m D1 ' ' i. q Z SMITHSONIAN"" INSTITUTION Z NOlinillSNI £ C — ^ Z 3 .'SzStfo' . Z! yf53*oX O NviNOSHiiiNS saiMvasn > 2 LIBRARIES SMITHSONIAN INSTITUTION NQIlf 2 m ^ m Xw>' ~ xMy' m ^ m TO ~ £ TO £ TO z y, NVINOSHIIIAIS S3JBVB0I1 LIBRARIES SMITHSONIAN INSTITUTION NOlinillSNI NVINOSHIIWS S3 lava 8 n z o in X h s > 2 in SMITHSONIAN INSTITUTION ■? = to i A00fo.. fk | NVINOSHillAIS^SBIBVHan UBRAR I ES^SMITHSONIAN 1/5 ^ 5 to * > x to Q 2 >' S INSTITUTION NOlif to SMITHSONIAN £ ___ 2 INSTITUTION 05 PB > rm to Ns,$i!JE£«/ £2 ntojwssJ/ to *2 xJOmsS^ TO to £ to £ ^ w \ £ to NouniiiSNi NviNOSHims saiavaan libraries Smithsonian institution noiii £0 2 CO 2 tO 2 W X CO o z i£ NVINQSHilWS 5 ”V > S tn *• 2 so S3IBVH8I1 LIBRARIES SMITHSONIAN to X to WJi ,n w TO 3 TO > ’9T- S ^ > Z to * 2 ^ INSTITUTION ^NOlinillSNI^ « =! « < ^fSo^5x O _ \\ P “ /laSx O o ‘institution 2 NOIII 2 o > NviNOSHiiws $3 1 eve an TO g XjVAS>sy TO II BRAR I ESWSMITHS0N!AN~INSTSTUTI0NW 2 to to p mrmi- x > ^ I” TO " INSTITUTION m NOlinillSNI TO W jfrostog NOIinillSNS OB 73 '<> -, -x 2 o CO X * V o 4 2 NVINOSHillNS n i ^SMITHSONIAN"' O |sj 1 v g 2 m.4 H to to i wi NVSNOSHIliNS^Sl S HVB S Slf LI B RAR I ES^SMITHSONIAN ^INSTITUTION NOIII ® 5 Ss^y** | 1 INSTITUTION N0II01IlSNl"J NVINOSHIIWS ' S3 1 ava ill LIBRARIES SMITHSONIAN 2 5® o — <»° w 73> '> TO LIBR SNST8TUII0N NOlinillSNI NVINOSHIIWS TO ^§S4£fe^ g TO ^ g TO ' S3iMVa8!iriieRARf£SWSMITH$0NlAN~!NSTITUTl0N NOIII ,i j&- I ^assv S ^S'. s yJITHSQNlAW INS1 1 1 U I IUN NOlinJLIXSNI NtflNUSHXIWIS S3iaVaan LI B KAR I £§ SMITHSONIAN INSIIIUIIUN NOI1MJLIJ *■“ (C/5J <“=■ t?r* ■=* it my “ cfi sr m/ 5 k h “ 0 Xl* o vc 2 o Xg'M. ocX ® ' CO ^ Z \ W < >T53T^v I < V 5 < V 2 s 2 I | M^sk § - &kt»K°A g 0 -g, CO 1 §■• - '■ § S%as^ §• * -sf > s - y- > VIN0SH1IWSOTS3 I HVU 8 nZU BRAR I ES^SMSTHSONIAN^INSTITUTION NOlinillSNI NVINOSHXWS^Sa I MVM 0 ll^LI B RAR to = tn ” to = > to to 0£ . < EC o “ XJVASBy O ' 03 J 2 — «J 2 =J 2 MITHSQNIAM INSTITUTION NOlinillSNI NVIN0SH1JWS S3IHVH8I1 LIBRARIES SMITHSONIAN INSTITUTION NOIlfllU 2 r5 ... z r* z r- _ z VIN0SH1IWS S38HVH0I1 LIBRARIES SMITHSONIAN INSTITUTION NOlifUllSNI NVINQSH1ISAIS S3IHVHail LIBRAR ^ 12 z w z to z ^ .. to 2 2 g P ..5 5 . -< > 2 ^ > NgUiisG^ S 'w,& > — 2 \4ITHS0NIAN_ INSTITUTION NOlinillSNI NVIN0SH1IWS^S3 I HVU S H^LI B RAR I ES^SMSTHSONIAN INSTITUTION NOilflll' so — m = to 5 to „ w XfSiwx f-, w 7k. w . to ^ to o -s “ 5 2 ^ o WM0SHiJWS__$3 S M¥H 0 11 LI B RAR I ES^SMITHSONIAN^INSTITUTION^OlinillSNI^NVINOSHimS S3 I BVB 3 11 LIBRAR m 73 'i> ■ 2 ^ m NZ»o$>\sy ^ m MITHSONIAN INSTITUTI0N%0linitlSNI~NVIN0SHllWSWS3 1 M VH a 11 " Z \ CO Z CO z 2 i I s /sotx I 2 z -f WK'/i>,. z HI z 03 > p 30 X o s v.^ >’ s xaypt«^y > w VINOSHXIWS^SB I eva a n2LI BEAR I ES ^SMITHSONIAN ^INSTITUTION "" N0liniSlSNl"NVIN0SHlllMS"S3 I B VB a n"*LI B R AR ^ m — « V C/) ~ __ m x^ost^i B RAR I ES SMSTHSON!AN”lNST!TUTtON%OIinilJ ^ z \ co z > z O “ ’ o ” XJVA5>^ o . s -J Z -J 2 SMITHSONIAN INSTITUTION NOlinillSNI NVINOSHiSWS S31HVHSI1 LIBRARIES SMITHSONIAN INSTITUTION NQUmi. “ Z r- ... z 5” z r- 2 jts |tj co rn Xg'v pcy X rn CO ™ ~ \ Z (/j ™ to IVIN0SH11WS S3 IHVa 8IT LIBRARIES SMITHSONIAN INSTITUTION NOlinillSNI NVIN0SH1IINS S3!HVHail LIBRAR ^ 2 ,,, W z CO z ^ ^ CO z ... ;■ — ,-^SVvt § 2 < 5 A NEW SPECIES AND SUBSPECIES OF BAT OF THE HIPPOS I DEROS BICOLOR-GRQUP FROM PAPUA NEW GUINEA, AND THE SYSTEMATIC STATUS OF HIPPOSIDEROS CALCARATUS AND HIPPOSIDEROS CUPIDUS (MAMMALIA: CHIROPTERA: HIPPOSIDERIDAE)1 James Dale Smith2 and J. Edwards HUP Natural History Museum of Los Angeles County * M:' . i? 900 Exposition Boulevard » Los Angeles, California 90007 1 at irregular intervals in three major series; the articles in each series are numbered individually, and numbers run consecutively, regardless of subject matter: • Contributions in Science, a miscellaneous series of technical papers describing original research in the life and earth sciences. • Science Bulletins, a miscellaneous series of monographs describing original research in the life and earth sciences. This series was discontinued in 1978 with the issue of Numbers 29 and 30; monographs are now published by the Museum in the Contributions in Science series. • Science Series, long articles on natural history topics, generally written for the layman. Copies of the publications in these series are sold through the Museum Bookshop. A NEW SPECIES AND SUBSPECIES OF BAT OF THE HIPPOSIDEROS B I COLOR-GROUP FROM PAPUA NEW GUINEA, AND THE SYSTEMATIC STATUS OF HIPPOSIDEROS CALCARATUS AND HIPPOSIDEROS CUPIDUS (MAMMALIA: CHIROPTERA: HIPPOSIDERIDAE)' James Dale Smith1 2 and J. Edwards Hill3 ABSTRACT. A new species of bat of the Hipposideros bicolor- group is described from New Guinea and the Bismarck Archipelago. The de- scription of this new taxon was prompted by the study of recently ac- quired specimens from the Bismarck Islands of New Britain and New Ireland. In addition, this new material has allowed for the clarification of a long-standing problem concerning Hipposideros calcaralus Dob- son 1877 and Hipposideros cupidus Andersen 1918. We find these lat- ter two species to be conspecific and referable to Hipposideros cal- caratus (by priority) and distinguishable from the new taxon on the basis of numerous qualitative and quantitative features. The geographic variation of the new taxon and Hipposideros cal- caratus was assessed by computing principal component and discrimi- nation analyses for 11 cranial and 10 wing variables. Sufficient geo- graphic variation was found to merit recognition of geographic races in both H. calcaratus and the new species. In the former, populations from the Bismarck and Solomon Islands, as well as the island groups off the southeastern coast of New Guinea, are assigned to Hipposideros calcaratus calcaratus and populations of this species from New Guinea are assigned to Hipposideros calcaratus cupidus (new combination). The nominate race of the new species is restricted to the Bismarck Is- lands, and a new subspecies is described to accommodate populations of the new species from New Guinea. INTRODUCTION For many years, two medium-sized species of the Hipposideros bicolor- group have been thought to coexist in New Guinea, the Bismarck Archipelago, and the Solomon Islands. One of these, Hipposideros calcaratus, was described from Duke of York Is- land in the Bismarcks by G.E. Dobson in 1877; the other, Hip- posideros cupidus, was described by Knud Andersen in 1918 from Papua (New Guinea). Andersen’s description of H. cupid- us (issued on his behalf by Oldfield Thomas) was brief and ba- sically defined it as being smaller than H. calcaratus. In his comprehensive review of the Indo-Australian species of Hipposideros, Tate (1941) introduced the notion of apparent sympatry of these two species in New Guinea and provided di- agnoses with additional characteristics to distinguish them. Tate’s assessment and conclusions regarding these two species were based primarily on collections in the American Museum of Natural History. Sanborn and Beecher (1947:390) first reported H. cupidus from the Solomon Islands; identification of this material was credited to Tate. In first recording the occurrence of H. cal- caratus in the Solomons (Rennell Island), Hill (1956) noted that, in this region, Tate’s criteria of size and height of the pos- Contributions in Science, Number 331, pp. 1-19 Natural History Museum of Los Angeles County, 1981 ISSN 0459-0113 terior canine cusp apparently were not reliable diagnostic fea- tures. Nevertheless, in his monographic treatment of the genus Hipposideros, Hill ( 1963) adopted most of Tate’s other diagnos- tic characteristics to distinguish between these two species. At the time. Hill’s assessment was limited to specimens in the col- lections of the British Museum (Natural History), which in- cluded the type material of H. cupidus and H calcaratus and some specimens from the Solomons. Since then, additional ma- terial has been obtained from the Solomons (Hill 1971, and BMNH) and Papua New Guinea (McKean 1972). This mate- rial was reported and identified primarily on the basis of Tate’s diagnostic features. In the summer of 1979, the Taylor South Seas Expedition from the Natural History Museum of Los Angeles County, led by one of us (Smith), conducted extensive field investigations on the bat fauna of New Ireland and New Britain Islands (Bismarck Archipelago). In the process of identifying the spe- cies of Hipposideros in this collection, it became apparent that there were two similar taxa that differed mainly in size; one small, presumed to be H. cupidus, and the other large, pre- sumed to be H. calcaratus. Subsequent examination and com- parisons with other material, including the holotypes of both species, clearly demonstrated that Tate (1941) wrongly as- signed specimens of his larger species from New Guinea and the Bismarcks to H. calcaratu's, which does not possess the sphenoidal characteristics he ascribed to it. Furthermore, it is even more apparent that Tate incorrectly identified specimens of H. calcaratus from the Solomons as H. cupidus. After considerable study, we have found that Tate’s sphenoi- dal character and other features that he employed to dis- tinguish H. calcaratus from H. cupidus do not, in fact, differ- entiate between these two species. His diagnostic features do, however, distinguish a generally larger species from a generally smaller one. The holotypes of calcaratus and cupidus both cor- respond with the smaller of these two species. We propose that. 1. Review committee for this contribution: G.B. Corbet, J. Knox Jones, Jr., Donald R. Patten, and Andrew Starrett. 2. Research Associate, Mammalogy Section, Natural History Mu- seum of Los Angeles County, Los Angeles, California 90007 and De- partment of Biology, California State University, Fullerton, California 92634. 3. Department of Zoology, British Museum (Natural History), Cromwell Road, London SW7 5BD, England. in the absence of compelling, qualitative characteristics to the contrary, there are no bases for maintaining Hipposideros cal- caratus and H. cupidus as separate and distinct species. The larger species, referred by Tate (1941) and others to Hippo- sideros calcaratus, represents an undescribed species. The geo- graphic variation and nomenclature of these taxa are consid- ered below. METHODS Five hundred and fifty-six specimens of Hipposideros cal- caratus (including cupidus) and the new species were examined in the preparation of this study. Lists of referred specimens are given under each subspecific account. Institutional abbrevia- tions are as follows: American Museum of Natural History, New York (AMNH); Bernice P. Bishop Museum, Honolulu (BBM); British Museum (Natural History), London (BMNH); Universitetets Zoologiske Museum, Copenhagen (CN); Field Museum of Natural History, Chicago (FMNH); Natural His- tory Museum of Los Angeles County, Los Angeles (LACM); Museum of Vertebrate Zoology, University of California, Ber- keley (MVZ); Natur-Museum Senckenberg, Frankfurt (SMF); United States National Museum, Washington (USNM); Mu- seum Fur Naturkunde der Humboldt-Universitat zu Berlin, Berlin (ZMB). In these lists of referred specimens, localities are reported within provincial boundaries or islands (arranged alphabetically) and within these, localities are listed from north to south, the westernmost listed first in cases where two lo- calities lie at or near the same latitude. As indicated in the lists, some localities were not plotted on the distribution maps be- cause undue crowding of symbols would have resulted. Also, localities where sympatry was noted between H. calcaratus and the new species described in this paper are noted. Specific and subspecific variation were assessed on the basis of qualitative features and quantitative measurements taken from the skull and wing. Eleven cranial measurements in- cluded: zygomatic breadth; breadth of the braincase; mastoid breadth; breadth of the interorbital constriction; condylocanine length; breadth across the canines; length of the maxillary toothrow; breadth across the last upper molars (M3/); length of the palatal bridge; length of the mandibular toothrow; and length of the mandible. Ten alar measurements included: length of the forearm; length of the third, fourth, and fifth metacar- pals; length of the proximal phalanx of the third, fourth, and fifth digits; and length of the distal phalanx of the third, fourth, and fifth digits (see Smith and Starrett 1979). These measure- ments were taken with dial calipers and were recorded to the nearest tenth of a millimeter. All measurements given in the text are in millimeters, and dental notations follow Miller (1907). Statistical analyses of the data were carried out at the com- puter center of the University of Southern California, Los An- geles, using an unpublished program of discriminant analysis (D1SANAL) written by Richard A. Pimentel (Department of Biology, California Polytechnic University, San Luis Obispo). This multivariate program assesses the morphometric variation among variables by computing principal components (PCA) and discriminant analyses with classification and matrices of generalized distance. For a discussion of the program and anal- yses, see Pimentel (1979) and Smith and Starrett (1979). SYSTEMATICS Order Chiroptera Blumenbach 1779 Suborder Microchiroptera Dobson 1875 Family Hipposideridae Miller 1907 Hipposideros calcaratus (Dobson 1877) Spurred leaf-nosed bat Figures 1-4, 7-8 and Table 1 (Synonymy under subspecies) DISTRIBUTION. Papua New Guinea (Bismarck Archipelago and mainland); West Irian; and Solomon Islands (Fig. 4). DIAGNOSIS. Size moderate (mean length of forearm, 51.74, range 55.8-46.2; mean condylocanine length, 18.16, range 19.5-15.9). Calcar relatively short, approximately one-half length of tibia. Tail with five equal to subequal tail vertebrae; fifth tail vertebra not especially shortened. Soft palate with nine interdental ridges, area between ridges divided into nu- merous small cells giving palate a honeycombed appearance. Bony palate with shallow depression on each side adjacent to lingual corner of M2/ and M3/. Mastoid process moderately developed and flangelike. Upper canine moderately large, not bulbous at base; posterior supplementary cusp well developed, but small, extending from base one third or less along length of tooth. First upper premolar (P2/) small, not crowded out of toothrow; canine and second upper premolar (P4/) not in con- tact. Hypocone on Ml / and M2/ distinct. Lower incisors trifid, equal or subequal in size, not overly crowded between lingual cingula of lower canines. Mandibular symphysis moderately broad. Lower molars relatively broad; trigonid and talonid of M/1 and M/2 equal to subequal in size. Hypoconid, hypo- conulid, and entoconid forming distinct cusps on M/ 1 -3. Coro- noid process moderately low; posterior margin markedly con- cave. Angular process long, slender, and deflected laterally. Sphenoid bridge relatively narrow and emarginated, not con- cealing large, round optic foramina. Basisphenoid depression shallow and troughlike. DESCRIPTION. Face (Fig. 1) moderately long, narrow, and rising gently to low crown of head. Muzzle, lateral and anterior to noseleaf, with moderate number of vibrissae, which emerge from small, round, and purplish epidermal glands; vibrissae on lower lip less numerous. Noseleaf (Fig. IB) simple, secondary leaflets absent. Lateral narial lappets not well developed, sub- tubular, and squared off above external nares. Intermediate noseleaf simple, medial ridge a large, prominent, conical bump, two lateral ridges smaller and less prominent. Each lateral ridge bears a long vibrissa seated in a purplish epidermal gland similar to those on muzzle and lower lip. Posterior noseleaf rela- tively long, with three strong vertical septa; middorsal portion of upper margin with a tiny, rounded bump. Small, wartlike 2 Contributions in Science, Number 331 Smith and Hill: New Hipposideros from New Guinea tubercle above each small eye and behind dorsolateral corner of posterior noseleaf; several long vibrissae with basal epidermal glands protrude from apex of each tubercle. Frontal sac of males relatively small, horizontal, and not much elevated above posterior noseleaf, which sometimes partially obscures it. Ears Figure 1. Lateral (A) and frontal (B) views of the head and face of Hip- posideros calcaratus. relatively long, broadly subtriangular, and funnel-shaped with rounded tips; 10 to 13 transverse ridges traversing inner surface of each pinna; short, round tragus concealed within ear conch. Antitragus relatively low, round, and moderately convex. Wing membrane attached to side of foot at level of proximal metatar- sus. Calcar short, approximately one-half length of tibia. Tail with five vertebrae, equal to subequal in length; fifth tail ver- tebra not noticeably short or absent. Pelage and coloration. Dorsal surface (including crown and nape) with long, silky pelage; sides of muzzle sparsely haired; cheeks below eye more densely so, with relatively long hairs (Fig. I ). Ventral pelage long, dense, and woolly. Inner surface of ear moderately hairy, inner margin and antitragus more strong- ly so (Fig. 1). Dorsal pelage dark brown, hairs unicolored, a slight paler “V” over the shoulders; ventral pelage grayish white; wing membranes naked, blackish brown. The pelage is occasionally bleached with a reddish-orange tinge. Soft palate. Nine interdental ridges traverse soft palate (Fig. 2). First ridge extends from posterior border of canine forward to a sharp point, then abruptly rearward to form a deep notch; it is fused on midline. Second ridge originates at anterior border of second premolar; middle portion of this ridge is roundly bowed anteriorly. Third ridge extends from near pro- tocone of second premolar and is gently bowed anteriorly. Fourth, fifth, sixth, and seventh ridges extend from sides of first molar; these are nearly straight with slight anterior bow near midline. Eighth ridge extends in a similar fashion from between first and second molars. Ninth ridge originates at anterior mar- gin of second molar and bows gently rearward to fuse on mid- line. Numerous cells between palatal ridges give soft palate a distinct and characteristic honeycombed appearance (may be less pronounced or absent in young individuals). There are no interdental ridges behind level of second molar; this portion of palate is flat and granular. Cranium. Skull (Fig. 7 C,D,G,H) teardrop-shaped (viewed dorsally), lightly built, and generally similar to those of other unspecialized members of the bicolor- group. Rostrum long, nar- row, subtubular, only slightly broader than interorbital constric- tion. Anterior nasal tubercles slightly inflated, but not exagger- ated above nasal profile (Fig. 8 C,D). Top of rostrum flat and lacking pits or depressions. Braincase slightly oblong (viewed dorsally), low and elongate in profile, not abruptly vaulted above facial plane. Some individuals with low sagittal crest on frontals; crest may be weak or absent on parietals; lambdoid crest weak. Infraorbital foramen relatively short and slitlike. Zygomatic arches moderately strong, not markedly arched dor- sally; malar flange well developed. Mastoid process well devel- oped and fiangelike. Premaxillaries long; anterior palatal emar- gination terminating posteriorly on or near line joining anterior margins of first molars. Palate moderate, its posterior emargina- tion near a line joining rear faces of last molars, flat, with a shallow depression on each side adjacent to lingual corner of M2/ and M3/. Pterygoid wings relatively long and hooked rearward. Sphenoidal bridge narrow and noticeably emarginate (Fig. 7 G,F1). Optic foramina large and rounded. When ventral aspect of skull is viewed, optic foramina extend laterally well beyond narrow sphenoidal bridge. [Tate (1941:362, 364) em- Contributions in Science, Number 331 Smith and Hill: New Hipposideros from New Guinea 3 Table 1. Means and ranges (in parentheses) of selected cranial and external measurements of Hipposideros calcaratus. Letters in parentheses following locality names coincide with centroids plotted on the canonical graphs in Fig. 10. Superscript numbers indicate sample size different from those given in left- hand column. 2 co 0 q q to CO O q q 00 04 q 04 O' P3 ON d ON ON ON 00 04 — ON — d NO q od d d q ON ** oo O’ NO 10 to NO NO O’ co to CO to to O’ q o- NO O’ O’ O’ oi to 1 tO 04 0 1 01 d to 1 04 (N to 1 OO C4 to 1 04 NO q CO to l 04 to 1 q 00 O’ 1 to ON | O’ q d O’ l 04 ON O 1 CO ON O’ 1 r- o d to to CO O’ to 04 ON od d Lu NO 0 to to to to to NO O’ to O to to w w w w ' — ' w w w w w w _ s s , y , N s ^ v ci NO 00 OO q q oc q OO NO °o NO q — < d CO CO CO CO O’ CO 04 _ 04 04 04 d H ON q co v— 1 •— 1 q — > CO --* q — q r- v-^ NO . — 1 q Z d 1 co O’ 1 q 1 04 1 q l q 'T 1 q O’ 1 q O’ 1 q co 1 q CO 1 to 04 q 04 1 q d 1 q UJ Q O’ ■d "7f O’ O’ O’ CO CO d 04 d w w w w w w w w w w oa NO 00 OO rf 00 CO q q NO CO q q to Q Z < ON OO O 00 00 OO q 00 q 00 •O’ OO | q 06 l 00 00 1 °0 r-~ ^ i d — d q od OO 04 ON CO OO q ON CO 00 q OO q 00 OO q r- 04 OO \Q OO q od CO od q ON ON 00 ON d 00 ON 00 00 OO od od od 2 z tu -j O; ro q q 00 q 00 q o q CO q q co NO co NO CO O) CO q CO q CO q CO 1 04 CO | — O’ q co q ^ q d O’ d q d 1 j d O co Jh CO q CO J- q co q O' q O’ q CO q d 00 d q d to d 04 3 O’ co CO q q H X < 5 7.5 d 1 00 7.3 NO q 6'9 NO q r- NO q OO NO sO 1 q 00 d r~- 1 q 00 r- 1 q 8.0 00 1 s d NO 1 q 6.5 8-6 NO d d q 6.5 d 1 q 6.8 d 1 q 3 3 UJ 3 *3 z Q Jk 00 q CO 04 co r» q q q 3 00 vO P q q q z ^3 4.3 co I r- 4.5 1 r- 4.5 •rt 1 q 4.5 Tf 1 OO 4.4 d 1 NO q O’ O’ 1 00 4.4 O’ 1 q 04 O’ O’ q Cj O’ CO 1 CO 04 O’ 3.9 d q 3.9 d 1 4.3 O’ 1 NO < 2 rr rf d O’ O’ O’ O’ O’ 0 O’ O’ u 03 ' — ^ ' — ^ ' — ^ ' — ■" ' — ^ w' O ' — ' ' — ' 0 V. CO pu1 73 CO CO CO CO CO CO to q to Q 04 04 q CO o 0 CO NO CO CO q co to CO q CO q CO 1 to CO 1 g. co CO co d CO d CO d 1 q d H §: d r-- CO r- CO Tf co OO CO q CO q CO q CO q .N4. CO q d q d q d to d to Z 5 d CO co CO CO CO CO CO CO d d d d s ’ ' ^ ' ^ ' ^ Q 00 — • q ON C4 ON — 0 q 00 S' q 04 R 0 04 ON I CO d ON to ON C4 d — ON | CO d 04 d q ON 1 ON . 1 I-— ON q ON | CO d C/5 d q d 1 NO d CO d q d 1 CO d NO d q d 1 CO d CO 0 rA ON 1 OO ON O d 1 to < d d 0 d d d d d d ^ d ON d d — w w w' ' — ' w < 0 00 r- q 00 q 04 0 1— i On 00 2 O — — — — — ; — q d d d d d N W w w w w w w w w w w z q —H 00 q ~— C4 q — r- to 04 q co < 0 d — OO r- pt 00 — OO r- q ON NO NO ^ d d 04 to NO d u z _ CO q rf NO 00 „ q OO 1 NO OO 1 to rC 1 06 1 to OO 1 q OO 1 q ON 1 to ON 1 q r~- d 1 co d 1 CO d 1 CO d 1 q o u OO 00 00 ON 06 00 ON ON OO d d d d w w W w z - 20 u c cd 33 C/3 ”0 co -TD C NO O’ to 7 r- O' «o to c cd "O T3 "5 < Cd C -O £ <0 z .c CQ -o c C3 -0 c — C/3 a CO s ~o c 0 £ cd 5 5 0 5 w cd J OX) c cd *TD u cd 3 JO C/3 c cd c a cd > cd o _1 2 O C/D d f2 C3 UJ a 2 UJ cd C/D U* 2 a 0 Qi X cd *“3 Q 2 cd •—3 2 cd -u Z Q z* < ^ U 2 S 5 — ■ 03 Cl •- ’S u oi) -C c *-* q. O' N < o > os < o O Sis 6*1 • - 03 03 ■5 5 E >.5 < •a ofi f— c c 7 8 — UJ z x 9. < p i U H o 0 C o I* O o6vd'0'Z)'--‘Z),t,to6a\,t OnM00On00O\'O00O>O'On 'OOOOO^O-'-OOOOOOO \G0G'S~)i/~)-rfO\cn\O,*f0C O\\do6(ridd,to6^- OsOsOsOsu’'iOsOssOOOvr> 03 u f— E £ u ^O — O — O — OO iti - 'O ' — < — < — — i/nrN — O — ^000^000^-000 'tMsO’tfsnoo-aooo sotj-socntj-o — vnoocoso -OO)OM0CMMMfN- 'OO'OO^ONOOTiON-OOOO h (N O' - rn OS ^ r" vnoo — O — OOmrfO VO VO — sd -'>000-fNls-'0-fN(N rn — OOOSfN — O^fNsO n-vOOOfO'tM^- — ^ rn i/S oo O l/~>SO fN- rnOOsr-^Os — t}- — r"OsTj-r<->Tj-Osm oo — m — o — — ©or^a'0‘ocv'0 — r' izirnio\o,odr'r^rn ir> O n M O (N - O ^ - mOsOO'O^tO^OO'tO 00(NXrt'Zina>^»Zlff) — (N — © Tf — — rS Tfr KD Uu i 0-4 i — fvq 1 — - N 16 Contributions in Science, Number 331 Smith and Hill: New Hipposideros from New Guinea CANINE, breadth across the canines; MAXTOH, length of maxillary toothrow; PALWID, breadth across the last upper molars (M3/); PALLEN, length of the palatal bridge; MANDIB, length of the mandibular toothrow; DENTAR, length of the mandible; CONCAN, condylocanine length; FORARM, length of the forearm; META-3, META-4, META-5, length of the third, fourth, and fifth metacarpal, respectively; DIG-31, DIG-41, DIG-51, length of the proximal phalanx of the third, fourth, and fifth digits, respectively; DIG-32, DIG-42, DIG-52, length of the distal phalanx of the third, fourth, and fifth digits, respectively. Table 3. Means and ranges (in parentheses) of selected cranial and external measurements of Hipposideros maggietaylorae. Letters in parentheses following locality name coincide with centroids plotted on the canonical graphs in Fig. 10. Superscript numbers indicate sample size different from those given in left-hand column. 2 O NO ON q Op 04 d ON q d d 0? r-J to. q O' to oo to O' oo o oo <0 ON to to to Sd oo Tf oo d q ■cT rf O oo OS ON 04 to z 2 to to r-i to O' bo <3 04 ■cr o — ^r 00 to CO 04 d" 2 — to ■3; d. •~ NO NO to to •o «o rf to to toi to to u to O CQ 5 , v Is. s y , y s , . _ v s y s c* c to 1 q q to o q CO >- N to to 04 04 04 04 cd cd cd 04 cd W w Z NO q OO ON as ON • — to 'Cj- q Os < u 04 CNj fN 04 04 o 04 NO OO 04 oo 04 OO Os OO d ON oo O 04 oo ON q oo — ON z 04 04 1 to rJ 04 1 to 04 1 ON ON 1 04 ON T O' ON 1 •q OO 1 CO o 04 CO d 04 1 CO ON T 04 ON 7 04 o 04 1 CO o on 04 — d On ON — o ON o o Ol 04 04 04 - — — 04 04 04 04 z to to C d co o NO o c ON nO NO to 7 CO >v | 03 C >v d CQ | <3 3 a. 3 03 S3 jd w 03 £ “O 3 C >. o -J 03 2 3 O C/3 w C/5 73 UJ o s ii > d —5 ~3 CO 2. 2 03 d P d in is 3 Q 3 on _ C J— o _ w 3 '£ j? 12- 8| a -- o •S JS 2 60 U C ,E- JiZ -Stu <. i o _] *o “s a£ “• O S2 f— 03 51 •s s -O Cl. T3 J2 ■ o o S 2-5 « 3 3J E Si? 9 O « ■£ ^ a -5 2 s; » 3 — "5 3 — X x> — 1 < o E ^ oSf N W 3 CE <2 <— = o O'Sx ^ C M N e c .. u " j= _c . 5 °(^ ,„-£z C OX) JjJ Sq 03 — ■ , , li! <. ^ zs « < . . 3 U SrE z.zn O = « U 2 E Contributions in Science, Number 331 Smith and Hill: New Hipposideros from New Guinea 17 130 140 150 160 Figure 11. Geographic distribution of Hipposideros maggietaylorae. See lists of referred material for key to plotted localities. erroneously to H. calcaratus, actually represented on New Guinea by the much smaller H. calcaratus cupidus. ACKNOWLEDGMENTS We are deeply indebted to the following persons and institu- tions for making available the material examined by us: K.F. Koopman, American Museum of Natural History [we are es- pecially grateful to Dr. Koopman, who quite apart from making the collection in New York freely available for study, has ad- vanced a variety of insights relevant to the views put forward in this paper]; A. Ziegler, Bernice P. Bishop Museum, Honolulu; H.J. Baagyle, Universitetets Zoologiske Museum, Copenhagen; P. Freeman, Field Museum of Natural History, Chicago; W.Z. Lidicker, Jr., Museum of Vertebrate Zoology, University of Cal- ifornia, Berkeley; H. Felten and D. Kock, Natur-Museum Senckenberg, Frankfurt; D.E. Wilson, United States National Museum, Washington; and H. Hackethal, Museum fiir Natur- kunde der Humboldt Universitat zu Berlin, Berlin, DDR. In addition, we wish to thank M. Raga (First Assistant Direc- tor), J. Wilmont, and F. Kimbag, Division of Wildlife, Depart- ment of Lands and Environment, Papua New Guinea, for granting permission to conduct field research and providing val- uable assistance to the Natural History Museum of Los An- geles County, 1979 Taylor South Seas Expedition to the Bis- marck Archipelago. Drs. L. Hill and J. Pernetta, Department of Biology, and M.A. Ghani, Department of Geology, Univer- sity of Papua New Guinea provided institutional affiliation for our field project and valuable assistance during our stay in Pa- pua New Guinea. Dr. J.I. Menzies, (formerly. Department of Biology, University of Papua New Guinea) provided valuable information and contacts concerning field investigations. The following persons (listed alphabetically) are only a few, among many, who provided Smith’s party with valuable assistance dur- ing our fieldwork on New Ireland and New Britain islands: Bob Beck; Bernie Gash; Jim Grose; Trevor Hodge; Peter and Anna LeMuth; Bruce Munday and family; Alastair Norrie; Roger Pocock; Sue and Chris Hurley; Graham Row and family; and Robert Seeto (Premier, New Ireland Prov.). Special thanks is extended by one of us (Smith) to his gradu- ate student, trusted field assistant, and tolerant companion Craig S. Hood. Susan E. Smith and Margaret Maas graciously contributed to the preparation of figures and typing of the man- 18 Contributions in Science, Number 331 Smith and Hill: New Hipposideros from New Guinea uscript, respectively. Some financial assistance to aid in prepa- ration of the manuscript was extended by the California State University, Fullerton. Finally, without the early interest, encouragement, and gen- erous support from Mrs. Reese Taylor, the new species de- scribed herein would still be foraging, unrecognized by science, through the green mansions of New Guinea. LITERATURE CITED Andersen, K. 1918. Diagnoses of new bats of the families Rhi- nolophidae and Megadermatidae. Ann. Mag. Nat. Flist., ser. 9, 2:373-384. Dobson, G.E. 1877. On a collection of Chiroptera from Duke- of-York Island and the adjacent parts of New Ireland and New Britain. Proc. Zool. Soc. London, pp. 1 14-127. 1878. Catalogue of the Chiroptera in the collection of the British Museum. British Museum, London, xlii 4- 567. Hill, J.E. 1956. The mammals of Rennell Island. Pp. 73-84, in The natural history of Rennell Island, British Solomon Is- lands, (T. Wolff, ed.), Danish Science Press, Ltd., Copen- hagen, 1 : 1 -288. 1963. A revision of the genus Hipposideros. Bull. British Mus. (Nat. Hist.), Zool. 1 1 : 1 -1 29. 1968. Notes on mammals from the islands of Rennell and Bellona. Pp. 53-60, in The natural history of Rennell Island, British Solomon Islands, (T. Wolff, ed.), Danish Science Press, Ltd., Copenhagen, 5:1 179. 1971. Bats from the Solomon Islands. Jour. Nat. Hist., 5:573-581. Jentink, F.A. 1888. Catalogue systematique de mammiferes (Rongeurs, Insectivores, Cheiropteres, Edentes et Mar- supiaux). Museum d’Historie Naturelle de Pays Bas., Leiden 12:1-280. 1906. Note XIV. On the New-Guinea mammals. Notes Leiden Mus. 28:161-212. 1908. Mammals collected by the member of the Hum- boldt Bay- and the Merauke River-Expeditions. Nova Guinea 5:361-374. Koopman, K.F. 1979. Zoogeography of mammals from islands off the northeastern coast of New Guinea. Amer. Mus. Novit. 2690:1-17. Laurie, E.M.O., and J.E. Hill. 1954. List of land mammals of New Guinea, Celebes and adjacent islands, 1758-1952. British Mus. (Nat. Hist.), London, vi + 175. McKean, J.L. 1972. Notes on some collections of bats (order Chiroptera) from Papua-New Guinea and Bougainville Is- land. Tech. Pap. Div. Wildlife Res. C.S.I.R.O., Australia 26:1-35. Miller, G.S., Jr. 1907. The families and genera of bats. Bull. U. S. Nat. Mus. 57:xvii + 282. Peters, W., and G. Doria. 1880. Enumerazione dei Mammiferi raccolti da O. Beccari, L.M. D’Albertis ed A. A. Bruijn nella Nuova Guinea propriamente detta. Annali del Mu- seo Civico di Storie Naturale di Genova 16:665-706. Pimentel, R.A. 1979. Morphometries: multivariate analysis of biological data. Kendall/Hunt Press, Dubuque, Iowa, x + 276. Sanborn, C.C. 1931. Bats from Polynesia, Melanesia, and Ma- laysia. Publ. Field Mus. Nat. Hist., Zool. Ser. 18:7-29. Sanborn, C.C., and W. J. Beecher. 1947. Bats from the Sol- omon Islands. Jour. Mamm. 28:387-391. Smith, J.D., and A. Starrett. 1979. Morphometric analysis of chiropteran wings. Pp. 229-316, in Biology of bats of the New World family Phyllostomatidae. Part III (R.J. Baker, J.K. Jones, Jr., and D.C. Carter, eds.). Spec. Publ. Mus., Texas Tech Univ. 16:1-441 Tate, G.H.H. 1941. A review of the genus Hipposideros with special reference to Indo-Australian species. Bull. Amer. Mus. Nat. Hist. 78:353-393. Topal, G. 1975. Bacula of some Old World leaf-nosed bats (Rhinolophidae and Hipposideridae, Chiroptera: Mam- malia). Vertebrata Hungarica 16:21-53. Vestjens, W.J.M., and L.S. Hall. 1977. Stomach contents of forty-two species of bats from the Australian region. Aus- tralian Wildlife Res. 4:25-35. Accepted for publication 7 October 1980. Contributions in Science, Number 331 Smith and Hill: New Hipposideros from New Guinea 19 p‘. Q 11 L52X NH Number 332 20 February 1981 tip rl*' Re. •’ CERATIOID ANGLERFISHES OF THE FAMILY GIGANTACTINIDAE: MORPHOLOGY, SYSTEMATICS, AND DISTRIBUTION E. Bertelsen, Theodore W. Pietsch, and Robert J. Lavenberg jam • -B-* ^ : i'- Natural History Museum of l.us Angeles County' • 900 Exposition Boulevard * Los Angeles, California 90007 m f series. » Science Series, long articles on natural history topics, generally written for the layman. Copies of the publications in these series are sold through the Museum Bookshop. CERATIOID ANGLERFISHES OF THE FAMILY GIGANTACTINIDAE: MORPHOLOGY, SYSTEM ARCS, AND DISTRIBUTION E. Bertelsen, Theodore W. Pietsch, and Robert J. Lavenberg Contributions in Science, Number 332 Natural History Museum of Los Angeles County 20 February 1981 Title page illustration: Gigantactis vanhoeffeni , from Chun 1903. ISSN 0459-0113 Natural History Museum of Los Angeles County 900 Exposition Boulevard Los Angeles, California 90007 CONTENTS GLOSSARY v ABSTRACT 1 INTRODUCTION 1 METHODS AND MATERIALS 2 GENERAL DESCRIPTION 2 Body Shape 3 Illicium and Esca 3 Fin Rays 5 Dentition 5 Gigantactis 5 Rhynchactis 7 Gigantactinid Males 7 Sensory Structures 7 Eyes 7 Olfactory Organs 7 Acoustico-Lateralis System 8 Oral Glands of Rhynchactis Females 10 OSTEOLOGY 10 Cranium 10 Mandibular Arch 10 Palatine and Hyoid Arches 11 Opercular Apparatus 11 Branchial Arches II Vertebrae and Caudal Skeleton 12 Illicial Apparatus 14 Pectoral Girdle, Pectoral Fin, and Pelvic Bones 14 Skin Spines 14 MYOLOGY OF THE FEEDING MECHANISM 14 Cheek Muscles 14 Lower Jaw 17 Floor of Mouth 17 Upper Pharyngeals 17 Illicial Musculature 18 FUNCTIONAL MORPHOLOGY OF THE FEMALE FEEDING MECHANISM 18 Gigantactis 18 Rhynchactis 22 REPRODUCTION 22 SYSTEMATICS 22 Family Gigantactinidae Boulenger 1904a 22 Keys to the Genera of the Gigantactinidae 23 Females 23 Males 23 Larvae 24 iii Synopsis of Gigantactinid Classification 24 Genus Gigantactis Brauer 24 Keys to the Species of the Genus Gigantactis 24 Adolescent and Adult Females 24 Males 25 Larvae 25 Gigantactis longicirra Waterman 26 Gigantactis kreffti new species 29 Gigantactis vanhoejfeni Brauer 31 Gigantactis meadi new species 33 Gigantactis gibbsi new species 36 Gigantactis gracilicauda Regan 38 Gigantactis paxtoni new species 39 Gigantactis perlatus Beebe and Crane 41 Gigantactis elsmani new species 43 Gigantactis golovani new species 44 Gigantactis gargantua new species 46 Gigantactis watermani new species 49 Gigantactis herwigi new species 49 Gigantactis macronema Regan 50 Gigantactis savagei new species 53 Gigantactis microdontis new species 54 Gigantactis ios new species 56 Gigantactis Species Incertae Sedis 57 Gigantactis filibulbosus Fraser-Brunner 57 Gigantactis ovifer Regan and Trewavas 57 Gigantactis Species Unidentified 57 Metamorphosed Females 57 Female Metamorphosal Stages 58 Males 59 Male Group I 59 Male Group II 60 Male Group III 60 Male Group IV 61 Male Group V 61 Males Not Referred to Group 61 Discussion 61 Larvae 62 Larval Group A 63 Larval Group B 63 Larval Group C 63 Larval Group D 63 Genus Rhynchactis Regan 63 Rhynchactis leptonema Regan 66 DISTRIBUTION 67 Cosmopolitan Forms 68 Atlantic and Eastern Pacific Forms 69 Tropical Atlantic Forms 69 Indo-West Pacific Forms 69 Eastern Pacific Forms 69 Southern Ocean Endemic 72 Comments 72 EVOLUTIONARY RELATIONSHIPS 72 ACKNOWLEDGMENTS 72 LITERATURE CITED 72 iv GLOSSARY ANATOMICAL TERMS A2 Adductor mandibulae section A2 L.1NT. Levatores interni muscle a3 Adductor mandibulae section A3 L.O. Levator operculi muscle An Angular Mp Metapterygoid Ar Articular Mx Maxilla Aw Adductor mandibulae section Aw O Opercle Bo Basioccipital OBL.D. Obliqui dorsales muscle Cbr Ceratobranchial P Parietal Ch Ceratohyal Pa Palatine Cl Cleithrum Pbr Pharyngobranchial Co Coracoid Pc Postcleithrum CR 2nd Cephalic ray Pmx Premaxilla D Dentary Po Prootic DEPR.D. Depressores dorsales muscle Pop Preopercle DO. Dilatator operculi muscle PR.HY. Protractor hyoidei muscle E Epiotic Ps Parasphenoid Ebr Epibranchial Pt Pterotic EC Ethmoid cartilage Ptm Posttemporal Eh Epihyal Ptp Pterygiophore of illicium EREC.D. Erectores dorsales muscle R Radials Ex Exoccipital R.DORS. Retractor dorsalis muscle F Frontal S Sphenotic H Hyomandibular SC Symphysial cartilage Hh Hypohyal Scl Supracleithrum IB Illicial bone SCAR. A. Supracarinalis anterior muscle IMD Intermandibularis muscle Se Supraethmoid INCL.D. ant. Inclinatores dorsales muscle, anterior So Subopercle subdivision STH. Sternohyoideus muscle INCL.D. post. Inclinatores dorsales muscle, posterior Su Supraoccipital subdivision Sym Symplectic lo Interopercle TR.D. Transversi dorsales muscle L.A.P. Levator arcus palatini muscle Q Quadrate LE Lateral ethmoid V Vomer L.EXT. Levatores externi muscle INSTITUTIONS AMS Australian Museum, Sydney. GMZA Goteborgs Musei Zoologiska Avdelning, Sweden. BMNH British Museum (Natural History), London. IOAN Institute of Oceanography, Academy of Sciences of BOC Bingham Oceanographic Collections, Yale Univer- the USSR, Moscow. sity. IOS Institute of Oceanographic Sciences, Surrey, En- CAS-SU Stanford University collections now housed at the California Academy of Sciences, San Francisco. gland (formerly the National Institute of Oceanography). ISH LACM MCZ NYZS SAM SIO UMML Institut fur Seefischerei, Hamburg. Natural History Museum of Los Angeles County. Museum of Comparative Zoology, Harvard Univer- sity. New York Zoological Society. South African Museum, Capetown. Scripps Institution of Oceanography, La Jolla. University of Miami Marine Laboratory. USNM National Museum of Natural History, Washington. ZIAN Zoological Institute, Akademia Nauk USSR, Leningrad. ZIFSUT Zoological Institute, Faculty of Science, University of Tokyo. ZMHU Zoologisches Museum der Humboldt-Universitat, Berlin. ZMUC Zoological Museum, University of Copenhagen. VI CERATIOID ANGLERFISHES OF THE FAMILY GIGANTACTINIDAE: MORPHOLOGY, SYSTEMATICS AND DISTRIBUTION' E. Bertelsen,1 2 Theodore W. Pietsch,3 and Robert J. Lavenberg4 ABSTRACT. The ceratioid anglerfish family Gigantactinidae is revised on the basis of all known material. The two recognized genera of the family, Gigantactis and Rhynchactis , are described and compared os- teologically. A description and discussion of the myology and functional morphology of the feeding mechanism is also provided. Of the nine nominal species of Gigantactis (species based on females), five are rec- ognized: G. longicirra, G. vanhoeffeni (with G. exodon as a synonym), G. gracilicauda (with G. sexfilis as a synonym), G. perlatus, and G. macronema. Gigantactis ovifer and G. filibulbosus are regarded as in- certae sedis. Twelve additional species of Gigantactis are newly de- scribed based on females recently collected from the three major oceans of the world. The species are distinguished on the basis of dif- ferences in the morphology of the illicium and esca, illicial length, the number, size, and pattern of placement of jaw teeth, caudal fin mor- phology, and fin-ray counts. The females of Gigantactis differ from those of Rhynchactis and all other known ceratioids in having the den- taries unfused on the midline and mobile relative to each other. Charac- ters that allow specific identification of Gigantactis males to species based on females have not been found. However, males have been di- vided into six groups based on differences in eye diameter, pigmenta- tion, presence or absence of skin spines, and fin-ray counts. As far as can be determined, the genus Rhynchactis contains a single species. Rhynchactis females differ from those of Gigantactis and most other ceratioids in lacking a photophore-containing escal bulb. Both gigantactinid genera are nearly cosmopolitan in geographic distribution and contain species that are for the most part wide-ranging forms. Vertically, gigantactinids are most commonly collected between approximately 1000 and 2500 m. Rhynchactis has undergone such a drastic reduction and loss of parts that clearly it is the more derived of the two gigantactinid genera. Within the genus Gigantactis, G. longicirra appears to be the least de- rived member of the family. Members of the G. macronema group, con- taining G. macronema and three newly described forms, are probably the most derived. The remaining species of the genus are more or less intermediate in specialization. INTRODUCTION The Gigantactinidae includes elongate, deepsea anglerfishes, the females of which are easily separated from those of allied families by having an exceptionally long illicium, five pectoral radials, and an elongate caudal peduncle. The first of two gi- gantactinid genera to be discovered was established by Brauer (1902) with the description of Gigantactis vanhoeffeni, based on two female specimens collected from the Indian Ocean during the German Deepsea Expedition of 1898-1899. Since that time, eight additional forms based on females have been de- scribed, each from a single specimen: G. macronema Regan 1925, G. gracilicauda Regan 1925, G. sexfilis Regan and Trewavas 1932, G. exodon Regan and Trewavas 1932, G. ovifer Regan and Trewavas 1932, G. flibulosus Fraser-Brunner 1935, G. longicirra Waterman 1939b, and G. perlatus Beebe and Crane 1947. A tenth species based on a single metamorphosed male not originally recognized as a gigantactinid, was described by Regan and Trewavas (1932) as Teleotrema microphthalmus and referred to Gigantactis by Bertelsen (1951). After examining the then known material of Gigantactis, Bertelsen (1951) concluded that the separation of nine nominal species based on a total of 1 1 metamorphosed females (30-100 mm) must be regarded as uncertain, that the five known meta- morphosed males probably represented at least two species and that the 233 larvae in the Dana Collections could be divided into three groups based on differences in pigmentation and fin- ray counts. In this paper, we propose that the 165 known female speci- mens (1 1.5-408 mm) of Gigantactis now available represent 17 well-defined species. Five of these are previously described forms: G. longicirra, G. vanhoeffeni, G. gracilicauda, G. mac- ronema, and G. perlatus. The remaining 12 species are recog- nized as new on the basis of recently collected material from the three major oceans of the world. The nominal species G. ovifer and G. filibulbosus, each represented only by a poorly preserved holotype, are regarded as incertae sedis. The recogni- tion and separation of species based on females is restricted to differences in the morphology of the illicium and esca, illicial length, the number, size, and pattern of placement of the jaw teeth, caudal fin morphology, and median fin-ray counts. 