nat. Sci. Phila- delphia, 122(6): 293-335. Robins, C. H. and C. R. Robins. 1971. Osteology and relationships of the eel family Macroceph- enchelyidae. Proc. Acad. Nat. Sci. Philadelphia, 123(6): 127-150. Rosenblatt, R. H. 1963. Some aspects of specia- tion in marine shore fishes. Systematics Assoc. Publ. No. 5: 171-180. Rosenblatt, R. H. 1967. The osteology of the con- grid eel Gorgasia punctata and the relation- ships of the Heterocongrinae. Pacific Sci., 21 CG) On-O7- Rosenblatt, R. H. and J. E. McCosker. 1970. A key to the genera of the ophichthid eels, with descriptions of two new genera and three new species from the eastern Pacific. Pacific Sci., 24(4): 494-505. Rosenblatt, R. H., J. E. McCosker and |. Rubinoff. 1972. Indo-west Pacific fishes from the Gulf of Chiriqui, Panama. Los Angeles County Mus., Contrib. in Sci., no. 234. 18 pp. Rosenblatt, R. H. and |. Rubinoff. 1972. Python- ichthys asodes, a new heterenchelyid eel from the Gulf of Panama. Bull. Mar. Sci. 22(2): 355- 106 364. Ruppell, E. 1826-1828. Atlas zu der Reise in Nordlichen Afrika: Fische des Rothen Meeres. 4 vols. (Not seen). Schultz, L. P. 1942. The first record of the oph- ichthyid eel Scytalichthys miurus (Jordan and Gilbert) from the Galapagos Islands, with notes on Mystriophis intertinctus (Richardson). J. Wash. Acad. Sci., 32(3): 83. Schultz, L. P. 1943. Fishes of the Phoenix and Samoan Islands. Bull. U. S. Nat. Mus., 180. 316 pp. Schultz, L. P. and O. Barton. 1960. A redescrip- tion of Letharchus pacificus, a snake eel from the Eastern Pacific Ocean. Copeia, 1960(4): 351-353. Schultz, L. P. and collaborators. 1953. Fishes of the Marshall and Marianas Islands. Families from Asymmetrontidae through Siganidae. Bull. U. S. Nat. Mus., 202, 1. 685 pp. Schultz, L. P. and L. P. Woods. 1949. Keys to the genera of echelid eels and the species of Muraenichthys of the Pacific, with two new species. J. Wash. Acad. Sci., 39(5): 169-174. Schwartz, E. and A. D. Hasler. 1966. Superficial lateral line sense organs of the mudminnow (Umbra limi). Zeitschr. Verglich. Physiol., 53(3): 317-327. Smith, A. 1847. Illustrations from the Zoology of South Africa; consisting chiefly of figures and descriptions of objects of natural history col- lected during an expedition into the interior of South Africa in 1834-1836. (Not seen). Smith, D. G. 1971. Osteology and relationships of the congrid eels of the western north Atlan- tic (Pisces, Anguilliformes). PhD diss., Univ. Miami. 163 pp. Smith, D. G. and P. H. J. Castle. 1972. The eel genus Neoconger Girard: systematics, osteol- ogy, and life history. Bull. Mar. Sci., 22(1): 196-249. Smith, J. L. B. 1957. The fishes of Aldabra. Part IX (with a new eel from East Africa). Ann. Mag. Nat. Hist., ser. 12, 10: 833-842. Smith, J. L. B. 1962. Sand-dwelling eels of the western Indian ocean and the Red Sea. Rhodes Univ. Ichthyol. Bull. 24: 447-466. Smith, J. L. B. 1964. The discovery in Mozam- bique of the little known eel Ophichthys ten- uis Gunther, 1870, a redescription of the type of Caecula pterygera Vahl, 1794, notes on other species and on generic relationships. Ann. Mag. Nat. Hist., ser. 13, 7: 711-723. SERIES 4, V. 41, #1 McCOSKER — EELS Smith-Vaniz, W. F. and V. G. Springer. 1971. Synopsis of the tribe Salariini, with descrip- tions of five new genera and three new spe- cies. Smithson. Contrib. to Zool., 73: 1-72. Sokal, R. R. and C. D. Michener. 1958. A taxo- nomic method for evaluating systematic rela- tionships. Univ. Kansas Sci. Bull., 38: 1409- 1438. Springer, S. and M. J. Allen. 1932. A peculiar snake-eel from the Gulf of Mexico. Copeia, 19322) 105: Springer, V. G. 1968. Osteology and classifica- tion of the fishes of the family Blenniidae. Bull. U. S. Nat. Mus., 284. 85 pp. Storey, M. H. 1939. Contributions toward a re- vision of the ophichthyid eels |. The genera Callechelys and Bascanichthys, with descrip- tions of new species and notes on Myrichthys. Stanford Ichthyol. Bull., 1(3): 61-84. Storms, R. 1896. Premiere note sur les poissons wemmeliens (Eocene supérieur) de la _ Bel- gique. Bull. Soc. Belge de Géol., de Paleont. et d’Hydrol., 10: 198-240. Swainson, W. 1838. The natural history of fishes, amphibians, and reptiles, or monocardian ani- mals. Longmans, London, vol. 1. 368 pp. Taylor, W. R. 1967. An enzyme method of clear- ing and staining small vertebrates. Proc. U. S. Nat. Mus., 122(3596): 1-17. Tilak, R. and S. K. Kanji. 1967. Studies on the morphology of the pectoral girdle of Pisood- onophis boro (Ham.) in relation to its habit. Anat. Anz., 120: 404-408. Tortonese, E. 1959. Contributo allo studio degli Ophichthidae del Mediterraneo (pisces anguil- liformes). Annal. Mus. Civ. Stor. Nat. Genova, eZ 3B =247 6 Tortonese, E. 1964. Contributo allo studio siste- matico e biogeografico del pesci della Nuova Guinea. Annal. Mus. Civ. Stor. Nat. Genova, 75: 13-98. Trewavas, E. 1932. A contribution to the classifi- cation of the fishes of the order Apodes, based on the osteology of some rare eels. Proc. Zool. Soc. London, part 3: 639-659. Turton, W. 1807. The British fauna, containing a compendium of the zoology of the British Islands; arranged according to the Linnaean system. Swansea, London. 230 pp. (Not seen). Vahl, M. 1794. Beskrivelse af en nye fiskeslaegt. Skrivt. Naturh. Selsk. Kiobenhavn 3(2): 149- 156. (Not seen). PROCEEDINGS CALIFORNIA ACADEMY OF SCIENCES Wade, C. B. 1946. Two new genera and five new species of apodal fishes from the eastern Pa- cific. Allan Hancock Pac. Exped., 9(7): 181- DAS: Weber, M. and L. F. de Beaufort. 1916. The fishes of the Indo-Australian Archipelago. III. Ostario- physi: Il Cyprinoidea, Apodes, Synbranchi. Leiden. 455 pp. Whitley, G. P. 1930. Five new generic names for Australian fishes. Austr. Zool., 6(3): 250-251. Whitley, G. P. 1934. Studies in Ichthyology. No. 8. Rec. Australian Mus., 19(2): 153-163. Whitley, G. P. 1940. Illustrations of some Aus- tralian fishes. Austr. Zool., 9(4): 397-428. Whitley, G. P. 1950. New fish names. Proc. R. Zool. Soc. New South Wales, 1948-1949: 44. Whitley, G. P. 1951. Studies in ichthyology. No. 15. Rec. Australian Mus., 22(4): 389-408. Whitley, G. P. 1968. A check-list of the fishes 107 recorded from the New Zealand region. Austr. Zool., 15(1): 1-102. Whitmore, F. J., Jr. and R. H. Stewart. 1965. Mio- cene mammals and Central American seaways. Science, 148(3667): 180-185. Williams, C. B. 1964. Patterns in the balance of nature and related problems in quantitative ecology. Academic Press, London. 324 pp. Wormuth, J. H. 1971. The biogeography, system- atics and interspecific relationships of the oegopsid squid family Ommastrephidae in the Pacific Ocean. PhD diss., Univ. Calif., San Diego. 189 pp. Ypiranga Pinto, S. 1970. Observacoes ictiologicas. VI — Antobrantia, novo genero de ofictideo do Brasil (Actinopterygii, Anguilliformes, Oph- ichthyidae). Atas da Sociedade de Biologia do Rio de Janeiro, 14 (1-2): 13-15. 108 Table |. Abbreviations are: M, molariform; V, villiform. serial; U, uniserial. +, presents = Shape: Rows in jaws and vomer: I-V signifies intermaxillary-vomerine. SERIES 4, V. 41, #1 Dentition of the Genera of Ophichthidae McCOSKER — EELS C, conical; F, fang-like; B, biserial; M, multi- Shape in Jaws Vomer I-V Gap Benthenchelyini Benthenchelys My rophini Ahita Muraent chthys Myrophts Neenchelys Pseudomyrophts Scehismorhynchus Schultztdia Callechelyini Aprognathodon| Callechelys Letharchus Leuropharus Paraletharchus Sphagebranchini Achtrophtchthys Apterichtus Caecula Cirriecaecula Hemerorhinus Ichthyapus Lamos toma Sttetorhinus Ytrrkala Bascanichthyini Alltps Bascanichthys Caralophia Dalophts Ethadophts Gorditchthys Leptenchelys Phaenomonas Ophichthini Aplatophtis Brachysomophis Cirrhimuraena Elapsoptis Echelus Echtophis Evips Leturanus Malvoltophis Myrichthys Mystriophis Ophtchthus Ophtsurus Phyllophtchthus Pisodonophis Pogonophts Quasst remus Seytaltchthys Xyrtas C,M (my Caz) (oe) (se) (se) (se) (sz) (sed se) Con) (ae) Ged ifs) aQan-e- eal GaGa (‘= Seo Sees Sse Sees o Sees eseSee SSeS iS SeSsesce See “eee crc U ie) Seste U Stet =giSg U,B te teterve¢t¢tr + tet te et 1. Intermaxillary teeth absent 2. Maxillary teeth biserial, dentary uniserial PROCEEDINGS CALIFORNIA ACADEMY OF SCIENCES Table 2. Number and Location of Branchiostegal Rays of the Species of the Ophichthidae Counts represent the right side only. Rays joined basally are counted sepa- rately. ''CH-EH'' represents the cartilaginous CH-EH interspace. ''Free'' rays are noticeably separated from the hyoid. Total Free Along CH CH-EH Along EH Benthenchelys cartiert 20 13 = = 7 Ahlia egmontts 47 4] - - 6 Muraentchthys chilenstis 32 25 = = 7 Muraentchthys gymopterus 43 36 = = 7 Muraentchthys macropterus 31 37 = = 4 Myrophis plumbeus 43 37 = = 6 Myrophis uropterus 32 26 = = 6 Myrophts vafer hg 42 - - 7 Neenchelys buitendijki' 30 24 - - 6 Pseudomyrophis micropinna 47 ho - - 7 Pseudomyrophts ntmtus 23 17 = - 6 Schismorhynchus labtalis 30 25 = = 5 Schultzidia johnstonensts 33 29 - - 4 Aprognathodon platyventris 28 = 14 2) 12 Callechelys btlinearis 27] = 21 ] 5 Callechelys ertstigmus 31 = 5 6 20 Callechelys galapagensts 27 = 15 2 10 Callechelys marmoratus 29 = 18 2 9 Callechelys melanotaentus 29 = 12 3 14 Callechelys nebulosus 29 10 8 3 8 Letharchus rosenblattt 34 = 16 3 15 Paraletharehus pact ficus 32 = 23 4 5 Aptertchtus flavicaudus 16 l I l 13 Caeeula pterygera 25 12 = 3 10 Ctrrteaecula johnsoni 18 - ] 3 14 Iehthyapus selachops 19 - 4 2 133 Lamostoma ortentalis 29 26 = = 3 Sttetorhinus potamius 29 - 5 4 20 Yirrkala Lwnbricotdes 27 = 5 2 20 Yirrkala misolensis 29 - 12 2 15 Yirrkala tenuis 25 > 4 3 18 Bascanichthys panamensts 30 ] 5 3 21 Caralophia loxochila 27 = 8 4 15 Phaenomonas cooperae 28 19 3 5 1 Phaenomonas pinnata 26 19 1 6 = Aplatophis chaultodus 22 = 2 | 19 Brachysomophts sauropsts 17 = 2 1 14 Cirrhtmuraena macgregort 24 = 7 2 15 Cirrhimuraena taentopterus 29 6 l 4 18 Echelus myrus 15 = 4 e 1] Echelus pachyrhynchus 18 - 4 1 13 Echtophis sp. 21 - 4 1 16 Elapsopts cyclorhinus 24 4 2 1 17 Leturanus semtctnctus 31 - 6 3 22 Malvoltophts pinguts 21 = 4 1 16 Myrichthys colubrinus 30 = 22 3 5 Myrtchthys maculosus 35 - 9 2 24 Myrtchthys xystrurus 31 3 3 ] 24 Ophtehthus cephalazona 27 2 2 2 21 Ophiehthus cruentifer 17 = 4 = 13 Ophiechthus frontalis 19 = 5 ] 13 Ophichthus triserialis 22 = 3 = 19 Ophtchthus zophochir 27 = 5 1 21 Ophisurus serpens 19 = 2 | 16 Phyllophichthus xenodontus 29 22 = = 7 Pisodonophts boro 31 I 6 2 22 Pisodonophis canerivorus 30 6 | I 22 Pisodonophis daspilotus 32 8 = 3 21 Quasstremus nothochir 21 2 2 = 18 Seytaltchthys miurus 25 - 4 2 19 1. From Nelson (1966a: fig. 2a) 109 110 SERIES 4, V. 41, #1 McCOSKER — EELS Table 3. Gill Arch Condition in the Ophichthinae Abbreviations are: B,_y4, basibranchials; H1-3, hypobranchials; Cyi-c, ceratobran- chials; Io_ , infrapharyngobranchials; UP 3-4, upper pharyngeal dermal tooth plates; 0, ossified; -O-, UP3-UP, fusion; C, cartilaginous; R, rudimentary; -, absent; *, from Nelson (1966a: Table 1) By Bo B3 Bh Hy Ho H3 C1-4 Cc 1) 13 UP3 UPL Ophichthus, zophochir o © &R © @ @ € 0 o © ®@ © 0 0. rhytidodermatoides OC. Se 0 Oba 0 Yo © @ @ 0 0. polyophthalmus* o ¢ Kk kK © @ € 0 OF O © 0) 0. altiptnnis Ws eS Oo © f 0 Oo @ © © 0 0. erabo yo ¢ & kK © © FE 0 OM oO © 0) O. eruentifer W es os = © @ ££ 0 oy O OM 0 0. cephalazona 0 R > = ® @ 0 9 OG © @ 0 Pisodonophis boro 2 Ss BR © oO —-E 0 XH @ 0) 0 P. canertvorus 0 ny GG go @ © 0 OF 0) 0) 0 0 Ophisurus serpens Oo © © & @ @ € 0) oO o OG © 0) Evips pereinetus 09 © FF kR O © CE 0 Oo © @ 0 Echelus myrus oy ¢€¢ © CG O © 0 O00 0 0 E. pachyrhynchus YO © sc © O © 0) J OO 0 0 Myrichthys colubrinus Oo € = = ®% © CE 0 Oo O @ @ 0) M. maculosus 09 € = &€ O OM E 0 oO OM @ 0) M. xystrurus Onc = C 0) (0) C 0 00) 0h a0 0 Aplatophis chaultodus QO «= =£ = M OO € 0) oO © @ 0 Brachysomophts sauropsts oy CC = © @ O CE 0 oO @ @ 0 B. henshawt* Y © oc & © 0 Oo © © @ 0 Ctrrhimuraena macgregort Oo © R KR O O € 0 QO © ® © 0 C. taentopterus ®O o> R > © © GE 0) 0 © @ 0 Echtophis tnterttnetus* Oo f RF C OM OO CE 0) 09 © @ @ 0) Echtophis sp. 0 ¢ C Gc @ @ CC 0 oO =(0)= Xyrtas revulsus oO © = ¢€ @® @ EC 0 0) 0) 0 -0- Malvoltophis pinguts o © R KR O O CEC 0 Oo @ @ -0- Elapsopts cyclorhinus 9 € KR = O DO © 0 = @ -0- Leturanus semictnctus Oo €©€ Ss € GO © 0 = 0 =(0)= Phyllophtchthus xenodontus 0 CC R - O0O O C 0 = © @ @ 0 Pogonophts fossatus 0 © = ss @ @ € 0 = @ @ 0) Quasstremus evtonthas 0 = - > (i Or (c 0 = 0) 0) (0) 0 Q. nothochir QO =o = => © © G 0 = @ © 0 Dalophis imberbts 0 Cc Kk © O @ 0) oo @) 0 Bascanichthys teres* 9 © = € M O CE 0 QO © @ @ 0 B. panamensis ) coc > € @O@ OM CE 0 >= @ @ © 0) Alltps concolor oo © = € OM © EC 0 = @ @ @ 0 Ethadophts byrnet 0 € os = OM @ E 0 > ®}) @ 0 E. merenda oO CC RR €C BO O CE 0 = © © 0 Phaenomonas ptnnata Oo © = € O O CE 0 > 0 @ © 0 Caralophia loxochila QO o © € OM © CEC 0 = 9 @ =(0)= Callechelys marmoratus 0) 6 >= @ @© € 0 > ®% @& @ 0 Aprognathodon platyventris 0 C - - 0 0 0 0 >= 0 © @ 0 Paraletharchus pact ficus 9 ¢€C = R © © EC 0 = 0 © © 0 Letharchus velifer* Oo GF os f 6 @ 0 > @ © @ 0) Letharchus rosenblatti oy © = R @ © CEC 0) > @ © © 0 Ichthyapus selachops Q@ oc = © ® © CE 0 QO @ @ @ 0 Cirrieaeeula johnsont 06 C= = rR © © € 0 OO 0 =(0)— Apterichtus flavicaudus Q@ =e ces =s @® © € 0 = @ @ @ 0 Caecula pterygera Oo © © € M ® FC 0 = @ © © 0) Yirrkala Lumbricotdes Ho CC C€ © 0 @ 0 = @ © @ 0 Y. tenuts OF Re ac C One e0) 1G 0 >= @ © 0 Y. misolensts oo € &€ & oO @ € 0 = @ © © 0 Y. kaupt OC Ge Gc @ @ € 0 = 00) 0 0 Lamostoma orientalis Oo © = = ® © Ef 0 2 9 @ @ 0 Stictorhinus potamtus O € C¢ & © @ CF 0 = @ © © 0 PROCEEDINGS CALIFORNIA ACADEMY OF SCIENCES 111 Table 4. Gill Arch Condition in the Myrophinae Abbreviations as in Table 3. By Bo B3 By, Hy Ho H C1-4 Cc Io 13 UP3 UPL Benthenchelys cartieri CoS ee ae Oh 050 0 =O 0 =) Ahlta egmontis QO os => = @ @ = 0 = = @ @ 0 Myrophis punctatus* 0 = = @ @ € 0 = @ @ © 0 M. vafer 0 S = 0 0 (¢ 0 = 0 0 0 0 M. uropterus Op eee ee ON Olan 0 > @ @ © 0 M. plumbeus QO ce = @ @ € 0 = @ © =0= Pseudomyrophis nimius 0 >= = @ @ € 0 = 7 © @ 0 P. mteropinna 0) = 5 @ @ € 0 = © @© @ 0 Neenchelys buttendijkt* R oc Ss & 0 © € 0 > 0 © @ 0 Sehultzidta johnstonensts = = =] = 0) 05 JC 0 - = 0 -0- Sehtsmorhynchus labialis = = @= 70 Oy = 0) = = @ =l0)= Muraenichthys chilenstis > =o 5 = @ © € 0 = © © © 0 M. macropterus = = = = 0) @) 0 - C 0 Oo 0 M. cooket* 3 oS OO UE 0) “6 © @ 0 M. gymnotus > = = = @ @ € 0 - =" 20) 30 0 M. latteaudata - = = = 0 0 C 0 = 0 0 0 0 M. schultzet = Oe OeeaC 0 = 2 @ =(0= M. gymnopterus > = & = @) @ 0 Seas =(0)= 112 SERIES 4, V. 41, #1 McCOSKER — EELS Table 5. Lateral Line and Cephalic Pore Conditions in Ophichthine Genera and Subgenera Cephalic pore locations are illustrated in Figure 24. Abbreviations are: *, condition of type species unknown; +, present; -, absent; C, lateral line ossicles continuous; M, lateral line ossicles moderately separated at pores; S, lateral line ossicles separated at pores. tp2 pop3 pop4 Lateral Line Ossicles Ophtchthus = Echelus - Ophisurus > Pisodonophts = Quasst remus = Ctrrhimuraena* Calamuraena Jenkinstella Echtophis = Mystrtophis = Aplatophis = Brachysomophts* = Xyrtas = = = Seytalichthys - - - Pogonophts - - - Evtps - - - Leturanus = = = Elapsopis - - - Phyllophtchthus = = = Malvoltophis = = = Myrtchthys = = - Leet ttt eee t ' Apterichtus <5 + Iehthyapus + + Cirrteaecula + + Stietorhinus + + Caecula + + Yirrkala +,- - - Lamos toma = = = Aprognathodon = = - Callechelys = = = Letharchus = - - Leuropharus = = = Paraletharchus = = = Bascantchthys* = = = Alltps = = - Phaenomonas - = = Ethadophis - = = Caralophta = = = Leptenchelys = = = Dalophts = = - = CS es x St Se (2) (eb ise) (op) (5p) QAANNDQNDAANAN FBNFTFTFBEMANNUUNUNUUNVYUNUUNUNNNYN PROCEEDINGS CALIFORNIA ACADEMY OF SCIENCES Table 6. Vertebral Counts of Various Ophichthid Species Data were obtained during the present study except as cited. The listing is arranged alpha- betically by genus and species within each tribe. Counts include the hypural. SUNY Ufayelti= cates number of individuals. ee Holotype Range Mean N Source Benthenchelys cartiert 156-174 168.3 34 Castle, 1972 Ahlta egmontis 152 l Muraentchthys aokt 137 Muraentchthys australis 152 Castle, 1965 Muraenichthys breviceps 164 Castle, 1965 Muraentchthys chilensts 149 =148-153. 150.9 12 £McCosker, 1970 Muraentchthys cooket 130 Gosline, 195la Muraentchthys gymnotus 129-130 129.5 2 Muraentchthys hattae 154 Muraenichthys tredalet 126-127. 126.5 2 Muraenichthys macropterus 127-130 128.2 4 Muraenichthys schultzet \772 2 Muraentchthys thompsont 128? 128-133 130.5 2 Myrophis plumbeus 142-154 Blache, et al., 1970 Myrophis punctatus 138-145 Eldred, 1966 Myrophis vafer 146-150 147.8 5 (holotype of Hesperomyrus fryt) 154? Neenchelys buitendijkt 145-148 Mohamed, 1958 Pseudomyrophis mtcroptnna 174 Pseudomyrophts ntmtus 212-216 214 2 Schismorhynchus labialis 136 134-138 136 6 Schultztdia johnstonensts 145-149 = 151.2 5 Schultztdia retropinnis (holotype of Muraenichthys malaita) 133 Aprognathodon platyventris 150-155 S275 2 Callechelys bilinearis 161-163 162 2 Callechelys elt ffi 155-149-158 154.9 14 Callechelys ertstigmus 159 154-163-157. 30 Callechelys galapagensis 172 170-174 172 4 Calleechelys holochromus 166 Callechelys Leucopterus 162-165 164 4 Blache and Cadenat, 1971 Callechelys marmoratus 176-183 179 4 (holotype of Callechelys guichenoti) 183 Callechelys melanotaentus 200-205 203 5 Callechelys muraena 141 41-144 142. 2 Callechelys nebulosus 158-159 158. 3 Callechelys perryae 178 Blache and Cadenat, 1971 Callechelys springert 170 166-170 168 2 Callechelys striatus 192 | Letharehus velifer 139 «135-143 s:«139. 14. McCosker, 1974 Letharehus rosenblattt 151 144-151 148. 20 McCosker, 1974 Leuropharus lastops 135 Paraletharchus opercularis 180 171-180 174. 9 Paraletharchus pact ficus 166 156-167 160. 15 Apterichtus ansp 123-132 - Bohlke, 1968 Apterichtus caecus 151 ] Aptertchtus equatorialis?| 146 1 (holotype has 53 preanal vertebrae) Apterichtus flavicaudus 145-157-149. 5 Aptertchtus gymnocelus 136 1 Apterichtus kendalli 137-144 Bohl ke, 1968 Apterichtus klazingat 140 Caecula pterygera 126 126-130 128. 8 Bohlke and McCosker, 1975 Ctrrtecaeecula johnsont 119 ] Iehthyapus acuttrostris 133 Blache and Bauchot, 1972 Iehthyapus ophtoneus 133 321378 lsse 3 Ichthyapus selachops 137-144 139. 15 1. The holotype of Caeeula equatorialis Myers and Wade lacks a tail. The specimen recorded here, ANSP 117436, is from 3015'S, 80°19'W, and was not compared with the type. we} SERIES 4, V. 41, #1 McCOSKER — EELS © Table 6. Continued Holotype Range Mean N Source Iehthyapus vulturis? 123 from Palau, Tahiti, and Seychelles N27, Wiis 12 from Hawaii and Kure 120-124 122. 4 from Easter Island 130-134 132. 9 Lamostoma kampent 143 ] Lamostoma mindora 144 Lamostoma orientalis 134-1376 135. 3 Lamostoma phtltppinensts 153? Sttetorhinus potamtus 140 3=—135=142 = 139. 11 = Bohike and McCosker, 1975 Ytrrkala lLumbricotdes 151-154; 152) 2 (paratype of Yirrkala chaselingt) 153 Yirrkala tenuis 153 ] Alltps concolor 174 Bascantchthys cecilae 226 225-226 225. 2 Blache and Cadenat, 197] Bascanichthys congoensts 189 189-190 189. 2 Blache and Cadenat, 1971 Bascantchthys longtssimus 212 - Blache and Cadenat, 197] Bascantchthys myerst 215 Baseantechthys panamensis 181 l Bascanichthys paulensts 191 Bascantchthys tenuts 203 Basecantchthys teres 181-184 - Blache and Cadenat, 197] Caralophta loxochila 139-145 142. 3 Dalophis tmberbis 152 148-159 152. 14. Blache and Bauchot, 1972 Ethadophts byrnet 189 Ethadophis merenda 159 Leptenchelys vermi formis 163 Phaenomonas cooperae 270 243-270 256 8 Phaenomonas pinnata 187 1175-194 1186. 14 Aplatophis chaultodus 110-111 110. 2 Brachysomophts henshawt 130 - Gosline, 195la Brachysomophis sauropsis NGS 125 124 4 Ctrrhtmuraena macgregort 181 - Gosline, 195la Ctrrhimuraena taentopterus 183 1 Echelus myrus from the Mediterranean S55) = Grassi, 1913 from the eastern tropical Atlantic 149-152 = Blache, et al., Echelus pachyrhynchus 157. 149-157 - Blache, 1968 Echtophis intertinectus 132 132-143 - Blache, 1971 Echtophis mordax 130 Echtophtis sp. 132-139 135. 14 Elapsopts cyeclorhinus 153-160 156. 2 Evips percinetus 132 Leturanus semicinetus 162-169 166. 3 (holotype of Machaerenchelys phoenixensis) 170 Myrichthys bleekeri 196 | Myrtchthys colubrinus 197-201 199 2 Myrichthys maculosus from Hawaii and Midway 174-182 178. 16 from western Pacific 190-199-193. 10 Myrichthys oculatus 170-171 7). 2 Myrichthys pardalts 158 151-159 - Blache and Cadenat, 1971 Myrichthys xystrurus 149-163 9152. 20 Mystrtophts blastorhinos 142 Blache, 1971 Mystrtophts crosntert 136-144 140. 57 Blache, 197] Mystrtiophts rostellatus 154-158 155. 13. Blache, 1971 Ophtchthus altiptnnis 173 | Ophtchthus callaensis 153 1 Ophichthus cruentifer 146 l Ophtchthus erabo 155 155 2 Ophichthus frontalis 144-157 149. 7 Ophtchthus gomest 141 - Jordan and Davis, 1892 2. Population differences in Ichthyapus vulturis are treated in Randall and McCosker (1974). 3. An undescribed species ranging from the Gulf of California to Panama. PROCEEDINGS CALIFORNIA ACADEMY OF SCIENCES Table 6. Continued a Holotype Range Mean N Source Ophichthus maecrochir 144 - Castle, 1965 Ophichthus ocellatus 134 - Jordan and Davis, 1892 Ophichthus ophis from Brazil 162 l from eastern tropical Atlantic 161-170 - Blache, in litt. Ophichthus rhyttdodermatotdes 191 1 Ophichthus triserialis 145-160 152.5 2 Ophtchthus untsertalis 152? Ophichthus urolophus 136 ] Ophtchthus zophochir 150-152 = 150.7 3 Ophisurus serpens 200-208 Blache, in litt. Phyllophtchthus xenodontus 169-170 169.5 6 Pisodonophis boro 171-173-172 2 Pisodonophits canerivorus 155 I Pisodonophis daspilotus 1372) 137=138) 13725 2 Pogonophts fossatus 166-171 168.5 2 Quasstremus evtonthas 153 1 Quasstremus nothochir 139-142 140.8 4 Seytaltchthys miurus 143-149 = 146 4 Xyrias revulsus 158 116 SERIES 4, V. 41, #1 McCOSKER — EELS Table 7. Characteristics of the Ophichthidae and Related Eel Families Abbreviations are: F, fused; 0, ossified; R, reduced; S, sutured; +, present; -, absent. Data are from this study and various sources, including McAllister (1968), Robins and Robins (1970, 1971), and Smith and Castle (1972). Frontal Frontal Temporal Gill Arch Branchiostegal Condition Commissure Pore Canal Ossification Rays (pairs) Ophichthidae F + + Ophichthinae 0 15-34 My rophinae R 20-49 Congridae F - + 0 8-17 Muraenesocidae F - +? 0 8-22 Macrocephenchelyidae F - + 0 8 Dysommi dae F - - R 9-16 Xenocongri dae S = ce R 12-21 Heterenchel yidae S - + 0 11-14 Muraenidae Ss = = R <16 Angui | lidae S = + 0 8-14 ee ——————————————eSSSSSSSSSSSSSSSSSeSSeeSeEeESEeEeEeEeEeEeEEeEEeEeEeEeEeESESESESESESeSESESEeSeses PROCEEDINGS CALIFORNIA ACADEMY OF SCIENCES Wala Table 8. Morphological and Meristic Characters of the Species of Callechelyini Adapted from and expanded upon McCosker and Rosenblatt (1972, Table 5). Abbreviations are as follows: 4, rounded mean value; *, from the holotype; +, present; ~-, absent; Ue .condii— tion not known. Postanal lengths in thousandths of total length. Postanal® Vertebrae Scapula Urohyal Branchiostegal Epihyal Ray Length Rays Broadening Aprognathodon platyventris 35] 1554 + simple 28 extreme Leuropharus lastops 405 135% + ? 32? no? Letharchus veltfer 400 149% + 2 32 extreme L. rosenblattt 428 151* + simple 34 extreme Paraletharchus pact ficus 352 166* + simple 32 yes P. opercularis 392 180* + simple 30 yes Callechelys nebulosus 408 1594 + simple 29 yes C. springert 350 170* + simple 25? yes C. btltnearts 364 1624 + forked 27 no Gs eiheyage 43h 155% + forked 26 no C. galapagensis 4h 172* + forked 27 no C. luteus 415 213 + forked 27 no C. muraena 385 141% + forked 23? no C. marmoratus 345 180° - simple 29 no C. ertstigmus 295 158% = simple 31 no C. melanotaentus 282 2034 - simple 29 no C. striatus 304 192 - simple 26 no C. perryae 319 178% - simple 24 no C. Leucopterus 453 1644 2 ? ? ? C. holoehromus 333 166% - simple ? yes C. bttaentatus 385 ? ? ? ? ? SERIES 4, V. 41, #1 McCOSKER — EELS Table 9. Characteristics of the Species of Callechelyini Used in Programs REGROUP and WVGM a 1]. Meristics Vertebrae: 130-139; 140-149; 150-159; 160-169; 170-179; 180-189; 190-199; 200-209; 210-219 Branchiostegal rays: 22-23; 24-25; 26-27; 28-29; 30-31; 32-33; 34-35 Supraorbital pores: 3 or 4 Il. External Morphology Postanal length: 275-299; 300-324; 325-349; 350-374; 375-399; 400-424; 425-449; 450-474 Anal fin: present or absent Gill opening ''pocket'': present or absent Underside of snout: grooved, slightly grooved or ungrooved Snout: blunt or conical; papillate or smooth Anterior nostril: tubular or not tubular Ill. Internal Morphology DFO: above supraoccipital or above epiotics Neurocranium: rounded or depressed Intermaxillary teeth: present or absent Urohyal: forked, simple and cartilaginous or simple and ossified Hyoid: inflexible along CH-EH suture, slightly flexible or well separated Hypohyals: present or absent Vomerine teeth: present or absent Scapula (?): present or absent Epihyal rays: broadened basally, moderately broadened or slender Third hypobranchial: ossified or cartilaginous IV. Coloration Mottled or weakly spotted Strongly spotted Longitudinally banded Uniform body (fins may contrast) mn PROCEEDINGS CALIFORNIA ACADEMY OF SCIENCES Table 10. Distribution of Certain Ophichthid Genera Refer to text for discussion of those genera marked with an asterisk. Apterichtus Iehthyapus Achtrophichthys Brachysomophis Ctirrt caecula Lammos toma Muraent chthys* Schultatdia Echtophts* Ethadophts Paraletharchus Ahlta Gorditchthys VI. Dalophis |. Circumtropical Bascanichthys Myrichthys Ophtchthus Il. Restricted to Indo-Pacific Alltips Caecula Elapsopts Leturanus Neenchelys Xyrtas Ill. Restricted to New World Letharchus Quasst remus IV. Restricted to Eastern Pacific Leptenchelys Pogonophis V. Restricted to Western Atlantic Aprognathodon Sttetorhinus Echelus Callechelys Myrophis Benthenchelys Ctirrhimuraena Evips Malvoltophis Sehtsmorhynchus Yirrkala Pseudomyrophis Leuropharus Seytalichthys Caralophta Restricted to Eastern Atlantic and Mediterranean Mystrtophts 119 120 SERIES 4, V.41, #1 McCOSKER — EELS INDEX TO GENERA AND SPECIES (Included are recognized species and valid and invalid genera) Acanthenchelys 80,81 Achirophichthys 65 acuminatus, Myrichthys 78 acutirostris, Ichthyapus 67,68 acutirostris, Muraenichthys 59 Ahlia 58 aliculatus, Letharchus 64 Allips 70 altipinnis, Ophichthus 80,81 Anepistomon 82 anguiformis, Apterichtus 66 Anguisurus 68 ansp, Apterichthus 66 Antobrantia 80 Aotea 58,59 apicalis, Ophichthus 80,81 Aplatophis 74 Aprognathodon 62 Apterichtus 59,65,68 asakusae, Ophichthus 81 ascensionsis, Ophichthus 81 ater, Ophichthus 81 atlanticus, Brachysomophis 74 australis, Muraenichthys 58,59 australis, Myrophis 60 Bascanichthys 70,78 bascanoides, Bascanichthys 71 Benthenchelys 57,85 bilinearis, Callechelys 63 bitaeniatus, Callechelys 63 blastorhinos, Echiophis 77,79 bleekeri, Myrichthys 78 bonaparti, Ophichthus 80,81 boro, Pisodonophis 82 boulengeri, Dalophis 72 Brachycheirophis 83 Brachysomophis 65,74,75,79 Branderius 65 breviceps, Muraenichthys 59 buitendijki, Neenchelys 60 byrnei, Ethadophis 72 Caecilia 65 Caecula 63,65,66,67,69,70,72,85 caecus, Apterichtus 65,66 Calamuraena 75 calamus, Cirrhimuraena 75 Callechelys 62,72 callaensis, Ophichthus 81 cancrivorus, Pisodonophis 82,83 Caralophia 71 cartieri, Benthenchelys 57 ceciliae, Bascanichthys 71 celebicus, Ophichthus 81 Centrurophis 79,81 cephalopeltis, Dalophis 71,72 cephalozona, Ophichthus 79,80,81 chauliodus,Aplatophis 74 cheilopogon, Cirrhimuraena 75 cheni, Myrophis 60 chilensis, Muraenichthys 59 chinensis, Cirrhimuraena 75 Chlevastes 78 Cirrhimuraena 75,85 cirrocheilos, Brachysomophis 74,75 Cirricaecula 59,67,68 cliffi, Callechelys 63 Coecilophis 80,81 Cogrus 79,81 colubrinus, Myrichthys 78 concolor, Allips 70 congoensis, Bascanichthys 71 cookei, Muraenichthys 59 cooperae, Phaenomonas 73 copelandi, Pisodonophis 82 crocodilinus, Brachysomophis 65,74 crosnieri, Mystriophis 79 Crotalopsis 76,77 cruentifer, Ophichthus 80,81,82 Cryptopterenchelys 80 Cryptopterus 80,81 Cryptopterygium 62,63 Cyclophichthys 77 cyclorhinus, Elapsopis 77 cylindricus, Bascanichthys 71 cylindroideus, Myrophis 59,60 Dalophis 62,68,71 daspilotus, Pisodonophis 82,83 derbeyensis, Ophichthus 81 devisi, Muraenichthys 59 dromicus, Pisodonophis 82 Echelus 75,76,81,85 Echiophis 76,77,79 Echiopsis 76 egmontis, Ahlia 58 Elapsopis 77 epinepheli, Apterichtus 66 episcopus, Ophichthus 81 equatorialis, Apterichtus 66 erabo, Ophichthus 81 eristigmus, Callechelys 63 Ethadophis 72 evermanni, Ophichthus 81 evionthas, Quassiremus 83,84 PROCEEDINGS CALIFORNIA ACADEMY OF SCIENCES Evips 77 filaria, Bascanichthys 71 flavicaudus, Apterichtus 68 foresti, Apterichtus 65,66 fossatus, Pogonophis 83 frio, Myrophis 60 frontalis, Ophichthus 81 fusca, Yirrkala 69 galapagensis, Callechelys 63 garretti, Ophichthus 81 Gisenchelys 80,81 gjellerupi, Yirrkala 69 godeffroyi, Muraenichthys 59 gomesi, Ophichthus 80,81 Gordiichthys 63,72 gracilis, Apterichtus 65 grandoculis, Ophichthus 81 gymnocelus, Apterichtus 66 gymnopterus, Muraenichthys 58,59 gymnotus, Muraenichthys 59 hattae, Muraenichthys 59 Hemerorhinus 67,70 henshawi, Brachysomophis 75 Herpetoichthys 80 Hesperomyrus 59 heyningi, Hemerorhinus 67 hijala, Pisodonophis 82 hoevenii, Pisodonophis 82 holochromus, Callechelys 62,63 hypselopterus, Pisodonophis 82 Ichthyapus 59,66,67,68 imberbis, Dalophis 71,72 inhacae, Cirrhimuraena 75 Innominado 79,81 intertinctus, Echiophis 76,77 iredalei, Muraenichthys 59 irretitus, Gordiichthys 72 Jenkinsiella 75,85 johnsoni, Cirricaecula 64 johnstonensis, Schultzidia 61,62 kampeni, Lamnostoma 65 kaupi, Yirrkala 69,70 kendalli, Apterichtus 65 kirkii, Bascanichthys 71 klazingai, Apterichtus 66 labialis, Schismorhynchus 61 Lamnostoma 65,67,68,69,85 lasiops, Leuropharus 64 laticaudata, Muraenichthys 58,59 Leiuranus 77 Leptenchelys 72,73,85 Leptognathus 82 Leptorhinophis 80 Leptorhynchus 82 lepturus, Myrophis 60 Letharchus 63,64 leucopterus, Callechelys 63 Leuropharus 64 limkouensis, Ophichthus 81 longissimus, Bascanichthys 71 loxochila, Caralophia 71 lumbricoides, Yirrkala 69,70 luteus, Callechelys 63 macgregori, Cirrhimuraena 75 Machaerenchelys 77 macrochir, Ophichthus 81 macrodon, Yirrkala 69 Macrodonophis 76 macrops, Ophichthus 81 macropterus, Muraenichthys 59 macrorhynchus, Ophisurus 82 macrostomus, Muraenichthys 59 macrurus, Phyllophichthus 82 maculata, Yirrkala 69 maculatus, Ophichthus 79,81 maculosus, Myrichthys 78,79 madagascariensis, Ophichthus 81 magnioculis, Ophichthus 80,81 Malvoliophis 77,78,85 manilensis, Ophichthus 81 marginatus, Ophichthus 81 marmoratus, Callechelys 62,63 melanochir, Ophichthus 81 melanoporus, Ophichthus 81 melanotaenius, Callechelys 63 merenda, Ethadophis 72 Microdonophis 75,80,81,85 micropinna, Pseudomyrophis 60,61 Microrhynchus 65-66 microtretus, Neenchelys 60 mindora, Lamnostoma 69 misolensis, Yirrkala 69,70 miurus, Scytalichthys 84 moluccensis, Yirrkala 69 monodi, Apterichtus 66 mordax, Echiophis 76,77 moseri, Apterichtus multidentatus, Dalophis 72 multiserialis, Ophichthus 81,82 muraena, Callechelys 63,65 Muraenichthys 58,59,61,62,73 Muraenophis 80 Muraenopsis 80,81 myersi, Bascanichthys 71 Myrichthys 78,79 Myrophis 57,58,59 Myropterura 58 Myrus 75,76 121 122 myrus, Echelus 75 Mystriophis 77,79,84 natalensis, Yirrkala 69 nebulosus, Callechelys 63 Neenchelys 60,85 nimius, Pseudomyrophis 60,61 nothochir, Quassiremus 84 obtusirostris, Dalophis 72 oculatus, Myrichthys 78 oliveri, Cirrhimuraena 75 omanensis, Ichthyapus 68 Omochelys 80,81,82,85 opercularis, Paraletharchus 64 Ophichthus 69,73,75,76,79,80,81,82,83,85 ophioneus, Ichthyapus 67,68 Ophis 79,80,81 Ophisuraphis 65 Ophisurapus 65 Ophisurus 74,76,77,78,79,80,81,83 Ophithorax 79,81 opici, Hemerorhinus 67 orientalis, Lamnostoma 68,69 Oxydontichthys 80 pachyrhynchus, Echelus 76 pacificus, Ophichthus 81 pacificus, Paraletharchus 64 panamensis, Bascanichthys 71 Pantonora 69,70 Paraletharchus 64 Paramyrus 59 pardalis, Myrichthys 79 parilis, Ophichthus 81 paucidens, Cirrhimuraena 75 paulensis, Bascanichthys 71 Pelia 71,72 percinctus, Evips 77 perryae, Callechelys 63 Phaenomonas 61,73 philippinensis, Muraenichthys 59 Phyllophichthus 78,82 pictum, Lamnostoma 67,68 pinguis, Malvoliophis 78 pinnata, Phaenomonas 73 Pisodonophis 79,80,81,82,83,85 platyrhynchus, Myrophis 60 platyventris, Aprognathodon 62 playfairi, Cirrhimuraena 75 plumbeus, Myrophis 59,60 Poecilocephalus 79,81 Pogonophis 83 polyophthalmus, Ophichthus 81 porphyreus, Mystriophis 79 potamius, Stictorhinus 69 productus, Quassiremus 84 SERIES 4, V. 41, #1 McCOSKER — EELS Pseudomyrophis 60,61 Pterurus 71 pterygera, Caecula 66 punctatus, Myrophis 59,60 puncticeps, Ophichthus 80,81 punctifer, Echiophis 76,77 pusillus, Bascanichthys 71 Quassiremus 83 reguis, Ophichthus 81 remiger, Ophichthus 81 retifer, Ophichthus 81 retropinna, Schultzidia 81 retropinnis, Ophichthus 62 revulsus, Xyrias 84 Rhinenchelys 67,68 rosenblatti, Letharchus 64 roseus, Ophichthus 81 rostellatus, Mystriophis 79 rufus, Ophichthus 81 rutidoderma, Ophichthus 81 rutidodermatoides, Ophichthus 81 sauropsis, Brachysomophis 75 Schismorhynchus 59,61 schultzei, Muraenichthys 59 Schultzidia 59,61 Scolecenchelys 58,59 Scytalichthys 84 Scytallurus 71 Scytalophis 80,81 selachops, Ichthyapus 68 semicinctus, Leiuranus 77,78 semicinctus, Pisodonophis 82 serpens, Ophisurus 81,82 sibogae, Muraenichthys 59 Sphagebranchus 65,66,67,68,69,70,71,72,77 spinicauda, Ophichthus 80,81 springeri, Callechelys 63,72 stenopterus, Ophichthus 81 Stethopterus 77 Stictorhinus 69 striatus, Callechelys 63 Syletophis 80 Syletor 80 taeniopterus, Cirrhimuraena 75 taylori, Lamnostoma 69 tenuis, Bascanichthys 71,73 tenuis, Yirrkala 69 teres, Bascanichthys 70,71 thompsoni, Muraenichthys 59 triserialis, Ophichthus 81 Typhlotes 65 typus, Achirophichthys 65 unicolor, Ophichthus 81 Uranichthys 80,81 PROCEEDINGS CALIFORNIA ACADEMY OF SCIENCES 123 urolophus, Ophichthus 81 woosuitingi, Ophichthus 81 uropterus, Myrophis 60 xenodontus, Phyllophichthus 82 vafer, Myrophis 59,60 xorae, Muraenichthys 59 velifer, Letharchus 63,64 Xyrias 84 Verma 65,66 xystrurus, Myrichthys 78,79 vermiformis, Leptenchelys 72 Yirrkala 67,69 vermiformis, Muraenichthys 59 Zonophichthus 80,81,85 versicolor, Elapsopis 77 zophistus, Pisodonophis 82 vulturis, Ichthyapus 68 zophochir, Ophichthus 75,80,81 Te ey eee ae | ry "hae Ss Ly dy. _ A | a ese f / f cis wry) i dy dN ek Fe Way J eS eT ce he Mol pail 4 po 2 C25SA% NH PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES FOURTH SERIES Vol. XLI, No. 2, pp. 125-160, 16 figs., 11 tables April 15, 1977 STUDIES ON EASTERN PACIFIC SAND STARGAZERS (PISCES: DACTYLOSCOPIDAE) 4. G/LLELLUS, SINDOSCOPUS NEW GENUS, AND HETERISTIUS WITH DESCRIPTION OF NEW SPECIES By C. E. Dawson Gulf Coast Research Laboratory Museum Ocean Springs, Mississippi, 39564 ABSTRACT: The dactyloscopid genera Gillellus and Heteristius are diagnosed and described and a new genus, Sindoscopus, is proposed for the Chilean endemic, Gillel/us australis Fowler and Bean. A key is provided for all Pacific genera and the seven species treated here. Gillellus includes one species, G. searcheri n. sp. (Islas Tres Marias and Nayarit, Mexico and Costa _ Rica to Panama), with some paired infraorbital pores. There are four more closely related species without paired pores. The latter group includes the generic type, G. semicinctus Gil- bert (Isla Guadalupe, Baja California peninsula and western Gulf of California, mainland shores from Nayarit, Mexico to Colombia, and the Revillagigedos and Galapagos Islands). Gillellus arenicola Gilbert occurs in the vicinity of Cape San Lucas, fails to enter the Gulf of California, and has been collected off mainland coasts of Colima and Oaxaca, Mexico. Gill- ellus ornatus Gilbert is apparently restricted to the Gulf of California where all but one speci- men has been taken along its western shores. Gillellus chathamensis n. sp. is an insular en- demic known only from Cocos Island, Costa Rica. The monotypic Heteristius occurs along the southern portion of Baja California. It is not recorded from the Gulf of California. This species also appears off Mexican mainland shores (Isla Tres Marias, Nayarit to Oaxaca) and is found off Costa Rica, Panama and Ecuador (to about 00°27’N). Heteristius jalisconis Myers and Wade, type species of Heteristius, is considered a junior synonym of Dactyloscopus cinctus Osburn and Nichols. Sindoscopus, also monotypic, is known only from the Chilean coast (ca. 23°S- 33°S). All treated species are illustrated and distribution maps are provided for species of Gillellus and Heteristius. This represents the final part of a review of Pacific dactyloscopids wherein a total of 7 genera and 29 species or subspecies are recognized. CALIFORNIA ACADEMY OF SCIENCES/GOLDEN GATE PARK/SAN FRANCISCO 126 INTRODUCTION This report, the last of a series reviewing Pacific dactyloscopids, treats known populations of Gillellus Gilbert and Heteristius Myers and Wade, and a new genus, Sindoscopus, is pro- posed for the Chilean endemic, Gillellus australis Fowler and Bean. As with other Pacific sand stargazers, these forms have never been adequately treated sys- tematically, existing illustrations are few and in- accurate, and identification has been difficult or impossible. Material from the Revillagigedos Is., mainland coast of South America and the Gala- pagos Is. is minimal and further study is needed in these areas. Nevertheless, present data Clarify the systematic status of all known museum hold- ings and establish a basis for identification of future collections. METHODS AND MATERIALS Counts and measurements, recorded in milli- meters (mm), follow methods of Dawson (1969, 1974, 1975). Standard length (SL) is used through- out; counts of meristic characters are total counts; preopercular canal pores are enumerated in the 1st (anterior) primary canal (Fig. 1). Counts of lateral-line scales were routinely made on the right side, but left side counts were substituted where right side squamation was incomplete. The posteriormost short spiniform caudal elements may occasionally be irregularly segmented in Gillellus and Sindoscopus, but these are omitted from counts of principal rays which include only the long, well-developed, segmented elements. Branching of principal segmented caudal rays is indicated by the formula: number of upper simple rays + number of branched rays + simple lower rays. Measurements are given for undamaged primary types except for Sindoscopus australis where a specimen of similar length has been substituted for the damaged holotype. Ob- servations on dentition and general osteological features are from one or more cleared and stained specimens of each species; vertebral counts are from radiographs. Materials examined include virtually all known museum holdings, but meristic data were not taken on all specimens. Abbreviations for repositories of examined material are: AMNH - American Museum of Nat- ural History; ANSP - Academy of Natural Sciences PROC. CALIF. ACAD. SCI., 4TH SER., VOL. 41, NO. 2 of Philadelphia; BC - Institute of Animal Resource Biology, University of British Columbia; CAS - California Academy of Sciences; CAS-SU - Stan- ford University specimens now housed at CAS; GCRL - Gulf Coast Research Laboratory Museum; LACM - Los Angeles County Museum of Natural History; MNHN- Museo Nacional de _ Historia Natural, Santiago, Chile; SIO - Scripps Institution of Oceanography; UCLA- University of Cali- fornia, Los Angeles; UCR - Universidad de Costa Rica, San José; USNM- National Museum of Natural History, Smithsonian Institution. KEY TO GENERA AND SELECTED SPECIES OF PACIFIC DACTYLOSCOPIDS 1a. Dorsal-fin origin on nape -_...---.---.--------- 2 1b. Dorsal-fin origin behind nape, near verti- cal from: anal-fin origin —- Se 10 2a. Dorsal fin usually discontinuous; princi- pal preopercular canals 3 or more than 6 2b. Dorsal fin continuous; principal preop- erqular canals: 4 22 eee ....Sindoscopus australis (Fowler and Bean) 3a. Anterior dorsal-fin elements a series of free spines; eyes on protrusile stalks __ __Dactyloscopus Gill (see Dawson, 1975) 3b. Anterior dorsal-fin elements united to form a distinct 3 or 2-spined finlet; eyes protruding but not stalked -__.......---... - 4a. Principal segmented caudal rays 10; pec- toral rays modally 12 or 13; upper lip without fimbriae; venter always naked... 5 4b. Principal segmented caudal rays more than 10; pectoral rays modally 14; fim- briae present on upper and lower lips; venter at least partly scaled in late juv- eniles and adults __............. ees 9 5a. Dorsal spines total 11-15; 18-29 scales in straight portion of lateral line; infraorbi- tal canal pores in single series ---........... 6 5b. Dorsal spines total 20-23; 9-13 scales in straight lateral line; some paired pores in infraorbital canal _... ee ede ete) et eT Gillellus searcheri sp. nov. 6a. Anterior dorsal finlet 3-spined -.-.-.....-..- 7 6b. Anterior dorsal finlet 2-spined ---........... 8 DAWSON: EASTERN PACIFIC SAND STARGAZERS 7a. Pectoral rays usually 12; segmented anal rays 30-35 (usually 32-34); 18-23 (usually 19-21) scales in straight lateral line __...... “oe he me Gillellus semicinctus Gilbert 7b. Pectoral rays usually 13; segmented anal rays 33-36 (usually 34-35); 23-25 (usually 24) scales in straight lateral line —......_. 2 Gillellus chathamensis sp. nov. 8a. Tip of lower jaw neither distinctly fleshy nor strongly protruding anteriad; 22-26 scales in straight lateral line; lateral line deflects between verticals from dorsal-fin elements 18-22 _...Gillellus ornatus Gilbert 8b. Tip of lower jaw fleshy and strongly protruding in undamaged fish; 27-29 scales in straight lateral line; lateral line deflects between verticals from dorsal- FIMPCICMMICE ES P4502 ooccaceoen oe ee 9a. Principal segmented caudal rays 11, some branched in adults; dorsal finlet 3-spined; narrow band of scales across venter and scales on pectoral-fin base in late juveniles and adults; principal pre- Spcetlanwednals: We. a ....Platygillellus Dawson (see Dawson, 1974) 9b. Principal segmented caudal rays 12, un- branched; dorsal finlet 2-spined; venter fully scaled; pectoral-fin base naked in late juveniles and adults; principal pre- opercular canals more than 6 _............ ...._Heteristius cinctus (Osburn and Nichols) 10a. Head truncate to broadly rounded in front; upper lip fimbriae 12-21, usually more than 13; infraorbital canal pores TOTS! TT ET LC) ee ea ace ce ea ee his Dactylagnus Gill (see Dawson, 1976) 10b. Head conically pointed in front; upper lip fimbriae 4-13, usually less than 13; infraorbital canal pores less than 10, typi- eal | Ve GaN ge et ee ee ere ————— Myxodagnus Gill (see Dawson, 1976) SYSTEMATIC DESCRIPTIONS Gillellus Gilbert, 1890: 98 (type-species by original designa- ) Genus Gillellus Gilbert | tion, Gillellus semicinctus Gilbert, 1890). DIAGNOSIS: With separate 2- or 3-spined dorsal finlet originating on nape, with or without iso- | 127 lated spines between finlet and remainder of dorsal fin; eyes not stalked, usually with a minute flap or papilla; pectoral fin-rays modally 12 or 13; principal segmented caudal rays 10, some branched in late juveniles and adults; lower lip with up to 7 fleshy fimbriae, upper lip entire; head, pectoral-fin base and venter naked; princi- pal preopercular canals 3, unbranched; without predorsal bones (interneurals). DESCRIPTION: Body rather slender, broader and deeper in front, laterally compressed and narrow- ing caudad; head moderately large, rounded dor- sad, broadest near posterior margin of pre- opercle, somewhat narrowed in front; lower jaw slightly to strongly protruding, rounded in dorsal profile; upper lip entire, smoothly rounded and devoid of broad emarginations or fimbriae; lower lip with up to 7 (usually 4 or 6) unbranched fleshy papilla-like fimbriae (Fig. 5); eyes dorso- lateral, protruding but not on long protrusile stalks, usually with a flap or papilla; anterior naris tubiform, the aperture simple or with valve- like emargination, located on preorbital; poster- ior naris, a simple pore with slightly elevated rim, located behind tubiform naris and adjacent to anterior rim of orbit; principal preopercular can- als 3, unbranched, and typically with but a single pore in the 1st (anterior) canal; infraorbital canal pores in single series or paired posteriad (Fig. 1); opercle with a fringe of up to 12 short, usually unbranched fimbriae overlying upper posterior margin, the frequency ontogenetic (Fig. 7); opercles broadly expanded, the membranous margins cover pectoral- and pelvic-fin bases and overlap on underside of head. Dorsal-fin origin on nape, in advance of upper opercular angle; dorsal fin with an isolated 2- or 3-spined anterior finlet, 1st and 2nd spines more closely spaced than 2nd and 3rd; when 2- spined, finlet is followed by one or two isolated spines; remainder of spinous dorsal continuous with a series of segmented rays behind, the fin slightly to deeply emarginate between; anal spines 2, the 1st about a third shorter than 2nd which is similarly shorter than 1st segmented ray; anterior anal-fin elements of mature (trans- formed) males modified, the spines angled cau- dad, 1st and 2nd segmented rays angled forward, modified segmented rays not appreciably thick- ened or fleshy but membranes stretched or elongated between. Posteriormost dorsal and anal rays usually free, occasionally bound to caudal peduncle by short membranes; last anal ray reaches to or beyond vertical through rear of 128 {> SPE EEEEEE Sd ae Spe eee ee )=)>>>>=) > ~ ine) Ww [Neos eee Ff We PROC. CALIF. ACAD. SCI., 4TH SER., VOL. 41, NO. 2 Configuration of infraorbital pores and preopercular canals. Top: Gillellus semi- cinctus (left), G. searcheri (right). Bottom: Sindoscopus australis (left), Heteristius cinctus (right). Principal preopercular canals numbered 1-4; note complete abdominal squamation in Heteristius. hypural, last dorsal ray usually shorter; caudal peduncle entire, without notch in ventral margin (Fig. 2); caudal fin rounded, upper and lower- most rays typically simple, lower ray usually the shorter, middle rays simple in early juveniles but branch ontogenetically in larger fish (Table 9). Pectoral fin-rays 11-14, most frequently 12 or 13; pectoral fin narrowly rounded to pointed, 5th-7th from dorsalmost rays the longest, rays gradually shorter ventrad but lower rays not ap- preciably swollen or thickened; tip of straight- ened and adpressed pectoral fin may reach to or slightly beyond descending portion of lateral- line arch, usually falls short of descending arch; pectoral fin somewhat elongate and cupped or outwardly convex in transformed males; pelvic fin with a short spine closely bound to 1st of three segmented rays, the tips free with emargin- ate membranes between; longest pelvic ray may reach anal-fin origin but usually falls short of anus; all fin-rays simple except for branched caudal rays of late juveniles and adults. Lateral line continuous, originates just above and in advance of upper opercular angle (near DAWSON: EASTERN PACIFIC SAND STARGAZERS vertical from 1st dorsal spine), arches upward to follow near dorsal-fin base, deflects (between 14th-32nd _ dorsal-fin supports) and continues midlaterally to terminate as the penultimate scale on caudal-fin base; canal of straight lateral-line scales parallel to longitudinal body axis, that of last scale not strongly angled ventrad (Fig. 2). Scales of upper portion of arched lateral line persistent, outwardly convex and more or less pointed posteriad; scales of descending arch and straight lateral line thin and often deciduous, rounded in outline, the canal straight and with- out a distinct terminal