1. Editorial committee for this contribution: Julian Badcock and Richard H. Rosenblatt. 2. Zoological Museum, University of Copenhagen, Copenhagen, Denmark. 3. College of Fisheries, University of Washington, Seattle, Wash- ington 98195; and Research Associate in Ichthyology, Natural History Museum of Los Angeles County, Los Angeles, California 90007. 4. Ichthyology Section, Natural History Museum of Los Angeles County, Los Angeles, California 90007. Contributions in Science, Number 332 Bertelsen, Pietsch & Lavenberg: Gigantactinidl Anglerfishes 1 Females of Gigantactis differ from those of Rhynchactis and all other known ceratioids in having the dentaries unfused on the midline and mobile relative to each other. Despite an eight-fold increase in the number of known meta- morphosed males, characters that will allow specific identifi- cation remain obscure. However, the material has been divided into six groups based on differences in eye diameter, pigmenta- tion, presence or absence of skin spines, and fin-ray counts: G. longicirra (recognized by the high number of dorsal-fin rays unique to this species) and five additional groups referred to as Groups I through V. Gigantactis male Group I includes “the naked type” of Bertelsen (1951) as well as Parr’s (1927) Laevoceratias liparis [tentatively included in the Diceratiidae by Bertelsen (1951)]. Group II includes Teleotrema microph- thalmus Regan and Trewavas (1932), and the remaining three groups are each based on one or two aberrant and previously undescribed specimens. Despite records of several large females (300 to 400 mm), no parasitic Gigantactis males have been found. The genus Rhynchactis was established by Regan (1925) with the description of R. leptonema based on a single female specimen collected by the Dana in 1921. Bertelsen (1951) re- described and figured the holotype and reported finding a single adolescent male and 23 larvae all referred to the same species. Material gathered for this study brings the total number of known metamorphosed specimens to ten (six females and four males) and has added a male metamorphosal stage and two lar- vae above the previously known size. The material is still in- sufficient to determine whether the genus contains more than one species. Females of Rhynchactis differ from those of Gigan- tactis and most other ceratioids in lacking a photophore- containing escal bulb. They are also unique in having large glands of unknown function lining the oral cavity and, in adult stages, completely lacking premaxillary and dentary teeth. METHODS AND MATERIALS Standard length (SL), measured from the tip of the pterygiophore of the illicium to the posteriormost margin of the hypural plate, was used throughout (except where noted) and was taken to the nearest 0.5 mm in specimens less than 100 mm SL, to the nearest millimeter in larger individuals. Measure- ments were taken on the left side whenever possible and rounded to the nearest 0.5 mm. Jaw-tooth counts are taken from one side only (left side whenever possible). Teeth in early stages of development and teeth in resorption could only be ob- served in cleared and stained material. In drawings of tooth patterns, the relative position and size of each tooth base was estimated as carefully as possible. To insure accurate fin-ray counts, skin was in most cases removed from the pectoral fins, and incisions made to reveal the rays of the dorsal and anal fins. Caudal-fin rays are numbered from dorsalmost to ventralmost. I llicial length is the distance from the articulation of the pterygiophore of the illicium and illicial bone to the tip of the escal bulb, excluding escal appendages. The escal bulb is the distal swelling of the illicium, excluding escal appendages. The basal portion of the escal bulb is that part proximal to the pho- tophore. Terminology used in describing the various parts of the an- gling apparatus follows Bradbury (1967). Definitions and terms used for the different stages of development follow Bertelsen (1951:10-11). Frequently used abbreviations are given in the glossary. Drawings were made with the aid of a Wild M-5 Cam- era Lucida. The osteological and myological evidence presented is based on the following list of specimens. Most of the material was cleared in KOH and stained with alizarin red S. In many cases dissections were made of uncleared specimens to confirm obser- vations made on cleared specimens and to determine ontogenet- ic changes. Bone and muscle terminology follows Pietsch (1974) and Winterbottom (1974), respectively. All specimens except Gigantactis meadi n. sp. were cleared and stained. • Females. Gigantactis longicirra Waterman: ISH 973/68, 209 mm. Gigantactis vanhoeffeni Brauer: ISH 802/68, 152 mm; ZMUC P921972, 67.5 mm. Gigantactis meadi n. sp.: LACM 1 1242-12, 353 mm. Gigantactis macronema Regan: MCZ 51255, 141 mm. Gigantactis metamorphosal stage of Larval Group D: MCZ 54041, 9 mm. Rhynchactis leptonema Regan: ISH 2332/71, 60 mm. • Males. Gigantactis longicirra Waterman: UMML 27412, 14mm. Gigantactis Male Group I: LACM 30199-26, 17.5 mm. Gigantactis Male Group 1LLACM 33324-1, 13.5 mm. Gigan- tactis Male Group II: ZMUC P921533, 14.5 mm. Gigantactis Male Group V: LACM 32775-1, 14.5 mm. Gigantactis uniden- tified metamorphosal stages: ZMUC P21534, 1 1.5 mm; LACM 32749-4, 13 mm; LACM 32773-3, 14 mm. Rhynchactis lep- tonema Regan: LACM 37519-1, 17 mm. • Larvae. Rhynchactis leptonema Regan: ZUMC P921751, 7.5 mm. The systematic revision is based on 165 females (11.5-408 mm), 50 males (10.55-22 mm), and 299 larvae (2.1-15 mm total length) deposited in 19 institutions, the names of which are given in the glossary. GENERAL DESCRIPTION Major problems in systematic studies of ceratioid anglerfishes have been the lack of adequately preserved material available for examination and the sparsity of characters that can be used for taxonomic purposes. Ceratioids are, for the most part, soft- bodied animals, subject to damage and deformation during capture, as well as considerable shrinkage due to loss of water during preservation (we have recorded reductions in standard length of as much as 11% over a 3-year period in large indi- viduals of Ceratias holboelli). For these reasons, nearly all tra- ditionally utilized morphometric characters are worthless in distinguishing the numerous and morphometrically similar spe- cies of gigantactinids. Values for head length, width, and depth, lower jaw and premaxillary length, and caudal peduncle length are difficult to take with accuracy and so highly variable intra- specifically that they are of little importance. Dentary tooth patterns (the number and size of teeth and their arrangement in Contributions in Science, Number 332 Bertelsen, Pietsch & Lavenberg: Gigantactinid Anglerfishes longitudinal series), analyzed by us in some detail, are rather irregular and change so considerably with growth that their use as diagnostic characters is limited. The separation of species based on females is thus confined primarily to differences in the morphology of the illicium and esca. Other important charac- ters include the length of the illicium, the shape of the caudal fin, the relative length of the caudal-fin rays, and the number of rays in the dorsal and anal fins (Tables 1 and 2). These and other external characters of general importance are discussed in more detail below. This is followed by an osteological de- scription, and a myological description of the feeding mecha- nism. BODY SHAPE Compared to most other ceratioids, gigantactinid females are slender, streamlined fishes (Fig. 22). Flead length and greatest depth of body are usually only 25% SL (compared to 40% or more in most other ceratioids). The caudal peduncle is un- usually slender, having a depth of 5 to 10% SL in most speci- mens. Because of the extreme anterior position of the pterygiophore of the illicium, the snout of Gigantactis females is unusually pointed and protruded forward beyond the lower jaw. In Rhynchactis females the illicium emerges slightly above and behind the tip of the snout so that the latter is considerably more blunt than that of Gigantactis. ILLICIUM AND ESCA In all gigantactinids (except for G. longicirra and R. lep- tonema) the illicium appears to reach its full relative length by the time females reach a standard length of approximately 30 mm. In the smallest specimens greater than 30 mm (32-38 mm), the illicial length falls within the variation observed in larger conspecific specimens but is shorter than this in all meta- morphosed specimens of 25 mm or less. In most of the best- represented species ( G . kreffti, G. vanhoejfeni. G. meadi, G. macronema, and G. microdontis ), as well as in the four species known only from two or three metamorphosed females (G. gibbsi, G. gracilicauda, G. elsmani , and G. golovani ), the ob- served intraspecific variation in illicial length is relatively small in specimens greater than 30 mm. However, large intraspecific variations in illicial length are found in larger individuals of some forms (G. longicirra, G. gargantua, and G. savagei). In Gigantactis the bulb of the esca is more or less club- shaped and more slender than in most other ceratioids (Fig. 1). Its basal portion tapers very gradually into the shaft of the illicium without a distinct margin (thus no exact measurement of the length of the bulb is possible). The greatest diameter of the bulb (usually at the photophore) is rarely more than two or three times that of the minimum diameter of the illicium. Dis- tal to the photophore, the bulb is more or less conically pro- longed in most species. Generally, the pore of the photophore protrudes as a short tube with a slightly inflated rim, situated, as in all other ceratioids, on the posterior side of the escal bulb. Like the shaft of the illicium, the skin of the basal portion of the bulb is pigmented and spiny in all species of Gigantactis. The pigment and spinule coverage increases with age. In some forms (members of the G. vanhoejfeni group) spines are present over the entire surface of the bulb except for a field around the pore of the photophore. As in most other ceratioids, most spe- cies of Gigantactis have a more or less distinct patch of pigment on the distal portion of the bulb. The escal appendages of Gigantactis species are in the form of filaments and papillae placed directly on the bulb except in G. longicirra and G. perlatus in which a number of filaments arise from the edge of a pair of peduncle-like lobes situated on the posterior-basal margin of the bulb (Fig. 1). Except for the internally-transparent, distal prolongation of the escal bulb, no appendages with internal light-transmitting tubes occur (as de- scribed in several other ceratioid families, Bertelsen and Pietsch 1977). Great interspecific differences in the length and arrangement of escal filaments are present among species of Gigantactis (Fig. 1 ). Some filaments are short and digitiform, and others are like thin threads or stout tentacles; most are simple, yet branching occurs, especially in larger specimens. In some spe- cies (G. kreffti, G. paxtoni, and members of the G. macronema group), the filaments are restricted to the distal part of the bulb; in others, distinct pairs or more or less irregular groups of filaments are also present on more proximal portions of the bulb, as well as on the shaft of the illicium. Generally, the num- ber of filaments (especially those arranged in more irregular groups) increases with standard length. Three kinds of escal papillae occur: (1) low, distally flattened papillae (unique among ceratioids) that are present in G. van- hoejfeni (Fig. 30), G. meadi (Fig. 32), G. paxtoni (Fig. 37), and G. gibbsi (Fig. 34); (2) more or less elongate, unpigmented pa- pillae that are present in juvenile G. perlatus and develop into short filaments, completely covering the distal prolongation of the bulb in larger specimens (Fig. 39); (3) one or two papillae that are present on the posterior basal margin of the bulb of G. kreffti (Fig. 28) and G. vanhoejfeni, respectively, and become distally bifid in larger specimens of the latter species (Fig. 30). In a histological study of the esca of G. vanhoejfeni, Brauer (1908) found nerve fibers entering the filaments and papillae of the esca, supporting his assumption that these structures are organs of touch [“Tastorgane” of Brauer (1908) misinterpreted by Waterman (1948) as “gustatory” in function]. Waterman (1948) showed that the esca of the holotype of G. longicirra is innervated by two pairs of large nerves, one trigeminal in ori- gin, the other spinal, both pairs largely sensory in function. In Rhynchactis females, the length of the illicium (in speci- mens in which it appears complete) varies between 118 and 177% SL (Table 20). The distal part of the illicium is well pre- served only in a 27-mm specimen (IOS uncatalogued). In con- trast to those of Gigantactis , it lacks a photophore-containing escal bulb (Fig. 2). Among other ceratioids, this condition is known only in the Caulophrynidae, in which the slightly club- shaped distal part of the illicium bears a cluster of branched filaments, and in the Neoceratiidae, in which only a sub- cutaneous rudiment of the illicium is present. In Rhynchactis , the slightly inflated distal portion of the illicium bears a series Contributions in Science, Number 332 Bertelsen, Pietscb & Lavenberg: Gigantactinid Anglerfishes 3 Table 1. Comparison of important systematic characters among species of Gigantactis (tentatively identified specimens excluded) 1 E n specimens greater than 30 mm SL ^Except in largest specimens of G. gargantua Table 2. Frequencies of fin-ray counts for metamorphosal and metamorphosed females of species of Gigantactis , including tentatively identified specimens. Dorsal Anal Pectoral (left and ri; ght sides) Species 4 5 6 7 8 9 4 5 6 7 14 15 16 17 18 19 20 21 22 G. longicirra 4 3 4 3 1 5 3 1 G. krejfti 4 4 3 3 1 G. vanhoejfeni 1 1 27 2 9 30 1 20 14 1 G. meadi 9 1 1 9 4 5 G. gibbsi 2 2 1 1 G. gracilicauda 1 2 1 2 2 1 G. paxtoni 6 2 1 7 2 6 5 1 G. perlatus 4 3 1 2 5 1 1 3 4 1 G. elsmani 2 1 1 2 1 G. golovani 3 2 1 1 1 1 G. gargantua 2 3 1 6 3 2 1 1 G. watermani 1 1 2 2 G. herwigi 1 1 1 1 G. macronema 8 2 8 2 1 5 4 1 G. savagei 1 3 2 1 3 2 4 3 1 G. microdontis 1 4 2 2 3 2 3 6 2 G. ios 1 1 1 of short, unpigmented branches (Figs. 2, 63). Histologically, the core of each branch is opaque (Fig. 3), consisting of dense con- centrations of cells with large nuclei surrounded by blood ves- sels. FIN RAYS A summary of fin-ray counts shows great similarity among most species (Tables 2, 21). The dorsal fin of Gigantactis con- sists of from four to ten rays, the anal fin from four to eight. Rhynchactis has considerably fewer rays in the unpaired fins, the dorsal consisting of from three to five rays, the anal from three to four. All of these rays are biserial, segmented, and un- branched. On the basis of dorsal-fin ray counts, only a single species, G. longicirra, can be distinguished from all other spe- cies of Gigantactis (this species is further unique in having the first and last rays of the dorsal fin distinctly prolonged). In most species the number of pectoral-fin rays varies be- tween 17 and 19, but especially low counts have been recorded in G. golovani (14-16) and G. longicirra ( 14-18), and especially high counts in G. gargantua (19-22). In contrast to other ceratioids (including Rhynchactis ), all the caudal rays of metamorphosed female Gigantactis are un- branched. The ninth caudal ray is very small and completely covered by skin. Extremely prolonged caudal rays (60 to 100% SL) are characteristic of G longicirra and members of the G. gargantua group (Table 1). In G. longicirra, the first and eighth rays are the longest; in the G. gargantua group, the 2nd and 7th rays are prolonged (Fig. 4). Furthermore, considerable dif- ferences in the development of the skin coverage of the caudal fin are found among the larger specimens of Gigantactis exam- ined (Fig. 4). In individuals of some species ( G . krejfti, G. meadi, G. perlatus, and G. elsmani), the proximal one-third to one-half of the fin is covered with undivided skin; in others (G. golovani and G. macronema), the fin rays are separated nearly to their bases. The black and spiny skin covering each fin ray may be compressed with broad extensions partially connecting the rays (as in G. krejfti, G. meadi, and G. golovani) or narrowly fitting the cylindrical bony rays (as in G. macronema). The di- agnostic value of these characters is limited because they are less distinctly developed in juvenile specimens and the number of larger specimens is insufficient at the present time to show the extent of intraspecific variation for most species. Gigantactinid larvae differ from those of other ceratioids (ex- cept for the Caulophrynidae) in the size of the pectoral fins, which vary in length between 40 and 50% SL, extending pos- teriorly to the caudal peduncle (Bertelsen 1951, figs. 99, 104; Figs. 26, 60). They differ from caulophrynid larvae in lacking pelvic fins (Bertelsen 1951, fig. II). DENTITION Gigantactis The premaxillary teeth of female Gigantactis are small, rela- tively few, and arranged in either one or two distinct series. In some specimens, a posterior series of anteriorly curved teeth is present (Fig. 13). Although considerable variation in numbers of premaxillary teeth was observed, no clear interspecific dif- ferences were found. One of the reasons for this seems to be that, in connection with the reduction of the upper jaw of Gi- gantactis, these teeth are feeble and loosely attached and there- fore lost in many preserved specimens. More distinct differences are found in the heavy dentition of the lower jaw of female Gigantactis (Fig. 5). In addition to vari- ation in numbers and in the relative length of the largest fangs, differences in their arrangement in more or less distinct longitu- dinal series are apparent. To examine to what extent these dif- ferences are due to ontogenetic and individual variation, the pattern present in the left dentary of each specimen was drawn. For comparison, dots indicating the bases of the teeth were Contributions in Science, Number 332 Bertelsen, Pietsch & Lavenberg: Gigantactinid Anglerfishes 5 Figure 1. Diagrammatic views of escae of Gigantactis showing range of morphological variation: A. G. longicirra; B. G. vanhoeffeni; C. G. perlatus; D. G. gargantua; E. G. macronema. drawn with the aid of a camera lucida using the same magnifi- cation and, as far as possible, the same dorsolateral angle of view for each specimen. The size and position of teeth partly covered by skin were estimated as exactly as possible. On the basis of a thorough intraspecific comparison as well as a com- parison of specimens of a similar size among the best repre- sented species (especially G. vanhoeffeni, of which a representa- tive series is shown in Fig. 6), the following statements can be made: 1. In Gigantactis, teeth develop externally and internally on the dentaries; in all other ceratioids (except perhaps for Neo- ceratias), replacement teeth of the dentary invariably arise be- hind and medial to the older teeth. 2. In most species, one to several distinct longitudinal series of teeth occur in the posterior one-half to two-thirds of the jaw, but no distinct pattern can usually be recognized in the anterior part of the jaw. During metamorphosis, all the close-set larval teeth are lost and replaced by a small number of well-spaced tooth rudiments forming a single (median) series that is fol- lowed shortly by the first tooth rudiments of the external and, in most species, of the internal series as well. With growth, teeth of gradually increasing size are added externally as well as internally, each series starting anteriorly on the jaw (Fig. 6). 3. The median tooth series, reaching furthest posteriorly and consisting of relatively small and close-set teeth, can be recog- nized in all specimens with the exception of those of G. golo- vani. Next in relative size are those of the first internal series followed by those of the second internal series, if present. The increase in tooth size (and a corresponding increase in number) is more rapid in the external series, so that the longest of the serial teeth of the lower jaw are found in the posterior part of the exterior-most external series present. Teeth of the irregular Figure 2. Esca of Rhynchactis leptonema, 27 mm, IOS uncatalogued. Drawn by K. Elsman. 6 Contributions in Science, Number 332 Bertelsen, Pietsch & Lavenberg: Gigantactinid Anglerfishes anterior group, however, may reach similar or even greater lengths. 4. Teeth in initial development can be recognized by having soft unossified bases (Fig. 5). Assuming that no further growth takes place after ossification is complete, the ontogenetic in- crease in size of the teeth can be explained only by tooth re- placement. This assumption appears to be correct at least for the smaller, anteriormost serial teeth, which are often found to be partially resorbed and embedded in the bone of the jaw. The possibility, however, that the larger fangs continue to grow throughout life cannot be excluded. 5. According to the position of the smallest and apparently oldest teeth, additions to each tooth series are mainly added in the posterior part of the jaw; but, in contrast to other ceratioids, teeth of increasing size are also added anteriorly along the border of the irregular group on the free anterior end of the dentary. 6. The number of series of lower jaw teeth varies among spe- cies of Gigantactis (Table 1 ). Gigantactis longicirra, G. krejfti, and G. meadi have the largest number (five to six series); G. elsmani and members of the G. gargantua group are intermedi- ate (four to five); G. perlatus, G. golovani, and members of the G. macronema and G. vanhoeffeni groups (excluding G. meadi) have the lowest number (two to three series). Gigantactis golo- vani, with two or perhaps three series, seems to be unique in lacking a distinct median series. With some exceptions, species with the largest teeth tend to be those that have the greatest number of series. However, G. perlatus, with only two series (representing the median and first external series), has very large teeth, whereas G. elsmani, with five series, has relatively short teeth. 7. The diagnostic value of characters of the dentition is lim- ited because the full number of series may not be developed in smaller specimens, and the series tend to become irregular and therefore difficult to interpret in the largest specimens. Further difficulties arise from the considerable amount of individual variation in both the numbers and size of teeth. Rhynchactis Metamorphosed Rhynchactis females lack dentary teeth; one or two pairs of premaxillary teeth are present in juveniles (27 and 60 mm SL), but the single larger known specimen (126 mm) is completely toothless. Regan’s (1925, 1926) descriptions of the holotype of R. leptonema as having “minute teeth in sev- eral series” in addition to a “pair of anterior canines” could not be confirmed. Cigantactinid Males As in other ceratioids, the jaw teeth of male gigantactinids are lost during metamorphosis and replaced by an upper and lower series of hooked denticles, loosely attached at the symphyses of the maxillae and dentaries, respectively. In Gigantactis, the 0.1 mm Figure 3. Cross section through single side-branch of esca of Rhynchactis leptonema, 27 mm, IOS uncatalogued. Courtesy of Geert Brovad. denticular teeth are all mutually free and proximally com- pressed, whereas in Rhynchactis , at least some of the teeth are paired and supported by relatively broad bases (Figs. 14, 15). SENSORY STRUCTURES Eyes As in other ceratioids, the eyes of metamorphosed female gi- gantactinids are very small. Their diameter increases during larval life to approximately 1 mm but then decreases relative to standard length from about 20% to roughly 10% (Bertelsen 1951). After metamorphosis, eyes grow very slowly, reaching a diameter of 2.5 to 3 mm in the largest specimens (300-400 mm) and decreasing to less than 1% SL. Histologically there is no evidence of ocular degeneration (Brauer 1908, Waterman 1948, Munk 1964), but as growth continues, the eyes sink be- neath a narrow transparent layer of skin, greatly restricting the visual field. Furthermore, since the lens is situated close to the retina, ocular function is probably reduced to mere light detec- tion. In contrast to other ceratioids (except Neoceratiidae), the eyes of gigantactinid males are somewhat reduced in actual size during and after metamorphosis. In the five meta- morphosal stages available for examination (specimens 11.5- 14.5 mm), eye diameter ranged between 0.75 and 1.0 mm (average 0.86 mm), yet in the 42 metamorphosed males avail- able (10.5-22.0 mm), eye diameter was 0.4 to 1.0 mm (average 0.62 mm, Table 19). Expressed as a percentage of standard length, eye diameter varied between 2.9 and 6.9. As in the females, there is no apparent ocular degeneration, but because shrinkage of the eye seems to occur mainly in the scleral and retinal portions, the visual abilities of the males may be even less than those of the females. Olfactory Organs In female gigantactinids, the small olfactory organs, each with two tiny nostrils, are raised on short, cylindrical stalks with a Contributions in Science, Number 332 Bertelsen, Pietsdi & Lavenberg: Gigantactinid Anglerfishes Figure 4. Caudal fins of species of Gigantactis: A. G. vanhoeffeni, 266 mm, ISH 2480/71; B. G. meadi, holotype, 306 mm, MCZ 52572; C. G. perlatus , 152 mm, ISH 1466/71; D. G. Kreffti, paratype, 185 mm, ISH 3236/71; E. G. elsmani, holotype, 384 mm, ISH 1360/71; F. G. longicirra, 221 mm. MCZ; G. G. herwigi, holotype, 262 mm, ISH 972/68; H. G. gargantua, holotype, 408 mm, LACM 6903-32; I. G. macronema, 232 mm, ISH 1596/71; J. G. golovani, holotype, 179 mm, ISH 2250/71. Courtesy of Geert Brovad. length of three to four times their width. They are situated an- teriorly on the snout very close to the edge of the upper jaw. No distinct olfactory lamellae are present and, as shown by Water- man (1948, fig. 5), the olfactory lobes of the brain are small. Like those of other ceratioids (except Ceratiidae and Neo- ceratiidae), gigantactinid metamorphosed males are macros- matic. The series of olfactory lamellae (11-12 in most specimens of Gigantactis , 13-15 in Rhynchactis) have a height of 5.2 to 1 1.8% SL (Table 19). The large, forward-directed an- terior nostrils are very close set and separated from the equally large posterior nostrils by a narrow bridge of skin. The distance between the eye and posterior nostril is two to three times the eye diameter. The olfactory lobes of the brain are well devel- oped. (Fig. 7). Acoustic©- Lateralis System The organs of the acoustico-lateralis system are raised on short stalks or “tags” of pigmented skin (Regan and Trewavas 1932, Waterman 1939b), more or less distinctly connected in series by narrow unpigmented grooves. On each side of the head, the following series can be distinguished (Fig. 8); (1) two supraorbi- Figure 5. Gigantactis vanhoeffeni, female, 152 mm, ISH 802/68: A. Dentary teeth, left lateral view; B. Diagrammatic representation of den- tary-tooth pattern. E = external series; M = median series; I = inter- nal series. 8 Contributions in Science, Number 332 Bertelsen, Pietsch & Lavenberg: Gigantactinid Anglerfishes mm 10 mm 5 mm I !__! l__l I 1 I . I I i I . I I i Figure 6. Diagrammatic representation of dentary-tooth patterns of females of Gigantactis vanhoeffeni arranged according to stage of development. Contributions in Science, Number 332 Bertelsen, Pietsch & Lavenberg: Gigantactinid Anglerfishes 9 tal series meeting anteriorly in front of the nostrils at the base of the illicium, (2) an infraorbital series following along the edge of the upper jaw and meeting posteriorly with (3) an oper- cular series that connects with the two supraorbital series and spreads out ventrally in an irregularly arranged group of organs below and in front of the gill opening, and (4) two somewhat irregular mandibular series that meet with the opercular series posteriorly and are continuous anteriorly with their counter- parts of the opposite side. No distinct cephalic lateral-line com- missures between left and right sides are apparent. The two supraorbital series continue posteriorly in a double lateral line from a very dorsal position in front of the dorsal fin descending to a mediolateral position below the dorsal fin, along the caudal peduncle, and finally out onto each ray of the caudal fin. Oral Glands of Rhynchactis Females The jaws and oral cavity of metamorphosed Rhynchactis females differ from those of other known ceratioid females in the extreme reduction of the jaw bones and lack of dentary and premaxillary teeth and in possessing numerous, large glands that line the inner walls of the jaws. On each side of the upper and lower jaw, a crescent-shaped internal lip is covered with a dense pavement of white papillae, each outlined by pigmented skin (Fig. 9). Each papilla has a more or less distinct central groove that represents the opening of a short tube; each tube is internally walled with large glandular cells that nearly fill the lumen (Fig. 10). Properly fixed material was unavailable for a more thorough histological investigation. No glands were found in the triangular pad of tissue covering the roof of the mouth between and in front of the pharyngeals (Fig. 9); this pad in- stead consists of tough connective tissue divided into large 2.5 mm Figure 7. Brain of juvenile male of Rhynchactis leptonema, 17 mm, LACM 37519-1, dorsal view. close-set papillae, each carrying a distal group of smaller pig- mented papillae. In Gigantactis females, this part of the roof of the mouth is covered with irregular, transverse folds of pig- mented skin. OSTEOLOGY CRANIUM, Figures 11-14, 17 The cranial osteology of gigantactinid genera is characterized most strikingly by an extreme reduction and loss of parts. With few exceptions the ethmoid complex of both Gigantactis and Rhynchactis consists only of widely separated lateral ethmoids. A supraethmoid ossification is present in the 14.5-mm Type V male (LACM 32775-1, Fig. 1 1 F) but greatly reduced in size in the 13.5-mm Type 11 male (LACM 33324-1). It is absent in larvae and in all metamorphosed male and female specimens examined osteologically except for the 152-mm preparation of G. vanhoeffeni, where it is represented by a small rudiment (Fig. 12). An ossified vomer is present in larvae and males (re- duced in the 17-mm Rhynchactis male. Fig. 1 IF) but absent in adolescent and adult females of both genera. The frontals are relatively short and broad in the larvae and males of both gen- era (Figs. 11, 14). However, in metamorphosed females of Gi- gantactis, the frontals are long, narrow, and widely separated throughout their length, approaching each other only slightly at midlength (anteriorly, each frontal lies in close proximity but does not meet with the dorsal process of the respective lateral ethmoid; posteriorly each is overlapped slightly by the respec- tive parietal bone). In Rhynchactis , the frontals and parietals are present in larvae and males but absent in the 60-mm female (Fig. 11). The pterosphenoid is absent in Gigantactis and Rhynchactis of both sexes. In both gigantactinid genera, the supraoccipital occupies a more anterior position than in other ceratioids. In females, most of the outer surface of this bone lies in the vertical plane, providing a posterior abutment for the pterygiophore of the il- licium. In larvae and males of both genera, the entire outer sur- face of the supraoccipital lies in the horizontal plane. MANDIBULAR ARCH, Figures 11, 13-15 In Gigantactis females, the premaxilla is long and narrow but well ossified, bearing teeth along most of its length (toothless in metamorphosed males and reduced to a small anterior remnant lying just in front of a small symphysial cartilage. Fig. 15). A small articular process is present, but ascending and postmax- illary processes are absent. The maxilla of Gigantactis females is greatly reduced, becoming further reduced with age. In the 141-mm specimen of G. macronema, the maxilla is represented by a short anterior portion and a filamentous posterior remnant; in the 146-mm and 209-mm preparations of G. vanhoeffeni and G. longicirra, respectively, only the posterior filament remains. The elements of the upper jaw are present in Rhynchactis lar- vae but become greatly reduced or lost in adolescents and adults (Figs. 11, 13). The maxilla is absent in the 60-mm female. A thin premaxilla bearing one large, recurved, anterior tooth and a series of minute teeth along its posterior extension is present in the 42-mm holotype of Rhynchactis leptonema, but 10 Contributions in Science, Number 332 Bertelsen, Pietsch & Lavenberg: Gigantactinid Anglerfishes Figure 8. Acoustical-lateralis system of female Gigantactis van- hoe ffeni, 152 mm, ISH 802/68. only a small, single-toothed, anterior remnant of the premaxilla remains in the 60-mm specimen (Fig. 13). A symphysial car- tilage is absent in larvae and metamorphosed females of both genera. Unlike the dentaries of most other ceratioids, those of gigan- tactinid genera are not forked posteriorly except in males and larvae. The posterior half of the ventral margin of these bones in females forms a relatively loose and mobile connection with the respective articular. Anteriorly, the dentaries curve to ap- proach each other on the midline where they are attached to each other by thick, elastic connective tissue. Compared to females of Rhynchactis, the bones of the lower jaw of Gigantac- tis females are thick and well-ossified (Fig. 13). Dentary teeth are absent in Rhynchactis; in Gigantactis, each dentary bears several rows of recurved, depressible teeth (see “Dentition”). As in other male ceratioids, jaw teeth of male gigantactinids are lost during metamorphosis. The upper and lower denticular teeth (Bertelsen 1951:21) are loosely attached at the symphysis of the maxillae and dentaries, respectively. In Gigantactis males, the denticular teeth are all mutually free and proximally compressed; in Rhynchactis, at least some of the teeth are paired and supported by relatively broad bases (Fig. 15). Unlike most other ceratioids, the articular of females of both Gigantactis and Rhynchactis does not extend back beyond its articulation with the quadrate. Futhermore, there is no liga- mentous connection between the angular bone and the reduced interoperculum. PALATINE AND HYOID ARCHES, Figures 13, 14, 16B The palatine arch of females of both gigantactinid genera con- sists of a triangular-shaped metapterygoid and elongate ectopterygoid and palatine bones; the mesopterygoid is absent except in males and larvae. The hyomandibular bone is large and unforked, forming a single, broad articulation with the cra- nium. The symplectic is conical in shape and short compared to that of other ceratioids, in females, not extending beyond the dorsal margin of the quadrate. In Gigantactis females, the pos- terodorsal margin of the symplectic is covered laterally by the posterior process of the quadrate; the latter bone is overlapped in turn by the ventral half of the reduced preopercle (Fig. 13 A). In Rhynchactis females, the posterior process of the quadrate is reduced dorsally so that the remnant of the pre- opercle lies directly on the symplectic (Fig. 13B). The epihyals and ceratohyals are narrow and elongate. Each pair bears the heads of six long and slender branchiostegal rays, the anteriormost two of which articulate on the medial surface, the remaining four on the lateral surface. Gigantactis females have a single, elongate hypohyal, whereas Rhynchactis females have retained both a dorsal and a ventral hypohyal. As is typical for ceratioids, nearly all examined gigantac- tinids have six branchiostegal rays. However, in the 14-mm G. longicirra male (UMML 27412), there are seven branchioste- gal rays on each side: two in an anteromedial position and five in a posterolateral position (Fig. 16B). The 14-mm male in met- amorphosis (unidentified, LACM 32772-3) has seven branchiostegal rays on the right side (three anteromedial and four posterolateral) but only six on the left side (two ante- romedial and four posterolateral). OPERCULAR APPARATUS, Figures 13-14 In both gigantactinid genera, the posterior margin of each oper- cle is deeply notched forming two narrow forks (except in males and larvae), the lower fork slightly longer than the upper. The two forks (the upper one of which is bifurcated on the left side of the 209-mm specimen of G. longicirra ) are slightly curved in Gigantactis but straight in Rhynchactis. The suboper- cle of juvenile females of Gigantactis consists of a relatively broad lower portion and a single, short, tapering posterior ex- tension. In larger female specimens, this broad lower portion becomes reduced, sometimes leaving behind a small anterior projection and as many as three filamentous, posterior pro- longations, the uppermost of which may be nearly as long as the lower fork of the opercle. The subopercle of the 60-mm female of Rhynchactis is a simple, slender, posteriorly tapering bone that is approximately equal in length to the lower fork of the opercle. In Gigantactis females, the preopercle is a small strut of bone that bridges the gap between the hyomandibular and quadrate bones. In Rhynchactis females, this element is further reduced, lying on the lateral surface of the symplectic (the latter ex- posed by the reduced quadrate) but not extending dorsally to the hyomandibular. BRANCHIAL ARCHES, Figures 13, 16 The elements of the branchial arches are similar in both gigan- tactinid genera. A pharyngobranchial I is absent. Phar- yngobranchials II and III are extremely well developed and heavily toothed (except in males), especially in Gigantactis. Epibranchial I is absent. The anterior halves of epibranchials Contributions in Science, Number 332 Bertelsen, Pietsch & Lavenberg: Gigantactinid Anglerfishes 11 I 5 mm | Figure 9. Mouth cavity of female Rhynchactis leptonema, 60 mm, ISH 2332/71, showing oral glands. Drawn by Elizabeth Beyerholm. Ill and IV are tightly held to each other by connective tissue. Ceratobranchials I through V are complete in males. In females, ceratobranchial I is reduced so that only a small pos- terior and one or two small ossified remnants of the anterior portion remain; ceratobranchial V is represented only by one or two tiny ossifications (completely lost in the 146-mm prepara- tion of G. vanhoeffeni) lying free in the connective tissue ma- trix. There are no ossified hypobranchials or basibranchials in females; males have a single basibranchial ossification. VERTEBRAE AND CAUDAL SKELETON, Figure 17 The vertebrae of gigantactinids do not differ substantially from those described for other ceratioids (Pietsch 1972a:36, fig. 16, 1974:12, fig. 12). In all cleared and stained female specimens of Gigantactis examined, there are 22 vertebrae. The posterior 15 of these bear complete haemal arches and are thus consid- ered caudal vertebrae. In the 60-mm Rhynchactis female there are 20 vertebrae, the posterior 14 of which are caudal. The caudal skeleton of gigantactinids is like that of other ce- ratioids in having the ural centra fused with the first pre-ural centrum to form a single, complex half-centrum which is fused Figure SO. Section through oral glands of female Rhynchactis lep- tonema, 60 mm, ISH 2332/71. Courtesy of Geert Brovad. 12 Contributions in Science, Number 332 Bertelsen, Pietsch & Lavenberg: Gigantactinid Anglerfishes A B C D E F Figure 11. Dorsal views of crania of gigantactinids: A. G. longicirra, female, 209 mm, ISH 973/68; B. R. leptonema, female, 60 mm, ISH 2332/71; C. R. leptonema, larval female, 7.5 mm, ZMUC P921751; D. G. longicirra, male, 14 mm, UMML 27412; E. R. leptonema, male, 17 mm, LACM 37519-1; F. G. male Group V, 14.5 mm, LACM 32775-1. Contributions in Science, Number 332 Bertelsem, Pietsch & Lavenberg: Gigantactinid Anglerfishes 13 to a single hypural plate. The hypural plate is unnotched pos- teriorly and bears the overlapping bases of nine principal cau- dal rays. These rays are all biserial and segmented. The ninth caudal ray is reduced and embedded within the skin surround- ing the adjacent ray. All caudal rays are unbranched in female Gigantactis, but rays three (from the top) through six are branched distally in male Gigantactis and in both sexes of Rhynchactis. In the females of both genera the outermost cau- dal rays are longer than the inner rays so that, unlike all other ceratioids, the tail fin is emarginate (except in largest females of some species of Gigantactis). ILLICIAL APPARATUS, Figures 11, 13, 14, 17 The pterygiophore of the illicium is exceptionally large and lat- erally compressed in female gigantactinids. It lies on top of the parasphenoid and ethmoid cartilage and passes between the proximal ends of the huge, dorsally projecting phar- yngobranchials II and III, the frontals (in Gigantactis females only), and the lateral ethmoids. In females of both genera, the cartilaginous posterior end of this element butts up against the vertical surface of the supraoccipital; the anterior end extends slightly beyond the tip of the upper jaw in Gigantactis but ter- minates just posterior to this point in Rhynchactis. A small os- sified remnant of the second cephalic ray is present in both genera. The illicial bone is strongly articulated to the distal tip of the pterygiophore. PECTORAL GIRDLE, PECTORAL FIN, AND PELVIC BONES, Figures 13, 14 The cleithrum of Gigantactis females is sharply angled at mid- length, the lower portion being considerably more slender than the upper. The cleithrum of Rhynchactis females, on the other hand, is much more like that of most other ceratioids, forming a smooth crescent-shaped curve, the lower portion gradually tapering anteriorly. In both genera, there is a single, elongate Figure 12. Lateral view of ethmoid region of Gigantactis vanhoeffeni, female, 152 mm, ISH 802/68. postcleithrum. In Gigantactis females, the coracoid consists of a small, conical, anterior ossification and an elongate, car- tilaginous, posteroventral process that extends backward along the postcleithrum; the scapula remains unossified. In Rhynchactis, both the coracoid and scapula are unossified. In all gigantactinids, there are five pectoral radials. In both genera, it appears that the fourth radial (from the top) tends to become reduced; the proximal end of this radial is absent in most specimens of Gigantactis examined. In the 60-mm Rhynchactis female, the uppermost pectoral radial extends con- siderably past the distal tips of the lower four radials. Gigantac- tis has from 14 to 22 pectoral rays, Rhynchactis has from 17 to 20 (Tables 2, 20, 21). These rays are all biserial, segmented, and unbranched. Pelvic bones appear to be absent in all Gigantactis examined but represented by small struts of bone in Rhynchactis. SKIN SPINES Numerous, close-set spines covering the entire body and fins are visually obvious in uncleared specimens of females of all species of Gigantactis without microscopic aid. These spines ex- tend out onto the illicium and in some species onto the escal bulb. In Gigantactis males, skin spines are present in members of groups II and III (Table 19) but absent in G. longicirra and members of groups I, IV and V. Minute spines are present in the skin of the largest known female (126 mm) of Rhynchactis but are absent in males and smaller females. MYOLOGY OF THE FEEDING MECHANISM The musculature of the feeding mechanism of gigantactinid females is similar to that described by Field (1966) for the much less derived, shallow-water anglerfish, Lophius piscatorius. Differences are mainly in the degree of develop- ment of various muscles, and in the reduction and loss of mus- cle segments due to a corresponding reduction and loss of bony parts. The muscles of the illicial apparatus (cephalic tentacle) of Gigantactis were studied by Brauer (1908). These muscles were redescribed by Waterman (1948), and discussion and fig- ures of those of the upper pharyngeals and superficial muscles of the cheek were added. These are reviewed below along with previously undescribed musculature of the lower jaw and floor of the mouth. CHEEK MUSCLES, Figure 18 Section A, of the adductor mandibulae, defined by its dorsal position and insertion on the maxilla (Winterbottom 1974:232), is absent in gigantactinids, corresponding to the severe reduc- tion or loss of this upper jaw bone. In Gigantactis, section A2 (adductor mandibulae superficialis of Waterman 1948:95, fig. 1 ) has a broad origin on the posterodorsal margin of the hyo- mandibular and a narrow insertion on the posterodorsal margin of the articular; it does not share a myocomma anteriorly with the posterior fibers of Section Aw (see “Lower Jaw” below). Section A3 (adductor mandibulae profundus of Waterman 1948:95, fig. 1) lies medial to A2, originating broadly on the 14 Contributions in Science, Number 332 Bertelsen, Pietsch & Lavenberg: Gigantactinid Anglerfishes Ex Figure 13. Lateral views of skulls of female gigantactinids with pectoral girdle, opercular apparatus and hyoid apparatus in place: A. G. longicirra , 209 mm, ISH 973/71; B. R. leptonema, 60 mm, ISH 2332/71. Contributions in Science, Number 332 Bertelsen, Pietsch & Lavemberg: Gigantactinid Anglerfishes 15 Figure 14. Lateral views of skulls of male gigantactinids with pectoral girdle, opercular appa- ratus and hyoid apparatus in place: A. G. longicirra, 14 mm, UMML 27412; B. R. leptonema, 17 mm, LACM 37519-1; C. G. male Group V, 14.5 mm, LACM 32775-1. 16 Contributions in Science, Number 332 Bertelsen, Pietsch & Lavenberg: Gigantactinid Anglerfishes quadrate, on the anteroventral margin of the hyomandibular and metapterygoid. A small dorsal subdivision originating on the sphenotic is thought to be part of A3. All of these muscles are essentially the same in Rhynchactis. The levator arcus palatini (L.A.P.) has its origin on the sphenotic and its insertion on the lateral face of the hyoman- dibular. A narrow dilatator operculi (D O.) originates on the sphenotic and inserts on the proximal tip of the opercle. The levator operculi (L.O.) is also narrow, originating on the ptero- tic and inserting on the upper fork of the opercle. These are all essentially the same in Rhynchactis. LOWER JAW, Figures 18, 19 In Gigantactis, section Aw of the adductor mandibulae muscle (articulodentary of Field 1966:54, fig. 4; see also Winterbottom 1974:242) is extremely well developed but has lost all connec- tion with section A2 becoming purely an intrinsic lower jaw muscle. It covers nearly the entire medial surface of the lower jaw, stretching between the dentary and articular, its oblique fibers running in a anterodorsal-posteroventral direction. In lat- eral view, it can be seen extending well below the ventral mar- gin of the lower jaw. In the material of Rhynchactis available to us, no trace of this muscle could be found. This absence corre- sponds with the narrow, extremely reduced and toothless bones of the lower jaw of this genus. FLOOR OF MOUTH, Figure 19A The intermandibularis (IMD) of Gigantactis is a relatively long, narrow muscle that has a broad insertion on the ventral margin of the dentaries just behind the symphysis. Its trans- verse fibers pass ventral to a thick, crescent-shaped pad of elas- tic connective tissue that lies between the two halves of the lower jaw preventing them from meeting on the midline. No trace of this muscle could be found in Rhynchactis. The protractor hyoidei (PR.HY.) of Gigantactis has a broad origin on the ventral-most margin of the dentary overlapping the insertion of the intermandibularis muscle. Each protractor hyoidei narrows posteriorly, approaching its counterpart from the other side and inserting on a fascia near the midline. From this narrow insertion, each protractor hyoidei widens further posteriorly, attaching broadly to the respective ceratohyal. Only traces of this muscle could be found in Rhynchactis. The sternohyoideus muscle (STH.) of Gigantactis, origina- ting on the cleithrum, splits into two sections: a considerably larger, lateral section inserts on the respective hypohyal; the medial section (not shown in Figure 19A) passes to the distal tip of ceratobranchial IV. This muscle has essentially the same morphology in Rhynchactis but is slightly larger than that of Gigantactis of a similar standard length. The hypaxial musculature forms a broad insertion along most of the posterior margin of the lower half of the cleithrum in both gigantactinid genera. UPPER PHARYNGEALS, Figures 18A, 20 As described by Waterman (1948:96, figs. 1, 8), the largest and most complex muscle system in Gigantactis, except for the body musculature, is the system that operates the hyper- trophied pharyngobranchials and epibranchials of the second and third branchial arches. These muscles are essentially the same in both gigantactinid genera; Rhynchactis is mentioned below only when known differences or additions occur. The levatores externi muscles (levatores arcuum branchialium externi of Waterman 1948:97, figs. 1, 8), of which only two can be differentiated, originate together on the parie- tal. The larger, more posterior of these divides distally to insert broadly on epibranchials 1 1 and III. The smaller muscle, in con- trast to the usual situation (Winterbottom 1974:250), and not mentioned by Waterman (1948), does not insert on an epi- branchial but passes down between epibranchials II and III to insert on the posterior margin of pharyngobranchial III just at the base of the pharyngobranchial teeth. The levatores externi muscles appear to be similar in Rhynchactis except that, in the absence of the parietal bones, origin takes place on the medial margin of the pterotic. The levatores interni (levatores arcuum branchialium interni of Waterman 1948), of which we can differentiate only one (most probably those that serve pharyngobranchials II and III have fused to form a single muscle mass; Waterman 1948:97), Figure 15. Anterior views of upper and lower jaws of male gigantac- tinids showing denticular teeth: A. G. longicirra , 14 mm, UMML 27412; B. R. leptonema , 17 mm, LACM 37519-1. Contributions in Science, Number 332 Bertelsen, Pietsch & Lavenberg: Gigantactinid Anglerfishes 17 Figure 16. Branchial arches of Gigantactis longicirra: A. Female, 209 mm, ISH 973/71; B. Male, 14 mm, UMML 27412, showing hyoid apparatus. originates on the prootic and sphenotic. It passes dorsal to the retractor dorsalis and medial to the levator externus muscles to insert on the proximal ends of pharyngobranchials II and III. Three obliqui dorsales muscles (OBL.D.) are present in Gi- gantactis. That serving the second arch (interarcualis dorsalis 2 of Waterman 1948, figs. 1, 8) is especially well developed, origi- nating on the dorsolateral surface of pharyngobranchial II and inserting on the proximal-lateral surface of epibranchial II. The transversi dorsales (TR.D.) originates partly on the para- sphenoid and partly on a fascia near the midline. It passes me- dial to the levator externus and under the levator internus to form a large, bulbous insertion on the dorsolateral surface of pharyngobranchial II. The retractor dorsalis (R.DORS.; retractor arcuum branchialium of Waterman 1948:97) is an enormous muscle that originates on the three anterior-most vertebrae and part of the fourth. Insertion is on the posteroventral margin of phar- yngobranchial III. In conflict with Waterman’s description (1948:97, figs. 1, 8), no fibers on the retractor dorsalis insert on pharyngobranchial II. 5LLICIAL MUSCULATURE, Figure 21 The illicial apparatus is controlled by five pairs of muscles: two intrinsic pairs — the depressor and erector dorsalis I (DEPR.D. and EREC.D., respectively; flexor and extensor, respectively, of Bertelsen 1951); and three extrinsic pairs — the supracarinales anterior (SCAR. A.: exertor of Bertelsen 1951), and anterior and posterior subdivisions of the inclinator dorsalis II (INCL.D.; inclinator and retractor, respectively, of Bertelsen 1951). The origins and insertions of these muscles have been previously described (Waterman 1948, Bertelsen 1951). Com- pared to most other ceratioids, the extrinsic illicial muscles of gigantactinids are small. In contrast, the two intrinsic muscle pairs are unusually large. FUNCTIONAL MORPHOLOGY OF THE FEMALE FEEDING MECHANISM GIGANTACTIS Despite its restricted mobility relative to the cranium (only a few millimeters, even in the largest individuals), the pterygiophore of the illicium of Gigantactis is equipped with moderately developed extrinsic muscles (the supracarinales an- terior and the anterior and posterior subdivisions of the inclina- tor dorsalis II; Fig. 21). The extrinsic illicial musculature of Gigantactis does not provide for gross movement in the antero- posterior and lateral directions, as is the case in most other ce- ratioids (particularly Ceratias and Oneirodes, see Bertelsen 1943 and Pietsch 1974, respectively). Instead, it is used to pro- duce vibration that passes out along the stiff illicial bone to the 58 Contributions in Science, Number 332 Bertelsen, Pietsch & Lavenberg: Gigantactinid Anglerfishes L. INI. 2*3 L. EXT. 2*3 ^ w\'- Figure 17. Diagrammatic representation of axial skeleton of Gigantac- tis longicirra , female, 209 mm, ISH 973/71. esca and escal filaments. That most of this vibration is trans- ferred to the surrounding water by the esca and its filaments rather than by the stem of the illicium was confirmed by one of us (R.J. Lavenberg) through direct observation. Immediately upon capture, the 408-mm holotype of G. gargantua was placed live in an aquarium. Several whip-like, backward and forward thrusts of the entire illicium were followed by moderately- strong vibrations of the esca with no apparent movement of the illicium, although quick, rapid contractions could be felt throughout its length. This vibratory action, combined with the bioluminescence of the bait, appears to be the most important mechanism of attracting prey; to what extent sweeping move- ments of the entire illicium are used in luring is unknown. Feed- ing on unattracted (or unattractable) prey items may occasionally occur. How female ceratioids detect their at- tracted (or unattracted) prey is still open to speculation. In Gi- gantactis. this is especially difficult to understand with the large predator-to-prey distance due to the exceptionally long fishing apparatus (Fig. 22). It is unlikely that the small, laterally posi- tioned eyes, lacking stereoscopic vision and an ability to pro- duce a well-defined image (Munk 1964:12), play any role in prey detection. The acoustico-lateralis system of the head, well developed in most ceratioids, may function in this way in those species in which prey is brought up close to the head, but it seems questionable that the system could provide sufficient dis- tance and directional signals at the distance required by Gigan- tactis. Although unsupported by direct observation, the nervous innervation of the esca described by Brauer (1908) and Water- man (1948) indicates that this organ and its filaments are sensi- tive to touch or to pressure waves produced by moving prey. It is hypothesized that once prey has been attracted to the bait, it is seized by a sudden, forward lunge, for which the streamlined body and powerful caudal of Gigantactis seem well adapted (Bertelsen 1951 ). The jaw mechanism of Gigantactis is similar to that of Thaumatichthys (family Thaumatichthyidae) in that it allows the seizure of prey to be made by means of long hooked teeth placed outside the mouth, rather than through suction as is the case in nearly all other anglerfishes (Bertelsen 1951:242; Ber- telsen and Struhsaker 1977:29; Grobecker and Pietsch 1979). Although it is the lower jaw of Gigantactis, in contrast to the upper jaw of Thaumatichthys, which is specialized in this way, the principle of both mechanisms is identical. In both, the bones of the right and left sides of the jaw are free at the sym- Figure 18. Muscles of the cheek, upper pharyngeals and lower jaw of Gigantactis meadi, female, 353 mm, LACM 11242-12; A. Superficial musculature; B. Portions of superficial musculature removed. physis, connected only by elastic ligaments; each bone can be twisted relative to the other in such a way that their long, curved teeth can be rotated inwards from a widely out- stretched, open position to a situation in which the teeth of the opposite side approach each other within the cavity of the mouth (Fig. 23). In both, the relatively small teeth in the oppos- ing jaw (the upper in Gigantactis, lower in Thaumatichthys) play a secondary role in seizing prey. It seems apparent that in Gigantactis the prey is snagged by the outstretched, recurved dentary teeth and brought into the buccal cavity within reach of the huge, upper pharyngeal teeth by a sudden, inward twist of the lower jaw (possibly aided by negative pressure created by a sudden expansion of the buccal and opercular cavities). The elastic connection between the rami of the lower jaw may allow for asymmetrical opening and closing of the mouth (much like the feeding mechanism of a snake); this alternating side to side adduction of the mandible, in association with the hooked and hinged jaw teeth, would facilitate the transport of prey items back towards the reach of the upper pharyngeal teeth. The extreme development and forward position of the upper pharyngeal teeth contribute to the efficiency with which the prey is seized and transferred to the stomach. Morphologi- cal evidence as well as direct observation on the living holotype Contributions in Science, Number 332 Bertelsen, Pietsch & Lavenberg: Gigamtactinid Anglerfishes 19 Figure 19. Musculature of the lower jaw and floor of throat of female Gigantactis vanhoeffeni: A. Ventral view; 67.5 mm, ZMUC P921972; B. Medial view of lower jaw showing intrinsic lower jaw muscle, 152 mm, ISH 802/68, left side reversed. 20 Contributions in Science, Number 332 Bertelsen, Pietsch & Lavenberg: Gigantactinid Anglerfishes Figure 20. Musculature of upper pharyngeals of Gigantactis meadi, female, 353 mm, LACM 11242-12, right side reversed. Figure 21. Illicial musculature of Gigantactis vanhoeffeni, female, 109 mm, 1SH 2331/71, slightly dia- grammatic: A. Dorsal view; B. Left lateral view. Contributions in Science, Number 332 Bertelsen, Pietsch & Lavenberg: Gigantactinid Anglerfishes 21 Figure 22. Gigantactis macronema, female, 232 mm, 1SH 1596/71. of G. gargantua shows that the upper pharyngeals work in pairs on each side, each pair alternating with the other, thrusting for- ward and pulling back with the teeth extended, like the pedals of a bicycle. Unlike other ceratioids, numerous individuals of Gigantactis have been captured with their stomachs everted; this evertion presumably occurs as a reaction to the stress of capture but could also indicate a mechanism by which Gigantactis is able to void the stomach of unwanted food. A stomach contents analysis of all available material of Gi- gantactis was largely unsuccessful. In the majority of speci- mens examined, the stomach was either empty or everted. The stomachs of only nine individuals contained organisms, most of which showed little or no evidence of digestion and could well have been swallowed while in the net: • Gigantactis vanhoeffeni: LACM 36046-10, 26 mm (one partly swallowed Cyclothone)', SIO 73-159, 39 mm (one copepod, one Sagitta)\ ZMUC P921972, 67.5 mm (one cephalopod-eye lens); ISH 872/68, 114 mm (one Phronima in barrel with a deformed cephalopod); IOAN uncatalogued, 196 mm (one Argyropelecus, ca. 40 mm, and one Nebaliopsisl, both apparently partly digested). • Gigantactis meadi: MCZ 52572, 306 mm (one Atolla, ap- parently partly digested). • Gigantactis gracilicauda (data taken from Bertelsen 1951:240): ZMUC P92132, 51 mm (eyes and beaks of two squid); ZMUC P92129, 82 mm (eyes and beak of a squid). • Gigantactis perlatus (data taken from Beebe and Crane 1947): CAS-SU 46487, 32.5 mm (one copepod, 6 mm long). RHYNCHACTIS As in Gigantactis, the restricted mobility of the pterygiophore of the illicium indicates that the primary function of the extrin- sic illicial muscles is to create vibratory movement of the illici- um and esca. In the absence of an escal bulb providing a site for the maintenance of bioluminescent bacteria, it seems appar- ent that the Rhynchactis esca cannot produce light. If so, per- haps the attraction of prey is based solely on pressure waves produced by pterygial vibration, in addition to other move- ments of the illicium. The possibility of luminescence, however, cannot be completely disregarded. In Rhynchactis, the reduction of nearly all the elements of the jaws, the loss of jaw teeth, and the reduced musculature reflect a very different kind of feeding mechanism than that found in Gigantactis. Prey attracted by the illicial apparatus of both genera is far beyond the reach of the jaws. However, the lack of effective jaw teeth and less slender body of Rhynchactis make it unlikely that this genus is able to reach out and seize prey by a sudden forward dart as is supposed for Gigantactis. It seems more likely that the curious oral glands are, in some way, involved in luring the prey the remaining distance from the bait to within reach of the jaws and powerful pharyngeal teeth. The inadequate preservation of the glands available to us does not allow conclusions as to their function; bioluminescence cannot be excluded, nor can the more likely function of secretion of some chemical attractant. If the latter is true, and if the esca has lost the ability to produce light, then perhaps Rhynchactis is adapted to feeding on prey that is not attracted by light. The stomachs of four Rhynchactis females examined were empty. REPRODUCTION Much of what is known about reproduction in gigantactinids was recently summarized by Pietsch (1976). Although a num- ber of the large Gigantactis females examined (353-408 mm) have relatively large ovaries, none approach maturity. Eggs larger than 0.5 mm in diameter have not been found. The largest of the known males in metamorphosis is 14.5 mm; metamorphosed males of G. longicirra and those of Groups II, III, and V are between 10.5 mm and 15.5 mm, whereas those of Groups I and IV are between 15 and 22 mm (Table 19). This relatively large, postmetamorphic increase in size indicates that Gigantactis males continue to grow after metamorphosis (beyond the usual change in shape), yet there is still no evidence that postmetamorphic males are able to feed (the stomachs of all metamorphosed males examined were empty). The absence of parasitic males and the relatively large number of large females known might indicate that gigantac- tinids are among those ceratioid families in which no perma- nent attachment of males occurs (Bertelsen 1951, Pietsch 1976). Compared to most other ceratioids, gigantactinid males have small eyes (diameter 4-5% SL in most specimens). It is as- sumed that they rely to a greater extent on the use of olfaction in locating potential mates. SYSTEMATICS Family Gigantactinidae Boulenger 1904a DIAGNOSIS. Metamorphosed gigantactinid females are dis- tinguished from those of all other ceratioid families by the following combination of characters: body elongate, head length less than 35% SL, caudal peduncle length greater than 20% SL; oral valves absent; upper jaw extending slightly be- yond lower jaw; supraethmoid extremely reduced, usually ab- sent; vomer absent; frontals reduced or absent; parietals reduced or absent; sphenotic spines absent; pterosphenoid ab- 22 Contributions in Science, Number 332 Bertelsen, Pietsch & Lavenberg: Gigantactinid Anglerfishes Figure 23. Anterior view of Gigantactis vahoeffeni, female, 306 mm, MCZ 52572, showing ability of rami of lower jaw to rotate relative to each other: A. Closed position; B. Open position. Courtesy of Geert Brovad. sent; ascending and postmaxillary processes of premaxillary ab- sent; symphysial cartilage absent; maxillary reduced to threadlike ossification or absent; dcntaries not ankylosed ante- riorly but held together by thick elastic connective tissue; ante- rior maxillomandibular ligament absent; articular and angular spines absent; mesopterygoid absent; hyomandibular with a sin- gle head; 6 branchiostegal rays (2 + 4); opercle reduced, bifur- cate; interopercle reduced, no ligamentous connection with angular; preopercle reduced; pharyngobranchial 1 absent; pharyngobranchials II and 111 heavily toothed; epibranchial and ceratobranchial teeth absent; epibranchial I absent; ante- rior half of epibranchials III and IV fused; ceratobranchial I reduced, represented only by posterior half and tiny isolated remnants of anterior half; ceratobranchial V absent except for tiny isolated remnants; no ossified hypobranchials or basibranchials; posterior margin of hypural plate entire; caudal fin emarginate (except in largest females of G. kreffti and G. niacronema), with 9 rays, ventralmost ray reduced and embed- ded within skin surrounding adjacent ray; pterygiophore of il- licium exceptionally large, compressed, its posterior end abutting up against supraoccipital; ossified remnant of second cephalic ray present; illicium greater than 60% SL; 5 pectoral radials; pelvic bones reduced or absent. Metamorphosed gigantactinid males are distinguished from those of all other ceratioid families by the following combina- tion of characters: eyes minute (diameter 3-5% SL in most specimens); olfactory organs large (depth 8-10% SL in most specimens); anterior nostrils close together, opening anteriorly; premaxillae degenerate; jaw teeth absent; denticular teeth all or nearly all mutually free; upper denticular teeth 3-6 (rarely 2), not connected to pterygiophore of illicium; lower denticular teeth 4-7 (rarely 3); hyomandibular with a single head; branchiostegal rays 6 (rarely 7); pectoral radials 5; pelvic bones absent; probably nonparasitic. In addition to the sexual dimorphism common to ceratioids, gigantactinid males differ from females of the family in having a symphysial cartilage, a vomer, and a basibranchial ossifica- tion. They also differ in having fully developed frontals, parie- tals, opercular bones, and ceratobranchials. Gigantactinid larvae differ from those of the other ceratioid families in having exceptionally large pectoral fins (length 45- 55% SL), comparable only to those of the Caulophrynidae (Bertelsen 1951, Pietsch 1979). They differ from caulo- phrynids, however, in the absence of pelvic fins. KEYS TO THE GENERA OF THE GIGANTACTINIDAE Females IA. Lower jaw teeth well developed in several rows; dorsal-fin rays 5-9, rarely 4 or 10; anal-fin rays 4-7, rarely 8; escal bulb present Gigantactis Brauer 1902 I B. Lower jaw teeth absent; dorsal-fin rays 3-4, rarely 5; anal- fin rays 3-4; escal bulb absent Rhynchactis Regan 1925 Males 1A. Three upper and 4 lower denticles; dorsal-fin rays 5-9, rarely 4 or 10; anal-fin rays 4-7, rarely 8; skin spinulose in some species Gigantactis Brauer 1902 Contributions in Science, Number 332 Bertelsen, Pietsch & Lavenberg: Gigantactinid Anglerfishes 23 1 B. Four upper and 6 lower denticles; dorsal-fin rays 3-4, rarely 5; anal-fin rays 3-4; skin naked Rhynchactis Regan 1925 Larvae IA. Dorsal group of pigment weakly developed or absent, when present extending posteriorly to base of dorsal fin but never contiguous with peritoneal pigment; dorsal-fin rays 5-9, rarely 4 or 10; anal-fin rays 4-7, rarely 8 Gigantactis Brauer 1902 IB. Dorsal group of pigment strongly developed terminating in front of base of dorsal fin but contiguous with peritoneal pigment; dorsal-fin rays 3-4, rarely 5; anal-fin rays 3-4 Rhynchactis Regan 1925 SYNOPSIS OF GIGANTACTINID CLASSIFICATION The listing of taxa below summarizes the taxonomic conclu- sions made in this study. Full descriptions of all taxa and complete generic synonymies follow. Genus Gigantactis Brauer 1 902 Gigantactis longicirra Waterman 1939b Gigantactis kreffti new species Gigantactis vanhoeffeni group Gigantactis vanhoeffeni Brauer 1902 Gigantactis meadi new species Gigantactis gibhsi new species Gigantactis gracilicauda Regan 1925 Gigantactis paxtoni new species Gigantactis perlatus Beebe and Crane 1947 Gigantactis elsmani new species Gigantactis golovani new species Gigantactis gargantua group Gigantactis gargantua new species Gigantactis watermani new species Gigantactis herwigi new species Gigantactis macronema group Gigantactis macronema Regan 1925 Gigantactis savagei new species Gigantactis microdontis new species Gigantactis ios new species Gigantactis ovifer Regan and Trewavas 1932, incertae sedis Gigantactis filibulbosus Fraser-Brunner 1935, incertae sedis Genus Rhynchactis Regan 1925 Rhynchactis leptonema Regan 1925 Genus Gigantactis Brauer Females: Gigantactis Brauer 1902:295-296 (type species Gi- gantactis vanhoeffeni Brauer 1902, by original designation). Males: Teleotrema Regan and Trewavas 1932:92-93, fig. 149 (type species Teleotrema microphthalmus Regan and Trewavas 1932, by original designation). Laevoceratias Parr 1927:33, fig. 13 (type species Laevoceratias liparis Parr 1927, by original designation). 24 Contributions in Science, Number 332 DIAGNOSIS. The genus Gigantactis is distinguished from Rhynchactis in lacking pelvic bones and by having D. 5-9 (rarely 4 or 10) and A. 5-7 (rarely 4 or 8). In addition, meta- morphosed females differ in having the following characters: frontals present; parietals present; premaxilla developed, with teeth present throughout length; maxilla represented by threadlike remnant; dentary with several rows of strong, re- curved teeth; 1 hypohyal; all caudal rays unbranched (in con- trast to males); skin spinulose; snout produced in front of mouth bearing illicium on tip; esca with luminous bulb. Metamorphosed males are distinguished from those of Rhynchactis by having the following characters: diameter of eye greater than 3% SL (usually 3.5-5% SL); olfactory lamellae 12 (rarely 10 or 11); depth of nostrils rarely greater than 9% SL; upper denticular teeth 3 (rarely 2 or 4); lower denticular teeth 4 (rarely 2, 3, or 5); all bases of denticular teeth compressed and mutually free; skin pigmented or unpig- mented, spinulose or naked (weakly pigmented and naked in Rhynchactis). Larvae of Gigantactis differ from those of Rhynchactis in having the following characters: dorsal group of subdermal pig- ment absent or only weakly developed, never contiguous with peritoneal pigment but, when best developed, extending pos- teriorly behind base of dorsal fin; length of pectoral fin approx- imately 45% SL. KEYS TO SPECIES OF THE GENUS GIGANTACTIS Adolescent and Adult Females Gigantactis ovifer and G. filibulbosus, each represented by only a poorly preserved holotype, and here referred to as incertae sedis. are omitted from the key. IA. Dorsal-fin rays 8-10, the first and last distinctly longer than intermediate rays; length of first and eighth caudal- fin rays 60-100% SL (esca, Fig. 24) G. longicirra Waterman 1939b, p. 26 IB. Dorsal-fin rays 4-7, all about the same length; length of first and eighth caudal-fin rays less than 40% SL .... 2 2A. Esca with a black, spiny, distal prolongation (G. vanhoef- feni group) 3 2B. Esca without a black, spiny, distal prolongation 7 3 A. Esca with filaments on base; illicium 70-120% SL ... 4 3B. Esca without filaments on base (Fig. 37); illicium 1 60— 200% SL G. paxtoni new species, p. 39 4A. Esca without distally flattened papillae (Fig. 35) G. gracilicauda Regan 1925, p. 38 4B. Esca with distally flattened papillae 5 5A. Illicium with a pair of small, papilliform or digitiform ap- pendages on posterior margin below esca; distal prolonga- tion of esca conical and confluent with escal bulb (Fig. 30) G. vanhoeffeni Brauer 1902, p. 31 5B. Illicium without a pair of appendages below esca; distal prolongation of esca constricted at base 6 Bertelsen, Pietsch & Lavenberg: Gigantactinid Anglerfishes 6A. Esca with distal prolongation nearly cylindrical, more than twice as long as wide, and covered with short fil- aments (Fig. 32) G. meadi new species, p. 33 6B. Esca with distal prolongation conical, about as long as wide, with filaments restricted to tip (Fig. 34) G. gibbsi new species, p. 36 7A. Illicial length 60-120% SL 8 7B. Illicial length 130-490% SL, rarely less than 200%. . 10 8A. Esca with a posterior pair of appendages at base, fringed in juveniles, divided into branched filaments in older specimens; conical, distal prolongation of esca longer than diameter of escal bulb (Fig. 39) G. perlatus Beebe and Crane 1947, p. 41 8B. Esca without posterior pair of appendages; distal part of esca shorter than diameter of escal bulb 9 9A. Base of escal bulb with a posterior, median papilla, with- out long filaments; conical, distal part of esca with lateral series of short filaments (Fig. 28) G. kreffti new species, p. 29 9B. Base of escal bulb without posterior papilla, but with a pair of long stout filaments; distal part of esca with two pairs of long filaments and several shorter ones (Fig. 41) G. elsmani new species, p. 43 IOA. Esca with distal filaments branched, and several fil- aments arising from base and below 11 IOB. Esca with distal filaments unbranched, and without pos- terior filaments on or below base 13 1 1 A. Esca with a single anteroproximal filament (Fig. 43) G. golovani new species, p. 44 1 IB. Esca with several anteroproximal filaments 12 12A. Esca with 4-5 pairs of distal filaments (Fig. 45) .... large specimens of G. gargantua new species, p. 46 12B. Esca with 8 pairs of distal filaments (Fig. 52) . . . large specimens of G. macronema Regan 1925, p. 50 13A. Esca with a group of anterior filaments arising from base; distal part of escal bulb bearing 4-5 pairs of stout fil- aments along posterior margin; second and seventh caudal-fin rays more than 50% SL (G. gargantua group) 14 13B. Esca without anterior filaments arising from base; fil- aments of distal part of bulb not as above; longest caudal- fin rays less than 40% SL ( G . macronema group) ... 16 14A. Esca with less than 10 (7 in a 262-mm specimen) fil- aments arising from base; distal part of escal bulb only slightly truncated, pigmented swellings at base (Fig. 50) G. herwigi new species, p. 49 14B. Esca with more than 10 filaments at base; distal part of escal bulb truncated, bearing 4-5 pairs of filaments with pigmented, swollen bases 15 15A. Esca with less than 15 (11 in a 99-mm specimen) fil- aments arising from base; distal filaments swollen and pigmented for one-half their length (Fig. 48) G. watermani new species, p. 49 1 5B. Esca with more than 1 5 filaments at base; distal filaments swollen and pigmented for less than one-fifth their length (Fig. 44) G. gargantua new species, p. 46 16A. Esca with distal prolongation 17 16B. Esca without a distal prolongation 18 17A. Esca with distal prolongation truncated, bearing several simple filaments arranged in nearly symmetrical pairs (Fig. 51) G. macronema Regan 1925, p. 50 17B. Esca with distal prolongation tapering to a point, fil- aments not arranged in pairs (Fig. 54) G. savagei new species, p. 53 18A. Esca with a U-shaped series of about 10 short, spatulate filaments surrounding a distal patch of pigment (Fig. 56) G. microdontis new species, p. 54 18B. Esca with a dense group of 16 short, lanceolate filaments (with tiny, internal bulbs) arising from a distal patch of pigment (Fig. 59) G. ios new species, p. 56 Males IA. Skin spinulose 2 IB. Skin naked 3 2A. Skin darkly pigmented; eyes relatively small, diameter 0.4-0. 6 mm Gigantactis Male Group II 2B. Skin unpigmented; eyes relatively large, diameter 0.7- 0.8 mm Gigantactis Male Group III 3A. Distinct V-shaped patch of subdermal pigment on throat (Fig. 62); number of dorsal and anal-fin rays 4 Gigantactis Male Group V 3B. No V-shaped patch of subdermal pigment on throat; number of dorsal and anal-fin rays 5-7 4 4A. Number of dorsal-fin rays 8-10 Gigantactis longicirra Waterman 1939b 4B. Number of dorsal-fin rays 5-7 5 5A. Number of pectoral-fin rays 18-22; eyes relatively large, diameter 0.6-0. 9 mm; number of olfactory lamellae 11- 12 Gigantactis Male Group I 5B. Number of pectoral-fin rays 15; eyes relatively small, di- ameter 0.5 mm; number of olfactory lamellae 8-9 Gigantactis Male Group IV Larvae 1A. Dorsal and peritoneal pigment well developed 2 1 B. Dorsal and peritoneal pigment weak or absent 3 2A. Dorsal and ventral series of three to four large, subder- mal melanophores on caudal peduncle Gigantactis Larval Group D Contributions in Science, Number 332 Bertelsen, Pietsch & Lavenberg: Gigantactinid Anglerfishes 25 2B. No large, subdermal melanophores on caudal peduncle Gigantactis Larval Group A 3A. Number of dorsal-fin rays 5-7, anal-fin rays 5-7 Gigantactis Larval Group B 3B. Number of dorsal-fin rays 8-10, anal-fin rays 5-8 (Fig. 26B) Gigantactis Larval Group C (= Gigantactis longicirra Waterman) Gigantactis longicirra Waterman Figures 1 A, 4F, 1 1 A,D, 1 3 A, 1 4A, 1 5 A, 16, 17, 24-26, 66; Tables 1-3, 19 Gigantactis longicirra Waterman 1939b: 82-85, figs. 1-2 (original description, single specimen); Waterman 1949:81- 149, figs. 1-10 (comparative anatomy; comparison with other species of the genus, other ceratioid genera and families); Clark 1950:6, 10, 18, 28 (problems in classifying ceratioid lateral line organs, references to Waterman 1948); Bertelsen 1951:1 50— 152, table 31 (comparison with all known material, comments); Grey 1956:269 (synonymy, vertical distribution). Gigantactis sp., Becker et al. 1975:327 (specimen tentatively referred to G. longicirra). MATERIAL. Eight females (seven metamorphosed [34.5-221 mm] and one in metamorphosis [19.5 mm]), two meta- morphosed males (14-14.5 mm) and eight larvae (4. 5-7. 5 mm). Holotype of Gigantactis longicirra : MCZ 35065, 39 mm, 39°06'N, 70° 16'W, closing net at 1000 m, bottom depth 2860 m. Females: IOAN uncatalogued, 34.5 mm, 18°29'N, 80°33'W, 0-1500 m; ISH 2561/71, 2 (108-118 mm), 7°32'N, 20°54'W, 0-1300 m; ISH 973/68, 209 mm, 4°43'S, 26°39'W, 0-2000 m (cleared and stained); MCZ 52570, 221 mm. Gulf of St. Law- rence; SIO 60-241, 19.5 mm, 7°26'N, 144°29'W, 0-2100 m; SIO 60-215, 72.5 mm, 13° 13'N, 127°06'W, 0-2300 m, bottom depth 4612 m. Males: UMML 27412, 14 mm, 11°40'N, 68°16'W (cleared and stained); UMML 2741 1, 14.5 mm, 4°56'N, 00°13'E. Larvae: ZMUC P92 1656-68, listed as Gigantactis “Type C” larvae by Bertelsen (1951:274). DIAGNOSIS. Gigantactis longicirra differs from all other spe- cies of the genus in having a relatively high number of dorsal- fin rays (8-10). Metamorphosed females are further distin- guished in having three or fewer distal, escal filaments, the first and last rays of the dorsal fin distinctly longer than the inter- mediate rays, and, in addition, the following combination of characters: illicial length less than 120% SL (39-105%); escal bulb without distal prolongation or papillae; dentary teeth rela- tively long (longest 3. 3-5.0% SL), in 5-6 longitudinal series; first and eighth rays of caudal fin prolonged (60-100% SL). Metamorphosed males of G. longicirra are further charac- terized by having the following combination of characters: eyes small, diameter 0.45 and 0.5 mm; olfactory lamellae 1 1; upper denticular teeth 3; lower denticular teeth 4; skin naked, unpig- mented (Table 19). Larvae of G. longicirra are further distinguished in lacking dorsal, subdermal pigment (See “Comments” on this species). Table 3. Counts and measurements in percent of SL of females of Gigantactis longicirra. Character S101 60-241 IOAN uncata- logued Holotype MCZ 35065 S10 60-215 ISH 2561/71 ISH 2561/71 ISH 973/68 MCZ 52570 Standard length (mm) 19.5 34.5 39 72.5 108 118 209 221 Length Illicium 13 + 62.3 + 38.5 45.1 104 96.6 broken 59.3 Longest premaxillary tooth 0.5 2.6 1.9 2.7 1.8 1.7 Longest dentary tooth 1.0 4.9 5.2 3.7 4.1 3.3 4.7 Longest caudal ray 38.0 broken 60.3 82.3 64.8 + 86.4 broken 98.6 Teeth Premaxillary 3 5 8 13 20 35 22 23 Dentary 5 9 14 17 41 52 60 22 Dorsal-fin rays 8 8 8 9 8 8 9 9 Anal-fin rays 6 7 6 6 7 6 6 7 Pectoral-fin rays 16 15 18 16 15-15 15-16 16 14-15 'Metamorphosal stage. 