branch anteriad, often with a minute dorsal pore on anterior third of canal. Squamation variable above |lateral-line arch, naked in early juveniles and some adults, fully or partially scaled in others; head, predor- sum, venter, sides below line between upper angle of pectoral axil and 1st to 3rd anal rays, and all fins (except caudal base) naked; 5-9 longi- tudinal scale rows cross vertical near origin of straight lateral line. Premaxillaries somewhat pro- tractile, the pedicels reach middle or posterior third of orbit but fall well short of posterior mar- gin; jaw teeth small, none distinctly enlarged, conical and more or less bluntly pointed, in 2-3 rows near symphyses but mainly uniserial pos- teriad; vomer and palatines edentate; tongue rounded in front; basihyal slender, barely spatu- late near tip. Three complete gill arches plus a hemibranch; no gill rakers on outer margin of 1st arch, often with a few rudimentary rakers that barely pick up stain on inner margin; pseudobranchiae present; branchiostegals 6; without predorsal bones (interneurals); proximal pterygiophore of 1st dorsal spine usually inserted in advance of 1st neural spine (Fig. 3); abdomi- nal vertebrae 10-12, modally 11 (from radio- graphs); caudal fin with 2-4 simple spiniform ele- ments above and below, upper and lower hypur- als each bear 5 segmented rays. Recognized species marked (in alcohol) with 6-10 principal dark saddle-like bands crossing dorsum between dorsal-fin origin and caudal- fin base, these often supplemented with second- ary bands between and seldom extend ventrad of lateral midline; body and fins elsewhere plain, _ flecked or spotted with brown; ground color pale to brownish. REMARKS: Gillellus is similar to Platygillellus Dawson, Leurochilus Bohlke and_ Heteristius Myers and Wade in possessing an isolated (or _ semi-isolated) dorsal finlet inserted on the nape 129 Fig. 2. Posterior body and fins, illustrating straight lateral-line scales and adult development of segmented caudal rays, in Gillellus (top), Sin- doscopus (middle) and Heteristius (bottom). (Table 1). In this feature it is immediately separ- able from Dactyloscopus Gill (continuous dorsal fin usually preceded by isolated spines) as well as from Dactylagnus Gill and Myxodagnus Gill where- in dorsal fin is continuous and originates well be- hind nape; these genera also have well devel- oped upper lip fimbriae (absent in Gillellus). PROC. CALIF. ACAD. SCI., 4TH SER., VOL. 41, NO. 2 130 GE tl + LL LL tf GL ‘vL ‘EL oO = exe} RIJIDed JO ‘ON AjjepOW aeIgayaA jeUIWOpgy jUaSaid sSajeds ejNdII9dO payeos aseq |e10}90d payeos 19}Ud/ OUI] Je19}e| }YB1e1}s JO Sajeds Ul PRlda}Ue PaydURIG sjeue>D aeuquily dij 1aMO7 eeuquiy dij saddpy yuasaid (eyjided) dey aAq poxjeys SaAq payouesq jeueD (JOIa}UR) JS] syeued Jejnd4adoaid jediduid AjjepOouw sAes-ut} [e10}D0q do} SQUIdS |eSIOP JSOWIOA}UY }Uuasaid jajUly |eSIOP JONA}UR JUNSIG snOnuljuod ul} |es1OG jUasaid (sjeinduUsa}Ul) S9UO [eSIOPId s}jnpe ul SARs yepned payduesq BWOS sAei jepned pajuawas jediouid SNI}S119}9H SNja//1dAVe/d sndoodsopuls SnffefliD snusepoxAw snusejAjaeqg sndoosojAjaeq YILIVAVHD “aepidoosojAyaeg Ajiwe} ay} JO eJaUad D1y!IDeq BUIYSINBUNSIP S49}IBIBYD Paj}I9]9S JO uosuedwoy “L JIdVL DAWSON: EASTERN PACIFIC SAND STARGAZERS TABEE 2: Gillellus, Sindoscopus and Heteristius. Genus and species N 11 12 18) Dorsal spines 14 15 16 17 18 Frequency distributions of dorsal spines, segmented dorsal rays and total dorsal-fin supports in Pacific 21 131 species of 22 23 Gillellus semicinctus chathamensis ornatus arenicola searcheri Sindoscopus australis Heteristius cinctus 190 30% © 87* 39 48 3 16 19 Segmented dorsal rays 21 22 23 24 25 26 27 32 33 Gillellus semicinctus chathamensis ornatus arenicola searcheri Sindoscopus australis Heteristius cinctus ZO 93" 74 3 54 40 4 42 43 4445 46 47 48 Gillellus semicinctus chathamensis ornatus arenicola searcheri Sindoscopus australis Heteristius cinctus 46* 127 58 16 12° Geyer she! 4 (0) 227/ *Primary type. 132 Flgee 3: (left), G. arenicola (right). Middle: G. ornatus (left), G. searcheri (right). Bottom: Sindosopus australis (left), Heteristius cinctus (right). PP - proximal pterygiophore; DSI - 1st dorsal spine. Platygillellus has 11 principal segmented caudal rays and modally 14 pectoral rays (caudal rays 10, modally 12 or 13 pectoral rays in Gillellus), fimbriae occur on both lips, and late juveniles and adults have scales on venter and _ pectoral- fin base (naked in Gillellus). The monotypic At- lantic genus Leurochilus differs from Gillellus in having 11 rather than 10 principal caudal rays, longer premaxillary pedicels (reach posterior edge of orbits), and in the complete absence of labial fimbriae. For comparison with Sindoscopus and Heteristius see Key and descriptions of these genera below. Gillellus is represented in Pacific waters by four closely related species and one somewhat aber- rant form (G. searcheri). This is apparently a stenohaline genus and no collections are re- corded from low salinity habitats. Maximum size is about 50 mm SL. Two species have been dredged in 137 m. The genus occurs from Isla PROC. CALIF. ACAD. SCI., 4TH SER., VOL. 41, NO. 2 Seat v =< (6) n LMU WEL Pp Anterior abdominal vertebrae and dorsal-fin supports. Top: Gillellus semicinctus Guadalupe, Baja California Norte, Mexico (ca. 28°53’N) south to the mainland coast of Ecuador and the Galapagos Is. (Archipiélago de Colon). Gillellus semicinctus Gilbert (Figure 4.) Gillellus semicinctus Gilbert, 1890: 98 (original description, Gulf of California [and Atlantic). DIAGNOSIS: Dorsal finlet 3-spined, without iso- lated spines between finlet and remainder of dorsal fin; pectoral fin-rays modally 12; seg- mented anal rays modally 33; 18-23 (usually 19- 21) scales in straight lateral line; eye flap usually present; without paired pores in_ infraorbital canal. DESCRIPTION: Measurements (mm) of 27.8-mm- SL male lectotype follow: depth of caudal pe- duncle 1.2; body depth 3.6; predorsal length 6.0; preanal length 9.1; pelvic to anal distance 3.6; DAWSON: EASTERN PACIFIC SAND STARGAZERS rs : oe , ie . ad > a 4 te Se ae Fig. 4. Gillellus semicinctus Gilbert. Top and middle: Strongly pigmented with indica- tions of secondary bands between primary bars crossing dorsum (GCRL 13244, 36 mm SL, female). Bottom: Pale pattern lacking secondary bands (CAS 31810, 32 mm SL, female). head length 6.6; maxillary to upper opercular angle 6.1; maxillary to upper preopercular angle 4.4: length of anterior naris 0.2; diameter of bony orbit 1.4; postorbital length 4.0; snout length 1.2; length of maxillary 2.1. See Tables 2-8 for meristic data. Lower jaw slightly protruding; lower lip typi- cally with 4 fleshy fimbriae or papilla-like emar- ginations, the inner pair the larger (Fig. 5); pre- orbital short, rather steeply inclined; tubiform anterior naris located close to anterior rim of preorbital; eye typically with a minute hyaline flap or papilla. Height of dorsal finlet equal to or slightly greater than height of remaining spin- ous dorsal fin, somewhat less (15-30%) than longest segmented dorsal rays (Fig. 6); last dorsal spine about 15-20% shorter than 1st segmented ray. Pectoral fin-rays 12 in 86% of 437 counts, 3 PROG. CALIF. ACAD, SGI_, 47H SER, VOL. 41, NO: 2 Fig. 5. Delineations of anterodorsal aspect of head. Top: Gillellus semicinctus (left), G. searcheri (right). Middle: G. arenicola. Bottom: Sindoscopus australis (left), Heteristius cinctus (right). Sensory canals and pores omitted. fish had 11 rays in each fin, whereas 9 had bi- lateral counts of 13; pectoral fin not greatly elongate or expanded in transformed males, tip of adpressed fin reaches vertical between dorsal- fin elements 16 and 19, between 14 and 19 in young and females; caudal fin formula modally 1 + 7 + 2 in examined material (Table 9); op- ercular fimbriae 4-12 (Fig. 7; Table 10); infra- orbital canal pores usually 7, none paired (Fig. 1). Arched portion of lateral line deflects between DAWSON: EASTERN PACIFIC SAND STARGAZERS Ls ore ie Pian ae vey, He Pete eee OFSTED. — a 4 Si yD: GOOD: / / Z he iS 135 a SSSSSSS SISSIES SSDS OOTP F RE. Fig. 6. Lateral and dorsal views illustrating lateral-line configuration and squamation above lateral-line arch in Gillellus semicinctus (top), G. arenicola (middle) and G. ornatus (bottom). verticals from dorsal-fin elements 17 and 24, usu- ally between 20 and 23; squamation variable above lateral-line arch, naked in early juveniles and some adults, usually with one or two scale rows anteriad above 2nd or 3rd lateral-line scale, narrowing to a single row for a distance of some 10-11 scales and replaced posteriad by the dor- salmost portions of 5-6 lateral-line scales in ad- vance of lateral-line deflection; anteriormost scales often embedded, occasionally isolated; area between nape, origin of lateral lines and base of 1st or 2nd dorsal spine naked (Fig. 6); 8-9 longitudinal scale rows cross vertical near origin of straight lateral line; scales present in arched portion of lateral line in 13-mm fish, squamation may be complete in fish about 20 mm SL. Lower jaw with a patch of teeth 2-3 rows broad and about 4-5 rows long near symphysis, uniserial posteriad; anterior teeth similar in up- per jaw but distinctly biserial behind; pseudo- branchiae 4-5 (7 specimens examined); proximal pterygiophore of 1st dorsal spine broad (Fig. 3). Color in alcohol. —Usually with an indistinct bar crossing interorbital and continued below eye to near posterior angle of gape; often with flecks of brown on snout, underside of lower jaw, upper portion of preopercle and predor- sum; lips and labial fimbriae usually pale. Dor- sum crossed by 6-8 (usually 7) principal saddle- like bars between nape and caudal-fin base; these primary bars equal to or slightly narrower than pale interspace, the 1st terminates ventrad near middle of opercle, last covers most or all of caudal base, whereas remainder terminate near lateral midline; well pigmented (usually large) specimens often with short secondary bars crossing dorsum between primary markings, nu- merous brown flecks on otherwise pale inter- spaces and two or more primary bars may be united ventrad by a narrow brown line; sides below midline mainly pale but occasionally with scattered brown flecks; pectoral fins usually pale but upper interradial membranes may be lightly streaked with brown, other fins typically im- maculate. Shape and pigmentation of lateral por- tions of primary bars (behind nape) variable; within a single sample, bars may be narrowed between lateral midline and dorsal-fin base, rec- tilinear, or narrowed below and somewhat tri- 136 SG. semicinctus SSS G. chathamesis OPERCULAR FIMBRIAE OPERCULAR FIMBRIAE O 20 60 100 0 20 STANDARD LENGTH (MM) G. arenicola G. searcheri G. ornatus STANDARD LENGTH (MM) PROC. CALIF. ACAD. SCI., 4TH SER., VOL. 41, NO. 2 S australis SS SS SS H. cinctus OPERCULAR FIMBRIAE (0) 60 100 O 20 60 100 STANDARD LENGTH (MM) Fig. 7. Regressions of frequencies of opercular fimbriae in Pacific species of Gillellus, Sindoscopus and Heteristius. See Table 10 for supporting data. angular; bars may have evenly distributed light to dark brown pigmentation or may have dark margins with light brown or pale between; inter- orbital bar and principal saddle-like markings are distinct in 14-mm fish. REMARKS: Gilbert (1890) failed to select a holo- type and based his description on specimens from both Atlantic (GRAMPUS Sta. 5108, 5112) and Pacific waters (ALBATROSS Sta. 2827, 2829). This ultimately resulted in uncertainty (Bohlke, 1968; Dawson, 1974) as to which of the two in- cluded species should be considered semicinctus, the type-species of Gillellus. | have recently found that no such problem exists and that the species must be based on Pacific material. Jordan (1897) referred to Gillellus semicinctus (p. 229) stating, “I present a figure of this interesting species from one of the types” and illustrated a specimen in Plate XXXII. The introduction (p. 20) indicates that Jordan’s material was in the Stan- ford collection and the list of plates (p. 244) states, ““Gillellus semicinctus: type. Gulf of Cali- fornia.’ Pacific specimens of the syntypic series include one from ALBATROSS station 2827 (US NM 126880) and two (CAS-SU 31) from ALBA- TROSS station 2829, located off Cabo San Lucas, Baja California Sur. Included in USNM 126880 is a handwritten label by Barton A. Bean (‘B. A. B.’’) dated January 1918 stating, ‘‘According to Dr. Jordan this is a cotype or paratype. The type is at Stanford. A third specimen (Florida) is in Nat. Museum.” There are two fish (25.8 and 27.6 mm SL) and a metal tag imprinted “drawn” in CAS-SU 31. The larger specimen has a total of 41 dorsal-fin sup- ports, whereas the smaller has 42. Jordan’s figure is very poor and shows a continuous rather than an anteriorly interrupted dorsal fin but dorsal-fin supports total 41. | conclude that Jordan figured the larger fish and thereby designated this speci- ment as the lectotype of Gillellus semicinctus in accord with Article 74b, International Code of Zoological Nomenclature. The paralectotype has been renumbered CAS-SU 67760. Considerable study material is available from Baja California waters, but representation from other areas is minimal and most specimens are small (under 30 mm SL) and in poor condition. Total material from the Revillagigedos consists of 17 specimens (10-22 mm SL) and there are but 8 (15-31 mm) from the Galapagos Islands. Available data (Table 11) suggest that counts of dorsal and anal rays, vertebrae, and lateral-line scales average slightly higher in the Baja Cali- fornia population (including Isla Guadalupe) than those of Revillagigedos or Nayarit, Mexico or Galapagos material. These differences may indi- cate distinct populations or merely clinal varia- DAWSON: EASTERN PACIFIC SAND STARGAZERS G. semicinctus G. chathamensis G. searcher! = G ornatus A G arenicola SG H. cinctus Fig. 8. General distribution maps of Pacific species of Gillellus and Heteristius. tion, but further treatment of this problem must await receipt of additional specimens. Among examined material, the smallest trans- formed male from Baja California waters was 28 mm SL, the smallest ovigerous female was 33 mm; a 26-mm transformed male and 7 ovigerous females (26-30 mm) occurred in a collection from Nayarit, Mexico. A 41-mm SL female contained 224 near-ripe eggs (maximum diameter about 0.7 mm) in the right gonad, together with about an equal number approximately 0.3 mm in dia- meter. Among Pacific dactyloscopids, the recorded bathymetric range of Gillellus semicinctus is matched only by that of G. arenicola and Dacty- loscopus lunaticus. Whereas the latter forms may occur in depths of a meter or less, | have no con- firmed records of Gillellus semicinctus from less than 5 m. Among collections with complete data, 29 were by dredge and 15 with SCUBA; 8 were 38 PROC. CALIF. ACAD. SClI., 4TH SER., VOL. 41, NO. 2 Fe, Fig. 9. Gillellus chathamensis n. sp. (GCRL 13215, female paratype, 23 mm SL). Inéz to Isla Angel de la Guarda: CAS 31809 and 31810; in confirmed depths of 5-15 m, 22 in 17-46 m, GCRL 13246 and 13341; LACM 23982 through 23984, 23990 and 9 in 50-137 m. Known from Islas Revillagig- edos, Isla Guadalupe (ca. 28°53’N) and Pacific coast of Baja California, western Gulf of Cali- fornia, various mainland and insular shores from Nayarit, Mexico to Isla Gorgona, Colombia and Galapagos Is. (Archipiélago de Colon), Ecuador (Fig. 8). MATERIAL EXAMINED: Two hundred eighty-two specimens, 9.6-48.3 mm SL, including lectotype and two paralectotypes. Lectotype. —CAS-SU 31 (27.6 mm SL, male), Mexico, Baja California Sur, 22°52'00'’'N, 109°55’'00’’W, 56.7 m, 1 May 1888, ALBATROSS Sta. 2829 Paralectotypes. —-CAS-SU 67760 (25.8 mm), taken with lectotype. USNM 126880 (ca. 31.2 mm), Mexico, Baja Cali- fornia Sur, 24°11'45’"N, 109°55’00""W, 18.3 m, ALBATROSS Stan 28277, Other Material. —MEXICO, Baja California Norte (Pacific). —Isla Guadalupe: SIO 50-31, 57-190, 58-497 and GCRL 13241. —Isla Cedros: SIO 63-168, 71-33 and LACM 23999. Baja Cali- fornia Sur (Pacific). —Bahia Magdalena to Cabeza Ballena: LACM 23985, 23986, 23996, 23997 and SIO 64-61. Baja Cali- fornia Sur (Gulf of California). —Punta de los Frailes to Punta Concepcion: CAS 32037; CAS-SU 5934; GCRL 13242 through 13245; LACM 9728-2, 23987 through 23989, 23994 and 23998: SIO 61-247, 61-249, 65-243, 65-270, 65-278, 65-295, 65-311, 65-317, 65-321, 65-335 and 65-337; UCLA W65-49 and W65- 86. Baja California Norte (Gulf of California). —Isla Santa through 23992, and 23995; SIO 65-306; UCLA W53-69 and W53-73. Islas Revillagigedos. —Isla Socorro: SIO 70-392. Isla Clarion: UCLA W55-136. Nayarit to Oaxaca: CAS 6923, 32035 and 32036; LACM 23993. COSTA RICA. —Puntarenas: UCR 664-6. —Isla del Cano: GCRL 13247 and 13248; LACM 32547-48 and 32562-46, UCR 675-11 and 685-1. PANAMA. —Islas Secas: USNM 101734. —Isla Jicaron: SIO 71-87. COLOMBIA. —LACM 23981 and USNM 101725. —Isla Gor- gona: USNM 101738. ECUADOR, Galapagos Islands (Archipiélago de Colon). —Isla Isabela (Albemarle Is.): LACM 23980. —Isla Santa Maria (Charles Is.): UCLA W67-43 and USNM 101735. —Isla Baltra (S. Seymour Is.): USNM 109426. -Isla Santa Cruz (Inde- fatigable Is.): CAS-SU 39962. —Isla Marchena (Bindloe Is.): USNM 205457. Gillellus chathamensis Dawson, new species (Figure 9.) DIAGNOSIS: Dorsal finlet 3-spined, without iso- lated spines between finlet and remainder of dorsal fin; pectoral fin-rays modally 13; seg- mented anal rays modally 34; straight lateral-line scales 23-25 (usually 24); eye flap usually obso- lete or vestigial; without paired pores in infra- orbital canal. DAWSON: EASTERN PACIFIC SAND STARGAZERS DESCRIPTION: Measurements (mm) of 25.8-mm SL transformed male holotype follow: caudal fin length 3.8; length of uppermost segmented caudal ray 3.0; length of lowermost segmented caudal ray 2.7; depth of caudal peduncle 1.1; body depth 2.8; predorsal length 4.5; preanal length 6.8; pelvic to anal distance 3.0; head length 5.2; head breadth 2.8; maxillary to upper opercular angle 4.8; maxillary to upper preoper- cular angle 3.4; length of anterior naris 0.2; eye diameter 1.2; postorbital length 3.0; snout length 1.2; maxillary length 1.8; length of 1st dorsal spine 1.5, 3rd spine 0.5, 4th spine 0.9, last spine 1.5; length of 1st segmented dorsal ray 1.9. See Tables 2-8 for meristic data. Lower jaw slightly protruding, rounded to sub- triangular in front; 4 fimbriae on lower lip; pre- orbital short, moderately inclined; tubiform an- terior naris located near anterior rim of preorbi- tal; eye flap usually vestigial or obsolete; height of dorsal finlet about equals that of remaining spinous dorsal fin, somewhat less than length of longest segmented dorsal ray. Pectoral-fin rays 13 in 87% of 134 counts, one fish had 12 rays in each fin, whereas three had bilateral counts of 14; pectoral fins not broadly expanded in trans- formed males, may reach past lateral-line de- flection; tip of adpressed fin reaches vertical be- tween dorsal-fin elements 17 and 22 in trans- formed males, between 14 and 18 in young and females; segmented caudal rays frequently simple, but some median rays branched in most material examined (Table 9); opercular fimbriae 4-8 (Fig. 7; Table 10); infraorbital canal pores 7, none paired. Arched portion of lateral line deflects between verticals from dorsal-fin elements 17 and 21, usu- ally between 18 and 20; squamation above lat- eral-line arch similar to that of G. semicinctus; 9 longitudinal scale rows cross vertical near origin of straight lateral line; squamation apparently complete in 16-mm fish. Dentition essentially as in| semicinctus; pseudobranchiae 4-5 in two specimens examined; proximal pterygiophore of 1st dorsal spine broad. Color in alcohol. —Ground color is pale in all material, markings tan to dark brown. Usually with only faint indication of an interorbital bar and without suborbital bar to angle of gape; 4-5 prominent spots about eye and with additional spots on upper sides and dorsum of head, some fish with a rather large brownish blotch on dorsum immediately behind each eye; usually with a faint narrow bar cross- 1139) ing upper and lower jaws on either side of sym- physis and 2-4 isolated spots on ventral margins of lower jaw; there is a small blotch dorsad be- fore 1st dorsal spine and 7-8 (usually 7) saddle- like primary bars cross dorsum from finlet to caudal-fin base, margins of bars often incom- pletely pigmented and median areas poorly marked; secondary bars often indicated by spots or faint blotches on upper portions of pale inter- spaces; head and sides elsewhere pale. Dorsal finlet usually with some brown pigmentation, this may be restricted to posterior membrane, to a few isolated spots, or may shade entire finlet; other fins immaculate. Specimens from SEARCH- ER Sta. 517 are more strongly marked and ap- proach G. semicinctus in general coloration. They have prominent interorbital and suborbital bars, there is little or no pigment on snout or jaws, the predorsal spot (if any) is included in the 1st (anterior) primary bar, primary bars are more or less evenly pigmented and some fish have a few brown flecks anteriad above anal fin base. ETYMOLOGY: The name chathamensis refers to the type-locality, Chatham Bay, Isla del Coco. COMPARISONS: This species is distinguished from the closely related G. semicinctus by modal counts of 13 pectoral rays and 24 straight lateral- line scales (12 and 19-20 in semicinctus). The combination of unpaired infraorbital canal pores, 3-spined dorsal finlet, and modally 13 pectoral rays and 24 straight lateral-line scales separates Gillellus chathamensis from all Pacific congeners. REMARKS: This species may readily be confused with G. semicinctus, and careful examination is required for identification. Among material ex- amined, the smallest transformed male is 18.9 mm SL, the smallest ovigerous female (19.6 mm) contained 104 developing eggs, about 0.5 mm in maximum diameter. All specimens have been taken with SCUBA in depths of 5-12 m. Gillellus chathamensis is known only from Isla del Coco, Costa Rica (Fig. 8). MATERIAL EXAMINED: Holotype and_ sixty-six paratypes, 15-32 mm SL. Holotype. —LACM 32256-23 (25.8 mm SL, male), Costa Rica, Isla del Coco, Chatham Bay, 05°33'12’’N, 87°02'14’'W, 6 m, SCUBA, R/V SEARCHER, sta. 511, 1 April 1972, R. J. Laven- berg and party. Paratypes. —COSTA RICA, Isla dei Coco, Chatham Bay: UCR 711-16, GCRL 13215 and 13216, 05°33'29’’N, 87°02'48''W, 11 m, SCUBA, R/V SEARCHER Sta. 509. LACM 32260-22, 140 PROC. CALIF. ACAD. SCI., 4TH SER., VOL. 41, NO. 2 Fig. 10. Gillellus arenicola Gilbert (GCRL 13240, 34 mm SL, young male). GGRE1382135705233121NF 87-°03/157W, 12.m; SGUBA, IR/V SEARCHER Sta. 517. LACM 32256-24, GCRL 13211 and 13212, taken with holotype. —Wafer Bay: LACM 32272-18, GCRL 13214, 05°32’45’’N, 87°03'49"W, 5 m, SCUBA R/V SEARCHER Sta. 532. Gillellus arenicola Gilbert (Figure 10.) Gillellus arenicola Gilbert, 1890: 99 (original description; Cape San Lucas, Baja California Sur). DIAGNOSIS: Dorsal finlet 2-spined; pectoral fin- rays modally 12; scales in straight lateral line 27- 29 (x = 28); lateral line deflects between verti- cals from dorsal-fin elements 18 and 22; lower jaw fleshy and strongly protruding (in undamaged fish); without paired pores in infraorbital canal. DESCRIPTION: Measurements (mm) of 41.1-mm SL neotype follow: caudal-fin length 7.5; length of uppermost segmented caudal ray 5.5; length of lowermost segmented caudal ray 5.0; depth of caudal peduncle 2.0; body depth 4.6; predorsal length 7.7; preanal length 11.5; pelvic to anal distance 3.9; head length 8.9; head breadth 3.6; maxillary to upper opercular angle 8.0; maxillary to upper preopercular angle 5.8; length of an- terior naris 0.2; eye diameter 1.6; postorbital length 5.4; snout length 2.1; maxillary length 2.8; length of 1st dorsal spine 0.9, 2nd spine 0.7, 4th spine 0.6, last spine 2.1. See Tables 2-8 for meristic data. Lower jaw distinctly protruding and produced anteriad as a conical fleshy prominence (Fig. 5); labial fimbriae 4; preorbital long, about equals diameter of pigmented eye, not steeply inclined; tubiform anterior naris located near middle of preorbital length, well removed from anterior rim; eye usually with a flap. Height of dorsal fin- let about 1/3 to 1/4 that of spinous portion of continuous dorsal fin; dorsal fin slightly notched or emarginate between last spines and anterior segmented rays. Finlet typically 2-spined (3- spined in two of 58 examined) and with one or two (40% of 35 examined) isolated spines be- tween finlet and continuous dorsal fin (Fig. 6). Pectoral fin-rays 12 in 75% of 89 counts, 7 fish with bilateral counts of 13; pectoral membranes expanded and the fin elongate in transformed males, may reach to or somewhat beyond origin of straight lateral line; tip of adpressed fin reaches vertical between dorsal-fin elements 15 and 20 in transformed males, between 12 and 17 in young and females; some median caudal rays branched in adults (Table 9); opercular fimbriae 3-8 (Fig. 7; Table 10); infraorbital canal pores usually 7, none paired, DAWSON: EASTERN PACIFIC SAND STARGAZERS 141 TABLE 3. Frequency distributions of segmented anal and pectoral fin-rays in Pacific species of Gillellus, Sindoscopus and Heteristius. S Segmented anal Gane : ea Genus and species N 30 31 32 33 34 35 36 37 38 39 40 41 x Gillellus semicinctus 219 4 17 64 100* 31 3 332,7/ chathamensis 56 So PD 7ET ROSS aei3 34.5 ornatus 75 2 Oe sbi 923 5 B5r2 arenicola 52 1 We eas 6 1 35.8 searcheri lO S22 25 48 Bile Sindoscopus australis 89 3 Bo se WC 28a, Heteristius cinctus 107 PANG I 7/ 2 34.6 Pectoral fin-rays N 1 12 13 14 15 x Gillellus semicinctus 437 U2 3A ANS 2 Pel chathamensis 134 5 Wee 13 13511 ornatus 150 4 139* 7 12.0 arenicola 89 O° 2D 12.2 searcheri 306 Assy |G 1301 Sindoscopus australis 185 37 148* 13.8 Heteristius cinctus 208 eA e2.7, 14.0 *Primary type. Arched portion of lateral line deflects between verticals from dorsal-fin elements 14 and 17, usu- ally between 15 and 16; area above arch usually fully scaled in adults, 3-4 scale rows between base of 1st dorsal spine and 1st or 2nd lateral- line scale, narrowing to a single row above pos- terior half of arch (Fig. 6); 9 longitudinal scale rows cross vertical near origin of straight lateral line; lateral-line scales present in 14-mm SL fish, squamation apparently complete in some 22-mm specimens. Dentition essentially as described for G. semicinctus; pseudobranchiae 4 (2 examined); proximal pterygiophore of 1st dorsal spine com- paratively slender (Fig. 3). Color in alcohol. —Coloration is poorly pre- served in study materials but well marked speci- mens have indications of an interorbital bar and a more or less distinct suborbital bar crosses posterior third of gape to terminate on ventral margin of lower jaw; a narrow bar crosses under- side of lower jaw; upper sides and dorsum of head plain, lightly spotted or blotched with brown. Dorsum crossed by 7 principal saddle- like bars between nape and caudal-fin base; late juveniles and adults often with secondary bars which may approach size and pigmentation of primary bars, the net effect being a series of 12- 14 saddles; pale interspaces of upper side some- times flecked with brown and a few scattered flecks may be present on lower portion of side; all fins immaculate. COMPARISONS: The 2-spined dorsal finlet and long, distally fleshy, lower jaw separate undam- aged Gillellus arenicola from all Pacific con- 142 PROG. CALIF; ACAD. SGI., 48H SER., VOL. 47,(\NO: 2 geners. Specimens with damaged lower jaw can be confused with G. ornatus, the only other spe- cies with 2-spined finlet (see ornatus for further comparisons). It is distinguished from searcheri by a number of meristic characters (Tables 2-8) and from both semicinctus and chathamensis by modal frequency of total dorsal spines (14 in arenicola against 13), number of straight lateral- line scales (27-29 against 18-25), and position of lateral-line deflection (between dorsal-fin ele- ments 14 and 17 versus 17 and 24 in semicintus and chathamensis). The several Pacific species of Myxodagnus (see Dawson, 1976) also have projecting lower jaws and are superficially similar to Gillellus arenicola. In Myxodagnus, however, there are both upper and lower labial fimbriae, dorsal fin is continuous and originates near verti- cal from anal-fin origin, and there are no promi- nent saddle-like blotches crossing the dorsum. x 29.0 28.2 28.8 25.4 yi? 33.0 21.4 62 Ces M4 SS 7 ace 1 1 1 1 REMARKS: Gilbert’s (1890) description was based on “A single specimen, 1-1/2 inches long, from Cape San Lucas.” This fish was not illustrated; there is no record of its receipt at the National Museum, and it is apparently lost. | therefore select a 41-mm SL specimen (CAS-SU 67821) as the neotype of Gillellus arenicola Gilbert. This fish, apparently an immature male, was collected near the type-locality; see measurements above, material examined and Tables 2-8 for additional data. There are relatively few specimens in collec- tions and most of these are in rather poor con- dition. Among material examined, the smallest transformed male is 32 mm SL, the smallest ovig- erous female 40 mm; a 49-mm female contained 129 developing eggs in the right gonad. Gillellus arenicola has been taken with seine, dredge or SCUBA to depths of 137 m. It is known from the southern portion of Baja California Sur and from single collections off the states of Colima and Oaxaca, Mexico (Fig. 8). sy eisk eb sky si yf sis) Ss tas} yh bi sh te) YS). Number of arched lateral-line scales 1 19 BO Bi) D2 PB BA Wh A Py/ Ph} VE) 30) Sil 1 ies (7 Ia) 24 Gt 810 9 4 9 3 M7 39) bye 2 3\* A137" 22 5 §) 70) 20 7/ 2 © 1 1 1 Frequency distribution of arched lateral-liné scales in Pacific species of Gillellus, Sindoscopus and Heteristius. 3} iI) 17/7 18 1 MATERIAL EXAMINED: Fifty-eight specimens, 13-50 mm SL, including neotype. Neotype. —CAS-SU 67821 (41.1 mm SL), Mexico, Baja California Sur, Bahia de San Lucas, 22°53’00’’N, 109°52’00’’W, Sefton-Stanford ORCA Expdn. Sta. 44-D-2, 19 May 1952. Other Material. —MEXICO, Baja California Sur. —San Jaime Bank to Bahia de las Palmas: AMNH 5398 and BC 60-513; CAS 2716, 14212 and 32038 through 32040; CAS-SU 18915; GCRL 13237 through 13240; LACM 23968, 23969 and 31774- 45; SIO 61-232, 61-247 and 62-704 or 59-210 (loc. uncer- tain); UCLA W59-251; and USNM 2533. Colima: LACM 23967. Oaxaca: CAS 32034. N 87 25 47 29 93 62 49 TABLE 4. cinctus semicinctus chathamensis ornatus arenicola searcheri australis *Primary type. Genus and species Gillellus Sindoscopus Heteristius 143 DAWSON: EASTERN PACIFIC SAND STARGAZERS ‘adA} Arewtid, CUS S) SIL ib Ob 47 6v Sn}DUID SN1}S119}9H JL e) boi G@ G t Jb 4 giz t & & c9 sijesjsne sndoosopuls 6°29 Ib [b es) 66 EG We G Y @ €6 layoseas OES 17 eS) 8 KE 6¢ P/OD/uale EGS bi ae eS) eel (OE eae LY snyeuso 6 LS L GD) VA ES 1 SG SISUBWUBYJEYD 88 ES lb TE [bea tb ae [LIL Z8 SN}JIUIDI| Was SNYPPT!D x SLCL LZ OZ69)5 89290599) SO POSTEO COL OOO SC Su ZS) OS eSGyG) eS) (CSnGeOGE Oh GhueZtanoy uGPautt, N saisads pue snuap Sa]eds duUl|-;e1aye] eJOL “SNi}S1J9}9H puke sndodsopuls ‘snjjajjiQ JO saidads 1I9eq Ul Sajeds aul|-je1a}e] |2}0} yO UOlZNQIysIp ADUaNbal4 “9 4IGVL sadA} Asewiid, CrCGeala ce aces aes 6v sn}aula snijsi49}9H ZOE [EG ASST RETE heir Bree I 79 sijesjsne sndoodsopuls SOL G SiIb OL GGAGIL (5 19YIICIS SG vy +OL SL 67 P/ODIuase LG ES cG SEES Lv sn}eulo SEG CG «Sl 8 cc sisuawey}yeyo 8'6L GS Sie ce cS, ¥ Z8 sn}dUIDI Was SN]/P/}!D x GE ifs fe Gs Me G8 4 fe dé Ge Se te be we Ie Oe GL OF 24 OL Si Dh Gh @h i OF G WN saisads pue snua5 Saj]eIS QuUl|-je1a}e] }YBIeI}S JO JaqUINN “sNijsi4a}9H Puke sndodsopuls ‘snjjajjid 4O saidads dij1I9eq Ul sajeds aulj-je19}ye] }YSIeW}s yO suOIyNqiiysIp ADUanbas4y “Ss FIGVL 144 PROG. GALE AGAD® SGI 4h SERY VOL 41 NOF2 Fig. 11. Gillellus ornatus Gilbert (GCRL 13234, 41 mm SL, transformed male). Gillellus ornatus Gilbert (Figure 11.) Gillellus ornatus Gilbert, 1892: 558 (original description, Gulf of California). DIAGNOSIS: Dorsal finlet 2-spined; pectoral fin- rays modally 12; scales in straight lateral line 22- 26 ( mean = 24); lateral line deflects between verticals from dorsal-fin elements 18 and 22; lower jaw neither fleshy nor strongly protruding anteriad; without paired pores in_ infraorbital canal. DESCRIPTION: Measurements (mm) of 36.2-mm SL transformed male neotype follow: caudal fin length 6.4; length of uppermost segmented caudal ray 5.0; length of lowermost segmented caudal ray 4.3; depth of caudal peduncle 1.7; body depth 3.6; predorsal length 6.2; preanal length 9.2; pelvic-anal distance 3.4; head length 7.5; head breadth 3.5; maxillary to upper oper- cular angle 7.1; maxillary to upper preopercular angle 4.7; length of anterior naris 0.2; eye dia- meter 1.5; postorbital length 5.1; snout length 0.6; maxillary length 2.0; length of 1st dorsal spine 1.3, 2nd spine 0.8, 3rd spine 0.6, 4th spine 1.7, last spine 1.8; length of 1st segmented dor- sal ray 2.2. See Tables 2-8 for meristic data. Lower jaw slightly protruding, not continued as a fleshy prominence anteriad; labial fimbriae 4; preorbital short, moderately inclined; tubi- form anterior naris slightly nearer anterior rim than middle of preorbital length; eye usually with a prominent flap. Height of dorsal finlet usually less than that of continuous portion of dorsal fin, segmented dorsal rays longer than spines. Finlet always 2-spined, with one (95%) or two isolated spines between finlet and con- tinuous portion of dorsal fin (Fig. 6); membrane rather deeply incised between 1st and 2nd spines of continuous dorsal fin, the fin emarginate be- tween last spine and 1st segmented ray. Pectoral fin-rays 12 in 93% of 150 counts, one fish with 11 rays in each fin, three with bilateral counts of 13; pectoral fin elongate in transformed males, DAWSON: EASTERN PACIFIC SAND STARGAZERS TABLE 7. Frequency distributions of caudal vertebrae and opercular fimbriae in Pacific species of Gillellus, 145 Sindoscopus and Heteristius. Caudal vertebrae i Genus and species N 30 31 32 33 34 35 36 37 38 39 40 41 42 x Gillellus semicinctus 190 Se MO 57 85s 24 88} 05 chathamensis 56 3 RH 3} 3 B55 ornatus 72 1 23 eS Ona 35.9) arenicola 47 O28 eo 1 36.0 searcheri 198 Yq Bil [ses SX) 2 32.0 Sindoscopus australis 93 26 43 Di | 3 40.0 Heteristius cinctus 134 (HO G0) 7 3535 Opercular fimbriae N 2 3 4 5 6 7 8 9 10 Th 12 Gillellus semicinctus 365 4 i 58* 109* 102 55 27 2 1 chathamensis 11333) 3 BS) ZIG ornatus 146 1 1 4 ai sty 32 13) 4 if arenicola 79 1 Ye DRY SNF] searcheri 318 1 SS OZ 658 8 Sindoscopus australis 180 2 1) LO 5.6 28 4 1 Heteristius cinctus DiV7 1 1 Sy (7/3 GF 1 *Primary type. the interradial membranes expanded, the fin scale, narrowing to a single crowded row behind somewhat cupped or outwardly convex; tip of adpressed fin falls short of origin of straight lateral line, reaches vertical between dorsal-fin elements 15 and 19 in transformed males, be- tween 12 and 15 in young and females; caudal fin formula modally 2 + 6 + 2 in examined mate- rial (Table 9); opercular fimbriae 2-10 (Fig. 7; Table 10); infraorbital canal pores 6-7, none Paired. Arched portion of lateral line deflects between verticals from dorsal-fin elements 18 and 22, usually between 18 and 21, squamation variable above lateral-line arch; naked in early juveniles and some adults; some with a few scattered scales anteriad (Fig. 6), others with about two rows originating above 2nd or 3rd lateral-line and replaced by impinging lateral-line scales over posterior half of arch; 7-9 longitudinal scale rows cross vertical near origin of straight lateral line; arch scales present in 12-mm SL fish, squa- mation apparently complete in some 19-mm specimens. Jaw teeth apparently in two rows over anterior half of upper jaw and in 2-3 rows near symphysis of lower jaw, teeth in single row posteriad; pseudobranchiae 3-4 (3 examined); proximal pterygiophore of 1st dorsal spine broad (Fig. 3). Color in alcohol. —An interorbital bar present and continued across suborbital as a broad blotch which may extend around and below posterior angle of gape to posteroventral margin of lower jaw; a faint narrow bar extends across 146 upper and lower jaw on either side of symphysis; upper sides and dorsum of head blotched and spotted with brown, head elsewhere mainly pale. Dorsum crossed by 8-10 (usually 9) principal saddle-like bars between nape and caudal-fin base; well marked late juveniles and adults have secondary bars, most distinct near dorsal-fin base, but frequently indicated only by brown flecks on upper sides; ventral margins of primary bars (behind nape) may be more or less united by irregular brown flecks along midline of sides; body elsewhere mainly pale, but some specimens lightly flecked with brown on lower sides; fins mainly pale but occasionally there is a trace of brown shading on dorsal finlet; primary bars are distinct in 12-mm fish. COMPARISONS: Among Pacific congeners, Gill- ellus ornatus shares the 2-spined dorsal finlet only with the sympatric G. arenicola. It differs from that species in average counts of arched and straight lateral-line scales (29 and 24 versus 25 and 28 in arenicola) and modal counts of total dorsal spines (13 against 14). In addition, the lateral-line arch deflects between verticals from dorsal-fin elements 18 and 22 in ornatus (between 14 and 17 in arenicola) and lower jaw is short and scarcely protruding (distally fleshy and protruding in undamaged arenicola). A num- ber of meristic characters (Tables 2-7) distinguish ornatus from the aberrant G. searcheri but the remaining 3-spined species (semicinctus and chathamensis) may be confused with ornatus wherein the finlet has been damaged. In such cases, Gillellus ornatus is best separated from chathamensis by the lower modal count of pectoral rays (12 against 13 in chathamensis), and from semicinctus by higher average values for total dorsal-fin elements and straight lateral- line scales (44 and 24 versus 41 and 20 in semi- cinctus). REMARKS: Gilbert (1892) based his description on a single specimen (“about 2 inches long’) from ALBATROSS Sta. 2828 located off the W side of Isla Cerralvo in the Gulf of California (24°11'30’"N, 109°55’00’W). This specimen was never received at the National Museum, has never been illustrated and is presumably lost. | therefore select a mature male (LACM W65-86- 1) as the neotype of Gillellus ornatus Gilbert (see measurements above, material examined and Tables 2-8 for additional data). Among material examined, the smallest trans- formed male is 28.0 mm SL, the smallest oviger- PROC. CALIF. ACAD. SCI., 4TH SER., VOL. 41, NO. 2 ous female 30.6 mm; a 51-mm female contained 214 near-ripe eggs, maximum diameter about 0.9 mm. Available data suggest that all collections have been made with dredge or SCUBA over a 3-55 m depth range. Gillellus ornatus is known only from the Gulf of California between Los Frailes and Isla Angel de la Guarda in the west and from a single eastern Gulf collection near Guay- mas, Sonora (Fig. 8). MATERIAL EXAMINED: Ninety-seven specimens, 12-51 mm SL, including neotype. Neotype. —LACM W65-86-1 (37.0 mm SL, male), Mexico, Baja California Sur, Isla Santa Cruz, 22 July 1965, B. W. Walker and party. Other Material. —MEXICO, Gulf of California. —Baja Cali- fornia Sur: CAS 24624 and GCRL 13234 through 13236; LACM 23964 and 23966; SIO 61-247 through 61-249, 65-295, 65-311 and 65-323; UCLA W65-71 and W65-86. —Baja Cali- fornia Norte: LACM 23965, UCLA W53-73 and W53-86. —Sonora: SIO 70-84. Gillellus searcheri Dawson, new species (Figure 12.) DIAGNOSIS: Dorsal finlet 3-spined, without iso- lated spines between finlet and remainder of dorsal fin; dorsal spines total 20-23; pectoral fin rays modally 13; scales in straight lateral line 9- 13; lateral line deflects between verticals from dorsal-fin elements 27 and 32; eye flap present; paired pores in posterior portion of infraorbital canal. DESCRIPTION: Measurements (mm) of 24.5-mm SL transformed male holotype follow: caudal fin length 4.4; length of uppermost segmented ray 3.8; length of lowermost segmented ray 3.6; depth of caudal peduncle 1.5; body depth 2.9 predorsal length 4.3; preanal length 6.7; pelvic to anal distance 2.7; head length 5.3; head breadth 3.2; maxillary to upper opercular angle 4.8; max- illary to upper preopercular angle 3.4; length of anterior naris 0.3; eye diameter 1.1; postorbital length 3.1; snout length 1.0; maxillary length 1.7; length of 1st dorsal spine 1.6, 3rd spine 0.6, 4th spine 1.1, last spine 0.5; length of 1st seg- mented dorsal ray 1.4. See Tables 2-8 for meristic data. Lower jaw but slightly protruding; lower lip with 4-7 papilla-like fimbriae, typically 6 in late juveniles and adults (Fig. 5); preorbital very short, much less than eye diameter, steeply in- clined; tubiform anterior naris located close to anterior rim of preorbital and with a distal valve- like emargination; eye with a distinctive flap lo- lated on posterior third of eye, narrow at base DAWSON: EASTERN PACIFIC SAND STARGAZERS 147 Fig. 12. Gillellus searcheri n. sp. (GCRL 13223, male paratype, 25 mm SL). but distally expanded. Height of dorsal finlet equal to or somewhat less than maximum height of remainder of spinous dorsal; last dorsal spines of continuous dorsal fin much shorter than the following segmented rays, the fin deeply notched or emarginate between (Fig. 13). Pectoral fin- rays 13 in 93% of 306 counts, three fish had bi- lateral counts of 14; pectoral membranes some- what expanded and the fin elongate in trans- formed males but always falls well short of origin of straight lateral line; tip of adpressed_ fin reaches vertical between dorsal-fin elements 14 and 18 (usually between 15 and 18) in trans- formed males, between 12 and 17 (usually 13 and 16) in young and females; caudal-fin formula modally 1 + 8 + 1 in adults (Table 9); opercular fimbriae 4-9 (Fig. 7; Table 10). Infraorbital canal typically with paired pores in lachrymal branch and in three posterior positions (Fig. 1), pores usually total 11 or 12. Archer portion of lateral line deflects between verticals from dorsal-fin elements 27 and 32, usu- ally between 28 and 30; area above arch mostly naked, adults often with 1-6 embedded scales anteriad on either side of dorsal-fin base (Fig. 13); 5-7 (usually 5) longitudinal scale rows cross vertical near origin of straight lateral line; scales present in lateral-line arch of 11-mm fish, squa- mation apparently complete in some 15-mm-SL specimens. Upper jaw teeth in about three rows near symphysis, narrowing to a single row pos- teriad; lower jaw teeth mainly uniserial with but 1-2 paired teeth on either side of symphysis; pseudobranchiae 4-5 (3 examined); proximal pterygiophore of 1st dorsal spine broad (Fig. 3). Color in alcohol. —Head usually without a dark interorbital bar but with a broad suborbital blotch which may include posterior portion of gape and continue on ventral margin of lower jaw; preorbital spotted or streaked with brown; usually with a narrow bar crossing upper lip and tip of lower jaw on either side of symphysis; up- per sides and dorsum of head spotted, blotched or mottled with brown. Early juveniles with about 148 PROC. CALIF. ACAD. SCI., 4TH SER., VOL. 41, NO. 2 ittecccccr rere yd 35x08 eae PII) ERD DELS ID 333 3339939232239 3999 339- 39°? 53279? passceeee ee QRS es (Qo ERPs U8 Fig. 13. Lateral and dorsal views illustrating lateral-line configuration and squamation above lateral-line arch in Gillellus searcheri (top), Sindoscopus australis (middle) and Heterist- ius cinctus (bottom). 6 primary saddle-like bars crossing dorsum from nape to caudal-fin base and secondary bars may be present in 15-mm-SL specimens. In late juven- iles and adults, primary and secondary bars are largely indistinguishable in lateral aspect (Fig. 12) and interspaces are often heavily spotted or flecked with brown. In dorsal aspect, 6 primary and 4-5 secondary bars are indicated for most specimens by darkly pigmented blotches along dorsal-fin base; secondary bars are represented by a single blotch, whereas primary bars are identified by two blotches separated by a short pale interspace; dark blotches along base of con- tinuous dorsal fin usually total 13-14; lower sides pale or with a narrow stripe of brown flecks just above anal-fin base. Dorsal finlet with 2-3 brown spots on 1st spine, elsewhere usually spotted or shaded with brown; continuous dorsal fin with 3-5 well spaced brown spots on most fin-rays, the membranes mainly pale; caudal fin-rays crossed by about 4 irregular vertical rows of spots; upper 5-6 pectoral rays crossed by a brown bar near their base and another near middle of pectoral length; pelvic and anal fins immaculate. Degree of pigmentation is variable but the individual character of the saddle-like bars is masked by flecks and spots in most adults and upper sides appear generally brownish to the naked eye. ETYMOLOGY: The name searcheri refers to the R/V SEARCHER. Collections from this vessel, owned by the Janss Foundation and destroyed by fire in May 1972, have contributed much to our knowledge of Pacific dactyloscopids. COMPARISONS: Gillellus searcheri differs from Pacific congeners in counts of total dorsal spines (20-23 against 11-15 in other species), segmented dorsal rays (16-19 against 25-33), arched lateral- 149 DAWSON: EASTERN PACIFIC SAND STARGAZERS *adA} Asewiig, L Melb [UG LY SN}JIUID $N1}S119}3H ar Al @ vZ sijesjsne sndoosopuls L Lé *«8Z 89 GL €8L layoseas € «vE 9 €r PjOd/Uale PAE GSeetz LZ sn}eulo bo 68C «SZ OE 09 sisuawey}eys JEL BYA fst e LR gy G Z6L SN}DUIDIWAS SNI[P][!!D es) SSN ieee Ts BOSE EO EG™ SNCS OC GE Rel) 5) amar eaten 7] N saizads pue snuay PEs WOG Oceans ce Sv Ste re (Ec, ee lee Or Gis BLN I OD REL vb ih OO CGC sJUBW9]a Ulj-|PSIOp UBaMmjaq adedsuajU] pue sndodrsopuls ‘snjjajji5 yO saivads diy19eg ul S}UsWA|a Ulj-;esIOP Uaamjaq adedssajul MO|9q UO!NDII}J9p aul] jesa}e] JO UOId0] BuNyedipu! suoNnqisysip “SNI}SIa}aH Aduanbasy °g FIGVL 150 line scales (52-62 against 23-33) and _ straight lateral-line scales (9-13 against 18-29). Other spe- cies typically have 4 labial fimbriae in adults, an- terior naris is a simple tube, the eye flap (when present) is located near middle of eye diameter and there are no paired pores in the infraorbital canal. There are usually 6 labial fimbriae in adult searcheri, the anterior naris has a distal valve- like emargination, the eye flap is located over posterior third of eye and there are paired pores in the infraorbital canal. This species appears most closely related to the Atlantic G. greyae Kanazawa which shares the emarginate naris and eye-flap position, has more than 4 labial fimbriae and also has paired infraorbital canal pores. There are, however, fewer scales in the lateral- line arch of greyae (about 28-34) and total lat- eral-line scales are more numerous in searcheri (64-73 versus about 53-58 in greyae). REMARKS: Although Gillellus searcheri (and G. greyae) is obviously distinct from Pacific con- geners in several features, close relationship is indicated by a number of common characters such as general dorsal-fin configuration, number of principal caudal rays, the naked venter, re- striction of labial fimbriae to lower lip, etc. | find no useful purpose in erecting a separate genus or subgenus to accommodate this somewhat ab- errant species (or species group). The paired in- fraorbital pores suggest that searcheri is inter- mediate between the more typical species of Gillellus and Sindoscopus (described below). Specimens from Mexican waters are somewhat more heavily pigmented than Costa Rican and Panamanian material, and there appear to be minor meristic differences in these populations (Table 11). Among material examined, the small- est transformed male is 21.6 mm SL, the smallest ovigerous female 21.5 mm; a 24-mm female con- tained 41 developing eggs in the right gonad. Recorded to a confirmed maximum depth of 15 m, Gillellus searcheri is known from the Islas Tres Marias and Nayarit, Mexico and from insular localities in Costa Rica and Panama (Fig. 8). MATERIAL EXAMINED: Two hundred fifty-six specimens, 11- 31 mm SL, including holotype and one hundred thirty-four paratypes. Holotype. —LACM 32546-45 (24.5 mm SL, male), Costa Rica, Isla del Cano, 08°43’15’’N, 83°53'07’’W, 9 m, SCUBA, R/V SEARCHER Sta. 471, 14 March 1972, W. A. Bussing and party. Paratypes. —COSTA RICA, Isla del Cano: UCR 676-9, 08°43'16’’"N, 83°52'50’’W, 14 m, SCUBA, R/V SEARCHER Sta. 474. LACM 32547-49 and GCRL 13217, 08°43'15’'N, PROG] GALIEVAGADSSGI>) 4 SER) VOL AlNOe2 83°53'07''W, 15 m, SCUBA, R/V SEARCHER Sta. 472. LACM 32546-46, GCRL 13222 and 13223, taken with holotype. Other Material. —MEXICO, Islas Tres Marias: GCRL 13219 through 13221; SIO 62-25, 62-56 and 62-58. Nayarit: GCRL 13218, SIO 62-49 and UCLA W58-5. PANAMA, Islas Secas: SIO 70-139. —Isla Saboga: SIO 67- 40. Sindoscopus Dawson new genus TYPE-SPECIES: Gillellus australis Fowler and Bean, 1923. DIAGNOSIS: Dorsal-fin origin on nape, the fin continuous but anterior membranes often deeply incised; eye not stalked and without minute flaps or papillae; pectoral fin-rays modally 14; principal segmented caudai rays 10, some branched in late juveniles and adults; well de- veloped fimbriae on both upper and lower lips; head, pectoral-fin base and venter naked; prin- cipal preopercular canals 4, unbranched; some paired pores in infraorbital canal; without pre- dorsal bones (interneurals). DESCRIPTION: The genus is monotypic; see de- scription of australis below. ETYMOLOGY: sindo plus scopus; in recognition of my Chilean field associate, Sr. Gumersindo Revuelta, who collected some of the present specimens and was instrumental in making all of the recent collections available for study; scopus - watcher, in reference to stargazers. COMPARISONS: The continuous dorsal fin is shared with Dactylagnus, Myxodagnus and some species of Platygillellus (Table 1). Dorsal-fin ori- gin is near vertical from anal-fin origin in Dacty- lagnus and Myxodagnus (on nape in Sindoscopus) and Platygillellus has 11 rather than 10 principal caudal rays. Closer relationship to Gillellus seems indicated by similarities in body configuration, caudal fin morphology, orientation and extent of straight lateral-line scales, and paired infraorbital pores (searcheri group only). Sindoscopus is dis- tinguished by the continuous rather than inter- rupted dorsal fin, by development of both upper and lower lip fimbriae (on lower lip only in Gillellus) and by the presence of 4 rather than 3 primary preopercular canals. Sindoscopus australis (Fowler and Bean) (Figure 14.) Gillellus australis Fowler and Bean, 1923: 23 (original de- scription; Valparaiso, Chile). DAWSON: EASTERN PACIFIC SAND STARGAZERS 151 Fig. 14. Sindoscopus australis (Fowler and Bean). Top: GCRL 13399 (64 mm SL, male) from Algarrobo, Chile. Middle and bottom: GCRL 13404 (71 mm SL, transformed male) from Antofagasta, Chile. 