26 Contributions in Science, Number 332 Bertelsen, Pietsch & Lavenberg: Gigantactinid Anglerfishes DESCRIPTION OF FEMALES. Illicium without filaments, nearly cylindrical throughout, length highly variable and unre- lated to standard length (Table 3). Escal bulb pear-shaped in juveniles, elongated and constricted below photophore in adults, without spines on distal portion; 1-3 short, distal fil- aments; an anterolateral, proximal group of short filaments; and a posteroproximal group of long filaments. Esca of 39-mm holo- type (described and figured by Waterman 1939b) with pear- shaped, unpigmented, and naked bulb; a single distal filament and a total of 20 proximal filaments; all filaments unpigmented and unbranched with slight distal swellings. Escal bulb of 118- Figure 24. Esca of Gigantactis longicirra, 118 mm, ISH 2561/71, left lateral view. Arrow indicates opening of pore of photophore. Drawn by K. Elsman. mm specimen elongate and darkly pigmented (Fig. 24), with a distinct constriction below distally placed photophore; skin of proximal part of bulb covered with small spines; a single, un- pigmented distal filament bearing a small terminal swelling; all of approximately 25 filaments of anterolateral-proximal group unpigmented and arranged in an anterior series (of which the proximal is branched) and a lateral, more scattered grouping of 6-7 single filaments; posteroproximal group darkly pigmented and arranged in two symmetrical, fan-shaped groups, each con- sisting of numerous filaments, long and short (some of which are branched), arising from a common base. Esca of 221-mm specimen similar in all major characters to that of 118-mm specimen but differs in having 3 distal filaments, and those of anterolateral-proximal group all unbranched and pigmented. Number of teeth in each premaxilla increasing with standard length from 5 (34.5-mm specimen) to 35 (118-mm specimen) but decreasing in largest specimens (22 and 23 teeth in the 209- mm and 221-mm specimens, respectively; Table 3); longest pre- maxillary tooth (present in the anterior portion of the jaw) 1.7- 2.7% of SL. Number of teeth in each dentary increasing with standard length from 9 (34.5 mm) to approximately 60 (209 mm), fewer (22 teeth) at 221 mm (this largest specimen, MCZ 52570, has lost many of the older and smaller teeth). Teeth in posterior part of dentary in 5-6 longitudinal series in specimens greater than 100 mm; number of series increasing from 2 (the median and first external) at 34.5 and 39 mm; first internal se- ries appearing at 72.5 mm; second and third internal series present in larger specimens (but many older and smaller lost in the largest specimen, MCZ 52570); longest dentary teeth pres- ent in anterior, irregular groups (Fig. 25). Anal-fin rays 6-8, pectoral-fin rays 14-18 (Table 2); first ray of dorsal fin longest (78 and 80% SL in 108- and 1 18-mm speci- Figure 25. Gigantactis longicirra, female, 118 mm, ISH 2561/71: A. Dentary teeth, left lateral view; B. Diagrammatic representation of den- tary-tootb pattern. E - external series; M = median series; I = inter- nal series. Contributions in Science, Number 332 Bertelsen, Pietsch & Lavenberg: Gigantactimid Anglerfishes 27 Figure 26. Gigantactis longicirra: A. Male, 14 mm, UMML 27412; B. Larval female, 4.7 mm, ZMUC P921668. mens, respectively; broken in other specimens); caudal rays free nearly to base without remains of connecting membrane, first and eighth rays prolonged 60-100% SL, becoming longer with increased SL (Fig. 4F, Table 3). Skin spines and lateral line organs on head and body slightly larger than in other members of genus. The 19.5-mm metamorphosal stage (SIO 60-241) with esca lost; jaw teeth few and rudimentary, those of dentary in a single (median) series; first caudal ray broken, eighth caudal ray 38% SL; seven anteriormost supraorbital lateral line organs enlarged on each side, the longest approximately 5% SL; skin faintly pig- mented; subdermal pigment absent. DESCRIPTION OF MALES AND LARVAE (Fig. 26). See “Di- agnosis” above, and “Males” and “Larvae” in the section on Gigantactis sp. unidentified. DISTRIBUTION. Gigantactis longicirra inhabits the tropical Atlantic from the Gulf of Guinea in the east to the Caribbean Sea in the west, the western north Atlantic as far north as ap- proximately 50°N, and the eastern tropical Pacific (Fig. 66). All metamorphosed female material of G. longicirra was col- lected by gear fished open at maximum depths of between 1000 and 2300 m. The 39-mm holotype was captured with a closing net at 1000 m. COMMENTS. The females of Gigantactis longicirra are simi- lar to those of a number of other Gigantactis species (Table 1) in having a relatively short illicium and long teeth in several rows in the lower jaw. Except for these similarities, the species shows no distinct affinity with any other species of the genus. The esca is unique in several characters: the low number of distal filaments combined with the lack of a distal prolongation of the bulb, and (in adults) the constriction of the bulb below the photophore as well as the posteroproximal pair of darkly pigmented, fan-shaped appendages. The prolongation of the first and last dorsal ray is found in no other Gigantactis species; greatly prolonged caudal rays also occur in G. gargantua. G. watermani, and G. herwigi, but in these species, the second and seventh rays are the longest. In both metamorphosed males, the testes are relatively short and narrow (0.7 and 0.8 mm in diameter), yet the specimens appear to have passed metamorphosis as evidenced by their well-developed denticulars, resorbed premaxillae, and close-set olfactory organs. Since G. longicirra is unique among the recognized species of the genus in having more than seven dorsal-fin rays, the eight larvae with such high dorsal-ray counts referred to Group C by Bertelsen (1951) no doubt represent this species. On the other hand, the two larvae with seven dorsal rays included in this group by Bertelsen (1951) should be removed to the uniden- 28 Contributions in Science, Number 332 Bertelsen, Pietsch & Lavenberg: Gigantactinid Anglerfishes Figure 27. Gigantactis kreffti , holotype, 252 mm, ISH 1099/71. tified larvae of Group B. The absence of dorsal subdermal pig- ment in G. longicirra is confirmed in the identified female metamorphosal stage (SIO 60-241) and in the two meta- morphosed males (UMML 2741 1, 27412). Gigantactis kreffti new species Figures 4D, 27-29, 65; Tables 1, 2, 4 MATERIAL. Four metamorphosed females, 44.0-252 mm. Holotype: ISH 1 099/7 1,252 mm, 39 ° 1 9'S, 3 ° 1 5'W, 0-2000 m. Paratypes: ISH 1262/71, 75 mm, 34° 12'S, 16°35'E, 0-1550 m; ISH 3236/71, 185 mm (data as for holotype); ZIFSUT 52706, 44 mm, 35°06'N, 139°24'E, 0-1000 m, bottom depth 1200-1400 m. DIAGNOSIS. Metamorphosed females of G. kreffti are dis- tinguished from those of all other species of the genus in having the following combination of characters: illicial length less than 120% SL (69-94%); distal escal prolongation with spines at base and several unpigmented, digitiform filaments on lateral margins; escal papillae absent; a posteromedial papilla at base of bulb and a second on illicium below bulb; proximal escal filaments absent; dentary teeth relatively long (longest 2.6- 5.2% SL), in 4-5 longitudinal series; rays of caudal fin less than 30% SL. Males and larvae unknown. DESCRIPTION. Illicium without filaments, nearly cylindrical throughout, only proximal part slightly compressed, length var- iable and unrelated to standard length (Table 4). Escal bulb pear-shaped with short, conical distal prolongation; spines cov- ering proximal part of bulb, reaching base of distal prolonga- tion on anterior margin. Escal bulb of 252-mm holotype (Fig. 28D) with distal prolongation and oval area surrounding pore of photophore, naked and unpigmented; remaining surface of es- cal bulb darkly pigmented and covered with spines; pore of photophore protruding from surface of bulb as a short tube; lateral margins of distal prolongation bearing approximately 20 digitiform, unpigmented filaments; an unpigmented, stout pa- pilla present posteriorly at base of bulb and a similar, but slightly smaller papilla on illicium about 65 mm (25% SL) be- low base of bulb (in the freshly caught, unpreserved specimen, the tips of the distal filaments and the papillae were red; be- neath the skin below the pore, a bright silvery, circular area was noted). Escae of paratypes (Fig. 28A-C) similar to that of holo- type except for some changes with growth: in smaller speci- mens (44 and 75 mm), escal bulb more spherical in shape; at 44 mm bulb without pigmentation, at 75 mm pigmentation reach- ing posterior part of base of bulb; number of distal filaments increasing from 4-5 on each side at 44 mm to 7-8 at 75 mm and to approximately 30 at 185 mm; in two smallest specimens, papilla on illicium situated at a distance of about 10% SL be- low base of bulb (papilla and part of the skin of the illicium are lost in the 185-mm specimen). Left premaxilla of holotype with 44 teeth of which approx- imately 10, posteriormost, curved forward; longest tooth in up- per jaw 1.3% SL. Number of premaxillary teeth of paratypes increasing from 7 at 44 mm to 17 at 75 mm and 30 at 185 mm; longest premaxillary tooth 1.3-2. 5% SL. Holotype with about 47 teeth in each dentary, posteriorly arranged in 5 longitudinal series, one external, one median, and three internal (Fig. 29). Number of dentary teeth of paratypes increasing from 15 at 44 mm to 27 at 75 mm and 42 at 185 mm; 4 posterior longitudinal series present except in the smallest paratype (44 mm) in which second internal series not yet developed. Dorsal-fin rays 7, anal-fin rays 6, pectoral-fin rays 16-18 (Table 2); skin covering caudal fin for some distance beyond fin base; skin of each ray broad, distally tapering, and basally con- nected by transparent membrane (Fig. 4D). DISTRIBUTION. Gigantactis kreffti, known from only four metamorphosed females, has an unusual disjunct distribution with three individuals captured in the South Atlantic off the tip of Africa and one in the western North Pacific off Japan (Fig. 64). Probably when this form becomes better known, it will Contributions in Science, Number 332 Bertelsen, Pietsch & Lavenberg: Gigantactinid Anglerfishes 29 Jill) A (I v*^” / << Cr ..•6«J*55^ - 1 mm 1 mm i 1 1 mm 1 1 mm Figure 28. Escae of Gigantactis kreffti: A. Paratype, 44 mm, ZIFSUT 52706, left lateral view; B. Paratype, 75 mm, ISH 1262/71, left lateral view; C. Paratype, 185 mm, ISH 3236/71, posterior view; D. Holotype, 252 mm, ISH 1033/71, posterolateral view. Drawn by K. Elsman. prove to be a wide-ranging if not cosmopolitan species. All ma- terial of G. kreffti was taken with open gear fished at maximum depths of between 1000 and 2000 m. ETYMOLOGY. Gigantactis kreffti is named for Gerhard Krefft of the Institut fur Seefischerei, Hamburg; his interest and energy have established an important ichthyological re- source that has immensely expanded our knowledge of the deep-sea fauna. COMMENTS. Gigantactis kreffti is similar to a number of other Gigantactis species (Table 1 ) in having a relatively short illicium and long teeth in several posterior series in the lower jaw. It is similar to members of the G. vanhoeffeni-gvoup (in- cluding G. vanhoeffeni, G. meadi, G. gibbsi, G. gracilicauda, and G. paxtoni ) in having a relatively spinulose esca with a dis- tal prolongation bearing short filaments but differs from these species in lacking both spines on the distal prolongation and filaments on the proximal part of the esca. The relatively short Figure 29. Gigantactis kreffti, holotype, female, 252 mm, ISH 1099/71: A. Premaxillary and dentary teeth, left lateral view; B. Dia- grammatic representation of dentary-tooth pattern. E = external series; M - median series; I = internal series. 30 Contributions in Science, Number 332 Bertelsen, Pietsch & Lavenberg; Gigantactinid Anglerfishes Table 4. Counts and measurements in percent of SL of females of 67- gantactis kreffti. Character Paratype ZIFSUT 52706 Paratype ISH 1262/71 Paratype ISH 3236/71 Holotype ISH 1099/71 Standard length (mm) 44 75 185 252 Length Illicium 94.0 81.3 68.6 86.5 Longest premaxillary tooth 2.5 1.3 1.4 1.3 Longest dentary tooth 5.2 3.3 2.6 3.4 Longest caudal ray 27.7 27.3 17.8 19.8 Teeth Premaxillary 7 17 30 44 Dentary 13 27 42 47 Dorsal-fin rays 7 7 7 7 Anal-fin rays 6 6 6 6 Pectoral-fin rays 18 16-16 17-17 16-17 caudal fin with well-developed skin coverage is similar to those of G. meadi, G. gibbsi, and G. perlatus. Gigantactis vanhoeffeni Brauer Figures IB, 4A, 5, 6, 8, 12, 19, 21, 30, 31, 64; Tables 1, 2, 5 Gigantactis vanhoeffeni Brauer 1902:296 (original descrip- tion, two specimens, lectotype hereby designated ZMHU 17712, paralectotype lost, used by Brauer 1908 for anatomical, histological description of the illicial apparatus and esca). Boulenger 1904b:720 (brief description); Brauer 1906:322, pi 15, figs. 8, 9 (description after Brauer 1902); Brauer 1908:103, 184, pi. 31, Figs. 18-20, pi. 32, figs. 1-5, pi. 34, fig. 14, pi. 44, fig. 1 (anatomical, histological description of illicial apparatus, esca, and eyes); Gill 1909:586, fig. 25 (description, figure after Brauer 1902, 1906, habits); Regan 1926:38 [three additional specimens, two subsequently described as new ( G . sexfilis and G. exodon) by Regan and Trewavas 1932]; Regan and Trewavas 1932:93, 94 (description after Brauer 1902, 1906, in key); Fowler 1936:1345, 1346, fig. 565 (brief description, figure after Brauer 1902, 1906, Regan 1926); Waterman 1939b:85 (lateral line organs on caudal rays, comparison with G. long- icirra)\ Beebe and Crane 1947:168 (comparison with G. per- latus)-, Bertelsen 1951:150-152, table 31 (comparison with all known material, comments); Grey 1956:267 (synonymy, verti- cal distribution); Parin and Golovan 1976:271 (in part, two ad- ditional specimens). Gigantactis exodon Regan and Trewavas 1932:93, 94, 95, fig. 151 (original description, single specimen, in key); Fraser-Brun- ner 1935:326 (comparison with G. filibulbosus)\ Waterman 1939b:84, 85 (comparison with G. longicirra ); Bertelsen 1951:150-152, table 31 (comparison with all known material, comments); Grey 1956:268 (synonymy, vertical distribution); Pietsch 1972a:42, 45 (holotype with five pectoral radials). Gigantactis perlatus, Parin et al. 1973:145-146 (misidentifi- cations, three specimens). Gigantactis “sp. 2,” Parin et al. 1973:146 (typical juvenile). MATERIAL. Forty-nine females: 46 metamorphosed (19-340 mm), and three in metamorphosis (16.5-21.5 mm). Lectotype of Gigantactis vanhoeffeni: ZMF1U 17712, 35 mm, 5°42'S, 43 ° 36'E, 2500 m. Holotype of Gigantactis exodon: ZMUC P92128, 25 mm, 1 8 ° 50'N, 79°07'W, 2500 m wire. Referred female material: IOAN uncatalogued, 16.5 mm, 9°09'N, 1 18°24'W, 0-1000 m; IOAN uncatalogued, 21.5 mm, 3°20'S, 82°02'W, 0-1300 m; IOAN uncatalogued, 23 mm, 3°2TN, 81°02'W, 0-1300 m; IOAN uncatalogued, 29 mm, 16°07'S, 53°39'E, 0-1000 m; IOAN uncatalogued, 37 mm, 32°01'N, 143°15'E, 0-5300 m; IOAN uncatalogued, 48 mm, Akademik Kurchatov Cruise 17, Station 1454; IOAN un- catalogued, 57 mm, 3°22'N, 120°07'W, 0-1000 m; IOAN un- catalogued, 96 mm, 2°03'S, 82°31'W, 0-1500 m; IOAN uncatalogued, 102 mm, 2°03'S, 82°31'W, 0-1000 m; IOAN uncatalogued, 196 mm, 5°03'N, 2°08'E, 0-4000 m; IOAN un- catalogued, 2 (198-205 mm), 10”36'N, 17°38'W, 0-1550 m; IOAN uncatalogued, 285 mm, 8°00'S, 59°27'E, 0-1300 m; IOAN uncatalogued, 340 mm, off Peru, 0-900 m; ISH 2188/- 71, 38 mm, 2°29'S, 18°58'W, 0-304 m; ISH 661/68, 57 mm, 12°07'N, 23 °08'W, 0-2000 m; ISH 2331/71, 3 (75, 109, 232 mm), T04'N, 18°22'W, 0-2100 m; ISH 872/68, 114 mm, 0° 14'N, 25°22'W, 0-600 m; ISH 802/68, 152 mm, 4°08'N, 24°41'W, 0-600 m (cleared and stained); ISH 721/68, 152 mm, 8°21'N, 24°10'W, 0-520 m; ISH 1959/71, 195 mm, 1 0° 57'S, 11°20'W, 0-1900 m; ISH 2480/71, 266 mm, 4°38'N, 19°41'W, 0-756 m; ISH 376/73, 270 mm, 62°39'N. 33°45'W, 0-2100 m; LACM 36031-1, 17 mm, 4°55'S, 129°48'E, 650- 810 m; LACM 31470-1, 22 mm, 00°00', 140°53'W, 0-1170 m; LACM 36039-6, 2 (24-31 mm) 5°08'S, 130°08'E, 650-1000 m; LACM 31526-1, 24 mm, 2°07'S, 169°43'E, 0-1200 m; LACM 36032-1, 25 mm, 4°47'S, 129°52'E, 0-1500 m; LACM 36046-10, 26 mm, 4°54'S, 129°43'E, 0-850 m; LACM 31521-2, 27 mm, 0°18'S, 169°57'E, 0-1200 m; LACM 36131-1, 32 mm, 4°56'S, 129°50'E, 550-815 m; LACM 31513-1, 28 mm, COO'S, 170°00'E, 0-1200 m; LACM 31497-2, 29 mm, 0°51'S, 170°04'E, 0-1030 m; LACM 36012-1, 30 mm. Central Tropical Pacific, 600-950 m; LACM 36034-5, 34 mm, 4°58'S, 130°12'E, 0-750 m; LACM 36013-1, 42 mm, 11°17'S, 142°47'W, 0-1200 m; LACM 36954-1, 305 mm, off southern California, 0-1070 m; LACM 37079-1, 315 mm, San Clemente Basin, California (no other data); MCZ 52571, 19 mm, 7°33'N, 64°41'E, 0-500 m; SIO 68-535, 29.5 mm, 4°59'N, 164° 1 4' W, 2550 m of wire out, bottom depth 4575 m; SIO 70-309, 35 mm, 29°32'N, 137°14'E, 0-1900 m, bottom depth 4277-4630 m; SIO 73-159, 39 mm, 31°02'N, 1 55 °04'W, 0-1450 m; UMML 11336, 33 mm, 25°02'N, 79°48'W, 0-320 m; UMML 23821, 92 mm, 4°00'N, 2°46'W, 0-490 m, bottom depth 1995-2233 m; UMML 23973, 232 mm, 6°54'N, 79°58'W, bottom depth 3184 m; ZMUC Contributions in Science, Number 332 Bertelsen, Pietsch & Lavenberg: Gigantacdnid Anglerfishes 31 P921972, 67.5 mm, 4‘14'S, 44°52'E, 0-1050 m (cleared and stained). DIAGNOSIS. Metamorphosed females of G. vanhoeffeni differ from those of all other species of the genus in having the follow- ing combination of characters: illicial length less than 120% SL (71-112% in specimens greater than 25 mm); escal bulb with an elongate, spinulose, and darkly pigmented distal prolonga- tion; escal bulb and distal prolongation bearing distally flat- tened papillae; short distal, and slender proximal escal filaments present; illicium with a posterior pair of papillae be- low escal bulb; dentary teeth relatively long (longest 2. 6-5.0% SL) in 3 longitudinal series; rays of caudal fin less than 45% SL. Males unknown (probably included in Gigantactis Male Group II). Larvae unknown (probably included in Gigantactis Larval Group A). DESCRIPTION. Illicium distinctly compressed in specimens greater than 50 mm SL, depth at base 2-3 times width, length variable and unrelated to SL. Escal bulb gradually tapering into a conical distal prolongation, darkly pigmented proximally, unpigmented and transparent around photophore, and less darkly pigmented distally; bulb covered to distal tip with spines and bearing several large, distally flattened papillae of varying pigmentation, some black, others, especially in skin covering photophore, nearly transparent; distal filaments short and re- stricted to tip of distal prolongation, increasing in number from one in specimens less than 30 mm to 10-30 in specimens greater than 100 mm; 2-3 thin filaments on each side of bulb immediately below photophore, the longest reaching to base of distal prolongation; a close-set pair of small appendages arising on posterior margin of illicium some distance below bulb, pa- pilliform and very small in juveniles, larger and somewhat com- pressed in adults, bifurcated in some of largest specimens (Figs. 30-31); distance from base of paired, illicial papillae to tip to distal filaments 10-22% SL in specimens less than 100 mm, 7- 14% SL in larger specimens; 1 to approximately 10 (increasing in number with standard length) short filaments present on pos- terior margin of distal portion of illicium of specimens greater than 70 mm SL (Figs. 30B, 31). In freshly captured, un- preserved specimens, the paired papillae and the tips of the dis- tal filaments are bright red. Number of teeth in each premaxilla increasing with standard length from 2-8 in smaller specimens to 22-53 in larger (Table 5); longest premaxillary tooth 1 .0-1.8% SL in specimens greater than 25 mm SL. Number of teeth on each dentary in- creasing with standard length from 3-15 in smaller specimens to 32-66 in larger (Table 5); median series of small dentary teeth distinctly developed (Figs. 5, 6). Dorsal-fin rays 5-7, anal-fin rays 5-6 (rarely 7), pectoral-fin rays 17-18 (rarely 19; Table 2); caudal fin divided between up- per and lower lobes nearly to base; caudal-skin coverage weakly developed, skin of fin rays only slightly compressed; remains of transparent membrane between bases of rays present in some specimens (Fig. 4A). Three metamorphosal stages (16.5 mm, IOAN uncata- logued; 17 mm, LACM 36031-1; 21.5 mm, IOAN un- catalogued) with illicial length 23.5-41% SL; escae with more or less distinct rudiments of papillae and paired posterior ap- pendages; jaw teeth rudimentary, premaxilla with 2-4 teeth, dentary with 3-5 teeth (representing the median series and one or two of the first external series); four to eight anteriormost supraorbital lateral line organs enlarged, the longest about 7% SL; dark subdermal dorsal and peritoneal pigment. DISTRIBUTION. Gigantactis vanhoeffeni has a cosmopolitan distribution between approximately 63°N and 15°S (Fig. 64; but see “Comments” concerning the northern-most record of this species, ISH 376/73). This species has been taken with open gear fished at maximum depths between 300 and 5300 m. 1mm 2mm Figure 30. Escae of Gigantactis vanhoeffeni, left lateral views: A. 38 nun, ISH 2188/71; B. 152 mm, ISH 721/68. Drawn by K. Elsman. 32 Contributions in Science, Number 332 Bertelsen, Pietsch & Lavenberg: Gigantactinid Anglerfishes but appears to be most commonly found between 700 and 1300 m. COMMENTS. Gigantactis vanhoejfeni is one of several Gigan- tactis species (Table 1 ) that have a relatively short illicium. It is the type of a group of species (the G. vanhoejfeni group, includ- ing G vanhoejfeni. G. meadi, G. gibbsi, G. gracilicauda, and G. paxtoni) characterized by having a distal, spinulose prolonga- tion of the escal bulb bearing short filaments and (except for G. paxtoni ) having at the base of the bulb a small number of more slender filaments that reach to the base of the distal prolonga- tion. Like G. meadi. G. gibbsi. and G. paxtoni. this species has distally flattened escal papillae but differs distinctly from these forms in the shape of the distal prolongation of the bulb and in having a posterior pair of papillae on the illicium below the bulb. Even though the characters that distinguish G. vanhoejfeni from its nearest relatives are not mentioned in Brauer’s (1902) original description (based on two specimens now lost except for badly damaged remains of one), his excellent figure of the esca (Brauer 1908, pi. 15, fig. 9) shows the characteristic coni- cal shape of its distal prolongation with the filaments restricted Figure 31. Esca of Gigantactis vanhoejfeni, 270 mm, ISH 376/73, left lateral view. Drawn by K. Elsmam. to the tip, as well as the paired appendages on the illicium be- low the bulb. Two of the specimens of the referred material have lost the esca, but in both, the characteristic paired papillae are present on the remaining part of the illicium. In one of them (IOAN uncatalogued), the loss of the esca seems to have happened be- fore capture. The esca has been broken off just below the pho- tophore, leaving a wound that has healed without sign of regeneration of the lost part. A second break just below the paired appendage has healed in such a way that the remains of the esca are attached to the illicium in a twisted and displaced position. The esca of the 270-mm specimen (ISH 376/73, represent- ing the most northerly capture [63 °N] of a gigantactinid) shows some aberrant features (Fig. 31 ). In addition to the clus- ter of filaments at the tip of the distal prolongation, it has nu- merous short, tapering filaments and papillae on the proximal part of the prolongation as well as on the proximal part of the bulb. A pair of filaments on the posterior margin of the illicium below the bulb probably represents the paired appendages characteristic of G. vanhoejfeni. but they are neither com- pressed nor particularly distinct from other filaments present on the illicium in this same region. Although these aberrant features are not present in the three specimens examined of similar or larger size, there is a general tendency among Gigan- tactis species for the number of distal filaments of the esca as well as those of the illicium to increase with increasing size. For this reason, we assume for the present that this variation is due to age, possibly combined with individual differences. Gigantactis meadi new species Figures 4B, 18, 20, 23, 32, 33, 64; Tables 1, 2, 6 MATERIAL. Fourteen females: ten metamorphosed ( 35.5— 353 mm), and three in late metamorphosis (19.0-21.0 mm). Holotype: MCZ 52572, 306 mm, 34°14'S, 64°56'E (depth unknown). Paratypes: ISH 415/76, 87 mm, 47°45'S, 40°04'W, 0-2000 m; ISH 1004/71, 2 (155-201 mm), 36°49'S, 12°17'W, 0-2000 m; ISH 965/71, 178 mm, 39°45'S, 17°40'W, 0-2000 m; ISH 571/76, 207 mm, 39°08'S, 40°00'W, 0-1850 m; ISH 883/71, 230 mm, 39°55'S, 26°02'W, 0-2000 m; ISH 1465/71, 290 mm, 33°00'S, 7°50'W, 0-2000 m; LACM 1 1242-12, 353 mm, 39°58'S, 150°31'W, 0-1900 m, bottom depth 5161 m; SAM 2781 1, 21.0 m, 30°06'S, 31°57'E, 0-750 m; USNM 208032, 35.5 mm, 33°06'S, 83°57'W, 0-1050 m, bottom depth 3731— 3822 m. Referred material: The following are tentatively referred to G. meadi: IOAN uncatalogued, 290 mm, 29°35'S, 14° 13'E, 0- 1300 m (esca lost, illicial length 93% SL, strong filaments on head at base of illicium); SAM 27807, 19 mm, 30°45'S, 30°42'E, 0-830 m; SAM 27808,20 mm, 31°14'S, 30°21'E, 0- 1120 m. DIAGNOSIS. Metamorphosed females of G. meadi differ from those of other species of the genus (except for G. paxtoni ) in having filaments on the dorsal surface of the head just behind Contributions in Science, Number 332 Bertelsen, Pietsdi & Lavenberg: Gigantactinid Anglerfishes 33 Table 5. Counts and measurements in percent of SL of females of Gigantactis vanhoeffeni. Z 2 "2 <* cd =J O c g> u ^ 3- ■— 1 m x ^ GO vO — r- X CO 7 cd-O < c« 3 O | g> c/i ^ On X C/3 X ^ C/3 (N X C/3 2 ™ is 2 ■ U- r i ^ J rd © (N y E x 3 U •- 00 C - 5 a § fi ■ op m —I O. , o o 3ll O ca 2P -a -5 £ a | o a 2Holotype of Gigantactis exodon Regan and Trewavas 1932. Table 6. Counts and measurements in percent of SL of females of Gigantactis meadi. Character SAM' 2781 1 Paratype USNM 208032 Paratype ISH 415/76 Paratype ISH 1004/71 Paratype ISH 965/71 Paratype ISH 1004/71 Paratype ISH 571/76 Paratype ISH 883/71 Paratype ISH 1465/71 Holotype MCZ 52572 Paratype LACM 1 1242-12 Standard length (mm) 21 35.5 87 155 178 201 207 230 290 306 353 Length Illicium 52.0 87.6 72.4 94.2 94.4 9.1 96.1 79.1 83.4 73.5 78.5 Longest premaxillary tooth 0.8 1.4 1.1 1.4 1.0 1.3 1.0 1.3 1.6 1.8 1.1 Longest dentary tooth 1.9 3.4 3.3 3.6 3.5 3.7 3.8 3.6 3.6 3.3 2.9 Longest caudal ray 33.0 28.2 20.7 + 23.9 + 23.6 27.4 20.9 + 23.0 22.1 28.4 24.9 Teeth Premaxillary 4 2 8 20 15 23 20 31 20 20 27 Dentary 13 14 24 51 33 51 53 63 58 51 81 Dorsal-fin rays 6 6 6 6 7 6 6 6 6 6 6 Anal-fin rays 6 6 6 6 6 5 6 6 6 6 6 Pectoral-fin rays 16 18 — 17 17 17 18 18 18 18 17 1 Metamorphosal stage. the base of the illicium. They are further distinguished in hav- ing the following combination of characters: illicial length less than 120% SL (72-96%); short filaments present along entire posterior margin of illicium; escal bulb with an elongate, spin- ulose, black, distal prolongation, slightly constricted at base; es- cal bulb and distal prolongation bearing distally flattened papillae; short distal and slender proximal escal filaments pres- ent; posterior pair of close-set illicial appendages absent; den- tary teeth relatively long (longest 2. 9-3. 8% SL), in 5-6 longitudinal series; rays of caudal fin short (less than 30% SL). Males unknown (probably included in Gigantactis Male Group II). Larvae unknown (probably included in Gigantactis Larval Group A). DESCRIPTION. Proximal part of illicium distinctly com- pressed, its depth at base more than twice width, length vari- able and unrelated to SL. Escal bulb club-shaped with cylindrical, distal prolongation more than twice as long as wide (about 3 times in the holotype), spinulose except for area sur- rounding pore of photophore (Fig. 32); small, distally flattened papillae present from tip of distal prolongation to some distance below bulb; papillae white or only weakly pigmented; 15 to ap- proximately 20 short filaments on distal prolongation, most con- centrated at tip, a few (only 2 in the holotype) below its base distal to photophore; a posterolateral group of filaments at base of bulb, some slightly longer than those of distal group; fil- aments along posterior margin of illicium (34 illicial filaments in holotype, 8-42 in paratypes) continuing onto head forming a cluster (16 cephalic filaments in holotype, 2-18 in paratypes). Number of teeth in each premaxilla of holotype 20, increas- ing with standard length from 2-8 in smaller specimens to 20- 27 in larger specimens (Table 6); longest premaxillary tooth 1.8% SL in holotype, 1.0-1. 8% in metamorphosed paratypes. Number of dentary teeth of holotype 51, increasing with stan- dard length from 14-24 in smaller specimens to 51-81 in larger specimens (Table 6); teeth in posterior part of lower jaw in 5-6 longitudinal series: 2 external series, a distinct median series of small teeth, and 1 or 2 internal series; anterior part of lower jaw of most specimens with a third external series (E3 in Fig. 33) containing largest teeth of jaw, 3.3% SL in holotype, 2. 9-3. 8% SL in paratypes (Table 6). Dorsal-fin rays 6-7, anal-fin rays 5-6, pectoral-fin rays 17-18 (Table 2); skin coverage of caudal fin well developed and com- plete for more than half length of rays; skin of each ray gradu- ally tapering distally and connected by transparent membranes (Fig. 4B). Two small females (19 mm, SAM 27807; 20 mm, SAM 27808), tentatively referred to this species, and the 21-mm par- atype (SAM 2781 1) represent an ontogenetic series of late met- amorphosal stages: illicial length 16, 16, and 52% SL, respectively; distal prolongation of esca short, unpigmented, and naked at 19 and 20 mm, elongate, cylindrical, and pig- mented, bearing papillae and short filaments at 21 mm; rudi- mentary dentary teeth in a single series at 19 and 20 mm. Contributions in Science, Number 332 Bertelsen, Pietscb & Lavenberg: Gigantactinid Anglerfishes 35 arranged in three series (a median series of eight teeth, a first external series of four, and a single tooth representing the sec- ond external series) at 21 mm; no enlarged lateral-line organs on head; dorsal and peritoneal subdermal pigment well devel- oped. A third specimen (290 mm, IOAN uncatalogued), tenta- tively referred to G. meadi, has lost the esca but has apparently retained a complete illicial bone (measuring approximately 90% SL). The cephalic filaments present at the base of the illicium, as well as characters of the teeth and caudal fin, agree with the description above. DISTRIBUTION. Gigantactis meadi is circumglobal in and about the subtropical convergence of the Southern Ocean (Fig. 64), where it appears to be a relatively deep-living form. All specimens 87 mm and larger were taken by gear fished open at maximum depths of between 1850 and 2000 m. ETYMOLOGY. This species is named for Giles W. Mead in recognition of his many contributions to ichthyology, but also for his service as Chief Scientist on the cruise during which the holotype of this new species was collected. COMMENTS. Gigantactis meadi belongs to the G. vanhoejfeni group characterized by having a black, spinulose, distal pro- longation of the escal bulb. In common with three members of H I 1 Figure 32. Escae of Gigantactis meadi, left lateral views: A. Paratype, 35.5 mm, USNM 208032; B. Holotype, 306 mm, MCZ 52572. Drawn by K. Elsman. this group (G. vanhoejfeni, G. gibbsi, and G. paxtoni), it has distally flattened papillae on the esca. It differs from G. van- hoejfeni in having a cylindrical, distal prolongation that is somewhat constricted at the base, a cluster of filaments on the head just behind the base of the illicium, and a shorter caudal fin that is covered to a greater extent by skin. It differs from G. gibbsi and G. paxtoni mainly in the shape and length of the distal prolongation. It further differs from G. paxtoni in illicial length. Gigantactis gibbsi new species Figures 34, 64; Tables 1, 2, 7 MATERIAL. Two metamorphosed females, 38 and 50 mm. Holotype; ZIAN 44262, 50 mm, 2”01'N, 3°56'W, 0-465 m. Paratype: USNM 218613, 38 mm, 33°00'N, 64“06'W, 0- 1000 m. DIAGNOSIS. Metamorphosed females of G. gibbsi differ from those of other species of the genus in having the following com- bination of characters: illicial length less than 120% SL (104- 1 18%); escal bulb with a short, conical, spinulose, black, distal prolongation; escal bulb bearing distally flattened papillae, and short distal and slender proximal filaments; illicium without posterior pair of papillae; dentary teeth posteriorly in 3 longitu- dinal series; rays of caudal fin less than 50% SL. Males unknown (probably included in Gigantactis Male Group II). Larvae unknown (probably included in Gigantactis Larval Group A). DESCRIPTION. Escal bulb of holotype pear-shaped with a short conical, distal prolongation, darkly pigmented except at base; skin covering area of photophore and base of distal pro- longation unpigmented; spines present on distal prolongation and bulb except around pore of photophore; distally flattened, unpigmented papillae present on proximal part of bulb; approx- imately 12 short filaments at tip of distal prolongation; a group of filaments of varying length lateroproximal to bulb, longest Figure 33. Gigantactis meadi, paratype, female, 207 mm, ISH 571/76: A. Dentary teeth, left lateral view; B. Diagrammatic representation of dentary-tooth pattern. E = external series; M = median series; I = internal series. 36 Contributions in Science, Number 332 Bertelsen, Pietsch & Lavenberg: Gigantactinid Anglerfishes about equal to width of bulb (Fig. 34A). Juvenile paratype with distal prolongation of escal bulb darkly pigmented and spin- ulose except at base, bearing a short, terminal filament (Fig. 34B); bulb and base of distal prolongation naked and unpig- mented; skin torn away from proximal part of bulb and distal part of illicium, with only a few remains of proximal filiaments. Flolotype with 13 teeth (the longest 1.4% SL) in each pre- maxilla, paratype with 6 (the longest 1.3% SL); holotype with 27 teeth (longest 2.6% SL) in each dentary, paratype with 23 (longest 2.4% SL; Table 7). Tooth pattern similar to that of G. vanhoeffeni of similar standard length, with 3 distinct, longitu- dinal series in posterior part of lower jaw; median series of small teeth well developed. Dorsal-fin rays 6, anal-fin rays 6, pectoral-fin rays 17-18 (Table 2); longest caudal-fin ray (second and seventh) 49% SL in holotype, 26-28% in paratype; skin coverage of caudal-fin relatively well developed. 2 mm 1 mm Figure 34. Escae of Gigantactis gibbsi, left lateral views: A. Holotype, 50 mm, ZIAN 44262; B. Paratype, 38 mm, USNM Z18613. Drawn by K. Elsman. DISTRIBUTION. Gigantactis gibbsi is known from two speci- mens collected from the Atlantic Ocean, the holotype taken be- tween the surface and 465 m from equatorial waters of the Gulf of Guinea, and a second individual from between the sur- face and 1000 m off Bermuda (Fig. 64). ETYMOLOGY. This species is named for Robert H. Gibbs, Jr., in recognition of his many contributions to deepsea biology. COMMENTS. Gigantactis gibbsi is a typical member of the G. vanhoeffeni group (Table 1) having a relatively short illicium, a spinulose, distal prolongation of the esca, and distally flattened papillae on the escal bulb. It differs from G. vanhoeffeni, G. meadi, and G. paxtoni in the shape and length of the distal pro- longation and in the length of the longest caudal-fin ray. It fur- ther differs from G. vanhoeffeni in lacking a pair of papilliform or flattened appendages on the posterior margin of the illicium below the escal bulb, from G. paxtoni in illicial length, and from G. gracilicauda in having distinct, distally flattened escal papillae. The two specimens on which G. gibbsi is based (38 and 50 mm) are juveniles; older specimens of this form could conceiv- ably undergo ontogenetic changes that would indicate con- specificity with some other species of Gigantactis. A compari- son of this material with a complete growth series of G. vanhoeffeni (16.5-340 mm) indicates that the differences ob- served in G. gibbsi cannot be explained as part of the variation within the more well represented G. vanhoeffeni. Some of the differences between G. gibbsi and the series of specimens repre- senting G. meadi could perhaps be explained in this way, but it seems unlikely that the latter species, known only from the Southern Ocean, would occur north of the equator. Finally, it can hardly be assumed that these two juveniles represent G. gracilicauda or G. paxtoni; besides involving great ontogenetic changes in the shape of the esca, this hypothesis would also Table 7. Counts and measurements in percent of SL of females of Gigantactis gibbsi. Paratype Holotype USNM ZIAN Character 218613 44262 Standard length (mm) 38 50 Length Illicium 104 118 Longest premaxillary tooth 1.3 1.4 Longest dentary tooth 2.4 2.6 Longest caudal ray 28.4 49.0 Teeth Premaxillary 6 13 Dentary 23 29 Dorsal-fin rays 6 6 Anal-fin rays 6 6 Pectoral-fin rays 17 18 Contributions in Science, Number 332 Bertelsen, Pietsch & Lavenberg: Gigantactimid Anglerfishes 37 imply an ontogenetic loss of either the escal papillae or the proximal escal filaments. Gigantactis gracilicauda Regan Figures 35, 64; Tables 1, 2, 8 Gigantactis gracilicauda Regan 1925:565 (original descrip- tion, single specimen); Regan 1926:18, 19, 38, fig. 12, pi. 10, fig. 2 (cranial osteology, description after Regan 1925); Regan and Trewavas 1932:93, 94 (brief description, in key); Water- man 1948:90, 93 (osteological comparison with G. longicirra, correction of errors made by Regan 1926); Bertelsen 1951:1 50— 152, table 31 (comparison with all known material, comments); Grey 1956:267 (synonymy, vertical distribution); Pietsch 1972a:42, 45 (holotype with five pectoral radials). Gigantactis sexfilis Regan and Trewavas 1932:38, 39, 93, 94, figs. 49, 50, 150, pi. 5, fig. 2 (original description, single speci- men, osteology of skull and pectoral lobe, in key); Waterman 1939b:84 (comparison with G. longicirra, holotype with four pectoral radials after Regan and Trewavas 1932); Waterman 1948:88 (ossification of skull of G. longicirra less extensive than that of G. sexfilis as figured by Regan and Trewavas 1932); 2 mm Figure 35. Escae of Gigantactis gracilicauda, left lateral views: A. Ho- lotype, 82 mm, ZMUC P92129, drawn by K. Elsman; B. Holotype of G. sexfilis, 51 mm, ZMUC P92132, after Regan and Trewavas 1932. Bertelsen 1951:150, table 31 (comparison with all known mate- rial, comments); Grey 1956:267 (synonymy, vertical distribu- tion); Pietsch 1972a:42, 45 (holotype with five pectoral radials). Gigantactis sp., Bertelsen 1951:150-152, fig. 100, table 31 (G. vanhoeffeni of Regan 1926). MATERIAL. Three metamorphosed females, 21-82 mm. Holotype of Gigantactis gracilicauda: ZMUC P92129, 82 mm, 13°47'N, 61 °26'W, 4500 m wire. Holotype of Gigantactis sexfilis: ZMUC P92132, 51 mm, 13°07'N, 57 ° 20'W, 4000 m wire. Referred material: ZMUC P921535, 21 mm, 25°H'N, 20° 57'W, 5000 m wire. DIAGNOSIS. Metamorphosed females of G. gracilicauda dif- fer from those of other species of the genus in having the fol- lowing combination of characters: illicial length less than 120% SL (104-107%, 86% in the 21-mm juvenile); escal bulb with an elongate, spinulose, and darkly pigmented distal prolongation; short distal and slender proximal escal filaments present; dis- tally flattened escal papillae absent; illicium without posterior pair of papillae; dentary teeth posteriorly in 3 longitudinal se- ries; rays of caudal fin less than 30% SL. Males unknown (probably included in Gigantactis Male Group II). Larvae unknown (probably included in Gigantactis Larval Group A). DESCRIPTION. Illicium slightly compressed laterally in larger specimens. Escal bulb gradually tapering to form a darkly pigmented and slender distal prolongation; median por- tion of bulb containing unpigmented photophore; skin of bulb and distal prolongation covered with spines but without distinct papillae; distal filaments short, restricted to distal part of taper- ing prolongation; 1 or 2 pairs of slender filaments below escal bulb not reaching base of distal prolongation. Escal bulb of ho- Table 8. Counts and measurements in percent of SL of females of Gigantactis gracilicauda. Character ZMUC P921535 ZMUC1 P92132 Holotype ZMUC P92129 Standard length (mm) 21 51 82 Length Illicium 85.7 104 107 Longest premaxillary tooth - 1.0 0.7 Longest dentary tooth — 2.4 2.4 Longest caudal ray — 30.4 28.0 Teeth Premaxillary — 7 10 Dentary — 23 43 Dorsal-fin rays 5 6 6 Anal-fin rays 5 6 6 Pectoral-fin rays 19 18 18 1 Holotype of Gigantactis sexfilis Regan and Trewavas 1 932. 38 Contributions in Science, Number 332 Bertelsen, Pietsch & Lavenberg: Gigantactinid Anglerfishes lotype of G. gracilicauda (Fig. 35A) with about 20 short, distal filaments, two pairs of slender filaments on posterior margin of illicium just below bulb, and below these, an additional short illicial filament. Bulb of holotype of G. sexfilis (Fig. 35B) with 4 short, distal filaments, and a single pair of slender illicial fil- aments at base of bulb. Esca of juvenile (21.0 mm) unpig- mented except for small spot on distal tip, with a few spines, but papillae absent; about 8 short, distal filaments and 2 illicial pairs of very thin, lateral filaments at base of bulb (some tiny filament-like structures, 0.2 to 0.4 mm long, present on the an- teroproximal margin of the bulb may be due to abraded skin). Flolotype of G. gracilicauda with 10 teeth in each premaxilla (longest 0.7% SL), holotype of G. sexfilis with 7 teeth (longest 1.0% SL), juvenile with 4 (all in development; Table 8). Holo- type of G. gracilicauda with 43 dentary teeth (longest 2.4% SL), holotype of G. sexfilis with 23 teeth (longest 2.4% SL), juvenile with approximately 10 (all in development). Tooth pat- tern similar to that of G. vanhoeffeni of similar size, with 3 dis- tinct, longitudinal series in posterior part of lower jaw. Dorsal-fin rays 5-6, anal-fin rays 5-6, pectoral-fin rays 18-19 (Table 2); longest caudal-fin rays (second and seventh) 26 and 28% SL in holotype of G. gracilicauda. both 30% SL in holo- type of G. sexfilis (both broken in juveniles); skin coverage of caudal fin well developed. DISTRIBUTION. Gigantactis gracilicauda is known from three individuals all collected in the tropical Atlantic Ocean (Fig. 64), where it appears to be a relatively deep-living form. All were taken by gear fished at maximum depths of between approximately 2000 and 2500 m. COMMENTS. Gigantactis gracilicauda is a member of the G. vanhoeffeni group (Table 1), having a relatively short illicium and a spinulose, darkly pigmented distal prolongation of the es- cal bulb bearing short filaments. It resembles G. vanhoeffeni and G. paxtoni in the shape of the distal prolongation but dif- fers from these species and other members of the G. van- hoeffeni group in that it lacks distally flattened escal papillae. The esca of the holotype of G. gracilicauda was not described originally by Regan (1925, 1926) or later by Regan and Trewavas (1932). However, its general shape and presence of some distal filaments is shown in Regan’s ( 1926) illustrations of the whole fish. Our figure (Fig. 35A) is to some extent a recon- struction, as the esca is now rather dehydrated and shrunken. Among the spines in its wrinkled, dark skin are some wartlike protuberances, which could be interpreted as the remains of distally flattened papillae but might well be artifacts. The esca of the holotype of G. sexfilis (Fig. 35B), well described and fig- ured by Regan and Trewavas (1932), is still in good condition; no papillae are present. These distally flattened escal papillae characteristic of several well-represented Gigantactis species ( G . vanhoeffeni. G. meadi) are developed at metamorphosis and indicate no distinct individual or ontogenetic variation. Gigan- tactis gracilicauda is recognized as a distinct species mainly be- cause these escal papillae are lacking. The differences observed between the holotype of G. gracilicauda and the two smaller specimens are within the expected range of variation for this species; for this reason, G. sexfilis is regarded as a junior syn- onym of G. gracilicauda. Gigantactis paxtoni new species Figures 36, 37, 38, 64; Tables 1, 2, 9 Gigantactis “sp. 1,” Parin et al., 1977:68-188 (single speci- men). MATERIAL. Eight metamorphosed females, 50-232 mm. Holotype: AMS 1.20314-018, 237 mm SL, 100 km east of Broken Bay, N.S.W., 33°28'S, 152°33'E, 0-900 m over 4200 m, 14 December 1977. Paratypes: AMS 1.20070-016, 124 mm SL, northeast of Cape Howe, N.S.W., 37°24'S, 150°30'E, 0-540 m over 3600 m, 1 November 1977. AMS 1.20306-007, 142 mm SL, 65 km east of Broken Bay, N.S.W., 33°31'S, 152"20'E, 0-900 m over 1800-2900 m, 12 December 1977. AMS 1.20314-018, 3 (175- Figure 36. Gigantactis paxtoni, holotype, 237 mm, AMS 1.20314-018. Drawn by R. Nielsen. Contributions in Science, Number 332 Bertelsen, Pietsch & Lavenberg: Gigantactinid Anglerfishes 39 228 mm SL), data as for holotype. IOAN uncatalogued, 50 mm SL, V1TYAZ Station 7288, 3”39'N, 131'22'E, 0-1500 m (Parin et al. 1977). IOAN uncatalogued, 210 mm SL, ZWIEZDA KRYMA Station 83, 34‘07'S, 44”50'E, 0-1260 m, 1976. 