52 AeA re ee H. cinctus S. australis W < fea a 2 ve a a or WJ a o = 0 20 40 60 80 100 STANDARD LENGTH (MM) Fikes 5) DIAGNOSIS: Diagnostic characters are those of the genus. DESCRIPTION: Measurements (mm) of 75.8-mm- SL male (GCRL 13399) follow: caudal fin length 10.7; length of uppermost segmented ray 10.3; length of lowermost segmented ray 9.4; depth of caudal peduncle 3.6; body depth 8.3; predorsal length 13.2; preanal length 19.9; pelvic to anal distance 9.0; head length 14.0; head breadth 6.8; maxillary to upper opercular angle 12.8; maxil- lary to upper preopercular angle 8.9; eye dia- meter 2.0; postorbital length 8.7; snout length 1.8; maxillary length 3.9; length of 1st dorsal spine 1.6, 2nd spine 1.6, 3rd spine 1.5, 4th spine 1.6, last dorsal spine 2.3; length of 1st segmented dorsal ray 2.7. See Tables 2-8 for meristic data. Body rather slender, broader and deeper in front, more or less oval in cross section and gradually narrowing caudad; head flattened to slightly rounded dorsad, broadest near posterior margin of preopercle, snout not strongly in- clined; head, in dorsal aspect, somewhat tri- angular anteriad, distinctly narrowed in front (Fig. 5); lower jaw moderately protruding, rounded in dorsal profile, not continued anteriad PROC. CALIF. ACAD. SCI., 4TH SER., VOL. 41, NO. 2 40 30 LOWER LIP FIMBRIAE DY) e) O 20 40 60 80 100 STANDARD LENGTH (MM) Regressions of frequencies of labial fimbriae in Heteristius cinctus and Sindo- copus australis. See Table 10 for supporting data. as a conical, fleshy projection; upper and lower lips with well developed simple fleshy fimbriae, most numerous below, frequencies ontogenetic (Fig. 15; Table 10); preorbital length about 2/3 eye diameter; eye dorsolateral, protruding but not on long protrusile stalk, without minute flaps or papillae; anterior naris a very short tube with distal valve-like emargination, located about midway between eye and anterior rim of pre- orbital; posterior naris a simple pore with slightly raised rim, located behind anterior naris and ad- jacent to anterior rim of orbit; principal preoper- cular canals 4 (Fig. 1), not branched, the 1st (an- terior) slightly shorter than remainder and usually with a single pore distad, occasionally with 1-3 smaller pores between this and base of canal; infraorbital canal typically with single pores an- teriad, single or paired pores in lachrymal branch, and paired (occasionally 3) pores in three posterior positions; opercle with a short fringe of up to 10 unbranched fimbriae overlying upper posterior margin, frequency ontogenetic (Fig. 7; Table 10); opercles broadly expanded, the membranous margins cover pectoral- and pelvic-fin bases and overlap on underside of head, DAWSON: EASTERN PACIFIC SAND STARGAZERS Dorsal-fin origin on nape, in advance of upper opercular angle (Fig. 13); dorsal fin continuous, anterior membranes often deeply incised; an- terior 4-5 spines short and rather flexible, 1st and 2nd more closely spaced than 2nd and 3rd; last dorsal spine somewhat shorter than 1st seg- mented ray, the fin somewhat notched or emar- ginate between; anal spines 2, about equal in length and a third shorter than 1st segmented ray; anterior anal-fin elements modified in trans- formed males, spines angled caudad, 1st seg- mented ray angled forward (about 45°), the 2nd ray more nearly vertical, modified elements somewhat swollen or fleshy and membranes are stretched or elongate between. Posteriormost dorsal and anal rays usually free from peduncle, occasionally bound by short membranes; last anal ray reaches past rear of hypural, last dorsal ray somewhat shorter; caudal peduncle entire, without notch in ventral margin (Fig. 2); caudal fin rounded, upper and lowermost rays typically simple, subequal in length or lowermost the shorter; 6-9 median rays branched in study mate- rial, modal formula 1 + 8 + 1 (83% of 84 speci- mens examined). Pectoral fin-rays 14 in 80% of 185 counts, 11 fish with bilateral counts of 13; pectoral fin rounded to somewhat pointed, 5th-7th rays from dorsalmost the longest, rays gradually shorten ventrad but not appreciably swollen or thick- ened; pectoral fin somewhat elongate in trans- formed males but usually falls short of origin of straight lateral line, the interradial membranes broad and fin cupped or outwardly convex; tip of straightened and adpressed fin reaches vertical between dorsal-fin elements 15 and 20 (usually 16 and 19) in transformed males, between 14 and 17 in young and females; pelvic fin I, 3; tips of segmented rays free with emarginate mem- branes between, longest ray seldom reaches anal- fin origin; all fin rays simple, except for branched caudal rays in late juveniles and adults. Lateral line continuous, originates above and in advance of upper opercular angle (on or in advance of vertical from 1st dorsal spine), arched upward to follow near dorsal-fin base, deflects between verticals from dorsal-fin elements 14 and 18, and continues midlaterally to terminate as penultimate scale on caudal-fin base; canal of straight lateral-line scales parallels longitudinal body axis, last scale not angled ventrad (Fig. 2). Scales thin, more or less rounded and persistent in upper portion of lateral-line arch; scales of descending arch and straight lateral line some- 153 what deciduous; canal of straight lateral-line scales without terminal branch anteriad, often with a short dorsal branch and terminal pore on anterior third of canal. Squamation variable above lateral-line arch, more or less fully scaled or with patches of embedded scales and naked areas between (Fig. 13); usually with small em- bedded scales beginning above 2nd or 3rd later- al-line scale about three rows between arch and base of 4th or 5th dorsal spine and with two crowded rows of larger scales above posterior half of elevated arch; head, pectoral-fin base, venter and sides (below line from upper pectoral angle to near 1st segmented anal ray) naked; 11- 13 longitudinal scale rows cross vertical near origin of straight lateral line. Premaxillaries somewhat protractile, the pedi- cels barely reach past middle of orbit; jaw teeth small, conical, somewhat curved inward, none distinctly enlarged; lower jaw teeth in two rows near symphysis, uniserial behind; upper teeth in three rows anteriad, two rows near middle of jaw and in single row behind; vomer and pala- tines edentate; tongue subtriangular, narrowly rounded in front; basihyal slender, barely spatu- late at tip. Three complete gill arches plus a hemibranch, a pore in membrane behind lower portion of hemibranch; no gill rakers on outer margin of 1st arch, 8-10 short rakers along inner margin; pseudobranchiae 4-6 (4 examined); branchiostegals 6; without predorsal bones; proximal pterygiophore of 1st dorsal spine not exceptionally broad, inserted slightly behind 1st neural spine (Fig. 3); abdominal vertebrae 10-11 (usually 11) in radiographs; caudal fin with 5 spiniform elements above and 3 below, upper and lower hypurals each bear 5 segmented rays (in 2 cleared specimens). Color in alcohol. —Ground color light tan to pale, markings tan to dark brown. Well marked fish with a dark interorbital bar divided below eye to form a short bar across posterior third of gape and a long narrow bar extending postero- ventrad to slightly beyond middle of preopercle; underside of lower jaw with a dark-edged blotch on either side of symphysis; upper sides and dorsum of head variously spotted or faintly blotched, remainder of head pale. Dorsum crossed by 9-10 (usually 10) dark-edged blotches or bars between nape and caudal-fin base; the anterior bar, variously solid or interrupted so as to form two bars with a pale area between (Fig. 14), extends ventrad to upper margin of opercle; remaining bars usually fall well short of lateral 154 midline, seldom extending more than 3-4 scale rows below dorsal-fin base; usually without dis- tinct secondary bars, pale interspaces and area between ventral margins of bars and midline ir- regularly streaked or flecked with brown; lower part of side pale in some fish, others with about 10 small blotches more or less equally spaced along Ist scale row above anal-fin base. Dorsal fin with pale membranes and 1-3 small spots on most fin rays; anal fin mainly pale, occasionally with a few isolated spots or brown shading on some rays and posterior membranes; pelvic fins pale; dorsalmost pectoral rays sometimes faintly spotted, the fin otherwise pale; caudal fin pale near base but crossed distad by about 10 vertical rows of small spots. REMARKS: Fowler and Bean (1923) described australis from two specimens collected by the Wilkes Exploring Expedition at Valparaiso, Chile. There have apparently been no subsequent rec- ords of this species and Myers and Wade (1946) suspected that there may have been an error in type locality. Present material, some of which was collected within 40 km of Valparaiso, clari- fies the matter and firmly establishes australis as the southernmost dactyloscopid. The holotype and paratype are now damaged, in very poor condition, and it is not possible to obtain mean- ingful measurements. The holotype was origin- ally described as having 26 dorsal spines and 21 segmented rays, 38 segmented anal rays and 15 pectoral rays; the lips were described as “en- tire.” | counted 21 spines and 25 segmented dor- sal rays, 39 segmented anal rays, and 14 rays in each pectoral fin. In addition, there are about 11 fimbriae on the upper lip and at least 6 on the undamaged right side of the lower lip. Fowler and Bean (op. cit.) described the 6 anterior dorsal spines as separate but this is understandable as the anterior membranes are often broken (as they are now in the holotype) and, even in un- damaged specimens, membranes are low and may be partly concealed in the somewhat de-’ pressed area along anterior portion of dorsal-fin base. The bulk of recent collections have come from the vicinity of Valparaiso and from Antofagasta, some 1100 km to the north. The southern mate- rial averages somewhat higher in a number of meristic characters (Table 11), but | find no other substantial differences and consider such varia-, tion to be clinal. The upper, lower or both of the posteriormost spiniform caudal elements were segmented in all Algarrobo specimens with the PROC. CALIF. ACAD. SCI., 4TH SER., VOL. 41, NO. 2 modal caudal formula of 1 + 8 + 1, and in 31% of the Antofagasta material. Segmentation of these typically simple elements also occurs infre- quently in Gille/lus but it has not been noted in other Pacific dactyloscopids. Among material examined, the smallest trans- formed male is 64 mm SL, the smallest ovigerous female 50 mm; a 74-mm female contained 419 well developed eggs in the left gonad, maximum diameter about 1.1 mm. An isopod, abundant in the stomachs of most Algarrobo specimens, has been identified as /socladus sp. by Dr. Peter Glynn, Smithsonian Tropical Research Institute. DISTRIBUTION: Most material has been seined in depths of 2 m or less; surface temperature range for 3 collections was 13-14°C. Sindoscopus australis is known only from coastal waters of Chile between about 23°34’S and 33°02’S. MATERIAL EXAMINED: Eighty-six specimens, 32-92 mm SE including holotype and paratype. Holotype. —USNM 83315 (ca. 76 mm SL), Chile, Valpar- aiso, Wilkes Expdn., no other data. Paratype. —USNM 83099 (60.7 mm SL), data as for holo- type. Other Material. —CHILE, Antofagasta: ANSP 131422, GCRL 13397 and 13400 through 13404, MNHN uncat. and USNM 214226. —Coquimbo: USNM 176588. -Valparaiso: GCRL 13398, 13399 and MNHN uncat. -Loc. uncertain; MNHN uncat. Genus Heteristius Myers and Wade Heteristius Myers and Wade, 1946: 160 (type-species by orig- inal designation, Heteristius jalisconis Myers and Wade, 1946 [ = Dactyloscopus cinctus Osburn and Nichols]). DIAGNOSIS: With 2-spined dorsal finlet origi- nating On nape and with 1 or 2 isolated spines between finlet and remainder of dorsal fin; eyes not stalked, without flaps or papillae; pectoral fin-rays modally 14; principal segmented caudal rays 12, unbranched; fimbriae present on both upper and lower lips; venter scaled in late juven- iles and adults, head and pectoral-fin base naked; principal preopercular canals more than 6, un- branched; some paired pores in_ infraorbital canal; without predorsal bones (interneurals). DESCRIPTION: The genus is monotypic, see de- scription of cinctus below. COMPARISONS: Heteristius is a specialized genus, differing from other sand stargazers in caudal-fin morphology and in configuration of the preopercular canal system (Table 1). Principal segmented caudal rays are 12 and unbranched in Heteristius, whereas there are 10 or 11 in DAWSON: EASTERN PACIFIC SAND STARGAZERS Fig. 16. Heteristius cinctus (Osburn and Nichols). SIO 62-720 (31 mm SL). other genera and (except for a specialized spe- cies-group within Dactyloscopus) there is onto- genetic branching of the median rays. There are either 3 or 4 simple or distally branched princi- pal preopercular canals in other genera, but these are replaced in Heteristius by about 7-9 simple canals on posteroventral portion of preopercle (Fig. 1). The scaled venter is shared only by Platygillellus, but scales are here confined to a narrow transverse band (venter completely scaled in Heteristius) and the pectoral-fin base is scaled in adults (naked in Heteristius). REMARKS: Osburn and Nichols (1916) described Dactyloscopus cinctus from ALBATROSS collec- tions at Baja California Sur. Their description and accompanying figure indicate a continuous dor- sal fin and there are other errors, but the venter is clearly scaled and characteristic bars are shown on the side of the illustrated specimen. Myers and Wade (1946) described jalisconis (type-spe- cies of Heteristius) from VELERO collections off Jalisco and Oaxaca, Mexico. Although these au- thors briefly reviewed described dactyloscopids and made generic comparisons in their descrip- tion of Heteristius, they apparently overlooked Osburn and Nichols’ publication. | have exam- ined the type-material of these nominal species and find them to be conspecific. Heteristius cinctus (Osburn and Nichols) (Figure 16.) Dactyloscopus cinctus Osburn and Nichols, 1916: 176 (origi- nal description; Cape San Lucas, Mexico). Heteristius jalisconis Myers and Wade, 1946: 160 (original description; Jalisco, Mexico). DIAGNOSIS: Diagnostic characters are those of the genus. DESCRIPTION: Measurements (mm) of 27.9-mm- SL lectotype follow: depth of caudal peduncle 1.5; body depth 4.1; predorsal length 5.7; pre- anal length 8.6; pelvic to anal distance 3.4; head length 6.1; head breadth 3.2; maxillary to upper opercular angle 5.7; maxillary to upper preoper- cular angle 3.9; eye diameter 1.5; postorbital length 3.4; preorbital length 1.0; maxillary length 2.2. See Tables 2-8 for meristic data. Body broad and deep in front, compressed and slender caudad; dorsum of head somewhat rounded, lateral profile a descending arch from nape to tip of lower jaw; maximum head breadth near posterior margin of preopercle, not strongly 156 narrowed in front, lower jaw slightly protruding and rounded in dorsal profile (Fig. 5); lips with well developed simple fimbriae, most numerous below, frequency ontogenetic, at least on lower lip (Fig. 15; Table 10); eye dorsolateral, protrud- ing but not on long protrusile stalk and without flaps or papillae; anterior naris a short tube with distal valve-like emargination, located close be- hind anterior rim of preorbital; posterior naris a pore with slightly elevated margin, located be- hind anterior naris and adjacent to anterior rim of orbit; preopercle with a closely set series of about 12-14 canals, usually unbranched, each with a distal pore and often with a few smaller pores between this and canal base (Fig. 1); infra- orbital canal typically with single pores in pre- orbital branch and with two or more pores in lachrymal branch and at most posterior posi- tions; opercle with a fringe of up to 12. slender unbranched fimbriae overlying upper posterior margin, frequency ontogenetic (Fig. 7; Table 10); opercles broadly expanded, membranous mar- gins cover pectoral- and pelvic-fin bases and overlap on underside of head. Dorsal-fin origin on nape, in advance of upper opercular angle (Fig. 13); dorsal fin with a 2- spined anterior finlet, usually with one (42%) or two (56%) isolated spines between finlet and continuous dorsal fin; height of spinous portion of continuous dorsal fin nearly thrice that of finlet and about 25% less than longest seg- mented dorsal ray, the fin notched or emargin- ate between spines and segmented rays. Posteri- ormost dorsal and anal rays usually free, some- times bound to peduncle by fragile membranes; last anal ray reaches beyond rear of hypural, last dorsal ray somewhat shorter, caudal peduncle entire, without notch in ventral margin (Fig. 2); caudal fin rounded; segmented caudal-fin rays 12, always simple, the upper and lowermost shorter than median rays. Pectoral fin-rays 14 in 79% of 208 counts, 6 fish had 13 rays in each fin, whereas 10 had bi- lateral counts of 15; pectoral fin somewhat pointed, 4th-6th rays from dorsalmost the long- est, rays gradually shorter but not enlarged ven- trad; tip of straightened and adpressed fin reaches well beyond deflection of lateral-line arch, reaches vertical between dorsal-fin ele- ments 14 and 18, usually between 14 and 16; pelvic fin |, 3, tips of segmented rays free with emarginate membranes between, longest ray usually falls short of anus; all fin-rays simple in examined material. PROC. CALIF; ACAD. SCI. 4TH SER., VOL: 41;;NOme Lateral line continuous, originates above and in advance of upper opercular angle (on or just before vertical from 1st dorsal spine), arches up- ward to follow close to dorsal-fin base, deflects between verticals from dorsal spines 8-12 and continues midlaterally to terminate as the last scale overlying caudal-fin base (Fig. 2); canals of straight lateral-line scales generally parallel to longitudinal body axis, canal of last scale may be angled slightly ventrad; scales of upper part of lateral-line arch persistent, posterior margins somewhat trilobate; scales of descending arch and straight lateral line somewhat deciduous; canal of straight lateral-line scales without a ter- minal branch anteriad, usually with a dorsal pore near middle or anterior third of canal. Area above lateral-line arch completely scaled in late juveniles and adults (Fig. 13), usually with three scales between base of finlet and 1st lateral-line scale and one scale between apex of arch and dorsal-fin base; venter, between anus and pelvic- fin base, and sides (except for pectoral axil) com- pletely scaled; head, pectoral-fin base and all fins, except for caudal base, naked; 9 longitudi- nal scale rows cross vertical near origin of straight lateral line. Premaxillaries somewhat protractile, pedicels fall just short of rear margin of orbit; jaw teeth coni- cal, somewhat curved inward, about three rows near symphyses but in single row behind; vomer and palatines edentate; tongue narrowly round- ed, basihyal spatulate; three complete gill arches plus a hemibranch with pore behind; no rakers on 1st gill arch; pseudobranchiae 7-9 (3 exam- ined); branchiostegals 6; without predorsal bones; proximal pterygiophore of 1st dorsal spine broad, inserted in advance of 1st neural spine (Fig. 3); abdominal vertebrae 10 in radio- graphs; caudal fin with 3 spiniform elements above and below, upper and lower hypurals each bear 6 segmented rays (in 3 cleared specimens). Color in alcohol. —Ground color is light tan to pale, markings tan to dark brown. Dorsum of head capped by a mottled brownish blotch be- hind eyes followed by a narrow pale bar cross- ing between upper margins of preopercles; short streaks of brown radiate from eye and cross up- per and lower jaw near symphysis; underside of lower jaw often with two small blotches anteriad; remainder of head mainly pale. From above, dor- sum crossed by about 12 brown bars from nape to caudal base, bar length irregular, about equal to pale interspaces. In lateral aspect, sides crossed by three prominent bars and there is an elongate DAWSON: EASTERN PACIFIC SAND STARGAZERS blotch on caudal base; these markings persist in long-preserved and otherwise faded material. The anterior bar is located near origin of continuous dorsal fin and extends slightly below middle of side; 2nd bar (below 15th-19th dorsal spines) and 3rd bar (below 9th-12th segmented rays) ex- tend ventrad to anal-fin base. Between these characteristic markings, bars crossing dorsum ter- minate on 1st or 2nd scale below dorsal-fin base and there are short stripe-like markings on mid- line and just above anal-fin base; pale areas else- where often irregularly flecked with brown. Dor- sal finlet mainly brown, tip of 1st spine pale; remainder of dorsal fin usually pale, occasionally with some brown shading on basal portions of rays located over dark bars; middle of caudal fin crossed by faint band of micromelanophores, other fins mainly pale. REMARKS: The short and partly inaccurate de- scription of Dactyloscopus cinctus is based on three specimens, ‘‘about the same size,”’ collected 23 March 1911. Neither figure nor description is specifically referred to the holotype which is numbered AMNH 5206. There are no specimens now shelved under this number at the American Museum of Natural History, but there is a lot of three fish (AMNH 5394) with an external jar label bearing the pencil notation, ‘‘Type and paratype.” The jar contains a pencil note stating, ‘‘Cape San Lucas Mch. 23/11’’ and a printed ALBATROSS label penciled, ‘““Dactyloscopus cinctus N............ INNAGGM Zoe ce eee oxen. Cape San Lucas.’ | am advised by J. E. Darovec, Jr. (AMNH) that catalog cards for AMNH 5206 and 5394 show the same data, and it appears that these numbers were intended to accommodate the three specimens mentioned by Osburn and Nichols (1916). It is not known whether these authors actually separated the selected holotype from the two remaining fish or whether the holotype was subsequently re- united with AMNH 5394, but there is little doubt that the three original specimens are now in- cluded in this single lot. In any event, it is now impossible to identify the originally intended holotype and | treat these fish as syntypes. | select the best preserved specimen (27.9 mm SL) as the lectotype of Dactyloscopus cinctus. This fish (AMNH 5206) has about 27 fimbriae on the lower lip, about 21 scales in lateral-line arch, scaled venter and the characteristic body bars are distinct on both sides (for additional data, see measurements above and Tables 2, 3 and 7). The two remaining specimens (29 mm SL) are paralectotypes (AMNH 5394). 1!57/ Labial fimbriae are distinct and characteristic body bars are present at 14 mm SL; squamation is apparently complete in some 16-mm fish. No specimens were found with modified anterior anal-fin rays or enlarged pectoral fins that char- acterize sexually mature (transformed) males of other sand stargazers. Examination of radio- graphs showed no evidence of the broadened anterior anal pterygiophores commonly associ- ated with the externally modified anal rays. A number of ovigerous females were present among examined material and it seems unlikely that mature males were not included. No further efforts were made toward sex determination, but it is highly probable that anal rays and pectoral fins are not modified in Heteristius cinctus. Ab- sence of such modifications would suggest that adult males do not carry incubating egg clusters beneath the pectoral fins as reported for other dactyloscopids (BOhlke and Chaplin, 1968; Petti, 1969). The complete squamation of venter and adjacent sides, naked or largely so in other dac- tyloscopids, would seem to support this assump- tion. Among material examined, the smallest ovigerous female is 22 mm SL; a 32-mm fish con- tained 171 developing eggs in the right gonad. | have collected specimens at depths of less than a meter, others have been taken with SCUBA in 15 m and maximum recorded depth is from a dredge sample in 27.4m. There are no collections from low salinity or estuarine habitats. Heteristius cinctus is known from Bahia Magdal- ena to the vicinity of Punta Arena on the Baja California peninsula. It is apparently absent from the Gulf of California but reappears in collections from a number of insular and mainland shores from Nayarit, Mexico south to Cabo San Fran- cisco, Ecuador (Fig. 8). MATERIAL EXAMINED: Three hundred fifty-eight specimens, 14-40 mm SL, including lectotype and two paralectotypes. Lectotype. —AMNH 5206 (27.9 mm SL), Mexico, Baja Cali- fornia Sur, Cape San Lucas, seine, 28 March 1911, Town- send ALBATROSS Expdn. Paralectotypes. —AMNH 5394 (2, 29 mm SL), taken with lectotype. Other Material. —MEXICO, Baja California Sur: CAS 32041 and LACM 31174-46; SIO 61-248, 62-706, 62-720, 64-42 and 64-43; UCLA W59-248. —Nayarit: GCRL 8500 and 13363; UCLA W58-5 and W58-8. —Islas Tres Marias: GCRL.13341, 13362 and SIO 62-25. —Jalisco: LACM 23185 (holotype of H. jalisconis). —Oaxaca: LACM 23186 and CAS-SU 39961 (paratypes of H. jalisconis). COSTA RICA, Puntarenas: GCRL 11813. —Isla del Cano: GCRL 13360 and 13361; LACM 32546-47, 32549-59 and UCR 674-5. PANAMA, Isla Coiba: SIO 71-211 and 71-213. ECUADOR: USNM 116221. PROC. CALIF. ACAD. SCI., 4TH SER., VOL. 41, NO. 2 158 TT ELE 8E-97 LY SLC Ov-6L £6 SN}IUID “YH Lig Sie €C-6 SiG: OLZ C6-CE 08 s1je4jsne °s eelquiy dij JaMO7 II > N te) Cy o XZELO+bSS' ZZ XOOL'0+€89'8 lI > w N ae =) XEEO'O-9ES EL = A L0i0m 6'L 9°8L €7-bL (a4 6°27 Ob-6L c6 SNJIUID *H xe9o00+sre9 =A 6b+'0 OL g°OL S1-9 SLL GL 76-ZE tL sipesysne “Ss eeuquiy dij saddq x060'0+7879 =A Z6€°0 OL £8 TL-S ob L°LZ OF-9L ELZ SN}DUID “H x6co'O+zes'p = A 69€'0 OL G9 OL-+ LL 6°69 T6-ZE BLL sijeszsne °S X{LLO+6r6'€ = A 89b'0 670 9°9 6-+ 6€ bE vE-LL 6LE Hayoseas “5 xoso'0+L99'¢ = A 609°0 lal: 9 g-€ £8 6'EE LS-8L 62 B/OD/Uase *D xe60'0+SLO'E = A Z8r'°0 Cal: 09 OL-Z 9°9 6'7TE LS-ZL bel SnyeUsO *D XLOL'O+O0Z@O'r = A 98€°0 80 £9 8-b Exe LGC TE-SL €€L SISUBWEYJEYD “FD XpLLO+zZe6e = A 7950 cal OL TL-¥ 99 vTE 6b-FL 8LZ SN}JIUIDIWAS *D aeluquiiy seynd1adO uoljenba uolssaisay 4 as x asuey as x asuey N sainads Jayoeieyy y}3ua| puepueys dayoeieyy *(UON}B]94109 JO 4UII91}j209 = 4 ‘UONRIAaP psepue}s = GS ‘ueaw = x) SNNSI9}9H Pue sndorsopuls ‘snjjajj!D 40 saidads djl9eg Ul aeluquiy dij Jamo] Pue saddn pue aeliquiy Jejno1ado yo AdUAaNbas YIM Y}BUa] puepue}s Jo UOSIUedWOD ‘OL F19VL *aGA} AleWiid, L’Sc ve-ZL «88 LiGYe NOSERG. WL OSE Ov-8c 69 VCC L OSE 8reSG BL ae ab 6CC = OE-6L ZL 0'6€ Ch-9E c LLE LS-9€ v EZ CE-bC s (As CW AVErE €€ 6 ars ae fh L672 L © a9) seal LC L (G4 L Pe OR SEC 9d-9L LL vSE vrre + OSES IGE-8G sie OIC IS Cee GG Mees, (SKEERG yl 6 ar) SS 6 EAL L CLE OSeZG v 6VC 8C-EC «8 Wels ere) or Gard L Ly ras iP oe 1? ar GES CLIAG i £6 L Goate (Sh aes hee PASRS 6 Lye MENTE UE QUZE L Sos ate 7€ SVL L LCV L Bi9G L (Aoi (Seam {2 L'97 L OSC It Vesta ate ab: v6C L Pete ate Si Tigah L Pie: Ie 9 Srl LL-EL c 6 EC CE-6L € ECE 9C-0C S VLC c-8L OL 8'6L CC-9L S ajdwis |v x asuey N x aduey N x aduey N x asuey N x asuey N P/NwWo} l4ayoseas Pjod/uase snjeulo sisuawey}eyd snjouioiwas Aes-Ul4 “ww ul (xX) URaW pue aduei yisua|-psepuejys ‘suawioads yo saquinu = Wy ‘snjjajji9n yo saidads a1y1deq Ul SAB jepNed pajyuawsdas jedidulid jo BuiydueJq jo yUaWdO|aAap DaUaZO}UQ “6 JIGWL 159 6°99 02-9 GL (9) [Le VE-6C GIE ClO Cran G6E Lv-6€ NE, VAY a OVEI VG oqoues|y 8°C9 SINS, (OS) vOE CE-8C OS L6E Lv-6€ L9 (Be (EAE eS) (OO APS AG eysedejo}uYy sijesjsne “Ss 0°89 LAA) (OE 8°6 CL-6 Of OLE €€-O€ v6 OLE CE-OE Z8 Z8E OV-BE 1g euweued - BdIY PSO) 8°29 €Z-9 €9 OLL €L-6 €9 IiGs VE-LE vOL G |bre CE-OE 80L O'6E Ly-8€ 60L O3IXoW 149 YIICIS “DY Gly b-Ly 7 = ll G Gle “eles 7 cee Bee-0ce 18 S'6€ Ob-6e 8 's] sodedejey O'8r 1S-9F LL Al vest (HI, EGE GE-Lew Ze Ciles BSE0E. 90 LOE Le-BE BZ eIQwojod - ed1y e}SOD Sly 6b-9b + Galen eCG- Co EGG MAvE-Gln Ah: Jatt, Asie II S6E€ Lr6e LL BDIeXeO - WIeAeN 68b CS-bb 8S €6L GGaail 8S 6 EE SREGe CCL OEE Siac 6rL LLY €b-6€ 6rL JB}OL C8 LS-vY LS Z6L €C-8L LS 6 CE SiaGe 80L OEE Site Ete SEL LLY €v-6€ SEL PlUsOJIJeD JO J[ND O06b OS-8h € 0°0¢ Lc-6L fe INES Sire texe 8 OEE WAES(Gs, 83 CLV cy-0Ob §=68 sean] ues ade) c 0S CS-8h OV O'LC tEGAUG v SEE Wee, ©) BCE Vice Comm Olv cr-Ob 9 JSPOD INfled BIusOyIeD eleg 9°0S €S-8P LL Z6L Lc-6L LL L ve Sitestete SL OEE Seare sith Glliv GF tote SL adnjepend e|s} came 8V L = LC L ae GS L Srllts, tebe v toate Ov-6€ © SOpasise||IAdY Se|s| SNJIUIDIWAS *) asuey N x oduey N x asuey N xs asuey N xi aduey N aje00] pue saiseds sajeos oul] sajeos aul} 9eIGI}JIA sAei jeue s}udWa|a -|esa}e| JRIOL -|e19}e] JYS1eI}S jepne pa}uawsas Ulj-]eSIOP [BJO] DAWSON: EASTERN PACIFIC SAND STARGAZERS “sijesjsne sndoosopuls pue saydieVas *D ‘snjIUIDIWaS SNjjajj!D JO Sia}eIeYD INjSisawW pPaydajas Ul UOHeUeA D1YydessoayH “LL IIAVL 160 ACKNOWLEDGMENTS Appreciation is expressed to curators of the various repositories for the loan of specimens and other courtesies. Special thanks are due W. A. Bussing (UCR), R. J. Lavenberg and C. Swift (LACM), R. H. Rosenblatt (SIO) and B. W. Walker (UCLA) for making their collections avail- able for study. Miss Pearl Sonoda (CAS) provided considerable help with problems concerning CAS-SU collections. For permission to work in Chile and for assistance during my visit, | thank N. Bahamonde, F. R. Manns, G. Pequeno-R., G. Revuelta and J. Tomicic. Drs. J. E. Bohlke (ANSP) and V. G. Springer (USNM) provided useful com- ment and suggestions on portions of this work. C. L. Smith (AMNH) conducted the search for data on the type of Dactyloscopus cinctus. Drawings are by Anne Langenfeld; computer services were provided by Paul Poole and Mrs. Lois Coquet; F. N. Jackson is thanked for general laboratory assistance and radiographs. Special acknowledgment is due Mrs. Betty Heal for ex- pert secretarial assistance and for her interest in this study and others now in progress. This work was supported in part by National Science Foun- dation Grant GB-31053X. LITERATURE CITED Bohlke, J. E. 1968. The descriptions of three new sand star- gazers (Dactyloscopidae) from the tropical west Atlantic. Notulae Naturae 414: 1-16. Bohlke, J. E. and C. C. G. Chaplin. 1968. Fishes of the Bahamas and adjacent tropical waters. Livingston Publ. Co., Wynnewood, Pa. 771 pp. Dawson, C. E. 1969. A new eastern Pacific sand stargazer (Pisces: Dactyloscopidae). Copeia 1969(1): 44-51. PROG. GALIF; AGAD. SGI 47H SER VOL. 41 NOme 1974. Studies on eastern Pacific sand stargazers (Pisces: Dactyloscopidae). 1. Platygillellus new genus, with descriptions of new species. Copeia 1974(1): 39-55. 1975. Studies on eastern Pacific sand stargazers (Pisces: Dactyloscopidae). 2. Genus Dactyloscopus, with description of new species and subspecies. Natural His- tory Museum, Los Angeles County, Science Bulletin 22: 1-61. . 1976. Studies on eastern Pacific sand stargazers. 3. Dactylagnus and Myxodagnus, with description of a new species and subspecies. Copeia 1976(1): 13-43. Fowler, H. W. and B. A. Bean. 1923. Descriptions of eigh- teen new species of fishes from the Wilkes Exploring Expedition, preserved in the United States National Museum. Proceedings of the United States National Museum 63: 1-27. Gilbert, C. H. 1890. Scientific results of explorations by the U.S. Fish Commission Steamer ALBATROSS. No. XII. —A preliminary report on the fishes collected by the steamer ALBATROSS on the Pacific coast of North America during the year 1889, with descriptions of twelve new genera and ninety-two new species. Pro- ceedings of the United States National Museum 13: 49-126. . 1892. Scientific results of explorations by the U.S. Fish Commission Steamer ALBATROSS. No. XXII. —De- scriptions of thirty-four new species of fishes collected in 1888 and 1889, principally among the Santa Barbara Islands and in the Gulf of California. Proceedings of the United States National Museum 14: 539-566. Jordan, D. S. 1897. Notes on fishes, little known or new to science. Proceedings of the California Academy of Sciences, ser. 2, 6: 201-244. Myers, G. S. and C. B. Wade. 1946. New fishes of the famil- ies Dactyloscopidae, Microdesmidae, and Antennariidae from the west coast of Mexico and the Galapagos Is- lands. Allen Hancock Pacific Expeditions 9: 151-178. Osburn, R. C. and J. T. Nichols. 1916. Shore fishes collected by the ‘Albatross’ Expedition in Lower California with descriptions of new species. Bulletin of the American Museum of Natural History 35(16): 139-181. Petti, J. C. 1969. Behavioral and morphological adaptations to burrowing of two species of dactyloscopid fishes from the northern Gulf of California. Unpbl. Master’s thesis, University of Arizona, Tucson. OF THE zi PROCEEDINGS LIBRANN sa CALIFORNIA ACADEMY OF SCIENCES FOURTH SERIES Vol. XLI, No. 3, pp. 161-168, 5 figs., 1 table. April 15, 1977 THREE NEW SPECIES OF INDO-PACIFIC MORAY EELS (PISCES: MURAENIDAE) John E. McCosker Steinhart Aquarium, California Academy of Sciences, Golden Gate Park, San Francisco, California 94118 John E. Randall Bernice P. Bishop Museum, Honolulu, Hawaii 96818 ABSTRACT: Three new species of Indo-Pacific muraenid eels are described and illustrated: Gymnothorax breedeni n.sp. an eel from moderate depth reefs, from the Comoro, Amirante, Maldive, Marquesas, and Line islands; Uro- pterygius kamar n.sp. a deep-reef, burrowing eel from the Comoro, Amirante, Society, Solomon, Caroline, and Pitcairn islands; and U. goslinei n.sp. a shallow-reef eel from New Guinea, and the Caroline and Solomon islands. INTRODUCTION Recent deep water collections using rotenone ichthyocides in the tropical Indian and Pacific oceans have resulted in the discovery of two new moray eels. With a grant from the Charline Breeden Foundation, the senior author was able to make the first extensive series of reef fish col- lections from the remote Archipel des Comores (Comoro Islands). Those specimens complement the extensive collections made by the junior author throughout Oceania, and extend the range of many Pacific species to include the western Indian Ocean. In the process of describing these morays, a third and distinctive new species was discovered among the collections of reef fishes made by the George Vanderbilt Foundation, now housed at the California Academy of Sciences. Measurements are straight-line, made either with a 300-mm ruler with 0.5-mm gradations (for total length, trunk length, and tail length) and re- corded to the nearest 0.5 mm, or with dial cali- CALIFORNIA ACADEMY OF SCIENCES / GOLDEN PATE PARK / SAN FRANCISCO 162 pers (all other measurements) and recorded to the nearest 0.1 mm. Body length comprises head and trunk lengths. Head length is measured from the snout tip to the posterodorsal margin of the gill opening; trunk length is taken from the end of the head to mid-anus; maximum body depth does not include the median fins. Vertebral counts (which include the hypural) were taken from radiographs. Materials used in this study are housed at the following institutions: Austral- ian Museum, Sydney (AMS); Academy of Natural Sciences of Philadelphia (ANSP); Bernice P. Bishop Museum (BPBM); California Academy of Sciences (CAS, including George Vanderbilt Foundation Collections, GVF); U.S. National Mu- seum of Natural History (USNM); J.L.B. Smith In- stitute of Ichthyology, Rhodes University (RUSI); and the Scripps Institution of Oceanography (SIO). Paratypes of Uropterygius kamar will be deposited at the Muséum National d'Histoire Naturelle Paris (MNHN), and the British Museum (Natural History). ACKNOWLEDGMENTS We wish to thank the following individuals: Maurice C. Giles for photographic assistance; William C. Ruark and James E. Gordon for the preparation of radiographs; Virginia M. Gregory (Figs 3 a-c), Katherine P. Smith (Fig. 5), and Mary H. Fuges (Figs. 1-2) for their artwork; Lillian J. Dempster and William N. Eschmeyer for their PROG) GAIA AGAD SS Gir. 4ltieS ERs eV © LertulaN © es critical reading of this manuscript; and the cur- ators of fish collections for allowing us to exam- ine specimens. James E. Bohlke kindly made available his specimens of the new species and the illustrations prepared by Mary Fuges. Ran- dall’s collections were made possible in part by grants from the National Geographic Society and the American Philosophical Society. A portion of McCosker’s work was supported by funds from the Charline Breeden Foundation. We are especially grateful to the people of the Archipel des Comores, and in particular, to Chief Moham- med Ali Chabane and Prince Nacr-Ed-Dine, for allowing us to collect specimens. TAXONOMY Gymnothorax breedeni McCosker & Randall, new species (Figures 1-3, Table 1.) Gymnothorax sp.: Randall & Helfman, 1972: Figs. 5-6 (photo- graphs of Marquesas specimens mentioned in text). Gymnothorax sp.: McCosker & McCosker, 1976: 26 (Comoran specimens mentioned in text). MATERIAL EXAMINED. Holotype: CAS 35250, 530 mm. Archi- pel des Comores, Isle Grande Comore, Itsandra, 0.5 km north of Hotel Itsandra. Captured in 20 m among coral and rock within a large cave, 50 m offshore, on 19 February 1975 by J. —. McCosker, S. McCosker, M. D. Lagios, L. Gunther, and D. C. Powell, using Pronoxfish ichthyocide. Paratypes: USNM 215283 (formerly BPBM 7763), 645 mm, Line Is., Washington Id., 15 m, J. E. Randall. BPBM 11903, 524 mm, Marquesas Is., Tahuata, 10 m, J. E. Randall. ANSP 134222, 496 mm, Amirantes Is., D’Arros Id., 4 m, J. E. Bohlke. TABLE 1. Counts, and proportions in thousandths, of the holotypes and paratypes of Gymnothorax breedeni (3 paratypes), Uropterygius goslinei (7 paratypes), and U. kamar (9 paratypes). TL = total length, HL = head length. ‘ie i G. breedeni Wnaestine! ; : U. bower “— Holotype Mean Range Holotype Mean Range Holotype Mean Range Total length 530 mm 549 496-645 483 mm_ 271 149-527 249 mm 232 164-345 Vertebrae 128 128.5 128-129 127, 127.9 124-130 13h 136.3 134-138 Head/TL 152 130 122-136 116 116 105-121 102 103 95-108 Trunk/TL 340 336 302-356 300 309 = 300-333 305 306 299-307 Tail/TL 528 533 512-562 584 574 547-587 592 SZ 571-606 Snout/HL 181 HS) 171-186 161 170 161-197 161 158 132-183 Jaw/HL 379 374 353-398 368 386 328-458 369 360 8279-423 Eye/HL 79 87 76-112 73 65 55- 76 U5) 69 63- 77 Interorbital/ HL 119 140 119-170 AS 11 105-133 110 97 79-119 Depth/HL 543 622 475-765 482 405 348-482 578 353 316-402 Width/HL 314 364 312-470 359 264 205-339 267 224 195-267 McCOSKER & RANDALL: NEW MORAY EELS 163 Figure 1. Gymnothorax breedeni, paratype, ANSP 134222, 496 mm. DIAGNOSIS. A moderate length species of Gym- nothorax with anus before mid-body, tubular an- terior nostrils and slightly raised posterior nos- trils, uniserial jaw teeth, vomerine teeth either absent or few, and a brown coloration with dis- tinctive black areas behind the eye, at the rictus, and over the gill opening. DESCRIPTION OF HOLOTYPE. (Counts and pro- portions of the holotype and the three paratypes are given in Table 1.) Greatest depth of body 14 times in total length (TL). Body width 1.7 times in depth. Tail longer than body, distance from snout to anus 2.12 times in TL. Head 7.6 times and trunk 2.9 times in TL. Dorsal fin origin ahead of gill opening, the predorsal distance ca. 8.8 times in TL. Dorsal fin height at gill opening ca. 3.5 times in body depth. Snout 5.5 times and upper jaw 2.6 times in head length (HL). Eye 12.7 times in HL and 2.3 times in snout, closer to rictus than tip of snout. Fleshy interorbital width 8.4 times in head. Gill opening nearly horizontal, its center slightly below mid-body, its length about equal to diameter of eye. Anterior nostril tubular, minute, not reaching snout tip. Posterior nostril well ahead of eye, a vertical line drawn from it falls nearly 0.5 eye diameter in front of eye; opening small, nearly circular, forming a small tube in larger individ- uals. Jaws subequal, the mouth closing completely. Teeth in jaws (Fig. 3) uniserial, nearly subequal, stout, pointed, and finely serrated. Two depres- sible long canines along mid-line of intermaxil- lary, the posterior the largest tooth in the mouth. Three pairs of depressible canines behind man- dibular symphysis. Number of vertebrae 128, 52 before anal fin. The first dorsal pterygiophore arises above the third vertebra. Head pores (Fig. 2) present but not obvious. A single pore anterior and proximal to, and a sec- ond pore below the base of the anterior nostril. Six pores along the mandible, the second through fifth the largest. Four equally spaced pores along the upper jaw, the first beneath the nostril base, the last beneath the center of the eye. A single pore between each anterior and posterior nostril. Color in isopropyl alcohol brown overlain with a darker speckling throughout, although slightly lighter along the belly. Anus and gill openings within black areas. A prominent black spot. be- hind each eye, with a black slash beginning at the ventral margin of the orbit and ending about an eye’s diameter behind the rictus. A black spot 164 PROC. CALIF. ACAD. SCI., 4TH SER., VOL. 41, NO. 3 ey aS eh ENN es MSGEe Mae Sep ete Figure 2. Gymnothorax breedeni, paratype, ANSP 134222, 496 mm. smaller than the eye present beneath the rictus of some specimens. Margin of caudal cream, yellowish in life. ETYMOLOGY. Named in honor of Vic E. Breeden, president of the Charline H. Breeden Foundation which made this collection and study of Comoran fishes possible. REMARKS. The closest relative of Gymnothorax breedeni appears to be G. flavimarginatus (Rup- pell, 1828), a wide-ranging species which was either observed or collected at each of the locali- ties where we have taken the new species. G. breedeni differs from G. flavimarginatus in_ its coloration (viz., the black markings associated with the eye and jaw and the restriction of the light fin edging to the tail) and in its reduced or lack of vomerine dentition. The new species is similar in appearance to G. monostigma (Regan, 1909) in having a black blotch behind the eye. G. monostigma is readily separable in having the lower jaw pores and pos- terior nostril lying within white spots and in lack- ing a black area around the gill opening. The posterior nostril of the new species be- comes more tubular with age (see Fig. 2). This condition occurs among several species of vari- Ous moray genera and is considered by us not significant as a generic character. A color photograph which nicely illustrates the facial coloration of living G. breedeni appeared in Randall and Helfman (1972). Besides the typi- cal material herein described, another specimen of this distinctive species was observed but not collected by Randall at Villingili Island, Maldives, at about 10 m depth. Uropterygius kamar McCosker & Randall, new species (Figures 3b & 4, Table 1.) MATERIAL EXAMINED. Holotype: CAS 35251, 249 mm. Archi- pel des Comores, Isle Grande Comore, Itsandra, 0.5 km north of Hotel Itsandra. Captured in 25 m over coral sand slope, 50 m offshore, using Pronoxfish ichthyocide, on 19 February 1975 by J. E. McCosker, S. McCosker, M.D. Lagios, L. Gunther, and D. C. Powell. Paratypes: CAS 35252, 20(73- 224 mm), and RUSI 4557, 5(112-215 mm), collected with the holotype. Collected by J. E. Randall using ichthyocides or a hand spear, generally associated with a coral rubble bottom between depths of 3 to 55 m: BPBM 12051, 345 mm, Tahiti, Popote Bay; BPBM 9442, 2(211-228 mm), Palau Is., Augelpela Reef; USNM 215281 (formerly BPBM 15665), 219 mm, Solo- mon ls., Florida Id.; BPBM 16471, 4(150-325 mm), Pitcairn Id., Oeno Id.; BPBM 16606, 297 mm, Pitcairn Id., Oeno Id.; BPBM 17005, 136 mm, Pitcairn Id.; BPBM 17047, 327 mm, Pitcairn Id. Collected by J. E. Bohlke et al.: ANSP 124794, 3(145-198), Amirantes Is., St. Joseph Id., 15-27 m. ANSP 134217, 11(124-365), Amirantes Is., D’Arros Id., 20-30 m. DIAGNOSIS. A small species of Uropterygius with vertical fins restricted to the tail tip, anus before mid-body, two prominent cephalic lateral McCOSKER & RANDALL: NEW MORAY EELS 165 2 2 Aas TR WAN oe b Whe SAdTALUN ISH "RABE a, SSS AS \se ea BLE =, ALA CLPYIP PVA IPP I 2p EEE S 5 Figure 3. Dentition of: (A) holotype of Gymnothorax breedeni, CAS 35250, 530 mm; (B) holotype of Uroptery- gius kamar, CAS 35251, 249 mm; and (C) paratype of Uropterygius goslinei, ANSP 117434, 527 mm. lines pores, posterior nostril not closely associ- ated with a supraorbital pore, triserial maxillary dentition, white