5mm 5mm 1 i Figure 37. Escae of Gigantactis paxtoni: A. Holotype, 237 mm, AMS 1.20314-018, left lateral view; B. Paratype, 50 mm, IOAN uncatalogued, left posterolateral view. Drawn! by R. Nielsen. DIAGNOSIS. Metamorphosed females of Gigantactis paxtoni differ from those of other species of the genus (except for G. meadi ) in having filaments on the dorsal surface of the head just behind the base of the illicium. The species is further dis- tinguished by the following combination of characters: illicial length 168-198% of SL; short filaments present on base of illi- cium; escal bulb gradually tapering into a conical, spinulose and darkly pigmented, distal prolongation measuring 12-28% of SL; escal bulb and distal prolongation bearing low unpig- mented papillae; short filaments present on distal prolongation, but absent on base of escal bulb; posterior pair of close-set illici- al appendages absent; dentary teeth long (longest tooth 3.4- 7.1% of SL), in 3-4 longitudinal series in posterior part of jaw; rays of caudal fin short (27.5-35% of SL). Males and larvae unknown. DESCRIPTION. Proximal portion of illicium distinctly com- pressed, depth at base more than twice width. Escal bulb grad- ually tapering into a distal prolongation several times as long as wide (approximately 9 times as wide in holotype), spinulose and pigmented except for area surrounding escal pore (Fig. 37); small papillae present from tip of distal prolongation to some distance below escal bulb; papillae low (height less than width), unpigmented, and transparent; approximately 10-30 short fil- aments on distal prolongation, none proximal to its base; fil- aments along posterior margin of proximal part of illicium (except in smallest known specimen) continuing onto head and forming an anterodorsal cluster on snout. Number of teeth in each premaxilla of holotype 15, increas- ing with standard length from 7 in 50-mm SL specimen to 10- 19 in larger specimens (Table 9); longest premaxillary tooth 1.3% of SL in holotype, 1 .0-1.8% in paratypes. Number of den- tary teeth of holotype 46, increasing from 10 in 50-mm SL specimen to 28-55 in larger specimens (Table 9); teeth in pos- terior part of lower jaw in 3 longitudinal series; anterior part of Figure 38. Gigantactis paxtoni, paratype, 210 mm, IOAN un- catalogued: A. Dentary teeth, left lateral view; B. Diagrammatic repre- sentation of dentary-tooth pattern. E = external series; M = median series; I = internal series. 40 Contributions in Science, Number 332 Bertelsen, Pietsch & Lavenberg: Gigantactinid Anglerfishes lower jaw of most specimens with a second external series con- taining largest teeth of jaw, 5.6% SL in holotype, 3.4-7. 1% SL in paratypes (Table 9, Fig. 38). Dorsal-fin rays 6-7, anal-fin rays 5-6, pectoral-fin rays 18-21 (Table 2); skin coverage of caudal fin relatively well developed and complete for about one-third of length of rays; skin of each ray gradually tapering and connected by transparent mem- branes. DISTRIBUTION. Six of the eight known specimens of Gigan- tactis paxtoni were caught off the southeast coast of Australia near the northern boundary of the Subantarctic Water Mass between 33° and 37 °S. A seventh specimen was collected from the western South Indian Ocean and an eighth from the west- ern tropical Pacific off the northwest coast of New Guinea (Fig. 64). With the exception of a single individual taken by bottom trawl at 1210-1260 m, the material was collected by pelagic gear fished open at maximum depths of between 540 and 1500 m (over bottom depth of 1800-4200 m). ETYMOLOGY. This species is named for John R. Paxton of the Australian Museum, Sydney, in recognition of his contribu- tions to deepsea ichthyology and in gratitude for providing most of the material on which this new form is based. COMMENTS. Gigantactis paxtoni belongs to the G. van- hoeffeni group characterized by having a darkly pigmented, spinulose distal prolongation of the escal bulb. It differs from all other members of this group in having a considerably longer illicium and a longer distal prolongation of the escal bulb and in lacking proximal escal filaments. In common with three members of this group (G. vanhoeffeni. G. meadi, and G. gibb- si ), G. paxtoni has distally flattened escal papillae; these pa- pillae differ from those of the other species, however, in being distinctly lower. Gigantactis perl at us Beebe and Crane Figures 1C, 4C, 39, 40, 65; Tables 1,2, 10 Gigantactis perlatus Beebe and Crane 1947:167-168, text fig. 13, pi. II, fig. 3 (original description, single specimen); Ber- telsen 1951:150, 151, table 31 (comparison with all known ma- terial, comments); Grey 1956:268 (synonymy, vertical distribu- tion); Mead 1958:133 (holotype transferred from NYZS to SU); Robins and Courtenay 1958:151 (depth distribution); Parin et al. 1973:145-146 (misidentifications, three specimens here referred to G. vanhoeffeni). MATERIAL. Eight females: seven metamorphosed (23-223 mm), and one in metamorphosis (20 mm). Holotype of Gigantactis perlatus: CAS-SU 46487 (originally NYZS 28621), 32.5 mm, 7°08'N, 81°57'W, 0-915 m. Referred material: ISH 1466/71, 152 mm, 33°00'S, 7°50'E, 0-2000 m; LACM 37518-1, 20 mm, Hawaii, off leeward Oahu, 0-800 m; LACM 36875-2, 36 mm, Hawaii, off leeward Oahu, 670-805 m; MCZ 51327, 23 mm, 39°38'N, 70°03'W, 0-1000 m; SIO 61-31, 222 mm, 11°57'S, 115°22'E; USNM 218614, 41 mm, 32°00'N, 64°00'W, 0-1340 m; USNM 208111, 223 mm, 35 °05'S, 91°59'W, 0-1650 m. DIAGNOSIS. Metamorphosed females of G. perlatus differ from those of other species of the genus in having an extremely large distal prolongation of the escal bulb (20% SL in 220-mm specimen), the entire esca densely covered with slightly elon- gated papillae (not distally flattened papillae as in members of the G. vanhoeffeni group). They are further distinguished in Table 9. Counts and measurements in percent of SL of females of Gigantactis paxtoni. Character Paratype IOAN uncata- logued Paratype AMS 20070-016 Paratype AMS 20306-007 Paratype AMS 20314-018 Paratype IOAN uncata- logued Paratype AMS 20314-018 Paratype AMS 20314-018 Holotype AMS 20314-018 Standard length (mm) 50 124 142 175 210 218 228 237 Length Illicium 168 185 190 187 192 190 158 + ? 198 Longest premaxillary tooth 1.0 1.5 1.1 1.4 1.8 1.6 1.3 1.3 Longest dentary tooth 3.4 6.1 7.0 7.1 6.9 5.0 5.0 5.6 Longest caudal ray 28 28.3 35 33 31.5 27.5 30.5 31.5 Teeth Premaxillary 7 15 10 17 17 19 19 15 Dentary 10 36 42 55 30 28 39 46 Dorsal-fin rays 6 6 6 6 7 7 6 6 Anal-fin rays 6 6 5 6 6 6 6 6 Pectoral-fin rays 19 18 19-19 20-21 19-19 18-19 20-20 20-20 Contributions in Science, Number 332 Bertelsen, Pietsch & Lavenberg: Gigamtactinid Anglerfishes 41 having the following combination of characters: illicial length less than 120% SL (74-111% in specimens greater than 30 mm); a pair of posterior escal appendages; dentary teeth large (longest 2. 2-5. 9% SL), arranged in 2 irregular, longitudinal se- ries; rays of caudal fin less than 35% SL. Males and larvae unknown. DESCRIPTION, lllicium somewhat laterally compressed in largest specimens, length variable and unrelated to SL. Escal bulb with a tapering, distal prolongation, increasing in size with standard length from 5-6% SL in specimens less than 50 mm to about 20% SL in specimens of approximately 220 mm; entire bulb and distal prolongation without spines, densely covered with unpigmented, slightly elongated papillae; a few, short, paired filaments along posterior margin of distal prolongation; a pair of wing-shaped appendages on posterior margin near base of bulb, with fringed edges in juveniles, divided into numerous branched filaments in older specimens; two pairs of long, proxi- mal filaments, larger pair on lateral margin slightly below base of wing-shaped appendages, smaller pair just below and on each side of pore of photophore (Fig. 39). Esca of three smaller specimens (23-45 mm) very similar to that of 34-mm holotype (described and figured by Beebe and Crane 1947), having a total length about 10% SL; slightly elon- gated papillae increasing in number and coverage of bulb with standard length; paired, wing-shaped posteroproximal appen- dages simple in 36-mm specimen, fringed on posterior edge in other specimens; 2-3 paired filaments along posterior margin of distal prolongation; 2 proximal pairs of larger filaments (Fig. 39A). Distal prolongation of three largest specimens (152-223 mm) greatly lengthened, total escal length 15% SL at 152 mm, 22 and 27% SL at 222 and 223 mm, respectively; bulb and distal prolongation densely covered with small, elongate pa- pillae except for narrow posterior area around tubular opening of pore of photophore; numerous short filaments on distal third of prolongation in addition to paired filaments on posterior mar- gin; all filaments of distal prolongation except proximal pair un- pigmented and simple, proximal pair branched in two largest specimens; paired, wing-shaped appendages divided into nu- merous short branches in 1 52-mm specimen (Fig. 39B), into nu- merous, thin filaments in largest specimens. Esca of fresh, unpreserved specimens “completely semi-translucent white” (except for black photophore) in holotype (Beebe and Crane 1947:167); opaque, pearly white except for pink and silvery area around pore of photophore in 1 52-mm specimen (E. Ber- telsen, personal observation); bright red in 223-mm specimen (according to color photo of fresh specimen provided by Anne Cohen). Teeth few; number of teeth on each premaxilla 10 in 223-mm specimen, 2-3 in all others, longest tooth approximately 1.0% SL; 5-10 dentary teeth in specimens 23 to 41 mm (longest about 6% SL), 14-18 in specimens 153-223 mm (longest 2.2- 3.0% SL), teeth arranged in 2 very irregular series throughout length of jaw (Fig. 40). Dorsal-fin rays 5-7, anal-fin rays 5-7, pectoral-fin rays 16-19 (Table 2); longest caudal rays (second and seventh) 22-31% SL; skin coverage of caudal fin and thin membranes between rays well-developed (Fig. 4C, Table 10). The 20-mm metamorphosal stage (LACM 37518-1) with il- licial length 19% SL; esca with relatively short distal prolonga- tion, no papillae, rudiments of posterior paired appendage and Table 10. Counts and measurements in percent of SL of females of Gigantactis perlatus. Character LACM1 37518-1 MCZ 52573 Holotype SU 46487 LACM 36875-2 USNM 218614 ISH 1466/71 SIO 61-31 USNM 208111 Standard length (mm) 20 23 32.5 36 41 152 222 223 Length lllicium 19.0 25.2 108 73.6 107 102 111 94.2 Longest premaxillary tooth Longest dentary tooth 5.9 5.0 2.2 3.0 3.0 Longest caudal ray 27.0 — — 26.4 31.2 22.4 27.0 — Teeth Premaxillary 0 — - 3 2 3 3 10 Dentary 0 — ■- 10 6 10 17 14 18 Dorsal-fin rays 6 5 5 7 6 5 6 5 Anal-fin rays 6 6 5 7 6 5 6 6 Pectoral-fin rays 19 17 16 18 18 17 18 17-18 'Metamorphosal stage. 42 Contributions in Science, Number 332 Bertelsen, Pietsch & Lavenberg: Gigantactinid Anglerfishes of three pairs of filaments; edge of premaxilla and dentary in resorption; jaw teeth absent; enlarged cephalic lateral line organs absent; skin and subdermal pigment absent (possibly bleached out). DISTRIBUTION. Gigantactis perlatus is known from eight specimens collected from seven widely distributed capture sites: two in the western North Atlantic, one in the South At- lantic off the tip of Africa, one in Indonesian waters, one in the Hawaiian Islands, one in the Gulf of Panama, and one in the eastern South Pacific at approximately 35 °S, 91 "W (Fig. 65). The 36-mm Hawaiian specimen (LACM 36875-2) was col- lected with a closing net between 670 and 805 m. The remain- ing material was captured by open gear fished at maximum depths of between 800 and 2000 m. COMMENTS. Among the Gigantactis species characterized by having relatively short illicia (less than 120% SL; Table 1 ), G. perlatus is distinguished by the characters of the esca: the numerous, small, slightly elongated papillae (not distal ly flat- tened as in members of the G. vanhoeffeni group), the extreme development of the distal prolongation, and the characteristic pattern of the filaments and appendages. The presence of paired, posteroproximal appendages is shared only with adult specimens of G. longicirra ; but this species, besides lacking a Figure 39. Escae of Gigantactis perlatus, left lateral views: A. 23 mm, MCZ 51327; B. 152 mm, ISH 1466/71. Drawn by K. Elsman. distal prolongation of the bulb, differs from G. perlatus in tooth pattern, median fin-ray counts, and caudal fin-ray lengths. A similar tooth pattern in which few, relatively large dentary teeth are arranged in two irregular series is found only in G. golovani. The latter species is very distinct from G. perlatus, however, in the length of the illicium (greater than 180% SL in G. golovani ) and in nearly all escal characters (Figs. 39, 43). Gigantactis elsmani new species Figures 4E, 41, 42, 65: Tables 1, 2, 1 1 MATERIAL. Two metamorphosed females (283-384 mm). Holotype: ISH 1360/71, 384 mm, 10°57'S, 11°20'W, 0- 1900 m. Paratype: LACM 10687-1, 283 mm, 63°00'S, 1 14°34'W, 0- 2932 m, bottom depth 5051 m. Referred material: The following specimen is tentatively re- ferred to G. elsmani: MCZ 51269, 1 1.5 mm, 34°26'S, 16° 1 l'E. DIAGNOSIS. Metamorphosed females of G. elsmani differ from those of other species of the genus in having a single prox- imal pair and two distal pairs of large escal filaments and in lacking a distinct distal prolongation of the escal bulb. They are further distinguished in having the following combination of characters: illicial length less than 120% SL (93-105%); escal papillae absent; dentary teeth relatively short (longest 2.8% SL), arranged in 5 longitudinal series; rays of caudal fin less than 30% SL. Males and larvae unknown. DESCRIPTION. Illicium without filaments, proximal part lat- erally compressed, depth at base about 3 times width. Escal bulb club-shaped without distal prolongation; proximal part of bulb darkly pigmented, covered with spines to (or slightly be- yond) photophore; distal part of bulb above photophore bearing numerous short filaments and 2 pairs of large filaments; a pair of large, proximal filaments reaching beyond tip of longest dis- tal filaments. Esca extremely similar in two known specimens (the proxi- mal and distal pairs of long filaments arise in exactly the same position and seem to be of very similar relative length, although some appear to be broken in the holotype); number of short, distal filaments somewhat greater, and longest slightly longer in holotype; pigmentation on posterior margin of bulb just below photophore deeply cleft in holotype, forming a rounded inden- B f Figure 40. Gigantactis perlatus, female, 222 mm, SIO 61-31: A. Den- tary teeth, left lateral view; B. Diagrammatic representation of dentary- tooth pattern. Contributions in Science, Number 332 Bertelsen, Pietsch & Lavenberg: Gigamtactinid Anglerfishes 43 tation in paratype; spines absent on unpigmented part of bulb in holotype, covering proximal part of bulb in paratype (Fig. 41) . Number of teeth on each premaxilla 32 in both specimens, arranged in 2-3 overlapping, longitudinal series, external and posteriormost turned forward; longest premaxillary tooth 1.2% SL in holotype, approximately 1.0% in paratype. Number of dentary teeth 56 in holotype, 35 in paratype, longest 2.8% SL in both; dentary teeth in posterior part of jaw arranged in 5 longitudinal series, an external, a median, and 3 internal (Fig. 42) . Dorsal-fin rays 5, anal-fin rays 4-5, pectoral-fin rays 16-17 (Table 2); longest rays of caudal fin (second and seventh) 21- Figure 41. Esca of Gigantactis elsmani, holotype, 384 mm, ISH 1360/71, posterolateral view. Drawn by K. Elsman. 27% SL; skin coverage of caudal fin and membranes between caudal rays well developed (Fig. 4E). The 11.5-mm metamorphosal stage tentatively referred to this species (MCZ 51269) with illicial length of about 10% SL; esca with naked, unpigmented skin, a short conical distal pro- longation and papilliform rudiments of a pair of filaments on base; edge of jaws in resorption, no rudiments of teeth; on each side, a series of seven or eight enlarged supraorbital lateral line organs, the longest about 10% SL; skin faintly pigmented; no distinct dorsal subdermal pigment (specimen somewhat bleached); differs from holotype and paratype in having four dorsal-fin rays and 18 pectoral-fin rays (Table 11). DISTRIBUTION. Gigantactis elsmani is known from two spec- imens: the holotype taken from between the surface and 1900 m in the central Atlantic at approximately 1 1 "S, and the para- type from between the surface and 3000 m in the eastern South Pacific at 63 "S. A third specimen, tentatively referred to this species, is from the South Atlantic off Cape Town, South Af- rica (Fig. 65). ETYMOLOGY. This species is named for the late Kai L. Els- man, whose superb illustrations have added immeasurably to this revision. COMMENTS. Among the Gigantactis species with an illicial length of less than 120% SL (Table 1), G. elsmani is clearly distinguished by the characters of the esca: the lack of a dis- tinct, distal prolongation of the escal bulb and the presence of a single proximal and two distal pairs of large filaments. Speci- mens lacking escae may be identified by using a combination of characters including a low number of fin rays, a relatively high number of posterior longitudinal series of dentary teeth, and well-developed skin coverage on the caudal fin. Gigantactis golovani new species Figures 4J, 43, 44, 65; Tables 1, 2, 12 Gigantactis vanhoeffeni, Parin and Golovan 1976:271 (in part, one specimen). Figure 42. Gigantactis elsmani , holotype, female, 384 mm, ISH 1360/71: A. Dentary teeth, left lateral view; B. Diagrammatic represen- tation of dentary-tooth pattern. E = external series; M = median se- ries; I = internal series. 44 Contributions in Science, Number 332 Bertelsen, Pietsch & Lavenberg: Gigantactinid Anglerfishes MATERIAL. Three metamorphosed females (25-179 mm). Holotype: ISH 2250/71, 179 mm, 2°27'S, 19°00'W, 0-660 m. Paratypes: ZIAN 44263, 153 mm, 10°36'N, 17°38'W, 0- 1550 m; MCZ 51272, 25 mm, 14°43'N, 25°27'W, 0-720 m. The following specimen, tentatively referred to this species, is not included in the description below (but data given in Table 12): LACM 37517-1, 29 mm, Hawaii, off leeward Oahu, 0- 2100 m. DIAGNOSIS. Metamorphosed females of G. golovani differ from those of other species of the genus in having branched, distal escal filaments, a single long filament below the anterior margin of the escal bulb, and several similar filaments on and below the posterior margin of the escal bulb. They are further distinguished in having the following combination of charac- ters: illicial length 180-200% SL (except in juveniles); escal pa- pillae absent; dentary teeth relatively long (largest 4.4-4. 5% SL), arranged in 2, possibly 3, very irregular, longitudinal se- ries; rays of caudal fin less than 35% SL. Males and larvae unknown. DESCRIPTION. Illicium nearly cylindrical throughout, length 183-199% SL, 72% in 25-mm juvenile. Escal bulb club-shaped without distal prolongation, naked and unpigmented except at base; a group of branched distal filaments; a single, long fil- ament on anterior margin of base of bulb; numerous filaments of different length (some reaching beyond the tip of the distal filaments) on posterior margin of proximal part of bulb and on illicium below bulb (Fig. 43). Esca of larger paratype less well preserved, but number and distribution of filaments quite simi- lar; filaments on proximal part of esca and on illicium shorter, not reaching beyond base of distal filaments. Esca of juvenile paratype (Fig. 43A) with distribution of filaments similar to Table 11. Counts and measurements in percent of SL of females of Gigantactis elsmani. Character MCZ1 51269 Paratype LACM 10687-1 Holotype ISH 1360/71 Standard length (mm) 11.5 283 384 Length Illicium ~10 92.6 105 Longest premaxillary — 1.1 1.2 tooth Longest dentary tooth — 2.8 2.8 Longest caudal ray 40.0 20.8 27.3 Teeth Premaxillary 0 32 32 Dentary 0 35 56 Dorsal-fin rays 4 5 5 Anal-fin rays 5 4 5 Pectoral-fin rays 18 17 16 'Metamorphosal stage tentatively referred to this species. that of holotype, but number somewhat less, considerably shorter, and unbranched. Premaxillary teeth 20 in holotype (longest 1.7% SL), 17 in larger paratype (longest 1.8% SL), 5 in smaller paratype (long- est 2.0% SL). Dentary teeth 17 in holotype, 14 in larger para- type, 10 in smaller paratype (Fig. 44). Dorsal-fin rays 6, anal-fin rays 6-7, pectoral-fin rays 14-16 (Table 2); longest caudal rays (second and seventh) 30-32% SL in holotype, 28-31% SL in larger paratype; caudal-fin rays con- nected by transparent membranes; pigmented skin covering median caudal rays, which are broad, tapering only at tip (Fig. 4J, Table 1 2). The 29-mm juvenile (LACM 37517-1, Table 12), tentatively referred to this species, differs from the type material in having only about four distal escal filaments, five proximal filaments on posterior margin, and none on anterior margin; it is similar to the type material in the arrangement of the escal filaments, in lacking a distal prolongation of the escal bulb, and in the Figure 43. Escae of Gigantactis golovani, left lateral views: A. Para- type, 25 mm, MCZ 51272; B. Holotype, 179 mm, ISH 2250/71. Drawn by K. Elsman. Contributions in Science, Number 332 Bertelsen, Pietseh & Lavenberg: Gigantaetinid Anglerfishes 45 dentary tooth pattern, which consists of a few (9) large (the six largest 3-5% SL) teeth arranged in two (3?) irregular series. DISTRIBUTION. Gigantactis golovani is known from three specimens all collected in the eastern tropical Atlantic Ocean (Fig. 65). The material was captured by open gear fished at maximum depths of between 660 and 1550 m. A fourth speci- men, tentatively referred to this species, is from Hawaii, col- lected between the surface and 2100 m. ETYMOLOGY. Gigantactis golovani is named in honor of George Golovan of the Institute of Oceanology of the Academy of Sciences of the USSR, in appreciation for his making large and valuable collections of ceratioids available to us. COMMENTS. The intermediate illicial length of G. golovani (180 to 200% SL, comparable only to G. paxtoni, G. gargantua and G. savagei) easily separates this from the eight species in which the illicium is less than 120% SL (Table 1), and, at the same time, distinguishes it from all remaining species in which the illicium is considerably longer. In escal characters, G. golo- vani is unique in having branched, distal filaments and a single anterior filament below the base of the escal bulb and several similar filaments on and below its posterior margin. Some branched distal filaments and several filaments on and below the base of the escal bulb are present in large specimens of G. gargantua and G. macronema; but, in these forms, the proximal filaments are concentrated anteriorly (Figs. 46, 52). In dentary tooth pattern (in which the teeth are relatively long, but few in number and irregularly arranged), G. golovani is similar only to G. perlatus. Gigantactis gargantua new species Figures ID, 4H, 45, 46, 66; Tables 1,2, 13 Holotype: LACM 6903-32, 408 mm, 32“ 16'N, 1 17“43'W, 0- 1250 m, bottom depth 1775 m. Paratypes: IOAN uncatalogued, 325 mm, 31“30.8'S, 95° 27.2'E, 0-1400 m; LACM 30415-27, 25 mm, 28“44'N, 118“ 10'W, 0-1850 m; LACM 32749-3, 49 mm, 21“20-30'N, 158“20-30'W, 0-1000 m; LACM 30997-2, 105 mm, 31“32'N, 1 18°29'W, 0-1300 m; LACM 30996-16, 106 mm, 31“54'N, 118'39'W, 0-500 m; LACM 9748-28, 166 mm, 32“13'N, 1 17°47'W, 0-835 m, bottom depth 1756 m. DIAGNOSIS. Metamorphosed females of G. gargantua differ from those of other species of the genus in having the following combination of characters; illicial length 134-354% SL; 4 to 5 pairs of large, distal escal filaments; 30 to 50 proximal fil- aments centered on anterior margin of escal bulb; escal papillae absent; dentary teeth relatively short (longest 2.3% SL), ar- ranged posteriorly in 4 longitudinal series; second and seventh caudal-fin rays extremely long (30-47% and 54-76% SL, re- spectively). Males unknown (probably included in Gigantactis Male Group 1). Larvae unknown. DESCRIPTION. Illicium usually without filaments (some fil- aments on distal part of illicium of 325-mm specimen, IOAN uncatalogued), distinctly compressed proximally, depth near base more than twice width in holotype; illicial length 1 34— 216% SL in five metamorphosed specimens from eastern Pa- cific, 354% SL in specimen from Indian Ocean (325 mm, IOAN uncatalogued). Escal bulb club-shaped with a short, dis- tal prolongation; posterior surface of distal prolongation darkly pigmented, with a distal and 2 or 3 lateral pairs of swellings forming bases for 4-5 pairs (2 on distal pair of swellings) of Gigantactis sp. n., Pietsch 1972a:42, 43, 45, fig. 24(2) (otolith described, figured). MATERIAL. Seven females: six metamorphosed (49-408 mm), and one in metamorphosis (25 mm). Figure 44. Gigantactis golovani, holotype, female, 179 mm, ISH 2250/71: A. Dentary teeth, left lateral view; B. Diagrammatic represen- tation of dentary-tooth pattern. Table 12. Counts and measurements in percent of SL of females of Gigantactis golovani. Paratype Character Paratype MCZ 51272 LACM' 37517-1 IOAN uncata- logued Holotype ISH 2250/71 Standard length (mm) 25 29 153 179 Length Illicium 72.0 93.0 183 199 Longest premaxillary 2.0 1.7 1.8 1.7 tooth Longest dentary tooth 4.4 5.0 4.4 4.5 Longest caudal ray — 32.0 31.4 31.8 Teeth Premaxillary 5 5 17 20 Dentary 10 9 14 17 Dorsal-fin rays 6 5 6 6 Anal-fin rays 6 6 6 7 Pectoral-fin rays — 17 16 15-14 'Only tentatively referred to G. golovani. 46 Contributions in Science, Number 332 Bertelsen, Pietsch & Lavenberg: Gigantactinid Anglerfishes 1 mm 5 mm Figure 45. Escae of Gigantactis gargantua, posterolateral views: A. Paratype, 105 mm, I. ACM 30997-2; B. Holotype, 408 mm, LACM 6903-32 Drawn by K. Elsman. Contributions in Science, Number 332 Bertelsen, Pietsch & Lavemberg: Gigantactinid Anglerfishes 47 Table 13. Counts and measurements in percent of SL of females of Gigantactis gargantua. Character Paratype1 LACM 30415-27 Paratype LACM 32749-3 Paratype LACM 30997-2 Paratype LACM 30996-16 Paratype LACM 9748-28 Paratype IOAN uncatalogued Holotype LACM 6903-32 Standard length (mm) 25 49 105 106 166 325 408 Length lllicium 50.0 208 216 190 189 354 134 Longest premaxillary tooth <0.5 0.8 0.8 0.9 0.7 0.9 0.8 Longest dentary tooth <0.5 1.6 1.7 2.1 1.9 1.3 2.3 Longest caudal ray 25.0 67.3 54.3 59.4 54.8 76.0 broken Teeth Premaxillary 2 7 15 11 17 34 59 Dentary 5 16 31 35 41 46 75 Dorsal-fin rays 7 6 6 5 6 6 5 Anal-fin rays 6 6 6 6 6 6 6 Pectoral-fin rays 19 19 19-20 20 21 20 22 1 Metamorphosal stage. large, unpigmented filaments, more or less branched in large specimens (Fig. 45); pore of photophore on a pigmented pa- pilla; proximal part of bulb spinulose, unpigmented, with nu- merous filaments of different length (longest reaching beyond tip of distal prolongation), longest and most dense on anterior margin. Esca of 408-mm holotype (Fig. 45B) differing from those of paratypes: proximal, right swelling of distal prolonga- tion small, deformed, with only a short filament (possibly in re- generation); division into six pairs of distal filaments less distinct, nearly all branched, some bifurcated near close-set bases; proximal filaments more numerous, group centered on anterior margins surrounding base of bulb. Most escal filaments of type material slightly swollen at tip. Premaxillary teeth increasing in number with standard length from 7 in smallest paratype to 59 in holotype (longest tooth 0.7-0. 9% SL). Dentary teeth increasing in number with standard length from 16 in smallest paratype (longest 1.6% SL) to 75 in holotype (longest 2.3% SL), arranged posteriorly in 4 longitudinal series, some teeth of second external series present anteriorly in the jaw of large specimen (Fig. 46, Table 13). Dorsal-fin rays 5-7, anal-fin rays 6, pectoral-fin rays 19-22 (Table 2); skin coverage of proximal part of caudal fin well- developed (Fig. 4H). The 25-mm metamorphosal stage (LACM 30415-27) with il- licial length 50% SL; esca with short distal prolongation and indistinct rudiments of distal filaments; teeth in early develop- ment; no enlarged lateral line organs on head; dorsal subdermal pigment faint (Table 13). DISTRIBUTION. Gigantactis gargantua is known from six specimens collected in the eastern North Pacific Ocean off southern California and the Hawaiian Islands, and a seventh individual collected in the eastern south Indian Ocean (Fig. 66). The material was taken by open gear fished at maximum depths of between 500 and 1300 m. ETYMOLOGY. Gargantua, a gigantic king, was the hero of a satirical romance written by Rabelais in 1535. COMMENTS. Gigantactis gargantua, G. herwigi, and G. wa- termani form a group of closely related forms here referred to as the G. gargantua group (Table 1). They are distinguished Figure 46. Gigantactis gargantua, paratype, female, 106 mm, LACM 30996-16: A. Dentary teeth, left lateral view; B. Diagrammatic repre- sentation of dentary-tooth pattern. E = external series; M = median series; I = internal series. 48 Contributions in Science, Number 332 Bertelsen, Pietsch & Lavenberg: Gigantactinid Anglerfishes from other members of the genus in having four to five pairs of large, distal filaments arising from swollen bases, a group of proximal filaments centered on the anterior margin of the bulb, and the especially prolonged second and seventh caudal fin rays. Gigantactis gargantua differs from G. watermani and G. herwigi in details of the escal morphology, particularly in hav- ing a greater number of proximal filaments (about 20 to 50 compared with 12 and 7 in G. watermani and G. herwigi, re- spectively). The material forming the G. gargantua group is quite similar in nearly all characters. However, we recognize three species within the group for the following reasons: the six meta- morphosed specimens from the Pacific and Indian Ocean, here described as G. gargantua, agree completely in the escal char- acters that separate them from the two Atlantic specimens; in light of the small intraspecific and ontogenetic variation in these characters observed in other Gigantactis species, it seems unlikely that the differences between the two Atlantic speci- mens could be explained in this way. The variation in illicial length is unusually great in G. gargantua compared to that in most other Gigantactis species, even within the representatives of the eastern Pacific population. For this reason, the extreme illicial length observed in the specimen from the Indian Ocean (IOAN uncatalogued) is not considered to express specific dis- tinction. For the same reason, the large difference in illicial length between the holotypes of G. watermani and G. herwigi may not be a diagnostic means of distinguishing these two forms. Gigantactis watermani new species Figures 47-49, 66; Tables 1, 2 MATERIAL. A single metamorphosed female, the holotype: ISH 2330/71, 99 mm, 1 °04'N, 18°22'W, 0-2100 m. DIAGNOSIS. Metamorphosed females of G watermani differ from those of other species of the genus in having the following combination of characters: illicial length 231% SL; 5 pairs of large, distal escal filaments; 12 proximal filaments centered on anterior margin of escal bulb; escal papillae absent; dentary teeth short (longest 2.5% SL), arranged posteriorly in 4 longitu- dinal series; second caudal-fin ray extremely long (70% SL), seventh caudal-fin ray broken in holotype (remaining portion measures 30% SL). Males and larvae unknown. DESCRIPTION. Illicium without filaments, proximal part not distinctly compressed. Escal bulb club-shaped with a darkly pigmented, distal prolongation approximately five times as long as bulb diameter, with proximal one-half greatly swollen; distal prolongation bearing 5 pairs of stout, tapering filaments, all but distal-most pair densely covered with black pigment except for narrow, tapering tip; opening of pore of photophore present in darkly pigmented skin at base of distal prolongation, not raised on a papilla; 12 narrow, unpigmented filaments on anterior mar- gin of base of bulb, longest reaching base of distal-most pair of distal filaments; base of bulb proximal to filaments pigmented and spinulose (Fig. 48). Premaxillary teeth 16 (longest 1.1% SL); dentary teeth 43, arranged in 4 longitudinal series in posterior part of jaw, some teeth of second external series present anteriorly (Fig. 49). Dorsal-fin rays 6, anal-fin rays 5, pectoral-fin rays 18-19 (Table 2). DISTRIBUTION. Gigantactis watermani is known only from the holotype collected from between the surface and 2100 m in the eastern tropical Atlantic at approximately 1°N, 18° W (Fig. 66). ETYMOLOGY. This species is named for Talbot Howe Water- man in recognition of his outstanding contribution to our knowl- edge of the anatomy of Gigantactis (see Waterman 1948). COMMENTS. Gigantactis watermani is a member of the G. gargantua group (Table 1). It is distinguished from the other two members of this group ( G . gargantua and G. herwigi) in having a more elongate, distal escal prolongation and larger and more heavily pigmented proximal portions of the distal es- cal filaments. It is further distinguished from G. gargantua in having fewer proximal filaments that are all restricted to the anterior margin of the escal bulb. Gigantactis herwigi new species Figures 4G, 50, 51, 66; Tables 1, 2 MATERIAL. A single metamorphosed female, the holotype, ISH 972/68, 262 mm, 4°43'S, 26°39'W, 0-2000 m. Figure 47. Gigantactis watermani, holotype, 99 mm, ISH 2330/71. Drawn by K. Elsman. Contributions in Science, Number 332 Bertelsen, Pietsch & Lavenberg: Gigantactinidl Anglerfishes 49 DIAGNOSIS. Metamorphosed females of G. herwigi differ from those of other species of the genus in having the following combination of characters: illicial length 373% SL; 4 pairs of large, distal escal filaments; 7 proximal filaments centered on anterior margin of escal bulb; escal papillae absent; dentary teeth short (longest 1.6% SL), arranged posteriorly in 4 longitu- dinal series; second and seventh caudal-fin rays prolonged (70 and 95% SL, respectively). Males and larvae unknown. DESCRIPTION, lllicium without filaments, proximal part slightly compressed laterally. Escal bulb club-shaped, without Figure 48. Esca of Gigantactis watermani, hoiotype, 99 mm, ISH 2330/71, left lateral view. Drawn by K. Elsman. distinct prolongation, bearing 4 pairs of large, unbranched, dis- tal filaments, slightly swollen and pigmented posteriorly at base; opening of pore of photophore on pigmented papilla; prox- imal part of bulb spinulose with 7 narrow, unpigmented fil- aments on anterior margin of base of bulb, longest reaching beyond base of distal filaments (Fig. 50). Premaxillary teeth 34 (longest 0.7% SL); dentary with ap- proximately 65 teeth, arranged posteriorly in 4 series, some teeth of second external series present anteriorly (Fig. 51). DISTRIBUTION. Gigantactis herwigi is known only from the hoiotype collected from between the surface and 2000 m in the tropical Atlantic at approximately 4°S, 26°W (Fig. 66). ETYMOLOGY. This new form is named after the German Re- search Vessel Walther Herwig, which has been used during nu- merous expeditions since 1966 to collect an outstandingly rich resource of deepsea fishes, including the hoiotype of G. herwigi as well as nearly all Gigantactis material catalogued at the In- stitut fiir Seefischerei, Hamburg (ISH). COMMENTS. Gigantactis herwigi is a member of the G. gargantua group (Table 1 ). It differs from other members of this group in lacking a distinct prolongation of the escal bulb. It further differs from G. gargantua in having all distal filaments of the esca unbranched and proximal escal filaments (7 com- pared to about 30 in G. gargantua) restricted to the anterior margin of the bulb (surrounding the bulb in G. gargantua). It further differs from G. watermani in having distal filaments that are less swollen and that are pigmented only near the base (swollen and darkly pigmented for more than half their length in G. watermani ), and in having unpigmented proximal fil- aments (pigmented in G. watermani). Gigantactis macronema Regan Figures IE, 41, 22, 52, 53, 67; Tables 1,2, 14 Gigantactis macronema Regan 1925:565 (original descrip- tion, single specimen); Regan 1926:38, pi. 11 (description after Regan 1925); Regan and Trewavas 1932:93, 94 (description after Regan 1925, 1926, in key); Waterman 1 939b:84— 85 (com- parison with G. longicirra, largest known gigantactinid); Water- man 1948:130 (comparison with G. longicirra ); Bertelsen 1951:150-152, fig. 101, table 31 (comparison with all known material, comments); Grey 1956:267 (synonymy, vertical dis- Figure 49. Gigantactis watermani, hoiotype, female, 99 mm, ISH 2330/71: A. Dentary teeth, left lateral view; B. Diagrammatic represen- tation of dentary-tooth pattern. E = external series; M = median se- ries; I = internal series. 50 Contributions in Science, Number 332 Bertelsen, Pietsch & Lavenberg: Gigantactinid Anglerfishes tribution); Robins and Courtenay 1958:151 (comparison with additional specimen designated Gigantactis sp ); Fitch and Lav- enberg 1968:135-137, fig. 74 (distinguishing characters, natu- ral history, fishery information, other family members olT California); Pietsch 1972a:29, 34, 35, 41, 42, 45 (comments on osteology, holotype with five pectoral radials). MATERIAL. Ten metamorphosed females, 34-354 mm. Holotype of Gigantactis macronema: ZMUC P92130, 98 mm, 31 °47'N, 41 °41'W, 5000 m wire. Referred material: 1SH 1596/71, 232 mm, 27" 14'S, 2°56'E, 0-2000 m; LACM 37516-1, 34 mm, Hawaii, off leeward Hawaii, 0-950 m; LACM 32745-2, 37.5 mm, 21°00'N, 158° 20' W, 0-930 m; LACM 9036-36, 93 mm, 33°26'N, 1 18°33'W, 0-1100 m, bottom depth 1270 m; LACM 6902-38, Figure 50. Esca of Gigantactis herwigi, holotype, 262 mm, ISH 972/- 68, left lateral view. Drawn by K. Elsman. 1 10 mm, 32°36'N, 1 18°05'W, 0-650 m; LACM 30599-20, 354 mm, 33°22'N, 118°02'W, 0-650 m; MCZ 51255, 141 mm, 23°04'N, 45°10'W, 0-1100 m (cleared and stained); SIO 73-156, 35 mm, 28°18.5'N, I55°14.5'W, 3000 m wire; SIO 57-46, 62 mm, 28°51'N, 118°11'W, 0-1041 m, bottom depth 1280-1830 m. DIAGNOSIS. Metamorphosed females of G. macronema differ from those of other species of the genus in having the following combination of characters: illicial length 340-447% SL; escal bulb with a lightly pigmented, truncated distal prolongation; 8-20 long distal filaments; escal papillae absent; proximal escal filaments absent; dentary teeth short (longest 1.3-3. 4% SL, average 2.3% SL), arranged posteriorly in 2 longitudinal series; length of caudal-fin rays less than 40% SL; skin coverage of proximal part of caudal fin weakly developed, caudal-fin rays free nearly to base. Males unknown (probably included in Gigantactis Male Group I). Larvae unknown (probably included in Gigantactis Larval Group B). DESCRIPTION. Illicium without filaments (except in the 354- mm specimen. Fig. 52B, in which distal sixth of illicial length covered with unpigmented, short filaments, each with white swollen base), nearly cylindrical throughout length, depth less than twice width; illicial length variable and unrelated to SL. Esca of holotype (not well preserved, see comments below) with remains of at least 8 filaments at tip of distal prolongation; es- cal bulb of 93-, 141- and 243-mm specimens (all well preserved and quite similar. Fig. 52) elongate and club-shaped, increasing gradually in width from illicium toward unpigmented area sur- rounding photophore; skin of bulb spinulose except area distal to photophore; pore of photophore raised on papilla, pigmented at tip; distal prolongation truncated, faintly pigmented on pos- terior margin, length more than twice width at base; 16 distal filaments, symmetrically arranged more or less in close-set pairs, each with a small swelling at tip; esca of 354-mm speci- men (Fig. 52B) with somewhat shorter distal prolongation bear- ing more numerous filaments (approximately 20) that are close- set and less distinctly arranged in pairs, some bifurcated; esca Figure 51. Gigantactis herwigi , holotype, female, 262 mm, ISH 972/68: A. Dentary teeth, left lateral view; B. Diagrammatic represen- tation of demtary-tooth pattern. E = external series; M = median se- ries; I = internal series. Contributions in Science, Number 332 Bertelsem, Pietsch & Lavenberg: Gigantactinid Anglerfishes 51 of 1 10-mm specimen with distal filaments lost; esca of 35-mm specimen with about 8 distal filaments. Number of teeth on each premaxilla increasing with stan- dard length from 3-5 in smaller specimens to 23 in largest, those in posterior part of jaw turned forward (Fig. 53, Table 14); longest premaxillary tooth 0.6-1. 6% SL. Number of den- tary teeth varying between 18 and 31 in 6 largest specimens, those in posterior part of jaw arranged in 2 longitudinal series (Fig. 53; in the larger specimens, teeth of the second external, and a few of the first internal series are present more anteriorly in the jaw). Dorsal-fin rays 5-6, anal-fin rays 5-6, pectoral-fin rays 17-20 (Table 2); skin coverage of proximal part of caudal fin less de- veloped than in most other species (Fig. 41), rays free nearly to base, skin nearly cylindrical without connecting membranes; longest caudal-fin rays (second and seventh) 26-37% SL, only slightly longer than intermediate rays. Figure 52. Escae of Gigantactis macronema, left lateral views: A. 232 mm, ISH 1596/71; B. 354 mm, LACM 30599-20. Drawn by K. Elsman. DISTRIBUTION. Gigantactis macronema is known from the north and south Atlantic, from the eastern Pacific just north and west of the Hawaiian Islands, and from off the coasts of southern California and Baja California (Fig. 67). The material was captured by open gear fished at maximum depths of be- tween 650 and 2500 m. COMMENTS. Gigantactis macronema is one of a number of Gigantactis species that have an illicium that is more than twice the standard length (Table 1). These forms, here referred to as the G. macronema group, are further characterized by the absence of filaments on the proximal part of the escal bulb and in having relatively small dentary teeth (those in the posterior part of the jaw in 2, rarely 3 longitudinal series) and a caudal fin with weakly developed skin coverage (the second and sev- enth caudal rays are the longest, usually 30-35% SL, but only slightly longer than the intermediate rays). Gigantactis mac- ronema differs from the other three members of the G. mac- ronema group in the length of the illicium (less than 275% SL in the other members) and in the size, shape and pigmentation of the distal prolongation of the escal bulb (dark and tapering in G. savagei; with a distal pigmented patch and only slightly raised in G. microdontis; distal prolongation absent in G. ios). It further differs from G. microdontis and G. ios in the length of the dentary teeth (the longest in these species is 1.1—1 .6% SL, average 1 .3% SL). The esca of the holotype of G. macronema was not described in the original description (Regan 1925). A well-developed dis- tal prolongation of the escal bulb and some slender distal fil- aments, however, are shown in an illustration provided by Regan (1926, pi. 1 1). The esca is now dehydrated and shrunken so that the position and the number of filaments of the distal B * * ‘ ' • • Figure 53. Gigantactis macronema , female, 354 mm, LACM 30599-20: A. Premaxillary and dentary teeth, left lateral view; B. Diagrammatic representation of dentary-tooth pattern. E = external series; M = me- dian series. 52 Contributions in Science, Number 332 Bertelsen, Pietsch & Lavenberg: Gigantactinid Anglerfishes prolongation is difficult to ascertain; the remains of at least eight filaments are present. Because of the poor condition of the holotype, the allocation of our material to this species is open to some doubt. However, all differ significantly from material of the other three species of the G. macronema group in illicial length (in addition to other differences discussed above). The 62-mm specimen (lacking an esca) is tentatively referred to this species because of the great length of the remaining part of the illicium (281% SL), which appears to have been broken some distance below the esca. Gigantactis savagei new species Figures 54, 55, 67; Tables 1,2, 15 MATERIAL. Three females: two metamorphosed (56-150 mm), and one in metamorphosis (19 mm). Holotype: LACM 9706-41, 150 mm, 31 °40'N, 120°23'W, 0- 650 m. Paratypes: LACM 37080-1, 56 mm, Hawaii, leeward Oahu, 0-1250 m; LACM 37520-1, 19 mm, Hawaii, leeward Oahu, 0- 985 m. The following material, only tentatively referred to this species, is not included in the description below (but data given in Table 15): LACM 9589-30, 33 mm, 29°01'N, 1 18°00'W, 0- 650 m; LACM 9652-32, 44 mm, 3P31'N, 118°26'W, 0-500 m; SIO H53-340, 38 mm, 41° 18'N, 168°21'E, 0-1640 m. DIAGNOSIS. Metamorphosed females of G. savagei differ from those of other species of the genus in having the following combination of characters: illicial length 165-268% SL; escal bulb with a darkly pigmented, tapering distal prolongation; 10- 18 narrow, distal filaments; escal papillae absent; proximal escal filaments absent; dentary teeth short (longest 1 .7—2. 1 % SL), ar- ranged posteriorly in 2-3 longitudinal series; length of caudal- fin rays less than 35% SL; skin coverage of proximal part of caudal fin weakly developed, caudal-fin rays free nearly to base. Males unknown. Larvae unknown (probably included in Gigantactis Larval Group B). DESCRIPTION. Illicium without filaments, length variable and unrelated to SL, nearly cylindrical throughout. Escal bulb club-shaped, proximal part pigmented and spinulose, without filaments; tapering, distal prolongation darkly pigmented (ex- cept area under swollen bases of filaments), distal filaments not arranged in a distinct pattern, largest and most numerous at tip, each with a series of small swellings throughout length; 18 distal filaments in holotype (8 longest present at tip), 10 in met- amorphosed paratype (5 longest at tip); tissue between oval swellings very thin and flexible (Fig. 54). Number of premaxillary teeth 7-14 (longest 0.8-1. 4% SL); number of dentary teeth 17-18 (Fig. 55, Table 15). Dorsal-fin rays 5-6, anal-fin rays 5-6, pectoral-fin rays 18-20 (Table 2); skin coverage of proximal part of caudal fin less de- veloped than in most other species, rays free nearly to base, skin nearly cylindrical without connecting membranes; longest caudal-fin rays (second and seventh) 30-35% SL, only slightly larger than intermediate rays (Table 15). The 19-mm metamorphosal stage (LACM 37520-1) has rela- tively shorter illicium; esca with conical, distal prolongation and four distal filaments; some remains of enlarged supraorbi- tal lateral line organs; skin and subdermal pigment absent (pos- sibly bleached away). Table 14. Counts and measurements in percent of SL of females of Gigantactis macronema. Holotype LACM SIO LACM SIO LACM ZMUC LACM MCZ ISH LACM Character 37516-1 73-156 32745-2 57-46 9036-36 P92130 6902-38 51255 1596/71 30599-20 Standard length (mm) 34 35 37.5 62 93 98 110 141 232 354 Length Illicium 447 340 184 + 281 + 372 350 444 443 432 343 Longest premaxillary tooth 1.2 1.4 0.8 1.6 1.0 0.8 1.2 1.5 1.3 0.6 Longest dentary tooth 3.2 3.1 2.7 3.4 1.9 1.4 3.2 2.7 1.9 1.3 Longest caudal ray 35.0 32.9 29.0 — 30.1 — 37.3 30.5 319 26.3 Teeth Premaxillary 5 5 3 7 13 1 1 16 15 16 23 Dentary 7 8 5 17 31 21 22 22 18 29 Dorsal-fin rays 5 5 5 6 5 6 5 5 5 5 Anal-fin rays 5 5 5 6 5 5 6 5 5 5 Pectoral-fin rays 19 17 19 18-19 19 18 18 18 20 18 Contributions in Science, Number 332 Bertelsen, Pietsch & Lavenberg: Gigantactimid Anglerfishes 53 DISTRIBUTION. Gigantactis savagei is known only from the north Pacific Ocean, from off the coast of southern California and the Hawaiian Islands (Fig. 67). This species appears to be a relatively shallow-living form: the holotype was captured in open gear fished above 700 m; the two paratypes were collected by open gear fished at maximum depths of between approx- imately 1000 and 1250 m. ETYMOLOGY. Gigantactis savagei is named in honor of Jay M. Savage of the University of Southern California in appre- ciation for his service as principal investigator of grants from the National Science Foundation in support of Ecological Stud- ies of Midwater Fishes in and about Southern California Bor- derland Region, 1960-1970. COMMENTS. Gigantactis savagei is a member of the G. mac- ronema group (Table 1). It differs from the other three mem- Figure 54. Esca of Gigantactis savagei, holotype, 150 mm, LACM 9706-41, left lateral view. Drawn by K. Elsman. bers of this group (G. macronema, G. microdontis, and G. ios), as well as from all other members of the genus, in escal pigmen- tation, in the size and shape of the distal prolongation of the escal bulb, and in the structure and position of the distal fil- aments. Gigantactis savagei further differs from G. macronema in having a significantly shorter illicium. The three juvenile specimens (33-44 mm), tentatively re- ferred to this species, represent the G. macronema group in il- licial length, in lacking proximal escal filaments, and in having relatively short dentary teeth in two longitudinal series (Table 15). In other escal characters, however, they are not in com- plete agreement with the material here placed within the G. macronema group. The largest (44 mm) has nine short distal filaments, each with a series of swellings as is characteristic of G. savagei, but the bulb is distally unpigmented and without a distal prolongation. In each of the two smaller specimens, the bulb has a somewhat pigmented, short, conical, distal prolonga- tion with short rudiments of distal filaments (approximately 7 in the 38-mm specimen, indistinct rudiments in the 33-mm specimen), and two additional filaments at the base. From a comparison with specimens of G. macronema and G. microdon- tis of a similar standard length, it seems less likely that they represent one of these species than that they represent develop- mental stages of G. savagei. Gigantactis microdontis new species Figures 56-58, 67; Tables 1, 2, 16 Gigantactis “sp. 2,” Parin et al. 1973:146 (19.5-mm meta- morphosal stage). MATERIAL. Seven females: six metamorphosed (25.5-127 mm), and one in late metamorphosis (19.5 mm). Holotype: MCZ 52574, 66 mm, 15'12'S, 75°44'W, 0-700 m, bottom depth 1060 m. Paratypes: IOAN uncatalogued, 19.5 mm, 12°30'S, 87”45'W, 0-100 m; LACM 32776-2, 25.5 mm, 2U20-30'N, 158°20-30 W, 0-925 m (juvenile); LACM 32791-3, 38.5 mm, 21 ° 20-30'N 1 58° 20-30' W, 0-1175 m; LACM 30284-29, 44 mm, 28°48'N, 118°10'W, 0-650 m (esca poorly preserved); LACM 9693-34, 118 mm, 3U45'N, 1 18°45'W, 0-600 m (esca previously drawn, but now lost); LACM 32204-2, 127 mm, 28°20'N, 118° 18'W, 0-650 m (esca poorly preserved). Figure 55. Gigantactis savagei, holotype, female, 150 mm, LACM 9706-41: A. Dentary teeth, left lateral view; B. Diagrammatic represen- tation of dentary-tooth pattern. E = external series; M = median se- ries; I = internal series. 54 Contributions in Science, Number 332 Bertelsen, Pietsch & Lavenberg: Gigantactinid Anglerfishes Table 15. Counts and measurements in percent of SL of Gigantactis savagei. Character Paratype1 LACM 37520-1 LACM2 9589-30 SIO2 H53-340 LACM2 9652-32 Paratype LACM 37080-1 Holotype LACM 9706-41 Standard length (mm) 19 33 38 44 56 150 Length Illicium ~65 273 165 260 268 265 Longest premaxillary tooth 1.0 1.2 1.3 1.4 0.8 Longest dentary tooth <0.5 2.0 2.0 2.5 2.1 1.7 Longest caudal ray 30.0 — 28.2 32.5 30.4 34.7 Teeth Premaxillary 0 6 8 6 7 14 Dentary 2-3 1 1 11 12 17 18 Dorsal-fin rays 6 4 5 5 6 5 Anal-fin rays 6 4 5 5 6 5 Pectoral-fin rays 20 18 18-19 19 18-19 18 1 Metamorphosal stage, tentatively referred to G. savagei. DIAGNOSIS. Metamorphosed females of G. microdontis dif- fer from those of other species of the genus in having excep- tionally short dentary teeth (longest 1.1-1. 6% SL). They further differ in having the following combination of charac- ters: illicia! length 216-240% SL (in specimens 38.5 mm and longer); escal bulb with a short, distal prolongation pigmented on distal surface; 8-10 short, distal filaments; escal papillae ab- sent; proximal escal filaments absent; dentary teeth arranged posteriorly in 2 longitudinal series; length of caudal-fin rays less than 45% SL; skin coverage of proximal part of caudal fin weakly developed, caudal-fin rays free nearly to base. Males unknown. Larvae unknown (probably included in Gigantactis Larval Group B). DESCRIPTION. I llicium without filaments, nearly cylindrical throughout length. Escal bulb club-shaped, pigmented only near base and on distal part of short, distal prolongation; distal patch of pigment oblong in shape, tapering posteriorly towards pore of photophore; 10 distal filaments in holotype, 9 in para- type, arising from surface of bulb just inside edge of distal pig- ment patch in 2 nearly parallel, lateral series that meet anteriorly; filaments short, approximately diameter of bulb in holotype, less in paratype, spatulate and compressed distally (Fig. 56); most distal filaments lost in 44-mm and 127-mm specimens (but in illicial length, in the shape of the escal bulb, and remains of the distal prolongation, they are in good agree- ment with the above description), those remaining differing in being compressed nearly from base; esca of 118-mm specimen (Fig. 56C, sketched before being lost) lacking a distinct distal prolongation, but bearing 10 filaments of similar shape; late metamorphosal stage and juvenile (19.5 and 25 mm, respec- tively) with illicium short (33% and 43% SL, respectively), and esca unpigmented with about 8 short, rudimentary filaments close-set on distal surface of bulb (Fig. 56). Number of premaxillary teeth 8-16 (longest 0. 8-1.0% SL). Number of dentary teeth 16-51 (Fig. 57, Table 16). Dorsal-fin rays 4-6, anal-fin rays 4-6, pectoral-fin rays 17-19 (Table 2); skin coverage of proximal part of caudal fin less de- veloped than in most other species (Fig. 4), rays free nearly to base, skin nearly cylindrical without connecting membranes; longest caudal-fin rays (second and seventh) 30-43% SL, only slightly longer than intermediate rays. The 19.5-mm IOAN specimen (Fig. 58) is a late meta- morphosal stage with nine to ten of anteriormost supraorbital lateral line organs enlarged, largest 10% SL; skin very faintly pigmented, no distinct subdermal pigmentation (possibly bleached). DISTRIBUTION. Gigantactis microdontis is known only from the eastern Pacific Ocean: three records from off southern Cal- ifornia, two from the Hawaiian Islands, and two off the coast of Peru (Fig. 67). It appears to be a rather shallow-living form; all material was collected by open gear fished above 1200 m; southern California and Peruvian specimens by gear fished above 700 m. Contributions in Science, Number 332 Bertelsen, Pietsch & Lavenberg: Gigantactinid Anglerfishes 55 Table 16. Counts and measurements in percent of SL of females of Gigantactis microdontis. Paratype Paratype Paratype Paratype Holotype Paratype Paratype IOAN LACM LACM LACM MCZ LACM LACM Character uncatalogued 32776-2 32791-3 30284-29 52574 9693-34 32204-2 Standard length (mm) 19.5 25.5 38.5 44 66 118 127 Length Illicium 33.0 43.2 240 226 217 216 226 Longest premaxillary tooth 1.0 0.9 0.9 0.8 1.0 Longest dentary tooth 1.2 1.6 1.2 1.1 1.3 Longest caudal ray 30.0 broken 31.4 33.2 33.3 31.8 42.5 Teeth Premaxillary — — 13 9 8 13 16 Dentary — — 28 16 22 28 51 Dorsal-fin rays 4 5 6 5 5 6 5 Anal-fin rays 4 5 6 4 5 6 5 Pectoral-fin rays 17 19 18 18 17 18 19 ETYMOLOGY. The name microdontis is derived from the Greek mikros, meaning small, and odontas, tooth, in reference to the exceptionally small dentary teeth of this species. COMMENTS. Gigantactis microdontis is a member of the G. macronema group (Table 1), separated from all other species of the genus by the characters listed for this group (see species account for G. macronema ); it differs from the other members of this group ( G . macronema, and G. savagei and G. ios) in pig- mentation, shape and size of the distal prolongation of the escal bulb, and structure and position of the distal escal filaments. It further differs from G. macronema in illicial length and from G. savagei in the length of the longest dentary tooth. Gigantactis ios new species Figures 59, 67; Tables 1, 2 MATERIAL. A single metamorphosed female, the holotype: BMNH 1977.9.13.1, 57 mm, 29'49'N, 23°00'W, 1005-1250 m. DIAGNOSIS. Metamorphosed females of G. ios differ from those of other species of the genus in having the following com- bination of characters: illicial length 256% SL; escal bulb with- out distal prolongation; a dense group of 14 short distal filaments; escal papillae absent; proximal escal filaments ab- sent; dentary teeth short (longest 1.1% SL), arranged posteri- orly in 2 longitudinal series; length of caudal-fin rays less than 35% SL; skin coverage of proximal part of caudal fin weakly developed, caudal-fin rays free nearly to base. Males unknown. 1 mm Figure 56. Escae of Gigantactis microdontis: A. Holotype, 66 mm, MCZ 52574, left lateral view; B. Holotype, 66 mm, MCZ 52574, dorsal view; C. 1 18 mm, LACM 9693-34, posterior view. Drawn by K. Elsman. 56 Contributions in Science, Number 332 Bertelsen, Pietsch & Lavenberg: Gigantactinid Anglerfishes Larvae unknown (probably included in Gigantaetis Larval Group B). DESCRIPTION. I llicium without filaments. Escal bulb club- shaped; pigment confined to base of bulb except for a small distal spot; distal escal filaments lanceolate with tiny internal, transparent bulbs (Fig. 59). Premaxillary teeth 16 (longest 0.9% SL); dentary teeth 32. Dorsal-fin rays 5, anal-fin rays 5, pectoral-fin rays 18 (Table 2); skin coverage of proximal part of caudal fin weakly devel- oped, rays free nearly to base, skin nearly cylindrical without connecting membranes; longest caudal-fin ray (seventh) 31% SL, only slightly longer than other rays. DISTRIBUTION. Gigantaetis ios is known only from the holo- type collected in a closing trawl between 1005 and 1250 m in the eastern North Atlantic Ocean just southwest of Madeira (Fig. 67). ETYMOLOGY. The name ios (to be pronounced ios) is an ac- ronym formed from the initial letters of the Institute of Oceanographic Sciences, Surrey, England, in recognition of important ichthyological contributions made by our colleagues of that institution. COMMENTS. Gigantaetis ios is a member of the G. mac- ronema group, distinct from other members of the genus in the characters mentioned for this group. It differs from G. mac- ronema and G. savagei in the same characters that distinguish G. microdontis. It is easily separated from G. microdontis by the structure, number, and position of the distal escal filaments. Gigantaetis Species Incertae Sedis Gigantaetis filibulbosus Fraser- Brunner Gigantaetis filibulbosus Fraser-Brunner 1935:326 (original description; single specimen, the holotype, BMNH 1934.8.8.92, 25 mm, 53°15'N, 12°28'W, 0-320 m); Waterman 1939b:85 (comparison with G. longicirra)\ Bertelsen 1951:150-152, table 31 (comparison with all known material; comments); Robins and Courtenay 1958:151 (depth distribution); Wheeler Figure 57. Gigantaetis microdontis, female, 118 mm, LACM 9693-34: A. Premaxillary and dentary teeth, left lateral view; B. Diagrammatic representation of dentary-tooth pattern. E = external series; M = me- dian series. 1969:585 (reference to original description); Maul 1973 (listed). The holotype and only specimen ever referred to this species is a juvenile stage with an undeveloped illicium (84% SL) that gives no indication of what its length might have been when fully developed. The escal bulb is shrunken with no remains of “a slender filament, expanded at tip, on each side of its distal end,” as described and figured by Fraser-Brunner (1935). Among the species recognized here, an esca somewhat similar to this description is found only in G. microdontis (Fig. 56); however, this species has about 10 escal filaments. Even if we assume that the pair of filaments present in G. filibulbosus rep- resents remains of a larger number, the length of the dentary teeth of the holotype (which according to the description and figure reach a length of at least 3% SL) fall well outside the range recorded for G. microdontis. Further, since the juvenile holotype has lost the distal portion of the caudal fin and is simi- lar to several Gigantaetis species in proportions and meristic characters, G. filibulbosus is regarded as incertae sedis. Gigantaetis ovifer Regan and Trewavas Gigantaetis ovifer Regan and Trewavas 1932:93, 95, fig. 152 (original description; single specimen, the holotype, ZMUC P92131, 30 mm, 14°37'N, 119°52'E, 2500 m wire); Fraser- Brunner 1935:326 (comparison with G. filibulbosus)-. Water- man 1 939b: 8 5 (comparison with G. longicirra)-, Bertelsen 1951:150-152, table 31 (comparison with all known material; comments); Grey 1956:268 (synonymy; vertical distribution); Pietsch 1972a:42, 45 (holotype with five pectoral radials). The esca of the holotype of G. ovifer is damaged. The “two very short terminal appendages” described by Regan and Trewavas (1932:95, fig. 152) represent fragments of torn tissue. The relatively long dentary teeth of this specimen (longest 3.4% SL), and the presence of a few teeth representing a third longi- tudinal series, only excludes it from the G. macronema group. Since it remains doubtful that its illicial length (90% SL) repre- sents the relative length as an adult, and as it in other mor- phometric and meristic characters is similar to several species of Gigantaetis recognized here, G. ovifer is regarded as insertae sedis. Gigantaetis Species Unidentified METAMORPHOSED FEMALES, Table 17 MATERIAL. Six specimens (23.5-300 mm) with escae and an unknown portion of the illicium lost: IOAN uncatalogued, 24 mm, 20°00'S, 76°42'W ( Gigantaetis “sp. 1” of Parin et al. 1973:146); IOAN uncatalogued, 32.5 mm, VITIAZ St. 7284; ISH 150/68, 177 mm, 33°42'N, 16*19'W, 0-840 m; ISH 1958/71, 292 mm, 10°57'S, 11°20'W, 0-1900 m; LACM 36076-8, 23.5 mm, 2°46'S, 127°54'E, 0-1500 m; LACM 3431 1-2, 47 mm, off leeward Oahu, 0-800 m; MCZ 52573 300 mm, landed at Gloucester, Massachusetts, ca. 42°30'N, 70 °W, 0-310 m. COMMENTS. All of these specimens lack the diagnostic fea- tures of the esca and illicium. The characters given in Tables 1 Contributions in Science, Number 332 Bertelsen, Pietsch & Lavenberg: Gigantactinid Anglerfishes 57 Figure 58. Gigantactis microdontis, female metamorphosal stage, 19.5 mm, IOAN uncatalogued. Drawn by Elizabeth Beyerholm. and 2, combined with observations on tooth patterns and cau- dal-fin shape, provide an insufficient basis for reliable identifi- cation. None of the individuals show a combination of characters that might indicate differences from the species rec- ognized above. FEMALE METAMORPHOSAL STAGES, Figure 60, Table 18 MATERIAL. Eight specimens, 9-19.5 mm: LACM 34297-2, 14 mm, 21 °20'N, 158°20'W, 0-1100 m; LACM 32768-3, 19 mm, 21 °20'N, 158°20'W, 0-900 m; MCZ 54041, 9 mm, 39° 33'N, 42°47'W, 0-390 m; SAM 27810, 17 mm, 30°17'S, 31 °25'E, 0-750 m; SIO 61-48, 14.5 mm, 8'14.5'S, 151° 36.5'W, 0-2500 m; SIO 73-158, 20 mm, 28°22'N, 155°02'W, 0-3000 m; ZMUC P921605, 9 mm, 17°55'N, 24°35'W, 300 m wire; ZMUC P921655, 19.5 mm, 1°06'S, 62°25'E, 4000 m wire. COMMENTS. Evidence obtained from the pigment patterns of larvae, and from those of identified metamorphosal stages and juvenile females, indicates that some of the unidentified females listed above probably represent the metamorphosal stages of species belonging to the G. vanhoeffeni group. Enlarged supraorbital lateral line organs, unique to Gigan- tactis, are present in some female metamorphosal stages (Fig. 58). In the material examined, they are found in nine speci- mens (11.5-21.5 mm): two unidentified (listed above), three G. vanhoeffeni, and one each of G. longicirra, G. elsmani, G. sav- agei, and G. microdontis. They are absent in ten meta- morphosal stages of a similar size range (14.5-29 mm): four unidentified (listed above), three G. meadi, and one each of G. perlatus, G. golovani, and G. gargantua, as well as in the two 9- mm specimens that represent Larval Group D (see Fig. 60). The occurrence of this type of organ thus seems unrelated to the supposed relationships of the species, being present in G. vanhoeffeni, but absent in G. meadi, the most similar species Figure 59. Esca of Gigantactis ios, holotype, 57 mm, BMNH 1977.9.13.1: A. Left lateral view; B. Dorsal view. Drawn by K. Elsman. 58 Contributions in Science, Number 332 Bertelsen, Pietsch & Lavenberg: Gigantactinid Anglerfishes within the G. vanhoeffeni group (see “Sensory Structures” in the general description). MALES, Figures 26A, 61, 62, 68; Table 19 Characters previously used to separate Gigantactis larvae into groups (the concentration of subdermal, dorsal, and peritoneal pigment and median-fin ray counts; Bertelsen 1951:148) were of limited use in separating the males. The two specimens here referred to G. longicirra on the basis of their high dorsal-fin counts agree with larval “Type-C” of Bertelsen (1951:150, fig. 99E-F) in having weakly developed pigmentation. In the re- maining material, the density of subdermal melanophores was relatively great but varied without significant relation to other characters. This variation is due to ontogenetic changes but also is probably caused by different preservation and storage times allowing for an unequal loss of pigment. The size of the olfactory organs, used to distinguish other ceratioid males (Bertelsen 1951), also proved to be of little im- portance in distinguishing Gigantactis males. The thin, fragile skin covering the olfactory lamellae is torn or deformed in most of the specimens so that the only reliable measurement of size is the length of the series of olfactory lamellae (this length var- ied from 0.8 to 1 .8 mm without any significant relation to other characters); but even in the best preserved males, the size of the nostrils of left and right sides differs significantly. There is the impression, however, that the anterior nostrils are slightly smaller and more narrow than the posterior nostrils and have a depth of approximately one-half the length of the series of olfactory lamellae. The final results of a thorough investigation provided five characters that can be used to separate Gigantactis males into groups: (1) eye diameter, (2) dermal pigmentation, (3) subder- mal pigmentation (used to separate a single specimen), (4) presence or absence of skin spines, and (5) fin-ray counts (the only character found in males that is also useful in distinguish- ing species based on females). In evaluating their significance, changes in these characters during metamorphosis were consid- ered in view of the fact that (1) in the larvae the skin is unpig- mented and naked, (2) the eyes of larvae are normally developed and relatively large, and (3) the olfactory organs of larvae are small (with few olfactory lamellae) and occupy a lat- eral position. To judge the relative stage of development, the testes of each specimen were examined and the maximum di- ameter of the larger of the pair was measured (since each testis is more or less pear-shaped, tapering into a stalk of variable length, no useful length measurement could be obtained). On this basis the material of Gigantactis was separated into six groups: G. longicirra (two specimens) and five additional groups that are for convenience called Group I through Group V. Group 1 includes the material referred to as the “naked type” by Bertelsen (1951:152, fig. 102C). Group II includes G. microphthalmus (Regan and Trewavas 1932:92, fig. 149; Ber- telsen 1951:152). The remaining three groups include pre- viously undescribed material. The characters of all known male specimens (including a number of metamorphosal stages that could not be assigned to a group) are compared in Table 19. Male Group I, Figures 61 A, 68; Table 19 Laevoceratias liparis Parr 1927:33, fig. 13 (original descrip- tion, single specimen); Regan and Trewavas 1932:93 (after Parr 1927); Bertelsen 1951:70, fig. 29 (tentatively referred to Di- ceratiidae, figure after Parr 1927). MATERIAL. Ten specimens, 1 5-22 mm. Holotype of Laevoceratias liparis: BOC 2013, 17 mm, 24 ’ 1 l'N, 75 ' 37' W, 2440 m wire. Referred material: GMZA, uncatalogued, 19.5 mm, Skagerak 1946, 38°25'N, 10°23'W, 5000 m wire; IOS un- catalogued, 22 mm, 31"51'N, 63°47'W, 1260-1500 m; LACM 33316-3, 16.5 mm, Hawaii, leeward Oahu, 0-1100 m (cleared and stained); LACM 35668-2, 19 mm, 21°20-30'N, 158°20- 30'W, 0-1000 m; LACM 30199-26, 17.5 mm, 31°48'N, 1 19°48'W, 0-1320 m (stained); SIO 68-490, 17 mm, 29°07'N, 178°05'W; SIO 70-336, 21.5 mm, 18“49'N, 124”22'E, 0-1525 m; USNM 218615, 15 mm, 32”27'N, 64°17'W, 0-1536 m; ZMUC P921536, 19 mm, 25” 1 l'N, 20“57'W, 5000 m wire. DIAGNOSIS. Males of Group I differ from other Gigantactis males in having the following combination of characters: eyes relatively large, diameter 0.6-0. 9 mm (average 0.79 mm); olfactory lamellae 11-12; upper denticular teeth 3 (rarely 2); Table 17. Counts and measurements in percent of SL of unidentified metamorphosed Gigantactis females. Character LACM 36076-8 LACM 34311-2 ISH 150/68 ISH 1958/71 MCZ 52573 Standard length 23.5 47 177 292 300 (mm) Length Illicium 60 + 85 + 155 + 147 + 83 + Longest premaxillary 2.1 2.1 1.4 0.8 0.6 tooth Longest dentary 3.8 3.7 3.2 5.2 0.9 tooth Longest caudal ray 32 28 32 27 40 Teeth Premaxillary 7 4 12 23 23 Dentary 12 18 46 63 48 Dorsal-fin rays 7 6 6 6 5 Anal-fin rays 7 5 6 5 5 Pectoral-fin rays 16 17 19-20 17 17 Contributions in Science, Number 332 Bertelsen, Pietsch & Lavenberg: Gigantactinid Anglerfishes 59 Table 18. Counts and measurements in percent of SL of unidentified female metamorphosal stages of Gigantactis. Character ZMUC P921605 MCZ 54041 LACM 34297-2 SIO 61-48 SAM 27810 LACM 32768-3 ZMUC P921655 SIO 73-158 Standard length (mm) 9 9 14 14.5 17 19 19.5 20 Illicial length 5 10 19 9 7 24 10 12 Larval teeth + + + + + — + — Post-metamorphic dentary teeth 0 0 0 0 0 1 1 0 0 Enlarged supraorbital filaments 0 0 7-8 0 0 0 4 0 Subdermal dorsal pigmentation + + faint + + + + + Dorsal-fin rays 6 6 5 5 5 5 6 — Anal-fin rays 5 6 5 5 6 6 5 — Pectoral-fin rays 18 19 17 21 21 17 18 — lower denticular teeth 4 (rarely 3); skin naked, pigmented; dor- sal-fin rays 5-6, anal-fin rays 5-7, pectoral-fin rays 18-22 (Table 19). COMMENTS. The material of Group I differs from other known Gigantactis males in length distribution: eight of the ten known specimens are larger than the males of all other groups, ranging from 17 to 22 mm. In two of the smallest known speci- mens of Group I (15 mm, USNM 218615; 17 mm, SIO 68-490), the anterior nostrils are somewhat separated, but oth- erwise their metamorphosis seems to be complete. The greatest diameter of the testes of the largest known specimen (22 mm, IOS uncatalogued) is 2.7 mm; in the remaining material, this measurement varies disproportionately with standard length from 0.8 to 2.2 mm. Male Group II, Figures 6 1 B, 68; Table 1 9 Teleotrema microphthalmus Regan and Trewavas 1932:93, fig. 149 (original description, single specimen). Gigantactis microphthalmus Bertelsen 1951:146, 152, 153, figs. 102C, 103E, table 32 (new combination; comparison with all known material). MATERIAL. Twenty-two specimens: 10.5-15.5 mm. Holotype of Teleotrema microphthalmus: ZMUC P92127, 16 mm, 8°26'N, 15° 1 l'W, 5000 m wire. Referred material: IOAN uncatalogued, 13 mm, 6°50'S, 103°28'E, 0-2000 m; IOS uncatalogued, 13 mm, 10°45'N, 20°10'W, 0-1250 m; IOS uncatalogued, 13.5 mm, 17°56'N, 24°59'W, 1250-1500 m; IOS uncatalogued, 15.5 mm, 17‘43'N, 24° 59'W, 1250-1500 m; LACM 36034-1, 2 (12-12.5 mm), 4° 58'S, 130‘12'E, 0-750 m; LACM 36033-1, 2 (13-14 mm), 4° 58'S, 129°43'E, 1000-1500 m; LACM 33324-1, 13.5 mm, 21 ° 20-30'N, 158°20-30'W, 0-620 m (cleared and stained); LACM 36032-1, 13.5 mm, 4°27'S, 129°52'E, 0-1500 m; LACM 32800-4, 14 mm, 21°20-30'N, 158°20-30'W, 0- 800 m; LACM 32778-1, 14 mm, 21°20-30'N, 158°20-30'W, 0-500 m; MCZ 54043, 10.5, 3°55'N, 60°08'E, 0-2000 m; MCZ 54042, 12 mm, 1°20'S, 27°37'W, 0-1100 m; SIO 63-560, 12.5 mm, 00°55'N, 1 1 °29'W, 0-2300 m; SIO 69-354, 13.5 mm, 17°48'N, 143°41'E, 0-525 m; SIO H52-409, 14.5 mm, 01 °02'N, 91°46'W, 0-1100 m; SIO 68-482, 14.5 mm, 22°03'N, 171 °46'E; SIO 60-232, 15.5 mm, 5°02'S, 135°03'W, 0-2750 m; ZMUC P921537, 11.5 mm, 17°56'N, 64°49'W, 3000 m wire; ZMUC P921533, 14.5 mm, 17°58'N, 64°41'W, 4000 m wire (cleared and stained). DIAGNOSIS. Group II males differ from other Gigantactis males in having the following combination of characters: eyes relatively small, diameter 0.45-0.60 (average 0.54 mm); olfac- tory lamellae 11-12 (two specimens with 10 and 13, respec- tively); upper denticular teeth 3 (rarely 4); lower denticular teeth 4 (rarely 3 or 5); skin densely covered with spines, darkly pigmented; dorsal-fin rays 5-6, anal-fin rays 5-7, pectoral-fin rays 16-18 (Table 19). COMMENTS. In the material of Group II, the diameter of the testes ranges from 0.4-1. 9 mm, with some tendency for larger males to have larger testes (diameter less than 1.0 mm in all specimens less than 14 mm SL, greater than 1.5 mm in all material more than 14 mm SL). All known specimens are postmetamorphosal stages. Male Group III, Figure 68; Table 19 MATERIAL. Two specimens: LACM 34291-3, 14 mm, leeward Oahu, 0-1150 m; LACM 34305-4, 15 mm, leeward Oahu, 0-995 m. DIAGNOSIS. Group III males differ from other Gigantactis males in having the following combination of characters: eyes relatively large, diameter 0.7 and 0.8 mm; olfactory lamellae 10-11; denticular teeth present in only the 15 mm male, 3 up- per and 4 lower; skin spinulose, unpigmented; dorsal-fin rays 5- 60 Contributions in Science, Number 332 Bertelsen, Pietsch & Lavenberg: Gigantactinid Anglerfishes Figure 60. Gigantactis sp. female metamorphosal stage, 9 mm, MCZ 54041. Drawn by K. Elsman. 6, anal-fin rays 5-6, pectoral-fin rays 16-18 (Table 19). COMMENTS. The diameter of the testes of the Group III males is 1.3 and 1.5 mm. Both specimens are postmetamor- phosal stages. Male Group IV, Figure 68; Table 19 MATERIAL. One specimen in metamorphosis: LACM 36030-1, 16.5 mm, 4°33'S, 129‘17'E, 0-1100 m. DIAGNOSIS. The Group IV male differs from other Gigantac- tis males in having relatively low pectoral-fin ray counts (15-15). In addition, this specimen differs in having the follow- ing combination of characters: eyes small, diameter 0.5 mm; olfactory organs well-separated with 8 to 9 lamellae; upper den- ticular teeth 3; lower denticular teeth 4; skin naked, darkly pig- mented; dorsal-fin rays 6, anal-fin rays 5 (Table 19). COMMENTS. The diameter of the testes of the Group IV male is 0.4 mm. The development of the olfactory organs and testes indicate that the specimen is a late metamorphosal stage. Male Group V, Figures 62, 68; Table 19 MATERIAL. One specimen in metamorphosis: LACM 32775-1, 14.5 mm, 2r20-30'N, 158°20-30'W, 0-1500 m. DIAGNOSIS. The Group V male differs from other Gigantac- tis males in having a distinct, V-shaped patch of subdermal pig- ment on the throat and in having relatively low dorsal and anal- fin ray counts (D. 4, A. 4). In addition, this specimen differs in having the following combination of characters: eyes large, di- ameter 1.0 mm; olfactory organs in development, not con- tiguous, with approximately 8 lamellae; denticulars in development with 2 upper teeth and 4 lower teeth; skin naked, unpigmented; pectoral-fin rays 21 (Table 19). COMMENTS. Despite the relatively large testes of the Group V male (approximately 2 mm long and 1.2 mm wide), the de- velopment of the premaxillae (unresorbed), eyes, olfactory organs and denticulars indicate that the specimen is an early metamorphosal stage. Males Not Referred to Group, Table 1 9 MATERIAL. Seven specimens, 11.5-14.5 mm: IOS un- catalogued, 13 mm, 30°14'N, 23°02'W, 1250-1500 m; LACM 32773-3, 14 mm, 21°20-30'N, 158°20-30'W, 0-820 m (cleared and stained); LACM 34271-2, 14.5 mm, leeward Oahu, 0-500 m; LACM 32749-4, 13 mm, 21°20-30'N, 158°20-30'W, 0-1000 m (cleared and stained); SIO 70-327, 11.5 mm, 18° 19'N, 133°41'E, 0-1500 m; SIO 70-343, 13.5 mm, 18°06'N, I19°08'E, 0-1850 m; ZMUC P921534, 1 1.5 mm, I7°58'N, 64°41'W, 4000 m wire. COMMENTS. The reduction of the premaxillae and develop- ment of the denticulars of the 13-mm specimen (LACM 32749-4) indicate that it is a postmetamorphosal stage. It is poorly preserved with torn olfactory organs. The unpigmented and naked skin, combined with its small size, make its refer- ence to any of the groups described above doubtful. The six remaining specimens are in metamorphosis. They have more or less transparent, naked skin and relatively small, well-separated olfactory organs containing eight to approximately ten lamellae. The eyes are unreduced and relatively large (0.75-1.1 mm in diameter), the denticulars are in development (not yet present in the 11.5-mm specimen, ZMUC P921534), and the premaxillae are complete or only slightly resorbed with remains of jaw teeth in some specimens. Fin-ray counts are dorsal 5-6, anal 5-6, pectoral 16-19. The greatest width of the testes ranges from 0.4 to 1.2 mm (Table 19). Discussion Except for G. longicirra, it is not possible to satisfactorily refer Gigantactis males to species based on females. However, there seems little doubt that the best represented Group II contains the males of the most common species, G. vanhoeffeni. having the same fin-ray counts. Probably the males of other members of the G. vanhoeffeni group are included in Group II as well. The high number of pectoral-fin rays characteristic of Group I indicates that these males correspond to members of the G. macronema and G. gargantua groups. The fin-ray counts of Group III males are shared with a number of species based on females, thus providing no indication of their identity. The ex- ceptionally low number of pectoral rays in Group IV males (15) corresponds only to G. golovani, but since both known females of this species are from the Atlantic and the single Group IV male is from the western Pacific, this evidence seems too slight Contributions in Science, Number 332 Bertelsem, Pietsdi & Lavenberg: Gigamtactinid Anglerfishes 61 10 mm h 1 Figure 61. Gigantactis sp. males: A. Group I, 22 mm, IOS uncatalogued; B. Group II, 14.5 mm, SIO H52-409. Drawn by K. Elsman. to conclude that these forms represent the same species. Simi- larly, the low number of dorsal and anal rays (4) in the Group V male corresponds only to G. elsmani, G. savagei, and G. micro- dontis. Females of G. savagei and G. microdontis and the Group V male are known only from the eastern Pacific, yet none of the known females have more than 19 pectoral-fin rays, while the Group V male has 21. The presence of a well-developed su- praethmoid, and the fact that the subdermal pigment of the Group V male differs significantly from that of all known gigan- tactinid larvae, males, and juvenile females, possibly indicate that it represents an undescribed genus. The most conspicuous difference between the groups of Gi- gantactis males is the presence or absence of skin spines. Since this difference is correlated with standard length (spines pres- ent in all the smaller specimens, but absent in all the larger specimens), it might be only an ontogenetic difference. This conclusion is clearly contradicted by the difference between Group I and Group II males in pectoral-fin ray counts (Table 19). Furthermore, a decrease in actual eye diameter from the metamorphosal stages (see below) to the small spiny specimens, followed by an increase in the larger naked specimens, is very unlikely. Finally, a comparison of the size of gonads (Table 19) shows that some of the naked specimens have smaller and prob- ably less advanced testes than most of the spiny specimens. LARVAE, Figure 68 MATERIAL. In addition to the 233 larvae (2.5-14 mm total length) listed by Bertelsen (1951), we have examined 66 speci- mens, 4-15 mm total length (18 MCZ, 48 in collections of the National Marine Fisheries Service, Southwest Fisheries Center, La Jolla; see also Maul 1962, 1973). COMMENTS. Bertelsen (1951) divided the Gigantactis larvae into three types (here referred to as groups): Group A, with dorsal and peritoneal pigment well developed, dorsal-fin rays 5- 7, anal-fin rays 5-7; Group B, with dorsal and peritoneal pig- ment weak or absent, dorsal-fin rays 4-6, and anal-fin rays 4-6; and Group C, with pigment as in Group B, but dorsal-fin rays 7-10, anal-fin rays 5-8. The groups, however, could not be sharply distinguished. Several specimens were found to be in- termediate in pigmentation, and the chosen limit between groups B and C in fin-ray counts was doubtful. In examination of much of the additional material now available, it was impos- sible to judge whether differences in the strength of pigmenta- tion in larvae obtained from different sources might be due to 62 Contributions in Science, Number 332 Bertelsen, Pietsch & Lavenberg: Gigantactinid Anglerfishes variation in the degree of bleaching caused by differences in preservation and storing time. Furthermore, much of the mate- rial is small (2.5-5 mm TL), representing young stages in which the separation between Groups A and B is especially un- certain. No larvae with more than seven dorsal rays were found. For these reasons, the attempt to separate Gigantactis larvae into distinct groupings failed. Nevertheless, combined with the much greater information on the genus now available, a reexamination of the larvae revealed some new facts. Larval Croup A Among the identified female metamorphosal stages (represent- ing seven of the 16 recognized species of Gigantactis), only those of G. vanhoejfeni and G. meadi have a well-developed dor- sal group of subdermal pigment. Furthermore, distinct remains of this pigment were found under the darkly pigmented skin of the juvenile specimens of G. gibbsi (38 mm, USMN 218613) and G. gracilicauda (21 mm, ZMUC P921535), the two re- maining species of the G. vanhoejfeni group. This indicates that Group A contains the larvae of species belonging to the G. van- hoejfeni group, and further infers that this type of pigmentation is unique to the group. If this is correct, five of the eight un- identified female metamorphosal stages listed in Table 18 (LACM 32768-3, SIO 61-48, SIO 73-158, SAM 27810, and ZMUC P921655) also belong to the G. vanhoejfeni group. Since approximately one-half of the identified Gigantactis be- long to this group, this possibility is not too unlikely. Larval Group B In the identified female metamorphosal stages representing G. perlatus, G. elsmani, G. gargantua. G. savagei, and G. micro- dontis, the dorsal subdermal pigment is either absent or (as in G. gargantua) very faint. Since none of the specimens have darkly pigmented skin, the possibility that this lack of subder- mal pigment is due to bleaching cannot be excluded. However, it seems probable that the larvae of these species are included among those of Group B (see also “Male Group I”). Larval Group C Since G. longicirra is unique among the recognized species of the genus in having more than seven dorsal-fin rays, the eight larvae with such high dorsal-ray counts referred to Group C by Bertelsen (1951) no doubt represent this species. On the other hand, the two larvae with seven dorsal rays included in this group by Bertelsen (1951) are removed to the unidentified lar- vae of Group B. The absence of dorsal subdermal pigment in G. longicirra is confirmed in the identified female metamorphosal stage and in the two metamorphosed males. Larval Group D Two small females (both 9 mm, MCZ 54041 and ZMUC P921605) differ slightly in subdermal pigmentation from other specimens examined. According to the presence of numerous tiny melanophores in the skin and a developing external illici- B 1 Figure 62. Gigantactis male Group V, 14.5 mm, LACM 32775-1: A. Lateral view; B. Ventral view of throat. um, they represent early metamorphosal stages. Besides having a well-developed dorsal group of subdermal pigment, they both have a short dorsal and ventral series of three to four large melanophores on the caudal peduncle (Fig. 60; see also Ber- telsen 1951, fig. 59c). All Gigantactis larvae examined lack pig- ment on the caudal peduncle, whereas metamorphosal stages and juveniles of Group A may have diffuse pigment extending posteriorly to the caudal peduncle. In no other specimens ex- amined is pigment concentrated in large, well-separated melanophores. For this reason, it seems probable that the Group D specimens represent one of the species for which no metamorphosal stages have been yet identified. Genus Rhynchactis Regan Rhynchactis Regan 1925:565-566 (type species Rhynchactis leptonema Regan 1925, by original designation and monotypy). DIAGNOSIS. The genus Rhynchactis is distinguished from Gigantactis by having pelvic bones and by having dorsal-fin rays 3-4 (rarely 5), anal-fin rays 3-4. In addition, meta- morphosed females differ in having the following characters: frontals absent; parietals absent; premaxilla represented by an anterior remnant bearing 0-2 teeth; maxilla absent (present in larvae); dentaries with minute teeth or toothless; 2 hypohyals; caudal-fin rays 3-6, branched; skin with minute spines in larger specimens, naked in juveniles; snout truncated, bearing illicium slightly behind tip; esca without bulb. Metamorphosed males are distinguished from those of Gi- gantactis by having the following characters: diameter of eye 2.4% SL in largest specimens ( 1 7-1 8.5 mm); olfactory lamellae 13-15; depth of nostrils 10-12% SL; upper denticular teeth 4- 6; lower denticular teeth 6-7; at least some denticular teeth paired with broad, conical bases; skin naked, weakly pigmented (subdermal pigment as in larvae, see below); pectoral-fin rays 17-19. Larvae of Rhynchactis differ from those of Gigantactis in having the following characters: dorsal group of subdermal pig- Contributions in Science, Number 332 Bertelsen, Pietsch & Lavenberg: Gigantactinidl Anglerfishes 63 Table 19. Counts and measurements in percent of SL of males of Gigantactis and Rhynchactis. ^ o 2 s* 9 •S? a O 3 IX) CO fN fN CO ON fN r- ON Tt CO © co NU •X) CO Ov Cl. •a 3 *T3 nJ 5 5 nJ .V) tj 9 cr OXJ ‘a -o S Z C/5 5 OO NO o © fN U o s <: u 3 S 1) -* 3 3 9 3*: 3 3 00 CQ 3 3 N N 3 C/3 O r- r- oo r- r- oo N0N0*XiiX)»X)N0N0»X)N0«X)»X)tX)»X)N0N0'O l l l 1X)nOnOnOnOnOnOnOnO'OnOnOvOnOnOnO l l l Tt‘Xir-'©©NO»'’‘)(Nco © Tt © »X) © © -^vd^oaiodvooc) — r- ~ r- o © — © 7 7 o 7 + fN fN */~) NO NO NO l/~! l/l NO NO NO t/~> no nO NO no NO no tJ- tJ- ■'-f Tf NO rf NO rn NO r~~ d d r-* d — ro O On O — 7 6 >^i oo — ri «ri ri m O O — — c«~> m no oo CL u D ON OO 2s Tt On oo CJ 2 O O 2 n 55 55 n O w 55 2 — »/-> ON o (N r^> ~ -o Q. W 2 ? 