elongate anterior nostrils, and a prominent white band in front of the eye. DESCRIPTION OF HOLOTYPE. (Counts and pro- portions of the holotype and nine paratypes are given in Table 1.) Body subrectangular in cross section, laterally compressed throughout, _ its greatest depth 26 times in total length (TL). Body width 1.4 times in depth. Tail longer than body, distance from snout to anus 2.45 times in TL. Head 9.8 times and trunk 3.3 times in TL. Dorsal fin restricted to tail, arising about one-half head length (HL) from tail tip. Snout 6.2 times and upper jaw 2.7 times in HL. Eye 13 times in HL and 2.2 times in snout, equidistant between snout and rictus. Fleshy interorbital width 9.1 times in HL. Gill opening small, ovate and hori- zontal, lying about mid-body, its length less than eye diameter. Anterior nostrils elongate, tubular, nearly equal in length to eye, extending noticeably beyond snout. Posterior nostril within the interorbital space, above and slightly in advance of middle of eye, lying within a short tube, the proximal edge flush with the head. Jaws subequal, the mouth closing completely. Teeth in jaws (Fig. 3b) pointed, recurved. Upper jaw teeth triserial, the outer row teeth small, 166 PROG (GALEIEVAGAD. SGlF 4a SERS VOL 41 NOs Figure 4. Head and anterior trunk region of paratypes of Uropterygius kamar. A. Uniform coloration, BPBM 9442, 228 mm. B. Mottled coloration, BPBM 12051, 345 mm. numerous and close-set, the middle row teeth somewhat irregular and intermediate in size, the inner row teeth the largest, comprised of widely- spaced depressible fangs. Lower jaw teeth bi- serial, except triserial anteriorly, the outer row teeth small, numerous and close-set, the inner row teeth widely-spaced, depressible fangs, larg- est anteriorly. A series of depressible fangs in the intermaxillary region, separated from the uniserial vomerine teeth by a gap. Number of vertebrae 137, 49 before anal opening. Median fins restricted to the end of the tail, appearing above the 115th vertebra on a radiograph. Head pores distinct, usually associated with white markings along snout and lips. Anterior nostril bases bordered laterally and ventrally by pores. Six pore pairs along upper and lower jaws, alternating in position. Two supraorbital canal pores. Two cephalic lateral line pores, fol- lowed by a minute series of lateral line pores passing along the mid-body to the tail tip. Color in isopropyl alcohol tan to chocolate brown, distinctively marked by a variable series of white bands along the chin, throat, snout and nape. A prominent band passes obliquely from the anterior edge of the eye through the poster- ior nostril and meets its partner along the inter- orbital space. Anterior nostrils white and con- nected between their bases by a white stripe. Inside of mouth white. The trunk of small speci- mens often has a faint series of lighter, broken bands. ETYMOLOGY. Named in honor of the Archipel des Comores, the location of capture of many of the type specimens. The newly independent Comoran Republic has derived its name from kamar, an Arab word for the moon, in reference to the moon-like quality of the volcanic surface of those lovely islands. REMARKS. This new species is most similar to Uropterygius xanthopterus Bleeker (1859), from which it is easily separable on the basis of its distinctive coloration. U. kamar also differs from specimens we have identified as U. xanthopterus (CAS 35254 from Kapingamarangi and SIO 73-205 from Australia) in having more vertebrae and a more anterior median fin origin (as viewed from radiographs). The new species is also similar to McCOSKER & RANDALL: NEW MORAY EELS 167 Figure 5. Head and mid-trunk region of holotype of Uropterygius goslinei, CAS 35253, 483 mm. U. alboguttatus Smith (1962), described from the western Indian Ocean, which is more closely related to U. xanthopterus (and possibly synony- mous) and differs markedly in its coloration. U. kamar differs from its other congeners particu- larly in its coloration, multiserial dentition, and the presence of two cephalic lateral line pores. Uropterygius goslinei McCosker & Randall, new species (Figures 3c & 5, Table 1.) MATERIAL EXAMINED. Holotype: CAS 35253, 483 mm. Palau Is., Kayangel Atoll. Captured in a lagoon 200 m east of the west side of the atoll, in 0-3 m depth over a sand and coral- line algal bed with coral heads, using Noxfish ichthyocide by R. Gaines and party on 8 October 1956. Paratypes: Collected in shallow water over sand and coral bottoms using rotenone ichthyocides. CAS 13972, 212 mm, and BPBM 19927, 149 mm, Kapingamarangi Atoll. CAS 29103, 213 mm, Papua, New Guinea, Egum Atoll. USNM 215282 (formerly GVF-1955, Sta. 25), 151 mm, Palau Is., Koror Id. AMS.1. 17600-001 (formerly CAS 13973), 252 mm, Palau Is., Kayangel Atoll. ANSP 117434, 527 mm, Solomon Is., Bougainville. ANSP 124812, 270 mm, Papua, New Guinea, Madang Harbor. DIAGNOSIS. A moderate length species of Uro- pterygius with vertical fins restricted to the tail tip, anus before mid-body, two cephalic lateral line pores, posterior nostril not closely associated with a supraorbital pore, short jaws, triserial jaw teeth, and a tan coloration overlain by narrow black bands and white spots and two distinct black stripes behind the eye. DESCRIPTION OF HOLOTYPE. (Counts and pro- portions of the holotype and seven paratypes are given in Table 1.) Body subrectangular in cross section, laterally compressed throughout, _ its greatest depth 18 times in total length (TL). Body width 2.5 times in depth. Tail longer than body, the distance from snout to mid-anus 2.4 times in TL. Head 8.6 times and trunk 3.3 times in TL. Dorsal fin restricted to tail. Snout 6.2 times and upper jaw 2.6 times in head length (HL). Eye 14 times in HL and 2.2 times in snout, its center slightly in advance of middle of jaw. Fleshy in- terorbital width 8.7 times in HL. Gill opening small, ovate and horizontal, lying about mid- body, its length less than eye diameter. Anterior nostril tubular, much smaller than eye, not extending beyond snout. Posterior nos- tril within the interorbital space, above and in advance of eye, lying within a short tube (Fig. 5). Jaws subequal, the mouth closing completely. Teeth in jaws (Fig. 3c) pointed, recurved. Upper jaw teeth triserial, the outer row teeth small, numerous and close-set, the middle row teeth intermediate in size, the inner row teeth the largest, comprised of depressible fangs. Lower jaw teeth triserial, the outer row teeth small, num- 168 erous and close-set, the middle row teeth inter- mediate and not extending beyond mid-jaw, the inner row teeth the largest, comprised of de- pressible fangs. Intermaxillary region is crowded by a group of large depressible fangs, separated from the uniserial vomerine teeth by a gap. Number of vertebrae 127, 48 before anal open- ing. Median fins restricted to end of tail, above the 114th vertebra on a radiograph. Head pores indistinct. Anterior nostril bases bordered laterally and ventrally by pores. Five pore pairs along upper and lower jaws. Two supraorbital canal pores. Two cephalic lateral line pores, followed by a minute series of lateral line pores passing along the mid-body to the tail tip. Color in isopropyl alcohol cinnamon, darker dorsally, overlain with narrow black lines which are broken along the lateral mid-line, with white punctations irregularly located between the lines. Two black lines originate at the posterior dorsal and ventral margins of the eyeball and extend 2-3 eye lengths posteriorly. ETYMOLOGY. Named in honor of William Alonzo Gosline in recognition of his numerous contributions to the study of apodal fishes. REMARKS. Following Gosline’s (1958) key to the central Pacific species of Uropterygius, the new PROC. CALIF. ACAD. SCI., 4TH SER., VOL. 41, NO. 3 species would be identified as U. marmoratus (Laceépede, 1803). U. goslinei differs in having two rather than a single cephalic lateral line pore and in its coloration. The new species is similar to U. xanthopterus and U. kamar in having two pores, but differs in its coloration and its com- plex dentition, and in attaining a larger size. LITERATURE CITED Bleeker, Pieter. 1859. Over eenige vischsoorten van de Zuid- kustwateren van Java. Nat. Tijdschr. Neder.-Indie, 19: 329-352. Gosline, William A. 1958. Central Pacific eels of the genus Uropterygius, with the descriptions of two new species. Pacific Sci., 12(3):221-228. Lacépede, (Comte) B. G. E. 1803. Histoire naturelle des pois- sons....deédiée au citoyen Lacépede. Vol. 5, Paris. McCosker, Sandra, and John £. McCosker. 1976. To the islands of the moon. Pacific Discovery, 29(1):19-28. Randall, John E., and Gene Helfman. 1972. Diproctacanthus xanthurus, a cleaner wrasse from the Palau Islands, with notes on other cleaning fishes. Trop. Fish Hobbyist, 20(11):87-95. Regan, C. Tate. 1909. Descriptions of new marine fishes from Australia and the Pacific Ann. Mag. Nat. Hist., ser. 8, 4:438-440. Ruppell, Wilhelm P. E. S. 1828. Atlas zu der Reise im nord- lichen Afrika. Zoologie. Fische des Rothen Meeres. 4 vols. Frankfurt-a-M., 1826-1828. Smith, J. L. B. 1962. The moray eels of the western Indian Ocean and the Red Sea. Ichthyological Bulletin, Rhodes University, 23:421-444. NH PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES FOURTH SERIES Vol. XLI, No. 4, pp. 169-182, 3 figs., 1 table. May 31, 1977 TIGER BEETLES OF THE GENUS C/C/INDELA IN THE SULPHUR SPRINGS VALLEY, ARIZONA, WITH DESCRIPTIONS OF THREE NEW SUBSPECIES (CICINDELIDAE — COLEOPTERA) by Norman L. Rumpp 3446 Bahia Blanca West Laguna Hills, California 92653 ABSTRACT: The Sulphur Springs Valley of southeastern Arizona contains a large tiger beetle fauna. Seventeen species of Cicindela (Cicindelidae - Coleoptera) are reported; of these, three are described as new subspecies: C. willistoni sulfontis, C. pulchra dorothea and C. nevadica citata. Ecological associations of each species are detailed and a key is provided for identification. The relationships with other subspecies are mentioned, with special emphasis on the three subspecies described. The Cicindela fauna of the valley is shown to be related, for the most part, to the fauna of the mid-Rio Grande River System, indicating that past physical geology of the lower Southwest allowed hydrographic flow in an eastward direction, at least as early as the Miocene. In late Pliocene or early Pleistocene, stream flow began going in a westerly direction allowing some migration of species in that direction. This had a limited effect on the tiger beetles of the Sulphur Springs Valley because this valley remained isolated during most of the Pleistocene. INTRODUCTION Many of the western United States tiger beetles of the genus Cicindela have undergone subspeciation. This is particularly evident in the Southwest where profound geological changes have occurred, especially during late Cenozoic. These changes have produced environmental dis- locations and have altered climate in such di- verse ways as to induce strong ecological pres- sures on the various populations. The Sulphur Springs Valley of southeastern Arizona has been totally isolated geologically from adjacent valleys at least through the latter part of the Pleistocene when it became an area of internal drainage. The Cicindela fauna of this valley contains a large number of species, some of which have uniquely subspeciated. The subspecies described in this paper are isolated from the main populations of their respective species, some by a considerable distance, thus extending our knowledge of the geographical limits of these species. Students and collectors of tiger beetles have been at- tracted to this valley for the past twenty years be- CALIFORNIA ACADEMY OF SCIENCES/GOLDEN GATE PARK/SAN FRANCISCO 170 cause of its rich fauna. During this same period the author has conducted fifteen field trips into the valley, and has accumulated over two thou- sand specimens for this study. ACKNOWLEDGMENTS The author is grateful for the assistance given by many persons who made it possible to dis- cover and accumulate specimens for this paper. These include: Dr. Mont A. Cazier of Arizona State University, Tempe, Arizona; Mrs. Patricia Vaurie of the American Museum of Natural His- tory, New York, N.Y.; Dr. Oscar A. Cartwright of the U.S. National Museum, Washington, D.C.; the Reverend Bernard Rotger of Pagosa Springs, Colorado; Mr. Douglas K. Duncan of Globe, Arizona and Mr. Vernon G. Clifford of Graham, Washington who sent specimens from their pri- vate collections; Mr. Jay M. Sheppard while a student at California State University, Long Beach, Dr. John Stamatov of Armonk, New York, Dr. deWendler-Funaro of Williston Park, New York, and Dr. David L. Pearson of Pennsylvania State University, who participated in the field; finally, and most importantly, my late wife Dorothy who shared enthusiastically in many field trips to col- lect specimens and take notes. PHYSIOGRAPHY OF THE SULPHUR SPRINGS VALLEY The Sulphur Springs Valley lies in the moun- tainous region of Arizona, south of the plateau region, and east of the desert region. The moun- tains are of the block-fault type, similar to those of the Basin and Range province of Utah and Nevada, but appear to be older. The valley lies across the center of Cochise County and extends northward into the southern portion of Graham County; it trends in a north-northwestward di- rection from the Mexican border for about 145 kilometers, with an average width of about 32 kilometers. The valley covers about 4700 square kilometers and drains another 2600 square kilo- meters of the bordering watershed. The 32nd parallel crosses the valley just south of the large central playa, while the 110th meridian traverses to the west of the playa. The valley is bounded on both sides by mountain ranges which sepa- rate it on the east from the San Simon Valley, and on the west from the San Pedro Valley; both of these valleys drain northward into the Gila River basin. The Sulphur Springs Valley is sepa- PROG. GALIF. AGAD:* SGI-, 41H SER:;, VOI AT) INO 34 rated on the southeast from the San Bernardino Valley by the Perilla Mountains, however, this smaller valley and the southern third of the Sulphur Springs Valley drain to the south into the Yaqui River system of Sonora, Mexico. The mountains that enclose the valley from the east include the Pinalino, Dos Cabezas, Chiri- cahua, Pedregosa, and Perilla mountains. The western chain includes the Galiuro, Winchester, Little Dragoon, Dragoon, and Mule mountains. The mean elevation of the valley is slightly over 1300 meters. The climate is semi-arid as the cen- tral portion receives an average yearly rainfall of less than 30 centimeters. There is a rainy season which begins in mid-July and lasts through Sep- tember. The center of the valley contains a_ large, roughly triangular playa that covers about 132 square kilometers and drains the upper two thirds of the valley. This playa lies a few kilometers south of the town of Willcox, at an elevation of 1257 meters. During pluvial times this basin of internal drainage contained a lake that was ap- proximately 32 kilometers long by 18 kilometers wide, thus covering an area of about 311 square kilometers to a depth of 13.5 meters. This Plei- stocene lake was named Lake Cochise by O. E. Meinzer (1913:34), and then as now, had no outlet. An extensive wind-built sand and clay re- gion lies south and west of Willcox. This region is characterized by scattered ridges and hills, some as high as 15 meters and several kilometers long, laid out in no specific pattern, but sepa- rated from each other by ponds or depressions which fill with water during the rainy season and are capable of holding this water for a con- siderable time thereafter. The isolation of the Sulphur Springs Valley, its unique drainage system which is partially south but mostly internal, the peculiarities of the top- ography where water can remain in blowouts for extended time are factors which contribute to the valley's ability to sustain a large diversified tiger bettle fauna. SYSTEMATIC ACCOUNT Genus Cicindela Cicindela willistoni sulfontis Rumpp, new sub- species. (Figure 1a.) DESCRIPTION.—Medium in size, narrow, con- vex, robust in form; color dark blue-green, green- RUMPP: TIGER BEETLES ish bronze, or brown; dull above, brilliant below; maculation broad, connected along edge of elytra HEAD: A few white hairs on vertex just above frons in interocular space. ELYTRA: The macula- tion consists of a broad C-shaped humeral lunule. The middle band is extended upward along the edge of the elytra to meet the humeral lunule, also considerably downward to join the apical lunule; the band enters the edge with only a slightly downward direction, then bends suddenly so that it is nearly parallel to the suture for a long distance before it ends in an inwardly di- rected bulb; the apical lunule is broad at the tip with its inner edge extending outwardly and down toward the edge where the lunule bulges without recurving. DIMENSIONS: Holotype male, green - length 13.0 mm, width 4.8 mm. Allotype female, brown, similar to the male but broader - length 13.0 mm, width 5.4 mm. ETYMOLOGY.—This subspecies is named for the Sulphur Springs Valley by combining the Latin sulfo (sulphur) with fontis (a spring). THE TYPE SERIES.—Until 1965 the type series consisted of only 4 specimens collected between 5 and 9 kilometers southeast of Willcox, in a re- gion of wind carved sinks. In August of 1969 Mr. Jay M. Sheppard discovered the primary habitat 171 at which time large series were obtained. The type location is 5.6 kilometers west-southwest of Willcox, on the playa, and a few hundred meters south of the tracks of the Southern Pacific Rail- road. All paratypes from the north shore of the playa, as all those from the blowouts to the east are considered paratypical topotypes; all speci- mens are assumed to be from the same general population. The type series consists of 217 specimens dis- tributed as follows: Holotype, male, type no. 12,530 in the collection of the California Academy of Sciences; allotype, female, in the N. L. Rumpp Collection (NLRC). There are 215 paratopotypes of which 68 are in the NLRC, 86 in the collection of Rev. Bernard Rotger, C.R., of Pagosa Springs, Colorado, 53 in the collection of Jay M. Shep- pard of Laurel, Maryland, and 2 each in the col- lections of the American Museum of Natural His- tory (AMNH), New York City, the U.S. National Museum (USNM), Washington, D.C., and the California Academy of Sciences (CAS), San Fran- cisco, California. VARIABILITY.—The elytral maculation is strik- ingly constant, but the color is either green or brown with few intermediates. This is shown in Table 1. TABLE 1. Distribution of color and sexes in the type series of C. willistoni sulfontis. green brown intermediate ponector Date male female male female male female F. H. Parker 13:X:1949 1 N. L. Rumpp 2:X:1961 1 1 N. L. Rumpp 13:X:1964 1 G. Gaumer VII:1969 1 J. M. Sheppard 10: VIII:1969 2 5 5 3 B. Rotger 17: VII1:1969 28 Di, 28 11 1 3 N. L. Rumpp 20:VII:1970 6 12 11 10 J. M. Sheppard 20:VII:1970 13 14 2. 10 3 N. L. Rumpp 8:VII1:1971 1 6 1 Subtotals 51 67 57 34 mA i Totals (217 specimens) aon 91 8 172 COMPARISON WITH OTHER SUBSPECIES OF C. WILLISTONI.—AII specimens of the type series are similar in marking and reasonably homogen- eous in size, but their color varies. In the majority, the color is blue-green, similar to ssp. pseudosen- ilis and ssp. praedicta of the Death Valley Hydro- graphic System of California and Nevada, but unlike these, ssp. su/fontis has a few hairs be- tween the eyes, but not nearly so many as are found on the more eastern subspecies willistoni, hirtifrons and estancia. A lesser number, making up nearly the balance of the population, is brown or bronze in color, similar to the eastern subspecies. The only known location for ssp. sulfontis is near Willcox where it is isolated by great distances from all other populations of C. willistoni. This isolation is especially signifi- cant because C. willistoni is usually not found in the Colorado River drainage system! as reported by Rumpp (1961: 168), although it has been located now in Arizona, a fact not known at that time. Cicindela willistoni shows a preference for lacustrine salt flats, although it may also invade marshy areas and semi-dry river beds where saline muds are exposed. Depending on eleva- tion and temperature, it may emerge early in the year at lower elevations, much later at higher altitudes. In the Sulphur Springs Valley the cli- mate is normally dry in May and June and re- mains that way until the start of the rainy sea- son in mid-July, at which time ssp. sulfontis sud- denly appears in numbers on the Willcox Playa where it remains in evidence throughout the rainy season. The larva builds a turret on the playa in the fashion of ssp. hirtifrons as reported by Willis (1967:176-177). DISTRIBUTION OF CICINDELA WILLISTONI. —Cicindela willistoni has an extensive distribu- 1ion in western United States. The species ranges throughout the Great Basin and that part of the Basin and Range province extending into south- eastern Arizona, but nearly exclusive of the Colo- rado River drainage system. East of the Rocky Mountains it occurs in Wyoming, Kansas, Okla- homa, Texas, and New Mexico. It may be found wherever there are wet playas, defined as sea- sonal salt flats or perennial lake beds. The various subspecies from the eastern side of the distribu- tion are a bright bronze color and bear a deeper 1An exception exists in the Ft. Bridger, Wyoming popula- tion of ssp. willistoni. This area is connected hydrographically to the Green River, the main tributary of the Colorado River. PROC. CALIF. ACAD. SCI., 4TH SER., VOL. 41, NO. 4 middle band, usually with very broad to conflu- ent maculation. In these the head always bears a number of long decumbent hairs (Willis, 1967: 303). They may be geographically distinguished as ssp. willistoni of Wyoming; ssp. hirtifrons of western Kansas, Oklahoma, Texas, and north- eastern New Mexico; ssp. estancia the endemic population of the Estancia Valley in central New Mexico, and finally ssp. funaroi of Jemez Creek in northcentral New Mexico. In the Great Basin, most populations are of ssp. echo. These may be subdivided into the eastern or Bonneville echo populations, and the western or Lahontan am- adeensis populations, with such little differences in habitus and color as to be considered synony- mous (Cazier, 1936:157). The name_ spaldingi was proposed by T. L. Casey (1924:14) for one bright green, broadly maculated specimen from Calleo, Utah. This is in a region encompassing widely variable populations of ssp. echo. These populations were sampled by the author on 29 May 1970 at Blue Lakes, Tooele County, Utah, at the edge of the Great Salt Lake Desert, ap- proximately 80 kilometers northwest of Calleo. Another location is at Fish Springs Wildlife Refuge, Juab County, Utah, also at the edge of the Great Salt Lake Desert and 25 kilometers east of Calleo. This last site was visited by the author on 4 May 1962, and by Lawton and Willis (1974:51) on 24 July 1971. These populations are extremely variable in that a few individuals are like Casey’s spaldingi, while others resemble ssp. estancia; however, the great majority of specimens is typical of ssp. echo. The third larg- est hydrographic entity of the Great Basin is the Death Valley system where C. willistoni is inter- spersed with populations of ssp. echo, ssp. pseudosenilis, and ssp. praedicta. Here colors trend to green or dark blue-green, with or with- out maculation. In all of the Great Basin popula- tions from Utah and Nevada, southern Idaho, and eastern parts of Oregon and California, the vertex bears only a few hairs or none at all. Cicindela pulchra dorothea Rumpp, new sub- species. (Figures 1b, 2.) DESCRIPTION.