0 E OX) ro rn 1 '§• S 2 2 3. y y § c J J ■SP’S. a 13 > 2 o | OX) £ ’5. o ^ S w <3 .2? O S > “ D. a 3 JJ 2 « o S- a . fs .§>« N5 5 2 Cl ^ ^ O a S- § U §25 Sj 3 2 3 r<= 7 The scientific publications of the Natural History Museum of Los Angeles County have been issued at irregular intervals in three major series; the articles in each series are numbered individually, and numbers run consecutively, regardless of subject matter. » Contributions in Science, a miscellaneous series of technical papers describing original research in the life and earth sciences. ® Science Bulletin, a miscellaneous series of monographs describing original research in the life and earth sciences. This series was discontinued in 1978 with the issue of Numbers 29 and 30; monographs are now published by the Museum in the Contributions in Science series. • Science Series, long articles on natural history topics, generally written for the layman. Copies of the publications in these series are sold through the Museum Bookshop. EDITORIAL BOARD Lawrence G. Barnes Robert L, Bezy Kenneth E. Campbell John M. Harris Robert. J. Lavenberg William B. Lee Roy R, Snelling Camm C. Swift Fred S. Truxal David P. Whistler Edward C. Wilson John W. Wright §wm ■ iPit.r&S mi THE HOLOCENTRID FISHES OF THE GENUS M YRI PRISTIS OF THE RED SEA, WITH CLARIFICATION OF THE MURDJAN AND HEXAGONUS COMPLEXES1 John E. Randall2 and Paul Gueze3 ABSTRACT: Three species of Myripristis Cuvier are characterized by having two pairs of symphysea! tooth patches outside the gape anteriorly on the lower jaw: M. hexagonus (Lacepede) from the western Pacific to the Indian Ocean, but not the Red Sea; M. xanthacrus new species, a Red Sea and Gulf of Aden endemic with yellow on the distal ends of the caudal lobes and elevated soft portions of the dorsal and anal fins; and M melanostictus Bleeker with a distribution similar to that of hexagonus. The first two have small scales in the axil of the pectoral fins, whereas these are lacking on M melanostictus. Ostichthys spiniceps Ogilby is a junior synonym of M. hexagonus, but M. australis Castelnau, previously regarded as synonymous with M. hexagonus, is a probable synonym of M. violaceus Bleeker. Apparently only one other Myripristis occurs in the Red Sea for which the name murdjan (Forsskdl) should be applied ( parvidens Cuvier is a synonym); it ranges to the western Pacific. The species M. seychellensis Cuvier and M. berndti Jordan and Evermann, both regarded as syn- onyms of murdjan by some authors, are valid. M seychellensis is known at present only from three localities in the western Indian Ocean — Seychelles, St. Brandon’s Shoals, and Reunion; it is distinctive in having a terminal mouth or the lower jaw slightly inferior and a patch of vomerine teeth with the posterior border rounded. M. berndti ranges throughout the Indo-Pacific and occurs in the eastern Pacific as well; it has a strongly jutting lower jaw (only slightly projecting in murdjan ), a narrow interorbital space (width 4.3 to 5.2 in head compared with 3.7 to 4.4 for murdjan ), and orangish-yellow on the outer part of the spinous dorsal fin. INTRODUCTION The senior author collected specimens of a soldierfish (Holo- centridae: Myripristis) in the Red Sea off Sudan in 1974 and 1975 and in Djibouti, Gulf of Aden, in 1977 that seemed unique in having the caudal lobes and elevated parts of the soft dorsal and anal fins broadly tipped with bright yellow. Only M chryseres Jordan and Evermann (1903), known from Hawaii, Japan, and Reunion,4 has solid yellow on these fins, but this color is not restricted to the distal part in this species. Further- more, M. chryseres has 32 to 38 lateral-line scales, whereas the Red Sea and Gulf of Aden species has 26 to 29. In his useful revision of the genus Myripristis, Greenfield (1974) recognized 15 species, of which three were recorded from the Red Sea: M. murdjan (Forsskdl 1775), M. parvidens Cuvier (1829), and M. hexagonus (Lacepede 1802). Our yellow-tipped specimens key to M. hexagonus in Greenfield’s paper in having two pairs of symphyseal tooth patches, one above the other, at the front of the lower jaw (just outside the gape — see Greenfield’s fig. 3). They also possess small scales in the axil of the pectoral tins (usually present on hexagonus, ac- cording to Greenfield). There is, however, no mention of yellow on the fins in Greenfield’s account of hexagonus, which caused us to compare our specimens with those identified as hexagonus in various museums. The collections of the British Museum (Natural History) contain three fish that Greenfield had identi- fied as hexagonus but which proved to be the same species as our Red Sea specimens. Our study also revealed two other spe- cies, one with scales in the axil of the pectoral fins and one without, among the specimens labelled hexagonus in these museums. To determine which of these species would bear Lacepede’s name hexagonus, the type specimen (Fig. 14) at the Museum National d’Histoire Naturelle in Paris was examined. This spec- imen (MNHN A. 5423) is a dried, varnished half skin 1 12 mm in standard length. The two pairs of symphyseal tooth patches may be clearly seen at the front of the lower jaw; there are small scales in the axil of the pectoral fins; the lateral-line scale count is 27. The type locality was not given, but Facepede ( 1 802) stated that the fish was a gift from the Dutch, thus mak- ing the eastern Indian Ocean or Indonesia, and not the Red Sea, the likely locality. Greenfield listed three nominal species as junior synonyms of M. hexagonus: M. melanostictus Bleeker ( 1863), M. macrolepis Bleeker (1873), and M. australis Castelnau (1875). The holo- type of M. melanostictus and the types of M. macrolepis at the Rijksmuseum van Natuurlijke Historic at Feiden were exam- ined. The former (RMNH 5157, 149 mm SF, Greenfield 1974, fig. 18) is a specimen of aberrant color pattern, variously blotched with black pigment. It has two pairs of symphyseal tooth patches on the lower jaw but no scales in the pectoral 1. Review Committee for this contribution: David W. Greenfield, Robert J. Lavenberg, John E McCosker, and Camm C. Swift. 2. Bernice P. Bishop Museum, Box 19000-A, Honolulu, Hawaii 96819, and Research Associate, Ichthyology Section, Natural History Museum of Los Angeles County. 3 B.P. 1, 97461 St. Denis, He de la Reunion. 4. The Indonesian record of this species given by Greenfield (1974) is based on a specimen at the Rijksmuseum van Natuurlijke Historie in Leiden (No. 5414). The true locality of this specimen, however, is Reu- nion. Recently M. chryseres was collected at Tutuila, American Samoa, by Richard C. Wass; the specimen is deposited in the Bishop Museum Phillip C. Heemstra (pers. comm.) reported the capture of a specimen at Aliwal Shoal off Durban, South Africa. Contributions in Science, Number 334, pp. 1-16 Randall and Gueze: Red Sea Myripristis 1 Natural History Museum of Los Angeles County, 1981 ISSN 0459-0113 axil. M. macrolepis is clearly a synonym of M. hexagonus. The type of M. australis is apparently lost (see discussion under the account of M. hexagonus ), but it appears to be a synonym of M. violaceus Bleeker (1851). Therefore, there are three species among the specimens iden- tified as M. hexagonus by Greenfield: the true hexagonus, melanostictus, and the Red Sea and Gulf of Aden species, which is described herein as new. Other than the new species, the senior author’s collections of Myripristis from the Red Sea seem to contain only a single spe- cies that is not clearly M. murdjan or M. parvidens as defined by Greenfield (solely on the relative width of the interorbkal space). As is discussed in the remarks under murdjan below, we determined that Greenfield was in error in recording both murdjan and parvidens from this locality. All of his Red Sea parvidens are referable to murdjan. A problem then arose as to what name to apply to the species from outside the Red Sea that Greenfield had called M. murd- jan. Our study has shown that two of the nominal species he referred to the synonymy of murdjan, M. seychellensis and M. berndti, are not M. murdjan but valid species. To permit differentiation of the three closely allied species of the hexagonus complex and the three we might term the murd- jan complex, we present a key to the ten large-scale species of Myripristis (which includes the last seven species listed in Greenfield’s table 1), followed by accounts of the six species of the hexagonus and murdjan complexes in the order they appear in the key. Color photographs of fresh specimens of these six species, underwater photos of five of them (plus M. violaceus ), and meristic data (Tables 1 and 2) are also provided. Green- field’s key and his counts of lateral-line scales, soft rays of the dorsal and anal fins, and gill rakers continue to be fundamental to the identification of the other species of the genus. METHODS AND MATERIALS The last two dorsal and anal soft rays were counted separately, even if closely spaced, as long as each had its own basal ele- ment. Pectoral-ray counts include the short upper ray. The up- per-limb gill rakers are listed first; the raker at the angle is con- tained in the lower-limb count. Rudiments were included in the gill-raker counts except for roundish, entirely sessile plates (sometimes rather large), which are occasionally seen at the end of the raker series. An important character in differentiating species of Myr- ipristis is the presence or absence of small scales on the fin side of the axil of the pectoral fins and the extent of this scalation when present. Standard length (SL) was taken from the medial anterior point of the upper lip to the midbase of the caudal fin (posterior end of hypural plate). Depth of body is the maximum depth, often just anterior to the base of the pelvic fins. Head length was measured from the median anterior point of the upper lip to the most posterior edge of the opercular membrane. The or- bit diameter is the maximum diameter of the bony orbit. The interorbital width is the minimum bony interorbital distance. The depth of the caudal peduncle is the least depth. The length of the caudal peduncle was measured horizontally between ver- ticals at the rear base of the anal fin and the base of the caudal fin. The length of the dorsal spines was taken from the edge of the groove into which the spines fold, whereas the anal spines were measured to their extreme bases. Caudal concavity is the horizontal distance between verticals at the tips of the longest and shortest caudal rays (with the fin in normal position). In the description of the new species, data in parentheses re- fer to paratypes when different from the holotype. More mea- surements are given in Table 3 than are summarized in the text. Specimens of Myripristis were examined at or obtained on loan from the following museums: Academy of Natural Sci- ences of Philadelphia (ANSP): Australian Museum, Sydney (AMS): Bernice P. Bishop Museum (BPBM): British Museum (Natural History), London [BM(NH)], California Academy of Sciences, San Francisco (CAS, SU): Hebrew University, Jeru- salem (HUJF): Natural History Museum of Los Angeles County (LACM); Museum of Comparative Zoology, Harvard University, Cambridge (MCZ); Museum National d’Histoire Naturelle, Paris (MNHN); Queensland Museum, Brisbane (QM); Rijksmuseum van Natuurlijke Historic, Leiden (RMNH); J.L.B. Smith Institute of Ichthyology, Rhodes Uni- versity, Grahamstown (RUSI); U.S. National Museum of Nat- ural History, Washington, D.C. (USNM); and Western Australian Museum, Perth (WAM). KEY TO THE SPECIES OE MYRIPRISTIS WITH LOW LATERAL-LINE SCALE COUNTS5 la. Margin of median fins black, broadest on lobes of caudal fin and elevated parts of soft dorsal and anal fins (black on distal third of highest part of dorsal and anal fins); spinous dorsal fin blackish with a submarginal unpigmented zone; scales in life rimmed with deep blue dorsally and salmon pink on sides and ventrally; a single scale (rarely two) in pectoral axil (lower half); largest species of genus (to 250 mm SL) (Indo- West-Pacific) adustus lb. Margin of median fins not black (although there may be a black blotch or streak distally on caudal lobes or on ele- vated parts of soft dorsal and anal fins); spinous dorsal fin not blackish; scales in life rimmed with brownish red, red, or pink (except anterodorsally on M. violaceus where the scales are dark purplish); either no scales or numerous scales in axil of pectoral fins 2 2a. Third anal spine distinctly longer (and much stouter) than fourth (fourth spine contained about 1.2 in length of third spine); third anal spine 1.5 to 1.8 in head; no scales in pectoral axil (southern and western Pacific)6 randalli 2b. Third anal spine shorter than or subequal to fourth spine; longest anal spine 1.9 to 2.65 in head; scales present or absent in pectoral axil 3 5. Average number of lateral-line scales less than 32. 6. M. randalli Greenfield was described from specimens collected in the Austral Islands and Pitcairn at a depth of 40 to 46 m. It has since been collected by Richard C. Wass in American Samoa. 2 Contributions in Science, Number 334 Randall and Gueze: Red Sea Myripristis Table 1. Fin-ray and lateral-line scale counts of species of Myripristis of the hexagonus and murdjan complexes. Dorsal Soft Rays Anal Soft Rays Pectoral Rays Lateral-Line Scales Species 13 14 15 10 11 12 13 14 14 15 16 25 26 27 28 29 30 31 32 melanostictus 1 15 1 1 16 1 15 1 1 15 1 hexagonus 3 56 7 46 20 10 53 3 1 3 48 14 xanthacrus 9 54 1 1 1 1 50 2 25 39 4 42 14 2 seychellensis' 1 9 11 3 8 11 berndti 5 90 2 3 87 7 6 88 3 3 63 28 3 murdjan 1 49 16 1 30 34 1 5 56 5 1 10 40 1 1 2 2 'One specimen with deformed dorsal soft rays; no dorsal-ray count made. 3a. Two pairs of symphyseal tooth patches, one above the other, at tip of lower jaw just outside gape (lower pair usually absent in specimens smaller than about 90 mm SL) 4 3b. A single pair of symphyseal tooth patches at tip of lower jaw just outside gape 6 4a. Small scales in axil of pectoral fins; no black on fins; total gill rakers 36 to 46; maximum size to 1 53 mm SL 5 4b. No scales in axil of pectoral fins; a prominent black blotch distally on caudal lobes and elevated portions of soft dor- sal and anal fins; total gill rakers 32 to 38; maximum size to 243 mm SL (western Pacific and Indian Oceans) melanostictus 5a. Soft dorsal, anal, and caudal fins not tipped with yellow; total gill rakers 36 to 43 (western Pacific and Indian Oceans, but not Red Sea or Gulf of Aden) hexagonus 5b. Soft dorsal, anal, and caudal fins tipped with yellow; total gill rakers 39 to 46 (Red Sea and Gulf of Aden) xanthacrus n. sp. 6a. No scales in axil of pectoral fins; lateral-line scales 30 to 34 (Oceania7) woodsi 6b. Small scales present in axil of pectoral fins; lateral-line scales 27 to 31 7 7a. Scales on upper part of body with broad blackish rims (deep purplish or blue to dark brown in life), those on in- terorbital and nape almost completely dark; upper-limb gill rakers 12 to 16 (modally 14); lateral-line scales 27 to 29 (modally 28) ( I ndo- West-Pacific) violaceus 7b. Scales on upper part of body without blackish rims; up- per-limb gill rakers 11 to 15 (modally 12 or 13); lateral- line scales 27 to 32 (modally 29 except seychellensis with 28) .’ 8 7. Will be reported from the Ryukyu Islands by Yamakawa and Shimizu. 8a. Mouth terminal or lower jaw slightly inferior when mouth fully closed; posterior border of patch of teeth on vomer rounded (Fig. 1A), at least on adults; dorsal soft rays 15 (western Indian Ocean) seychellensis 8b. Lower jaw projecting when mouth closed (except juve- niles); posterior border of patch of teeth on vomer straight (Fig. 1 B); dorsal soft rays modally 14 9 9a. Interorbital space relatively narrow, the bony width 4.3 to 5.2 in head length; lower jaw of adults strongly projecting when mouth fully closed; a broad zone of orangish-yellow on outer part of spinous dorsal fin in life (eastern Pacific and Indo- West-Pacific, but not Red Sea) berndti 9b. Interorbital space relatively broad, the bony width 3.8 to 4.4 in head length; lower jaw of adults slightly projecting when mouth fully closed; outer part of spinous dorsal fin red in life (Indo- West-Pacific, but absent from Hawaii and French Polynesia) murdjan SPECIES ACCOUNTS Myripristis melanostictus Bleeker Figures 2,8; Tables 1,2 Myripristis melanostictus Bleeker 1863:237 (type locality, Ternate). Myripristis murdjan ( non Forsskdl) Fourmanoir and Laboute 1 976: 1 56, col. figs. DIAGNOSIS. Dorsal fin rays X-I,13 to 15 (usually 14); anal fin rays IV, 1 1 or 12 (usually 12); pectoral fin rays 14 to 16 (usually 15); lateral-line scales 27 to 29 (usually 28); rows of scales be- tween lateral line and middle of spinous dorsal fin 2Vi\ gill rakers 1 I to 1 3 + 21 to 26. Depth of body 2.12 to 2.45 in SL; head length 2.67 to 2.98 in SL; snout 4.21 to 4.69 in head; orbit diameter 2.09 to 2.48 in head; interorbital width 4.49 to 5.64 in head; maxilla extending posteriorly to a vertical at hind edge of pupil, its length 1.63 to 1.78 in head; least depth of caudal peduncle 3.15 to 3.75 in head; fourth (rarely third or fifth) dorsal spine the longest, 2.05 Contributions in Science, Number 334 Randall and Gueze: Red Sea Myripristis 3 to 2.67 in head; longest dorsal soft ray (usually the second) 1.34 to 1.52 in head; fourth anal spine usually slightly longer than third, 2.25 to 2.69 in head; longest anal soft ray (second or third) 1.38 to 1.57 in head; caudal fin 1.32 to 1.41 in head; pectoral fins 1.50 to 1.65 in head; pelvic fins 1.35 to 1.53 in head. Two pairs of tooth patches, one above the other, at symphysis of lower jaw just outside gape; vomerine teeth in a triangular patch with rounded corners, the posterior border slightly indented; no scales in pectoral axil. Color in alcohol light brown with a silvery cast, the edges of the scales darker than centers; upper opercular membrane blackish, the pigment disappearing about half the distance from opercular spine to level of upper pectoral base; only a trace of dark pigment in a small area in upper axil of pectoral fins; a blackish spot on about outer third of soft dorsal fin be- tween third and sixth rays, and a comparable dark spot on about outer fourth of anal fin between second and fourth or fifth rays; caudal lobes tipped with a blackish spot. Color when fresh as illustrated in Figure 2. REMARKS. The proportional measurements given above are based on 15 specimens, 127 to 243 mm SL. The holotype (RMNH 5157, 149 mm SL) was examined by the senior author in Leiden. It has been illustrated by Green- field (1974: fig. 18). As pointed out by Weber and de Beaufort (1929) and Greenfield (op. cit ), this specimen is melanistic. It has large black areas on the fins and random black blotches over the head and body. Greenfield stated that he found 37 specimens of four other species of Myripristis that exhibit simi- lar aberrant coloration M. melanostictus seems to be a relatively rare species. The Bishop Museum has only five specimens: one from a fish mar- ket at Negombo on the west coast of Sri Lanka (Ceylon), two speared by the senior author in dead reef area at 30 m in the lagoon of North Male Atoll, Maidive Islands, and two from the Cebu City fish market, Philippines. In other museums, we found 12 additional specimens from the following localities: Ternate and Bourou, Indonesia; Bulan Island and Zamboanga, Philippines; Sri Lanka; and Tanegashima, Japan. Fourmanoir and Laboute (1976) have two color illustrations of this species (as M. murdjan) in their book on fishes of New Caledonia and the New Hebrides. The senior author observed and photo- graphed an adult individual in 28 m off Sodwana Bay, Kwazulu, South Africa (27°S). The species from southern Japan identified as M. melanostic- tus by Masuda, Araga, and Yoshino (1975:194, pi. 306) is M. O Os ABC Figure 1. Approximate shape of the patch of vomerine teeth of species of Myripristis. A. seychellensis B. berndti and murdjan C. hexagonus and x anthacrus. adustus. Although not mentioned by Greenfield (1974), adustus often has a second pair of tooth patches below the first pair at the front of the lower jaw. These lower patches are gen- erally smaller than the upper ones; however they can be larger, as is the case for a specimen from Assumption Island, Sey- chelles (RUSI 7385, 247 mm SL). Specimens of M. adustus can generally be identified by the broad black zone distally on the soft dorsal, anal, and caudal fins (although the black is broader at the tips of these fins, it is not confined to the tips as on M melanostictus). If the fins are damaged or faded, another helpful character is the presence on adustus of one or two mod- erately large scales on the lower half of the pectoral axil (a single scale found in axil on one side of one of 14 specimens of melanostictus). M. melanostictus has most often been confused with M. hex- agonus because of the double pair of symphyseal tooth patches on the tip of the chin of both species and the incomplete meris- tic separation. The presence of numerous small scales in the axil of the pectoral fins of M. hexagonus is the most useful means of separation. Because M. hexagonus is a relatively small species (maximum about 155 mm SL), the larger size of M. melanostictus can also be helpful in distinguishing the two species. The largest specimen of M. melanostictus (BPBM 22437) measures 243 mm SL. Thus it is the second largest spe- cies of the genus (after M. adustus). Five of the seventeen specimens of melanostictus examined (USNM 218390-93, 218457), all adults, were taken by bottom trawling at Wadge Banks (7 to 8 °N; 77 to 81 °E) off Sri Lanka in 44 to 71 m in 1969, and a sixth specimen (USNM 216719) was taken from the same bank (depth and gear not recorded). Since species of Myripristis usually remain close to coral reefs or rocky areas (they tend to hide in caves or crevices during daylight hours), the capture of these specimens by trawling was unexpected. The label with one specimen (USNM 218391) taken at a station in 64 m, however, stated that the haul was terminated when the net “hung up on rocks.” Myripristis hexagonus (Lacepede) Figures 1C, 3, 14; Tables 1,2 Lutjanus hexagonus Lacepede 1802:213-214 (type locality, Sumatra). Myripristis macrolepis Bleeker, 1873:181,195 (type locality, Java, Nias, Celebes, Buro, and Ambon). Ostichthys spiniceps Ogilby 1908:31 (type locality. Great Barrier Reef). DIAGNOSIS. Dorsal fin rays X-1,13 to 1 5 (usually 1 4), anal fin rays I V, 1 2 or 13 (usually 12); pectoral fin rays 14 to 16 (usually 15); lateral-line scales 25 to 28 (modally 27); scale rows be- tween lateral line and middle of spinous dorsal fin 2Vr, gill rakers 12 to 15 4- 24 to 29. Depth of body 2.10 to 2.51 in SL; head length 2.66 to 3.00 in SL; snout 4.59 to 5.07 in head; orbit diameter 2.04 to 2.39 in head; interorbital width 4.10 to 5.13 in SL; maxilla extending to a vertical at or slightly beyond hind edge of pupil, its length 1.64 to 1.75 in head; least depth of caudal peduncle 3.38 to 3.89 in head; third to fifth (usually the fourth) dorsal spine the longest, 2.10 to 2.42 in head; second or third dorsal soft ray 4 Contributions in Science, Number 334 Randall and Gueze: Red Sea Myripristis longest, 1.43 to 1.60 in head; longest anal spine (usually the third) 2.39 to 2.65 in head; second or third anal soft ray the longest, 1.42 to 1.60 in head; caudal fin 1.22 to 1.50 in head; pectoral fins 1.38 to 1.52 in head; pelvic fins 1.37 to 1.52 in head. Two pairs of tooth patches, one above the other, at symphysis of lower jaw just outside gape [lower pair absent in juveniles (may be seen in individuals as small as 75 mm SL, but one of 105 mm lacked them)]. Vomerine teeth in a broad “V”-shaped patch (Fig. 1C). Small scales on fin side of pectoral axil, usu- ally extending almost to dorsal edge of fin base. Color in alcohol silvery brown, the edges of the scales nar- rowly brown (darker dorsally); opercular membrane blackish, the dark pigment nearly reaching ventrally to level of upper pectoral base; upper part of pectoral axil blackish; no dark pig- ment in fins. Color when fresh as illustrated in Figure 3. REMARKS. The proportional measurements given above are based on 18 specimens 90 to 153 mm SL. The holotype of M. hexagonus (MNHN A. 5423, 1 12 mm SL) was examined in Paris. It is a dried varnished skin of the right side (Fig. 14). In order to determine if there are small scales in the axil of the pectoral fin, a bit of varnish was re- moved with slender forceps from the axil and dissolved in acetone; a small cycloid scale was found in the residue. Bleeker (1873) described M. macrolepis from 25 specimens, 130 to 190 mm in total length, from five different islands in Indonesia: Java, Nias, Celebes, Buro, and Ambon. The Rijks- museum van Natuurlijke Historie in Leiden has three lots of macrolepis collected by Bleeker, the lectotype (RMNH 24910, 134 mm SL) that was selected by D. W. Greenfield, 1 1 paralec- totypes (RMNH 5421, 109 to 138 mm SL), and two other spec- imens (RMNH 24912, 103 to 1 1 1 mm SL). These specimens were examined by the senior author, as was another Bleeker specimen from Nias in the Museum National d’Histoire Natu- relle in Paris (MNHN 2577, 126 mm SL). All are M. hexagonus. The holotype of Ostichthys spiniceps Ogilby, a name over- looked by Greenfield (1974), QM 1.1203, 101 mm SL, was kindly sent on loan by R.J. McKay of the Queensland Museum. It proved to be M. hexagonus. Ogilby gave the type locality as “South Sea Islands”; however, the locality on the original label, written in Ogilby’s handwriting, is “Great Barrier Reef” (McKay, pers. comm.). Greenfield (1974:25) placed Myripristis australis Castelnau (1875) in the synonymy of M. hexagonus. He did not list the type of M. australis , a four-inch specimen from Cape York, Australia, among his “Material Examined” of M. hexagonus. Castelnau’s description of M. australis is brief, and there is no illustration. Nothing in the description indicates that M. aus- tralis is synonymous with M. hexagonus. There is, however, one remark in the description on color, “each scale of the back has a broad edge of dark purple,” which strongly suggests that Cas- telnau had a specimen of M. violaceus Bleeker (illustrated herein as Fig. 9). An effort was made to locate Castelnau’s type specimen. R.J. McKay of the Queensland Museum in Brisbane, Joan M. Davis of the National Museum of Victoria in Mel- Table 2. Gill-raker counts of species of Myripristis of the hexagonus and murdjan complexes. Species 11 12 Upper 13 Limb 14 15 16 21 22 23 24 Lower Limb 25 26 27 28 29 30 31 melanostictus 6 9 2 2 1 6 6 1 1 hexagonus 6 28 30 2 3 7 13 21 18 4 xanthacrus 1 1 38 13 2 6 17 20 10 7 4 seychellensis 1 5 4 1 2 2 4 2 1 berndti 8 51 33 4 3 14 33 26 19 2 murdjan 6 40 17 3 2 11 30 18 3 2 Total Gill Rakers 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 melanostictus 1 2 1 7 4 2 hexagonus 1 1 8 8 15 16 13 4 xanthacrus 4 8 11 13 17 4 5 2 seychellensis 1 1 1 2 4 1 1 berndti 5 14 20 26 18 10 3 1 murdjan 1 4 9 20 19 7 4 1 1 Contributions in Science, Number 334 Randall and Gueze: Red Sea Myripristis 5 bourne, Douglass F. Hoese of the Australian Museum in Syd- ney, and Marie-Louise Bauchot of the Museum National d’Histoire Naturelle in Paris all wrote that the specimen is not in the fish collections of these museums. McKay stated in his letter, “1 have made a careful search of the collection and find no trace of the type of Myripristis australis Castelnau. The type may have been the specimen registered 1206 Ostichthys australis (cast), Cape York, ‘Show Collection’ (no other data) destroyed 20.1 1.1950, as no other O. australis from Cape York is registered in the early register.” The most useful character in separating M. hexagonus and allied species melanostictus and xanthacrus from other species of the genus (except occasional adustus — see remarks on mel- anostictus) is the presence of a double pair of tooth patches at the tip of the chin. It should be noted, however, that juveniles of hexagonus less than about 90 mm SL usually lack the lower pair of tooth patches. The smallest hexagonus with detectable lower tooth patches that we have seen is QM 65241, 75 mm SL. M. hexagonus is a small species. The largest specimen we have seen, USNM 192511, from Mindoro, Philippines, mea- sures 153 mm SL. We have examined specimens of M. hexagonus from the fol- lowing localities: many islands of Indonesia and the Philippines (the specimen illustrated in Fig. 3 was collected in 40 m off Batangas, Luzon); Gulf of Thailand (10-12°N; 99-101 °E); Yirrkala near Cape Arnhem, Northern Territory, Australia (re- ported as M. macrolepis by Taylor, 1964); Palm Islands and Lizard Island, Great Barrier Reef; Cairns and Bowen (20°S), Queensland (Queensland Museum specimens were identified as M. australis by T.C. Marshall 1964:132, pi. 30, fig. 146); Ken- drew Island, Dampier Archipelago (20°28'S; 116°32'E); Mahe, Seychelles; Nossi Be, Madagascar; Inhaca Island, Mozambique; Zanzibar; Bird Island, Suva, Fiji Islands; and Tu- tuila, American Samoa. The specimens from Fiji (AMS I 18354-028, 2: 111-140 mm SL) and Samoa (BPBM 26368, 2: 147-147.5 mm SL), the only known from Oceania, differ from the other M. hexagonus in their high gill-raker counts (15 + 30 or 31); meristic data from these four fish are not included in Table 2. In other re- spects, they seem typical of M. hexagonus. T. Yamakawa and T. Shimizu (MS) will record M. hex- agonus from the Ryukyu Islands and southern Japan. Myripristis xanthacrus new species Figures 1C, 4, 10; Tables 1-3 Myripristis hexagonus ( non Lacepede) Greenfield, 1974 (in part): 24 (Red Sea). HOLOTYPE. BPBM 19784, 100.9 mm SL, Red Sea, Sudan, Suakin Harbor, S side near harbor entrance, fringing reef front, caves in 15-17 m, J.E. Randall, P.J. and P. Vine, 15 October 1975. PARATYPES. BM(NH) 1960.3.15.100-101, 2: 86.6-93.8 mm SL, Red Sea, Mersa Sheikh Ibrahim, “Manihine,” 1950-51; BM(NH) 1960.3.15.111-1 13, 3: 45.3-1 18.5 mm SL, Red Sea, Khor Inkeifail and Sanganeb, “Manihine,” December 1950; USNM 216608, 17: 75.8-121.3 mm SL, Red Sea, Ethiopia, N end of Isola Delemme just E of Ras Coral (15°30.5'N, 39°54'E), depth to 3 m, rotenone, V.G. Springer, 7 August 1969; USNM 216609, 12: 103.3-134.1 mm SL, Red Sea, Ethi- opia, Vi mile off SW shore of Sciumma Island (15°32'31"N, 40°0'E), depth to 7 m, rotenone, V.G. Springer, 9 August 1969; USNM 216610, 5: 98.4-1 15 mm SL, Red Sea, Ethiopia, Ethiopian Naval Base, Massawa, depth to 1.8 m, rotenone, V.G. Springer, 12 August 1969; USNM 216611, 117 mm SL, Red Sea, Ethiopia, Sheikh el Abu, off lighthouse just W of S end of Harat Island (16°08'N, 39°26.5'E), depth to 4 m, ro- tenone, V.G. Springer, 14 August 1969; BPBM 20381, 2: 71.0- 71.5 mm SL, Red Sea, Sudan, Suakin Harbor, off marine lab. jetty, rocks in 1 m, rotenone, J.E. Randall and P.J. Vine, 12 October 1974; BPBM 20443, 2: 101-109.3 mm SL, Red Sea, Sudan, 100 m S of entrance to Port Sudan Harbor, cave in 18 m, spear, J.E. Randall, 17 October 1974; BPBM 20469, 6: 87.7-107.5 mm SL, ANSP 137869, 88.2 mm SL, CAS 38534, 105.2 mm SL, HUJF 8374, 90.2 mm SL, LACM 36278-1, 96.2 mm SL, MNHN 1977. 