—Large in size, form robust; head, prothorax and elytra brilliant metallic-cerise or wine-red. Elytra with definite maculation, some- times fully connected along sides. HEAD: Brilliant red on vertex from eye to eye, fringed with gold and green before shading into the bright deep RUMPP: TIGER BEETLES 73 [On Figure 1. Dorsal views of: a. Cicindela willistoni sulfontis; b. C. pulchra dorothea; and c. C. nevadica citata. Scale indicates 10 mm. blue that covers the underside. Slightly rugose on the vertex, wrinkled near the eyes; deep setiger- ous punctures on the frons. Vertex with sparse erect white hairs, frons covered with long white decumbent hairs, genae bare. First four segments of the antennae bright green; first segment with over twenty stiff hairs on the outer edges, sec- ond segment bare, third and fourth with a sparse row of white hairs on the outer edges, with one setigerous puncture on the inner tips. Clypeus bright blue-green, with two setigerous punctures. Labrum white with a narrow dark edge, tridentate, with six setigerous punctures bearing long white hairs close to the outer edge. THORAX: Pronotum as broad as the head ex- cluding the eyes, broader than long by a ratio of 4 to 3; broadest at the forward one third, then slightly tapering inward along a slight curve to the rounded basal angle. Surface finely rugose, brilliant red fringed with gold and green. Anterior and basal impressions deep, bright blue fore and aft respectively of these impressions; median impression very narrow, shallow, colored green. ELYTRA: Brilliant metallic red throughout, except for the very narrow outer edge which is blue with a row of green punctures. Surface punctured, these punctures deepest nearest the base, very shallow toward the apex. Suture golden-green. Maculation complete but reduced in extent; central band triangular, broadest at the edge of the elytra, very slightly recurved downward, reaching nearly half way to the su- ture; humeral lunule shallow at its tip, some- times connected to the central band; apical lun- ule narrow, ending in a shallow dot, then nar- rowly connected to the middle band. UNDER- SIDE: Bright deep blue. Proepisterna covered with long decumbent hairs; meso- and metaepis- terna with shorter decumbent hairs. Upper ab- dominal segments with sparse white hairs on the sides, and long white hairs at the bottom edge of the 3rd, 4th, and 5th sternites; pro- and meso- coxae with long recumbent hairs, metacoxae bare; trochanters bare except for one setigerous puncture at the tip of those of the front and middle legs; femora blue with long hairs, more so on the anterior ones than on the middle and back ones; tibiae blue-green with rows of stiff white bristles, apical spurs about two thirds the length of the first tarsal joints; tarsal joints with several erect hairs, especially at tips; tarsal claws unpigmented, short, about half the length of the fifth tarsal joints. DIMENSIONS: Male - length 15.2 mm, width 6.0 mm. Female - length 15.6 mm, PROC. CALIF. ACAD. SCI., 4TH SER., VOL. 41, NO. 4 Figure 2. Habitat of Cicindela pulchra dorothea south of Williston, Arizona on 26 July 1964. In the distance is a water filled pond. width 6.4 mm. ETYMOLOGY.—This insect is named in mem- ory of my late wife, Dorothy, who collected many of the specimens in the initial large series. THE TYPE SERIES.—There are 78 specimens in the type series. This includes the holotype male, the allotype female, and 44 topotypical paratypes collected by N. L. and D. H. Rumpp at the type location 5.2 to 6.2 kilometers southeast of Will- cox, on 26 July 1964, and an additional specimen collected on 11 September 1965. Five paratopo- types were captured on 21 May 1970, three on 16 July 1974, twelve on 17 July 1974, and one more on 6 August 1975, all by the author. Two para- topotypes were collected by Dr. John Stamatov on 8 August 1971. The series also includes one paratype from 11.3 kilometers west of Douglas, Arizona, dated 6 August 1965. Distribution of the type series: Holotype male, type no. 12529 in the California Academy of Sci- ences; allotype female, 66 paratopotypes and one paratype in the NLRC. Two paratopotypes deposited in each of the collections of the AMNH, the USNM and the CAS. Two are in the private collection of Dr. John Stamatov. COMPARISON WITH C/ICINDELA PULCHRA PULCHRA.—From the time it was described by Thomas Say in 1823, Cicindela pulchra was con- sidered a monotypic species from midwestern United States, ranging from Montana to Texas. Subspecies pulchra is larger, generally immacu- late, with a color varying from dark red to deep violet. Some specimens, and these are few, have a triangular marginal middle spot not extended at the sides, sometimes with two dots representing a vestigial humeral lunule; otherwise the border of the elytra is broadly blue. Subspecies dorothea has a brighter red color, often with strong green reflections when viewed from an angle. In many specimens the white maculation is connected along the outer edge of the elytra, while in a few the maculation is reduced to dots. Less than ten percent of the specimens in the type series have a narrow middle band with blue connec- tions along the outer edge of the elytra, and whereas these resemble the rarely maculated specimens of ssp. pulchra, the color is typical of ssp. dorothea, as is the smaller size. Although ssp. dorothea is now isolated in southeastern Arizona, it was probably derived from the ancestral Rio Grande system during the Mio-Pliocene when the flow of the ancestral upper Colorado River drained east, and the Con- tinental Divide was west of the valley, as will be discussed later. Now that this subspecies occu- pies a region that drains into the Yaqui River RUMPP: TIGER BEETLES system, it will probably be found in northern Sonora, especially since it has been located a few meters from the Mexican border west of Douglas, Arizona. Some New Mexico specimens of C. pulchra from the USNM, AMNH, and private collections were examined. Those from the northwest or Farmington district can be identified with ssp. pulchra, whereas those from the southwest or Silver City district appear to identify with ssp. dorothea, as do those from the trans-Pecos re- gion of western Texas. Cicindela nevadica citata Rumpp, new subspecies (Figure 1c.) DESCRIPTION.—Medium size, body narrow, color brownish bronze with green micro pits, maculation of elytra narrowly connected along the outer edge. HEAD: Small relative to body size but with salient eyes; finely granulate vertex, frons very narrowly striate. Color green suffused with bronze. Pilosity of medium density on vertex, frons, and clypeus. First 4 joints of antennae green with bronze reflections on last 3 joints; first joint with 12 to 15 white recumbent hairs, 2nd joint bare, 3rd and 4th with 2 to 4 short hairs on outer edges and 2 setigerous punctures at tips. Clypeus white, barely produced in front, unidentate, ratio of width to length is 3 to 1; it bears 16 to 18 setigerous punctures located near anterior edge in two irregular rows. THORAX: Pronotum narrow, about as broad as the head excluding the eyes, cylindrical, length equal to width, finely rugulose, color greenish bronze. Anterior impressions of medium depth, basal impressions deeper, median impression shallow; sparse white decumbent hairs cover the top sur- face, more so at the edges, top, and bottom leav- ing the disc nearly bare. ELYTRA: Long, narrow, widest at the middle and twice as wide as the pronotum. Male - sinuous near tip, finely serrul- ate in this region; tips end in spines. Female - more sinuous and indented; terminal spines are smaller. Color brown except for center half from suture Outward which is bronze; pitted through- out, pits green. A row of small green punctures descends near the suture, two rows of small punctures are located in the humeral impres- sions. The maculation is white and of typical nevadica design, except that the middle band is more narrow where it enters from the outer edge, and the central extension is longer; all markings are narrowly connected along the sides. WHS UNDERSIDE: Mostly shiny brown with green re- flections, however the genae are bright green, narrowly striate and covered with dense decum- bent white hairs. The proepisterna are bronze with green reflections, densely covered with long white decumbent hairs; meso- and metaepi- sterna similarly clothed with long white hairs; pro- and mesocoxae also clothed with long white hairs, metacoxae bare; trochanters unpigmented, bare except for a setigerous puncture at tip of pro- and mesotrochanters. All abdominal sternites clothed with decumbent hairs on outer halves, with a few hairs at forward edges centrally. Fore and middle humeri with several rows of dense white hairs on the outer and inner sides; pos- terior humeri sparsely covered with non-over- lapping hairs, none on outer sides. Tibiae bronze with greenish reflections, sparsely bristled. DI- MENSIONS: Male - length 10.3 mm, width 4.2 mm. Female - length 11.0 mm, width 4.4 mm. ETYMOLOGY.—The name citata alludes to the insect’s fast running over moist sandy lacustrine shores. THE TYPE SERIES.—The type series consists of 48 specimens. The holotype male was collected at the edge of the central playa 8 kilometers west- southwest of Willcox on 20 July 1970. The allo- type female is from 2.9 kilometers south of Will- cox, captured on 8 August 1955. Forty-six para- topotypes were collected near Willcox at sites such as Twin Lakes, nearby ponds, or at the northern edge of the Willcox Playa; all were col- lected from July through August between the years 1955 and 1975. The holotype is deposited in the collection of the California Academy of Sciences as type no. 12,528. The allotype and 29 paratopotypes are in the NLRC. Paratopotypes in private col- lections are: Sheppard 3, Clifford 6, and Pearson 2. Two each were deposited in the AMNH, the USNM, and the CAS collections. VARIABILITY: Most of the specimens in the type series are similar to the described type except for two specimens. One is a large female 12.4 mm long, with more cupreous coloration, the other a very small male only 9.1 mm long. COMPARISON WITH OTHER SUBSPECIES OF CICINDELA NEVADICA.—The nominate subspe- cies nevadica of the Great Basin region of Nevada and California is of a coppery bronze with green reflections, somewhat similar in color to. ssp. citata. The typical form of ssp. nevadica is reputed to be from Ash Meadows, Nye County, Nevada, 176 although LeConte (1875:159) merely reported it from Nevada. In the Nevada populations of ssp. nevadica the middle band seldom has any extension along the border, whereas in California populations there is usually a narrow extension along the border, sometimes connecting all the markings. The elytral pattern of the California populations more nearly matches the woodcut accompanying the description by LeConte, but he also states in the notes following his descrip- tion that the markings were not connected along the margin. The markings of ssp. nevadica, where there is a connection along the outer border, differ from ssp. citata in that the middle band, where it enters from the edge, is broader in the former than in the latter; furthermore, the extension of this band is much deeper and nar- rower in ssp. citata. Subspecies tubensis differs from ssp. citata in that it has a red color, and al- though its maculation is always connected along the margin, these markings are much broader. Subspecies knausi of the upper Rio Grande basin is more nearly like ssp. citata in markings, but it also has slightly broader markings and _ less greenish coloration. Subspecies citata fits best between ssp. nevadica in color, and ssp. knausi in markings. The range of ssp. citata is the Sul- phur Springs Valley, possibly into the Yaqui River system of Sonora. Those from the north- ern shores of the Gulf of California, specifically from Puerto Penasco, can be identified with this subspecies. ECOLOGICAL ASSOCIATIONS OF CICINDELA SPECIES OF THE SULPHUR SPRINGS VALLEY There are seventeen species of the genus Ci- cindela in the Sulphur Springs Valley. These are separated below according to their ecological as- sociations. Comparisons of related species and subspecies are given for some of the seventeen. On wet flats, at the edge of water 1. C. praetextata erronea. This bright green or blue-green subspecies is endemic to the valley. It may be found on the northern shore of the Willcox Playa, and from the wind-formed ponds to Croton Springs. It is strikingly different in color from the bronze ssp. praetextata of the Colorado River drainage system, and from the red ssp. fulgoris of the Rio Grande system and upper regions of the Little Colorado River. 2. C. marutha marutha. When individuals of PROC. CALIF. ACAD. SCI., 4TH SER., VOL. 41, NO. 4 this subspecies emerge at the start of the sum- mer rains, they form the largest populations along the mud flats of the ponds near Willcox. They will also invade marshy grasslands for they are not restricted to open areas, and at night they will come to light. This population is made up of nearly all pure grass-green individuals. Only 4 percent of the total are red. Most other pop- ulations of C. marutha in Arizona and New Mex- ico contain a high percentage of red individuals, but the northern ones contain the highest per- centage of this color. In southeastern Utah there are pure red populations which should be re- ferred to ssp. rubicunda. 3. C. haemorrhagica. The valley’s population contains rather small, well marked individuals which have most often been identified with ssp. woodgatei. There is a resemblance to C. cartha- gena carthagena, especially in the median widen- ing of the elytra and the angle so produced where the outer edges converge toward the tip. Individuals are found at the edge of water, usu- ally on the darker spots where they segregate from other species. 4. C. nevadica citata. Individuals are relatively scarce in the valley, most being found immedi- ately south of Willcox. This scarcity may be at- tributed to the high salinity of the local flats, be- cause elsewhere this species seems to prefer damp sand, free from salt incrustations. In grassy marshlands and very wet spots 5. C. tenuisignata. This is a brown monotypic species common to the Rio Grande system, the lower Colorado River drainage system, and Son- ora. Individuals prefer damp as well as wet sandy soil. 6. C. punctulata chihuahuae. This is a grassland tiger beetle that is relatively rare in the valley. When found it will be in wet places. Individuals exhibit a variety of colors, from a bronzy green to a bright blue, with or without maculation. Bluish green is the most common color. 7. C. nigrocoerulea nigrocoerulea. This sub- species is polymorphic with individual members exhibiting a variety of colors that range from dull green to deep blue to black; blue-black is the predominant color. Individuals may be found most frequently in the thicker marshy areas and muddy spots, sometimes in great abundance. 8. C. horni horni. This is also a polymorphic subspecies with colors varying from metallic green or bright blue, to black. It is widely dis- RUMPP: TIGER BEETLES tributed in the valley from Douglas to Willcox, from the lowest grassy areas to upland meadows. 9. C. sedecimpunctata sedecimpunctata. This is a very common subspecies in southern Ari- zona, Sonora, and the Rio Grande system. It has a tolerance for high altitude, having been located in the Pinalino Mountains at elevations of 2700 meters, but always near ample water supplies. Although it is related to, and resembles C. haemorrhagica, individuals are noticeably smaller than in that species. On muddy flats of the Willcox Playa 10. C. willistoni sulfontis. Members of this subspecies emerge in great numbers on the playa south-southwest of Willcox after the rains begin in mid-July and the surface of the playa becomes mucky. This form was first discovered at one of the ponds in the wind-built area south of Willcox, an area which it seldom invades. One individual was captured here following a long flight from the edge of a pond over a dry clay spot populated by C. pimeriana. Individuals are strong fliers, making frequent flights of as long as 10 meters, with very brief rests between flights. On dry flats, not far from water 11. C. pimeriana. This monotypic species is the most common one found in early spring and again in autumn. Its characteristic color is brilliant metallic green, although the few spring individ- uals that survive until October become dull black. This tiger beetle is found only in the Sulphur Springs Valley and the San Bernardino Valley to the southeast. This last area contains the type locality at the San Bernardino Ranch (the old Slaughter Ranch) on the Mexican border. The species is related to C. pulchra, with which it shares a general area south of Willcox, but its preferred habitat is on bare clay banks near water. An individual’s flight is long and singular, and it makes such flights between long rest periods, possibly as a defense against asilid rob- ber flies that prey upon it. The fly’s attack strategy is to swoop upon the beetle as the beetle takes to flight. The fly then embraces its prey with its legs while still flying and inserts its beak at the junc- ture of the open elytra, directly into the meso- sternum, stunning the beetle almost immediately. The flight of the asilid is so slowed by this load that fly and prey can be netted with ease. Five captures were made with C. pimeriana as the fly’s WAZ victim, and once with a C. marutha as victim. 12. C. ocellata ocellata. This small dark-brown subspecies belongs to the Sonoran fauna and is widely dispersed throughout southern Arizona. It differs most noticeably from ssp. rectilatera of the Chihuahuan fauna by its rufous abdomen. In grasslands, away from water 13. C. pulchra dorothea. This strikingly beauti- ful tiger beetle is found exclusively on open flat- lands covered with clumps of grass. The beetle emerges from these clumps to fly onto bare spots in one long flight. There it will either re- main in the open where it will seek the shade of grass, or it will rapidly run into the nearest clump of grass and disappear from view. 14. C. lemniscata lemniscata. This small red tiger beetle with pale legs and longitudinal macu- lation seeks open spots far from water. When dis- turbed it quickly takes to wing, but it is a weak flier with an erratic flight. Although often seen in the daytime, its habits are more nocturnal than diurnal, and at night it will come to light in great numbers. 15. C. debilis. This small dull-green tiger beetle is related to C. /Jemniscata, but unlike that species it rarely flies. It prefers heavy grass where it seeks the densest parts for refuge. It is rarely en- countered in the daytime and may be nocturnal in habit. The species will usually not fly, but when coaxed to do so its flight is erratic and very short, only a few decimeters in distance. The green C. debilis segnis was described as a variety from southern Arizona by Harris (1913:69), but Cazier (1954:287), after studying large samples from southern Arizona and Durango, Mexico did not favor retaining this subspecies. A population of darker individuals of this species was found near Willcox, and one with brighter green indiv- iduals was found near Douglas. However, these color variations may have been due to the age of the individuals sampled, the older possibly being the darker. 16. C. obsoleta santaclarae. This is the largest tiger beetle in the valley, often measuring 17 mm in length. It too is polymorphic for it may be green, blue, brown, black or red, although the predominant color seems to be green. The elytra are spotted, and a vestigial middle band is in- dicated by two dots on each elytron. The species is robust and its members are excellent fliers. Its flight is long, sometimes on the order of 100 meters, and often very high. It is found in grass- lands on well-drained alluvial slopes, particularly 178 along the edges of the valley. In the grasslands west of Douglas it is encountered with C. nigro- coerulea, C. pulchra, C. horni, C. lemniscata, and C. debilis. Tree areas near water 17. C. viridisticta arizonensis. This is another very small tiger beetle related to C. debilis and C. lemniscata. Near Douglas it is found among grasses on the slopes of drainage ditches in the chaparral-covered plains, while at Kansas Settle- ment it is found in grass near willows and tules that grow at the sides of irrigation canals. The species is a weak flier with a short and erratic flight. It prefers to remain on the ground where it will move with great speed and agility through the grass. KEY TO THE SPECIES OF CICINDELA IN THE SULPHUR SPRINGS VALLEY Tas Abdominall Segments MUfOUS jn. cccnccesncec-scneqeececreenreneneceec- 2 1b. Abdominal segments black, dark blue or greenish .... 4 2a. Elytra gradually widening to apical fourth; color [eA] cecsceRerceckeunbe Pe GeS-er acaeeccaen ce USO Are rae asc te paneer cocECeREMURESE aS 3 2b. Elytra widest at middle or basal third; color black haemorrhagica 3a. Post, median, lateral and apical dots present ............ 2 ID PRIOR RE Ren ee ee ae ere sedecimpunctata 3b. Post, median, lateral and apical dots absent ....ocellata 4a. Front trochanters only with seta (small species, less aetna, G3 Inenn) (ROG) coetee cto coeceoscecet tebe ccaceeeeeren eran ceeeeEceoceeed nec 5 4b. Front and middle trochanters with seta .................... 6 5a. Elytra bright red, shiny, deeply punctate; markings consisting of a longitudinal line .................... lemniscata 5b. Elytra dull grayish brown with a longitudinal row of green foveae; middle band and apical lunules present Blea eater) PEE EES Se eee eae eee rrere ren ere viridisticta 6a. Small species, less than 9 mm long (velvety green in color, elytral maculation consisting of a wide partially submarginal border) ................--.-...------ debilis 6b. Larger species, 10 mm long or longer ................--..--.- 7 7a. Genae covered with decumbent hairs .................2....-.-- 8 IMMA Clic POR ARG Mesto ittees cee ccaceseec eee ceree scesuere wcrenneecnsesteccenersct nae 9 8a. Antennal scape with subapical sensory hairs only; bright green, occasionally red (elytra with feeble SVanieelt Gpollats)) aecceescocoreececucecceasoreo-c- pees eneeceercReencec ance marutha 8b. Antennal scape with decumbent hairs below the sub- apical setae; brown or bronze with greenish reflec- HONS ons cc cecec cs fara sone cee ter aseene errno eee etecenedesesecteencs nevadica 9a. Frons or vertex to some extent pilose ..........0....0..00--- 10 9b. Frons and vertex bare except for ocular setae ............ 12 10a. Pilosity of vertex sparse, frons bare; color dark green or brown; (maculation complete and joined Ate MVANBIN) | sssecesees-aceecece ceases cecsevonteccontorenecscteeeteeres willistoni 10b. Pilosity of frons and vertex medium to dense; color yrs ren ie ees ae noe re Ae ee 11 PROC. CALIF. ACAD. SCI., 4TH SER., VOL. 41, NO. 4 41a. Green or blue-green; immaculate or with very small apical dots, rarely any median dots ................ pimeriana 11b. Red; maculation marginal consisting of humeral, median, and apical spots, sometimes connected; (Keren Si Ze) ett cs cvcessceseese reese stcecseceeceeotese meneeeneee pulchra 12a. Under surface sparsely to densely pilose laterally......13 12b. Under surface bare except for a few scattered hairs, especially on sides of metasternum ...............-.....+ horni 13a. Proepisterna densely clothed with long white decum- bent hairs; maculation complete .................. praetextata 13b. Proepisterna bare or sparsely clothed with hairs ...... 14 14a. Elytra with subsutural row of blue foveae from base fO!) AD OX cece. ccnceccedncnc-caeceenss-cessnesse