1 , 92 mm SL — all collected with holo- type; MNHN 1977-452, 7: 81-1 18 mm SL, Gulf of Aden, Djibouti, Gulf of Tadjoura, Maskali Island, W side, 10-15 m, rotenone. J.E. Randall and L.A. Mauge, 14 May 1977; BPBM 21568, 2: 65-127 mm SL, Gulf of Aden, Djibouti, Seven Brothers Islands (Sawabi Is.), Tolka Island (lie Basse) 12° 27.7' N, 43°24.9'E), N side, cave in 12 m, rotenone, J.E. Randall and L.A. Mauge, 19 May 1977. DIAGNOSIS. Dorsal fin rays X-I,13 to 15; anal fin rays I V, 1 0 to 13; pectoral fin rays 14 or 15; lateral-line scales 26 to 29 (modally 27); scale rows between lateral line and middle of spinous dorsal fin 2Vi\ gill rakers 13 to 16 + 26 to 31. Depth of body 2.22 to 2.45 in SL; head length 2.68 to 2.83 in SL; snout 4.43 to 4.97 in head; orbit diameter 2.13 to 2.69 in head; interorbital width 4.30 to 4.77 in head; maxilla reaching posteriorly between verticals at hind edge of pupil and hind edge of orbit; least depth of caudal peduncle 3.22 to 3.92 in head; fourth dorsal spine longest, 2 15 to 2.41 in head; second dorsal soft ray longest, 1.32 to 1.80 in head; second or third anal spine longest, 1.92 to 2.59 in head; caudal fin 1.17 to 1.28 in head, pectoral fins 1.51 to 1.77 in head; pelvic fins 1.41 to 1 .69 in head. Two pairs of tooth patches, one above the other, at symphysis of lower jaw just outside gape; vomerine teeth in a broad “V”- shape; small scales in axil of pectoral fins. Color in alcohol silvery brown, the edges of the scales darker brown, especially dorsally; opercular membrane dark brown to about level of lower edge of pupil; only a little dark pigment dorsally in pectoral axil, but a dark spot present above pectoral base; fins pale. Color in life silvery pink, the edges of the scales brownish red dorsally, red on sides and ventrally; opercular membrane brownish red; median fins red except distal tips of soft dorsal, anal, and caudal lobes, which are yellow; upper and lower mar- gins of caudal fin, leading edge of soft dorsal and anal fins dis- tal to initial spines, and lateral edge of pelvic fins white. DESCRIPTION. Dorsal fin rays X-I,l 4 ( 1 3 to 15); anal fin rays 6 Contributions in Science, Number 334 Randall and Gueze: Red Sea Myripristis I V, 1 2 (10 to 13); pectoral fin rays 14 ( 14 or 15); pelvic fin rays 1,7; principal caudal rays 19 (upper and lower unbranched); up- per procurrent caudal rays 5 (the first four spinous); lower pro- current caudal rays 4 (the first three spinous); lateral-line scales 27 (26 to 29); scales above lateral line to origin of dorsal fin 3; scales above lateral line to middle of spinous portion of dorsal fin 2Vi\ scales below lateral line to origin of anal fin SVi\ predor- sal rows of scales about 9 1/2 ; circumpeduncular scales 12; diago- nal rows of scales on cheek 4; gill rakers 14 + 28 (13 to 16 + 26 to 31); pseudobranch lamellae 32 (18 to 36, generally more on larger specimens); branchiostegal rays 8; vertebrae 11 + 15. Depth of body 2.33 (2.22 to 2.45) in SL; width of body about 2 in depth; head length 2.79 (2.68 to 2.83) in SL; snout 4.79 (4.43 to 4.97) in head; orbit diameter 2.40 (2.13 to 2.69) in head; interorbital space flat, the width 4.58 (4.30 to 4.77) in head; maxilla reaching posteriorly between verticals at hind edge of pupil and hind edge of eye, its length 1.72 (1.67 to 1.77) in head; depth of caudal peduncle 3.35 (3.22 to 3.92) in head. Lower jaw slightly protruding when mouth closed; front of Table 3. Proportional measurements of Myripristis xanthacrus new species, expressed as a percentage of the standard length. Holotype BPBM 19784 BPBM 20381 BPBM 20469 Paratypes BPBM 20735 USNM 216609 USNM 216609 Standard length (mm) 100.9 71.0 87.9 98.5 1 17.9 134.1 Depth of body 42.8 43.8 40.8 43.3 44.9 42.7 Head length 35.9 35.3 37.3 35.5 35.5 36.3 Snout length 7.5 7.1 7.7 7.6 7.9 8.2 Orbit diameter 14.9 16.2 17.5 14.4 13.6 13.5 Bony interorbital width 8.0 8.2 8.2 8.2 7.9 7.6 Length of upper jaw 20.9 21.1 21.8 21.3 21.1 20.5 Depth of caudal peduncle 10.7 10.0 9.5 10.4 11.0 10.6 Lenghtof caudal peduncle 14.2 14.3 13.5 13.6 13.7 13.9 Predorsal length 41.6 42.5 42.4 40.4 42.8 42.1 Preanal length 70.3 69.7 70.8 71.2 70.0 73.0 Prepelvic length 41.3 39.4 40.6 40.7 40.1 42.0 Length of 1 st dorsal spine 10.9 10.6 10.4 8.9 11.5 11.2 Length of 2nd dorsal spine 14.5 13.1 13.7 12.4 13.7 13.3 Length of longest dorsal spine 16.7 16.2 16.1 14.0 15.7 15.1 Length of 1 0th dorsal spine 5.4 4.8 5.0 5.3 5.6 5.1 Length of 1 1 th dorsal spine 11.5 abnormal 10.4 10.5 11.4 10.4 Length of longest dorsal ray 24.6 26.6 broken broken 22.8 20.2 Length of 1st anal spine 2.2 2.5 1.4 1.4 1.6 2.0 Length of 2nd anal spine 6.9 6.8 7.4 5.6 7.5 7.1 Length of 3rd ana! spine 17.3 18.4 17.0 14.8 16 3 14.0 Length of 4th anal spine 19.6 17.2 17 0 15.7 16.1 15.0 Length of longest anal ray 25.3 25.8 25.1 24.0 23.7 22.1 Length of caudal fin 29.3 29.5 29.2 broken 30.2 29.1 Caudal concavity 15.9 16.5 15.9 broken 15.9 16.0 Length of pectoral fin 23.8 22.8 23.0 23.0 22.9 20.6 Length of pelvic fin 24.8 25.0 22.9 23.4 23.6 21.5 Length of pelvic spine 18.4 17.1 16.3 15.3 16.5 15.4 Contributions in Science, Number 334 Randall and Gueze: Red Sea Myripristis 7 Figure 2. Myripristis melanostictus , 195 mm SL, BPBM 19044, Sri Lanka. I i Figure 3. Myripristis hexagonus, 159 mm SL, BPBM 23473, Luzon, Philippine Islands. Figure 4. Myripristis xanthacrus, holotype, 100.9 mm SL, BPBM 19784, Sudan, Red Sea. 8 Contributions in Science, Number 334 Randall and Gueze: Red Sea Myripristis Figure 5. Myripristis seychellensis, 159 mm SL, BPBM 21155, Seychelles. Figure 6. Myripristis berndti, 153 mm SL, BPBM 8456, Guam, Mariana Islands. Figure 7. Myripristis murdjan, 128 mm SL, BPBM 19804, Sinai Peninsula, Red Sea. Contributions in Science, Number 334 Randall and Gueze: Red Sea Myripristis 9 upper lip and front of lower jaw nearly straight when viewed from above; two pairs of protuberant tooth patches at front of upper jaw, one directly above the other, the symphyseal gap be- tween each pair approximately equal to the diameter of one of the patches (one of the two paratypes of BPBM 20381,71.0- 71.5 mm SL, lacks the lower pair of tooth patches, and they are very small in the other paratype); a foramen in lower jaw imme- diately behind each lower tooth patch. Villiform teeth in a band in jaws, the band of the upper jaw notably broader than that of the lower; outer row of teeth in jaws enlarged, nodular, with nearly truncate ends, those at anterior corner of the upper jaw the largest and those on midside of lower jaw the largest (teeth progressively larger anteriorly within the enlarged outer row of jaws); villiform teeth in bands on vomer and palatines, those on vomer in a broad “V”-shape (Fig. 1C); tongue moderately rounded, the upper surface finely papillate. Longest gill fila- ment of first gill arch contained about 1.7 times in longest gill raker. Pattern of longitudinal bony ridges and intervening mucous channels on the top of the head typical of the subgenus Myr- ipristis as illustrated by Greenfield (1974: fig. 1 1C). Opercular spine not very large, the adjacent serrae of some specimens nearly as large. Lower margin of maxilla with a few truncate serrae near posterior edge. Margins of suborbital, opercle, sub- opercle, and double margin of preopercle with numerous sharp serrae, a few at angle of preopercle slightly enlarged. Nasal fossa triangular, usually with a few serrae on posterior edge. Scales strongly ctenoid, as is characteristic of the Holo- centridae, with as many as 38 ctenii on scales on midside of body of holotype. Specialized scales forming a sheath along base of soft dorsal and anal fins. Small scales extending well out on caudal fin (variously missing on type specimens); small pointed scales basally on pectoral fins; small scales in axil of pectoral fins; pointed axillary scale of pelvic fins about, one- third length of pelvic spine. Fourth dorsal spine the longest (but third and fifth may be nearly as long), its length 2.15 (2.18 to 2.41) in head; second dorsal soft ray the longest, 1.46 (1.32 to 1.80) in head; third anal spine much stouter than fourth, but the two subequal in length, the third 2.07 (1.92 to 2.59) in head, and the fourth 1.83 (2.05 to 2.59) in head; second anal soft ray the longest, 1.42 (1.37 to 1.65) in head. Caudal fin forked, its length 1.22 (1.17 to 1.28) in head, the caudal concavity 2.25 (2.13 to 2.34) in head; pectoral fins pointed, the second or third rays the long- est, 1.51 (1.54 to 1.76) in head, the upper two and lowermost rays unbranched, the uppermost ray less than half length of second ray; pelvic fins nearly reaching anus, 1 .45 (1 .41 to 1 .69) in head. Color in alcohol silvery with a brownish yellow cast, the scales on upper part of body rimmed with brown, particularly those above lateral line; opercular membrane brown, the dark pigment diminishing ventrally at about level of lower edge of pupil; only a trace of dark pigment at upper end of pectoral axil, but a moderate amount of pigment above pectoral base forming a broad blotch, which aligns with dark color of opercu- lar membrane; fins entirely pale. Color in life, as illustrated in Figure 4, silvery pink, the edges of the scales red (brownish red dorsally); opercular mem- brane and a zone below leading to pectoral base brownish red; median fins red, the color concentrated in the rays of the soft portions, the tips of the soft dorsal and anal fins and caudal lobes bright yellow; distal half of first soft rays of soft dorsal and anal fins and most of the uppermost and lowermost princi- pal caudal rays whitish. ETYMOLOGY. Named xanthacrus from the Greek “xanthos” for yellow and “akros” for tip or at the end, in reference to the yellow areas distally on the soft dorsal fin, anal fin, and caudal lobes. REMARKS. M. xanthacrus is known thus far from the south- ern half of the Red Sea and the Gulf of Aden. Only one other species of Myripristis, M chryseres, has yellow on the median fins, but this color is not confined to the tips of the fins. M. chryseres is not a close relative; it differs notably from M. xanthacrus in having 32 to 38 lateral-line scales and only a single pair of symphyseal tooth patches at the front of the lower jaw. M. xanthacrus is most closely related to M. hexagonus and may be an allopatric derivative of this spe- cies. It differs from M. hexagonus in the yellow on its median fins, higher average gill-raker counts (see Table 2), and shorter pectoral fins. Myripristis seychellensis Cuvier Figures 1A, 5, II; Tables 1, 2 Myripristis seychellensis Cuvier in Cuvier and Valenciennes 1829: 172 (type locality, Seychelles). Myripristis murdjan ( non Forssk&l) Greenfield 1974 (in part): 20. DIAGNOSIS. Dorsal fin rays X-I,14 or 15 (rarely 14); anal fin rays I V, 1 3; pectoral fin rays 14 or 15 (usually 15); lateral-line scales 28; rows of scales between lateral line and middle of spinous dorsal fin 2Vi\ gill rakers 12 to 15 + 25 to 29 (counts based on 1 1 specimens). Depth of body 2.19 to 2.38 in SL; head length 2.58 to 3.02 in SL; snout 4.75 to 5.28 in head; orbit diameter 2.27 to 2.48 in head; interorbital space 4.19 to 4.50 in head; maxilla extending to between verticals at hind edges of pupil and orbit, the max- illary length 1.71 to 1.82 in head; least depth of caudal pedun- cle 3.39 to 3.83 in head; third or fourth dorsal spines longest, 2.29 to 2.49 in head; second or third dorsal soft ray longest, 1 .44 to 1 .73 in head; third and fourth anal spines subequal, 2. 1 2 to 2.46 in head; second or third anal soft ray longest, 1.31 to 1.58 in head; caudal fin 1.16 to 1.29 in head; pectoral fins 1.47 to 1.56 in head; pelvic fins 1.46 to 1.66 in head. Lower jaw terminal or inferior when mouth fully closed; a single pair of tooth patches at symphysis of lower jaw just out- side gape; vomerine teeth of adults with the posterior border distinctly rounded (Fig. 1A) (the shape of the tooth patch of juveniles seems to resemble that of M. murdjan)', pectoral axil with small scales except on approximately the upper one- eighth. Color in alcohol brown, the edges of scales of body darker than centers; opercular membrane blackish to or nearly to level 10 Contributions in Science, Number 334 Randall and Gueze: Red Sea Myripristis of upper base of pectoral fins; unsealed dorsal part of pectoral axil blackish; fins pale except for broad blackish streaks on each caudal lobe centered on second and third branched rays (darker distally), and on about outer three-fourths of elevated portions of soft dorsal and anal fins (mainly on third and fourth rays and adjacent membranes, but continuing faintly to basal anterior part of fins); longest pelvic ray (second) tipped with blackish. Color when fresh as illustrated in Figure 5. The outer part of the spinous dorsal fin is mainly red (a small amount of yellow pigment may be present). REMARKS. The proportional measurements given above are based on nine specimens, 92 to 176 mm SL. The holotype, as reported by Bauchot (1970:23), is in the Museum d’Histoire Naturelle in Paris; this specimen (MNHN 9518), which was collected by Dussumier, measures 159 mm SL. She listed one paratype, MNF1N 7579, 176 mm SL, also collected by Dussumier, though this fish was not mentioned by Cuvier in the original description. The name Myripristis seychellensis has been used by very few authors. Sauvage (1891) applied it to specimens from Madagascar, but his illustration and lateral-line scale count of 38 clearly indicate that he did not have the true seychellensis (though it doubtless occurs in Madagascar). Gunther (1859), Weber and de Beaufort (1929), and Herre (1953) all erroneous- ly regarded M. seychellensis as a synonym of M. pralinius Cuvier in Cuvier and Valenciennes (1829). Greenfield (1974) placed M. seychellensis in the synonymy of M. murdjan. The senior author speared two specimens of this species (BPBM 21155, 156 169 mm SL) on the S and W sides of North Islet, Mahe, Seychelles, in caves in 15 m on 11 June 1977. Observed in the same caves were M. murdjan, M. adustus, M. kunlee, M. violaceus, and M. berndti, all of which were more common than the M. seychellensis. We have examined only 1 1 museum specimens that we can identify with confidence as M. seychellensis: the two types, the two mentioned above in the Bishop Museum, four from St. Brandon’s Shoals (Cargados Carajos, 16°28-45'S; 59°34- 37'E, USNM 218412-13 and USNM 218460, 131-175 mm SL) collected in 4 to 21 m in April 1976 by V.G. Springer and associates, and three more from the Seychelles collected during the International Indian Ocean Expedition by James E. Bohlke and associates (ANSP 106575, 164 mm SL) and Margaret G. Bradbury and party (CAS 35356, 92-117 mm SL). A color photo of a fresh specimen from Reunion by the junior author leaves little doubt that the species occurs there as well as in the Seychelles and St. Brandon’s Shoals. Two small specimens (CAS 35356) from the Seychelles, which we identify as M. seychellensis, have the posterior border of the patch of vomerine teeth only slightly rounded. Other small specimens, mostly less than 90 mm SL, from the same locality, have the correct meristic data for M. seychellensis and an inferior lower jaw but a straight border to the patch of vomerine teeth. Since M. murdjan of this siz.e may have a ter- minal mouth or slightly inferior lower jaw, these Seychelles specimens would seem to be murdjan. However, it seems likely that the vomerine teeth of seychellensis begin in a triangular pattern in juveniles and develop the rounded posterior border with age. If this is the case, young seychellensis and murdjan, as here diagnosed, may be indistinguishable. Myripristis berndti Jordan and Evermann Figures 1 B, 6, 1 2, Tables 1,2,4 Myripristis berndti Jordan and Evermann 1903: 170 (type locality, Honolulu). Myripristis murdjan (non Forsskdl) Randall 1973: 181. Myripristis murdjan (non ForsskM) Greenfield 1974 (in part):20. Myripristis amaenus Fourmanoir and Laboute 1976:1 55, col. figs. DIAGNOSIS. Dorsal fin rays X-1, 1 3 to 15 (usually 14); anal fin rays I V, 1 1 to 13 (usually 12); pectoral fin rays 14 to 16 (usually 15); lateral-line scales 28 to 31 (modally 29, often 30); rows of scales between lateral line and middle of spinous dorsal fin 2Vi\ gill rakers 1 1 to 1 4 + 23 to 28. Depth of body 2.32 to 2.62 in SL; head length 2.67 to 2.84 in SL; snout 4.62 to 5.30 in head; orbit diameter 2.25 to 2.73 in head; interorbital space narrow, the bony width 4.30 to 5.21 in head; maxilla extending to between verticals at hind edge of pupil and posterior edge of orbit, the maxillary length 1.66 to 1.78 in head; least depth of caudal peduncle 3.27 to 3.68 in head; third to fifth dorsal spines the longest, 2.15 to 2.59 in head; longest dorsal soft ray (usually the second) 1.53 to 1.74 in head (relatively longer on smaller individuals); third and fourth anal spines subequal, 2.14 to 2.42 in head; longest anal soft ray (usually the second) 1.50 to 1.62 in head; caudal fin 1.19 to 1.39 in head; pectoral fins 1.41 to 1.57 in head; pelvic fins 1.49 to 1.73 in head. Lower jaw of adults prominently projecting when mouth is closed; a single pair of tooth patches at symphysis of lower jaw just outside gape; vomerine teeth in a triangular patch with rounded corners; lower one-half to three-quarters of pectoral axil with small scales. Color in alcohol brown, the edges of the scales darker (es- pecially dorsally on body); opercular membrane blackish to a short distance (about half a pupil diameter) below opercular spine, becoming dusky ventrally; unsealed portion of pectoral axil black, some of this pigment generally extending above pec- toral base; fins pale or with dark pigment submarginally near ends of caudal lobes and elevated parts of soft dorsal and anal fins, the pigment on some specimens developed as a distinct band that parallels the rays. Color when fresh as shown in Figure 6. The most significant color feature is the outer yellow or orangish-yellow part of the spinous dorsal fin. REMARKS. The proportional measurements given above are based on 12 specimens, 99.6 to 231 mm SL. This species has been confused with M. murdjan and M. seychellensis. Usually it has been misidentified as murdjan. Randall ( 1 955) gave a detailed color note of the species from a Gilbert Islands specimen Underwater photographs in color have been published by Bagnis et al. (1972, pi. on p. 251) and Contributions in Science, Number 334 Randall and Gueze: Red Sea Myripristis I ! Figure 8. Underwater photo of Myripristis melanostictus, Maldive Islands, 30 m. Figure 9. Underwater photo of Myripristis violaceus, Maldive Is- lands, 10 m. Figure 10. Underwater photo of Myripristis xanthacrus, Sudan, 12 m. Figure II. Underwater photo of Myripristis seychellensis, Seychelles, 14 m. Figure 12. Underwater photo of Myripristis berndti, Kwajalein, Figure 13. Underwater photo of Myripristis murdjan, Sudan, 5 m. Marshall Islands, 10 m. 12 Contributions in Science, Number 334 Randall and Gueze: Red Sea Myripristis Table 4. Comparison of gill-raker counts of Myripristis berndti from the Pacific and Indian Oceans. Upper Limb Lower Limb 1 1 12 13 14 23 24 25 26 27 28 Eastern Pacific Ocean 3 14 4 2 7 9 3 Central and Western 5 30 19 1 7 21 16 9 Pacific Ocean Indian Ocean 8 10 4 3 7 10 2 by Fourmanoir and Laboute ( 1976, right hand figures of p. 1 55, as M. amaenus). M. berndti has the broadest distribution of all the species of the genus: East Africa to the eastern Pacific (Clipperton, Co- cos, and Galapagos Islands); it is not known, however, from the Red Sea or Persian Gulf. In the western Pacific, it ranges from southern Japan to Australia; the senior author observed it at Lord Howe Island (31.5°S). It is one of the most common spe- cies of Myripristis at islands of Oceania and the Indian Ocean. The Bishop Museum has specimens from the Hawaiian Islands, Line Islands, Society Islands, Pitcairn Group, Rapa, Cook Is- lands, Samoa Islands, Gilbert Islands, Marshall Islands, Mar- iana Islands, Wake, Marcus, Solomon Islands, New Hebrides, Mauritius, Reunion and Seychelles. These have been collected in the depth range of 1 to 30 m, but the species occurs at least as deep as 45 m. We have identified other specimens from Mozambique, Kwazulu in South Africa, Indonesia, Borneo, Philippines, Ryukyu Islands, Caroline Islands, Clipperton Is- land, Pacific coast of Costa Rica, and Cocos Island, Costa Rica. The largest specimen (BPBM 3755), 237 mm SL, was collected at French Frigate Shoals, Hawaiian Islands. There is a difference in the average number of gill rakers between the specimens of M. berndti from the Pacific and In- dian Oceans (Table 4). A problem exists with respect to specimens of Myripristis from the Marquesas Islands; these fish have only a few small scales at the extreme ventral part of the pectoral axil and a less projecting lower jaw than typical specimens of berndti. The life color of these specimens is unknown. If normal specimens of berndti were shown to occur in the Marquesas, then the ones with few axillary scales probably represent an undescribed spe- cies. More material from the Marquesas is needed to provide a full analysis of the apparent population differentiation of berndti in this archipelago. Myripristis murdjan (Forsskal) Figures 1 B, 7, 1 3, Tables 1 , 2, 5, 6 Sciaena murdjan Forsskdl 1775:48 (type locality, Jeddah, Red Sea). Myripristis parvidens Cuvier 1829:151 (type locality. Port Praslin, New Ireland). Myripristis axillaris Valenciennes in Cuvier and Valen- ciennes 1831:491 (type locality, Mauritius). Myripristis melanophrys Swainson 1839:207 (new name for M. murdjan Riippell, 1828). Myripristis bowditchae Woods in Schultz et al. 1953:202, pi. 18B (type locality, Bikini Atoll, Marshall Islands). DIAGNOSIS. Dorsal fin rays X-l,13 to 15 (usually 14, rarely 13)'; anal fin rays IV, 1 1 to 13 (rarely 1 1 ); pectoral fin rays 14 to 16 (usually 15); lateral-line scales 27 to 32 (modally 29); rows of scales between lateral line and middle of spinous dorsal fin 2I/2; gill rakers 12 to 15 + 24 to 29. Depth of body 2.30 to 2.49 in SL; head length 2.68 to 3.06 in SL; snout 4.89 to 5.30 in head; orbit diameter 2.1 1 to 2.49 in head; interorbital space relatively broad, 3.78 to 4.43 in head; maxilla extending to between verticals at posterior margins of pupil and orbit, the maxillary length 1.68 to 1.89 in head; least depth of caudal peduncle 3.09 to 3.72 in head; third or fourth dorsal spines longest, 1.96 to 2.48 in head; second dorsal soft ray longest, 1.38 to 1.68 in head; third and fourth anal spines subequal, 2.07 to 2.49 in head; second anal soft ray longest, 1.38 to 1.66 in head; caudal fin 1.12 to 1.33 in head; pectoral fins 1.31 to 1.49 in head; pelvic fins 1.50 to 1.73 in head. Lower jaw of adults slightly projecting when mouth fully closed; a single pair of tooth patches at symphysis of lower jaw just outside gape; vomerine teeth in a triangular patch with rounded corners; lower one-fourth to three-fourths of pectoral axil with small scales. Color in alcohol light brown with a faint longitudinal linear pattern due to concentration of dark pigment along the centers of the scales; upper part of opercular membrane dark brown, the pigment diminishing a short distance below principal oper- cular spine, but some scattered dark dots persisting to level of pectoral base; upper unsealed portion of pectoral axil dark brown to blackish, with some pigment continuing above pec- toral base along edge of gill opening; fins pale or with dusky soft rays (particularly on Pacific specimens), with or without a concentration of dark pigment anteriorly in the soft dorsal and anal fins and the caudal lobes. Color when fresh as illustrated in Figure 7. The outer part of the spinous dorsal fin is bright red. REMARKS. The proportional measurements given above are based on 24 specimens, 92 to 177 mm SL, twelve of which are from the Red Sea. As indicated by Klausewitz and Nielsen (1965:19, pi. 15), the holotype of M. murdjan in the Zoological Museum, Copenhagen, consists only of a part of the skin of the body and portions of the median fins; the head, nape, abdomen, and paired fins are missing. Table 5. Comparison of fin-ray counts of Myripristis murdjan from the Red Sea and other localities. Dorsal Soft Anal Soft Pectoral Rays Rays Rays 13 14 15 11 12 13 14 14 15 16 Red Sea 29 3 1 20 1 1 5 26 1 Other localities 1 20 13 10 23 1 30 4 Contributions in Science, Number 334 Randall and Gueze: Red Sea Myripristis 13 Table 6. Comparison of proportional measurements' of Myripristis murdjan from the Red Sea and other localities. Red Sea Other Localities Range Mean Range Mean Depth of body 2.34-2.46 2.41 2.30-2.49 2.40 Head length 2.68-3.03 2.89 2.68-3.06 2.83 Snout length 4.89-5.30 5.10 4.94-5.26 5.1 1 Orbit diameter 2.1 1-2.49 2.26 2.13-2.39 2.27 Interorbital width 3.88-4.43 4.16 3.78-4.43 4.02 Maxillary length 1.68-1.82 1.76 1.76-1.89 1.83 Depth of caudal peduncle 3.12-3.72 3.40 3.09-3.54 3.34 Longest dorsal spine 1.96-2.48 2.21 2.01-2.48 2.20 Longest dorsal ray 1.39-1.68 1.57 1.38-1.62 1.52 Longest anal spine 2.07-2.49 2.27 2.19-2.36 2.28 Longest anal ray 1.38-1.66 1.53 1.38-1.60 1.50 Caudal fin 1.12-1.29 1.20 1.15-1.33 1.21 Pectoral fins 1.37-1.49 1.44 1.31-1.42 1.39 Pelvic fins 1.50-1.61 1.57 1.57-1.73 1.65 'The first 2 measurements (depth of body and head length) are given as a ratio of the standard length, and the last 1 2 as a ratio of the head length. Based on I 2 Red Sea specimens front 1 03.5 to 1 54.3 mm SL and 1 2 specimens extralimital to the Red Sea, 92 to 1 77 mm SL. Figure 14. Holotype of Myripristis hexagonus (Lacepede), 112 mm SL, MNIIN A. 5423. 14 Contributions in Science, Number 334 Randall and Gueze: Red Sea Myripristis Greenfield (1974) recognized three species of Myripristis in the Red Sea: M. hexagonus (which we have referred to xanthacrus ), M. murdjan, and M. parvidens. He distinguished the latter two on the basis of the relative width of the interorbi- tal space, that of murdjan supposedly contained 4 times or more in the head and that of parvidens less than 4. Since the interorbital width of the holotype of murdjan cannot be deter- mind from Forsskdl’s fragmentary description or from the frag- mentary holotype, Greenfield arbitrarily selected the species with the broad interorbital to bear the name murdjan in his revision. Apart from M xanthacrus, the senior author has collected numerous specimens of Myripristis in the Red Sea that seem to be one species. The interorbital width of these fish varies from about 3.8 to 4.4 in the head. If separated into two groups strictly on whether the interorbital is more than or less than 4 in the head, no other basis for separation can be found. Noting that Greenfield listed only one lot of four specimens of M. murdjan from the Red Sea in his “Material Examined,” we asked for a loan of these four specimens and some Red Sea specimens he identified as parvidens from the British Museum (Natural History). We conclude that all represent one species. The name murdjan must therefore be applied to this species. The specimens which Greenfield identified as parvidens are here referred to as murdjan (along with the four Red Sea speci- mens he called murdjan ). Greenfield should nonetheless be credited with noticing the difference in interorbital width among specimens of closely re- lated species of Myripristis. As shown in the Key in this paper, the interorbital width is the principal character that dis- tinguishes the true murdjan from berndti, even though there is some overlap (interorbital width of berndti 4.3 to 5.2 in head). The point of division in interorbital width of the two species is not 4.0, however, but about 4.35. The authorship of Myripristis parvidens is usually attributed to Cuvier in Cuvier and Valenciennes; however Bauchot (1970) has shown that the description of this species in Cuvier’s Regne Animal appeared in March 1829, hence prior to the ap- pearance of Cuvier and Valenciennes’ volume 3 of Histoire Naturelle des Poissons in April 1829. The holotype of M. axillaris (MNHN A. 70, 153 mm SL) was examined by the junior author at the Museum National d’Histoire Naturelle in Paris. We can confirm the following localities for M. murdjan: Red Sea, Gulf of Aden, Comoro Islands, Seychelles, Maidive Is- lands, Reunion, Mauritius, Indonesia, New Guinea, New Ire- land, Solomon Islands, New Hebrides, Philippines, Japan, Caroline Islands, Samoa Islands, Mariana Islands, Marshall Is- lands, and Pratas Reef in the South China Sea [specimens of M. murdjan at the California Academy of Sciences identified as M. parvidens and M. murdjan by Greenfield (1974) and listed by him as from Taiwan are actually from Pratas Reef]. The species is not known from the Hawaiian Islands and evi- dently is absent also from the Line Islands and French Poly- nesia. The record of M. murdjan from the Society Islands by Randall (1973:181) refers to M berndti. M. murdjan is generally taken in less than 10 m, at times in as little as 1 m. However, Red Sea collections made by the se- nior author include specimens from one station in 37 to 49 m. As indicated in Table 5, there is a slight difference in the count of the soft rays of the anal fin of M. murdjan in the Red Sea and outside the sea. No differences were apparent, how- ever, in the lateral-line scale and gill-raker counts. In a search foe other possible differences, a comparison was made of 14 porportional measurements of specimens of this species from within and outside the Red Sea (Table 6). With the possible exception of the interorbital width, the length of the maxilla, the depth of the caudal peduncle, and the length of the pelvic fins, there are no significant differences.8 The largest specimen examined is BPBM 19643, collected by the junior author in Mauritius; it measures 177 mm SL. ACKNOWLEDGMENTS The opportunity to collect fishes in the Red Sea was provided by Grant 439 of the U.S. -Israel Binational Science Foundation. Assistance in collecting fishes in Sudan by Peter J. and Paula Vine, in the Gulf of Aqaba by Adam Ben-Tuvia, Ofer Gon, and others of the Hebrew University, and in Djibouti by L.A. Mauge is gratefully acknowledged. Special thanks are due Vic- tor G. Springer for sending on loan four lots of our new species, which he collected in Ethiopia, as well as other material from the U.S. National Museum of Natural History. We also thank the following for providing pertinent information and/or loan- ing specimens: Gerald R. Allen, Marie-Louise Bauchot, Mari- nus Boeseman, Joan M. Davis, David W. Greenfield, Karsten E. Hartel, Phillip C. Heemstra, Douglass F. Hoese, Susan J. Karnella, Beverley Marsh, R.J McKay, Norbert Rau, Takeshi Shimizu, Margaret M. Smith, William F. Smith-Vaniz, Pearl M. Sonoda, Camm C. Swift, Richard C. Wass, and A.C. Wheeler. The photographs of Myripristis were taken by the senior au- thor, except for the photo of the holotype of M. hexagonus, which was supplied by the Museum National d’Histoire Naturelle in Paris. Funds for the color separations were pro- vided by the Charles Engelhard Foundation. LITERATURE CITED Bagnis, R., P. Mazellier, J. Bennett, and E. Christian. 1972. Poi- ssons de Polynesie. Papeete: Les Editions du Pacifique, 368 pp. Bauchot, M L. 1970. Catalogue critique des types de poissons du Museum National d’Histoire Naturelle. Publ. Div. Mus. Nat. Hist. Natur. 24:1-55. Bleeker, P. 1851. Bijdrage tot de kennis der ichthyologische fauna van de Banda-eilanden. Nat. Tijdschr. Ned. Ind. 2:225-261. 8. The junior author believes that additional material of M. murdjan for this comparison may reveal sufficient differences to warrant nomen- clatorial recognition of the Red Sea population. Further study is planned by him. Contributions in Science, Number 334 Randall and Gueze: Red Sea Myripristis 15 Bleeker, P. I 863. Onzieme notice sur la faune ichthyologique de 1’ile de Ternate. Ned. Tijdschr. Dierk. 1:228-238. 1873. Revision des especes indo-archipelagiques du genre Myripristis. Ned. Tijdschr. Dierk. 4:178-197. Castelnau, F. De. 1875. Researches on the fishes of Australia. Intercol. Exhib. Essays, no. 2, 52 pp. Cuvier, G. 1829. Regne Animal, ed. 2, vol. 2. Paris: Deterville, xv + 406 pp. Cuvier, G., and A. Valenciennes, 1829. Histoire naturelle des poissons, vol. 3. Paris: F.G. Levrault, xxviii + 500 pp. 1831. Histoire naturelle des poissons, vol. 7. Paris: F.G. Levrault, xxix + 531 pp. Forsskal, P. 1775. Descriptiones animalium avium, amphibi- orum, piscium, insectorum, vermium; quae in itinere orien- tali. Molleri, Haunie, 164 pp. Fourmanoir, P, and P. Laboute. 1976. Poissons de Nouvelle Cal- edonie et des Nouvelles Hebrides. Papeete: Les Editions du Pacifique, 376 pp. Greenfield, D.W. 1974. A revision of the squirrelfish genus Myripristis Cuvier (Pisces: Holocentridae). Bull. Natur. Hist. Mus. Los Angeles County 19:1-54. Gunther, A. 1859. Catalogue of the acanthoptergian fishes in the collection of the British Museum. London: Taylor and Francis, xxxi + 524 pp. Herre, A.W. 1953. Check list of Philippine fishes. U.S. Fish Wild!. Serv. Res. Rep. 20:1-977. Jordan, D.S., and B.W. Evermann. 1903. Descriptions of new genera and species of fishes from the Hawaiian Islands. Bull. U.S. Fish Comm. 22:163-207. Klausewitz, W. and J. G. Nielsen. 1965. On Forsskdl’s collec- tion of fishes in the Zoological Museum, Copenhagen. Spolia Zool. Mus. Laun. 22:1-29 Lacepede, B.G. 1802. Histoire naturelle des poissons, vol. 4. Paris: Plassan. xliv + 728 pp. Marshall, T.C. 1964. Fishes of the Great Barrier Reef. Sydney: Angus and Robertson, 566 pp. Masuda, H., C. Araga, and T. Yoshino. 1975. Coastal fishes of southern Japan. Tokyo: Tokai University Press, 379 pp. Ogilby, J.D. 1908. New or little known fishes in the Queensland Museum. Ann. Queensland Mus. 9:1-41. Randall, J.E. 1955. Fishes of the Gilbert Islands. Atoll Res. Bull. 47, xi + 243 pp. _ _. 1973. Tahitian fish names and a preliminary check-list of the fishes of the Society Islands. Occ. Pap. B.P Bishop Mus. 1 1:167-214. Riippell, E. 1828. Atlas zu der Reise nordlichen Afrika, Fische des rothen Meeres. Frankfurt: Heirn. Ludw. Bronner, 141 pp. Sauvage, M.H. 1891. Histoire physique, naturelle et politique de Madagascar, vol. XVI: Histoire naturelle des poissons. Paris: lTmprimerie Nationale, 543 pp. Schultz, L.P., and collaborators. 1953. Fishes of the Marshall and Marianas Islands. Bull. U.S. Nat. Mus. 202, vol. 2, xxxii + 660 pp. Swainson, W. 1839. The natural history of fishes, amphibians, and reptiles or monocardian animals, vol. 2. London: Orme, Brown, Green, and Longmans, 452 pp. Taylor, W.R. 1964. Fishes of Arnhem Land. Rec. Amer.-Aus- tral. Sci. Exp. Arnhem Land, 4:43-307. Weber, M. and L.F. de Beaufort. 1929. The fishes of the Indo- Australian Archipelago, vol. 5. Leiden: E.J. Brill, xiv + 458 pp. Submitted 26 May 1976; accepted for publication 26 Novem- ber 1980. 16 Contributions in Science, Number 334 Randall and Gueze: Red Sea Myripristis 3lli n SERIAL PU BLI CATIONS OF THE NATURAL HISTORY MUSEUM OF LOS ANGELES COUNTY The scientific publications of the Natural History Museum of Los Angeles County have been issued at irregular intervals in three major series; the articles in each series are numbered individually, and numbers run consecutively, regardless of subject matter. • Contributions in Science, a miscellaneous series of technical papers describing original research in the life and earth sciences. ® Science Bulletin, a miscellaneous series of monographs describing original research in the life and earth sciences. This series was discontinued in 1978 with the issue of Numbers 29 and 30; monographs are now published by the Museum in the Contributions in Science senes. • Science Series, long articles on natural history topics, generally written for the layman. Copies of the publications in these series are sold through the Museum Bookshop. EDITORIAL BOARD fail •’ Lawrence G. Barnes Robert L. Bezy Kenneth E. Campbell John M. Harris Robert J. Lavenberg William B. Lee Roy R. Snelling Camm C. Swift EDITOR Robin A. Simpson Fred S. Truxal David P. Whistler Edward C,. Wilson John W. Wright • jOSf . *•* r . W’M in mmm THE CRANE FLIES (DIPTERA: TIPULIDAE) OF COCOS ISLAND, COSTA RICA, WITH DESCRIPTIONS OF FOUR NEW SPECIES1 George W. Byers2 ABSTRACT. Eleven species of Tipulidae, all in subfamily Limo- niinae, are recorded from Cocos Island, Costa Rica. Four new species, Limonia (Rhipidia) hoguei, Limonia (Caenoglochina) paniculata, Helius brunneus, and Orimarga (Diotrepha) fla- vescens, are described and illustrated. The ecology and zoo- geography of the tipulid fauna of Cocos Island are briefly considered. RESUMEN. Once especies de la familia Tipulidae, todas que pertenecen a la subfamilia Limoniinae, se reporta de la Isla del Coco, Costa Rica. Se describe y ilustra cuatro especies nuevas: Limonia ( Rhipidia ) hoguei, Limonia ( Caenoglochina ) pan- iculata, Helius brunneus y Orimarga ( Diotrepha ) flavescens. Se considere, ademas, en breve, la ecologia y zoogeografia de la fauna de tipulidos de la isla. INTRODUCTION According to legend, there is a fortune in pirate gold buried somewhere on Cocos Island. Since the days of the con- quistadores, however, one treasure-seeker after another has come away empty-handed from this tiny tropical island. For those content with less than gold, the remote island, some 500 km (300 mi) southwest of Costa Rica, still has much to offer. For more than 300 years, it served as a source of coconuts and fresh water for passing vessels. Cocos Island, at 5°32'57"N, 86°59'17"W, is a rugged vol- canic island of about 46.6 square km in area and 23.3 km in circumference, with elevations up to about 850 m (2790 ft) (Hertlein 1963). It receives heavy rainfall throughout the year and has temperatures from 20 to 33 °C (68 to 92 °F). As a re- sult, it has numerous small streams and waterfalls and is cov- ered by a luxuriant growth of palms, ferns, cecropia trees, and other tropical rain-forest plants. Stewart (1912) judged the flora to be that of an oceanic island to which mainland plants had been transported by ocean currents, wind, or birds. It is the consensus of biogeographers who have studied the island that it is a true oceanic island of relatively recent (probably Pleis- tocene or Pliocene) origin, and that its fauna, other than species evidently introduced by human visitors, was received by trans- port over the open ocean (Hertlein 1963). That is, although Co- cos Island rises from an undersea ridge, there is no evidence to indicate it may ever have had a land connection to Central or South America. Long ago, biologists took an interest in the island because of the contrasts it offers to the near-desert Galapagos some 630 km (350 mi) on to the southwest. While the climate of the Ga- lapagos is affected largely by the cold Peru, or Humboldt, Cur- rent that turns westward from the coast of South America as the South Equatorial Current, the Cocos Island climate derives mainly from the warm, eastward-flowing Equatorial Counter- current. The presence of the Equatorial Countercurrent helps to explain the abundant rainfall on verdant Cocos Island but contributes little to an understanding of the origins of the is- land’s insect fauna. Periodically, however, there have been shifts in the current, which could transport insects living in driftwood, or rafting on floating vegetation, from the coasts of Central America to the island. While July average surface winds over Cocos Island are southerly or from the southeast, those of Janu- ary blow generally from Central America (i.e., from the north- east) and could introduce mainland insects, especially those that are not strong fliers. Two recent trips to Cocos Island from the Natural History Museum of Los Angeles County, California, particularly by Charles L. Hogue (Curator of Entomology) and Scott A. Miller (Research Associate), have yielded some 14,000 specimens of insects and arachnids. These add greatly to the small, pre- viously existing collections of arthropods from the island, mostly made incidentally during studies of the vertebrate or plant biota. Hogue and Miller have assembled all extant collec- tions and are attempting to identify the entire entomofauna as part of a biogeographical study. All collections that include Tipulidae were made at Wafer Bay (named for early explorer Lionel Wafer), where there is safe anchorage and a beach (in contrast to the steep, rocky shoreline of most of the rest of the island). This is at the mouth of the Rio (or Arroyo) Genio. Virtually all the crane flies were 1 Review committee for this contribution: Paul H. Arnaud, Jr., Charles L. Hogue, Henry Knizeski, and William J Turner. 2. Department of Entomology, University of Kansas, Lawrence 66045. University of Kansas Contribution No. 1774. Contributions in Science, Number 335, pp. 1-8 Byers: Cocos Island Crane Flies 1 Natural History Museum of Los Angeles County, 1981 ISSN 0459-0113 obtained by use of Malaise traps or were attracted to 15-watt ultraviolet or fluorescent light traps. It should be pointed out that, for some reason, lights attract particularly female crane flies. A small number of Tipuiidae collected on Cocos Island by botanist Ira L. Wiggins in 1967 were recently studied by Pro- fessor Charles P. Alexander. From among these, Alexander (1978) described two new species, Limonia (Geranoniyia) co- coensis and Limonia (G.) wigginsi. This collection also included a few species widely distributed in tropical America. All the species found by Wiggins were collected again by Hogue and Miller, who also secured several additional species, of which four — all apparently endemic to Cocos Island — are described as new in the following account. SYSTEMATICS Tipuiidae of Cocos Island The crane fly fauna of Cocos Island is restricted to representa- tives of the subfamily Limoniinae, as is usual for tropical Pa- cific islands. In the large, virtually cosmopolitan genus Limonia (tribe Limoniini), the subgenera Rhipidia, Geranomyia, and Caenoglochina are represented by numerous individuals among the specimens at hand, but curiously there are no flies of the subgenus Dicranomyia. Rhipidia, Geranomyia, and Dicram omyia have abundant species in tropical America; the smaller group Caenoglochina is wholly neotropical. Species of Helius and Orimarga (also Limoniini) are also present. Species of the tribes Pediciini and Hexatomini are apparently not found on the island, but the first of these tribes is only poorly represented anywhere in the Neotropical region. Of the Eriopterini, two widely distributed species of Gonomyia have reached Cocos Island. Holotypes, allotypes, and most of the paratypes of the follow- ing new species are in the collection of the Natural History Mu- seum of Los Angeles County, Los Angeles, California. A small number of paratypes have been deposited in the Snow Ento- mological Museum, University of Kansas, Lawrence. Limonia ( Rhipidia ) domestica (Osten Sacken) Three subspecies of this unusually widespread species have been recognized, one in Brazil, one in Ecuador and Peru, and the typical form recorded from eastern North America as far north as Iowa and Maryland, from many parts of the Antillean region, and from northern South America (Alexander and Alexander 1970). Specimens from Cocos Island do not conform exactly to the descriptions of any of these subspecies. Rather than describe them as an insular subspecies, I choose to recog- nize their specific affinity only. Cocos Island records: Wafer Bay, 17-22 April 1975, C.L. Hogue (1 2); in Malaise trap. Sta- tion 3 (Rio Genio), Wafer Bay, 24 March 1978 (1 6), same but 25 March (1 6,1 2), same but 26 March ( 1 6). Limonia (Rhipidia) hoguei new species Figures 1-5 DESCRIPTION. Description based on one 6, two 22, pinned. Head. Dorsum dark brown grading into brown at sides, with silvery pollinosity and black bristles. Eyes narrowly separated dorsally by about width of two rows of ommatidia, widely sepa- rated ventrally. Rostrum and palps dark brown. Antennal scape and subspherical pedicel blackish brown in both sexes; flagellum dark brown, with 12 flagellomeres. Flagellum of male (Fig. 1): first flagellomere wider than long, second with single, thick ventral lobe over half length of segment; flagellomeres 3- 8 bipectinate, or biflabellate, with inner (mesal) flabellations shorter and thicker than outer ones on segments 3, 4, and 8; flagellomere 9 with single ventral pectination; flagellomeres 10- 11 not modified, 12 as long as 10 and 11 together; setae on flagellomeres 2-9 conspicuously longer than respective seg- ments. Flagellum of female nodulose, except first flagellomere conspicuously wider than long, as in male; setae short. Thorax. Pronotum brown. Mesonotal prescutum dull orange with three dark reddish brown longitudinal stripes; broad me- dian stripe and narrower, more lateral stripes connected above pseudosutures in holotype and allotype (somewhat so in para- type), producing four orange spots, two short ones before level of pseudosutures, two longer ones behind. Scutum and scutellum unevenly brown, grading into light brown at sides. Halteres light brown throughout. Pleural surfaces mostly brown, yellowish brown on lower sternopleurite (katepister- num) and meron. Upper ends of coxae brown, lower ends yel- lowish brown. Femora sordid yellowish brown, slightly expanded and darkened to blackish brown at tips. Tibiae nar- rowly yellowish brown at base, otherwise dark brown; tarsi dark brown; claws slender, with basal tooth. Wings patterned approximately as in L. willistoniana Alex- ander (Alexander 1914), with diffuse clouding of grayish brown in all cells, six more distinct darkened areas near anterior border of wing: ( 1 ) centered on Sc2 and R2, (2) over fork of Rs, (3) at fork of Sc, (4) at origin of Rs, (5) midway between ar- culus and origin of Rs, and (6) over the humeral cross-vein, arculus, and origin of M (this last spot more diffuse than other five). Clear spots near tip of 2A in cell 1A, in cell Rl+2 and extending into cell R3, in middle of cell 1st M2, and between darkened spots listed above. Abdomen of Male. General color sordid yellowish brown; terga 2-5 with darker brown posterior margins; sterna slightly paler than terga; pleural membranes blackish brown. Tergum 8 (Fig. 2) broadly emarginate, tergum 9 much narrower than 8, rounded at sides, with shallow, rounded posterior indentation. Proctiger nearly transparent, two-thirds as wide as ninth 2 Contributions in Science, Number 335 Byers: Cocos Island Crane Flies tergum at base, abruptly narrowed near midlength, extending backward beyond rostral tips of ventral dististyles. Basistyles (Figs. 2, 4) each with conspicuous, setiferous ventral lobe. Dor- sal dististyle evenly curved, generally blackened but slightly paler near base. Ventral dististyle (Fig. 2, vd) with moderately elongate rostrum bearing two thick spines, their bases well sep- arated, outermost on slight elevation. Gonapophyses (Figs. 3, 5, gon) blackened in upturned apical half, sharply pointed. Aedeagus (Fig. 5) bilobed at apex, with short apicolateral spines and two setae at base on each side. Abdomen of Female. Terga and sterna sordid brown, indistinctly darker posteriorly; cerci and hypovalves amber-colored. Cerci short, evenly upcurved to sharp tips, closely appressed to each other in dorsal aspect. Hypovalves extending approximately to midlength of cerci. Length of body, excluding antennae, male (holotype), 4.6 mm; female, 4. 3-4. 7 mm (allotype 4.7 mm). Wing length, male, 5.0 mm; female, 4. 6-5.0 mm (allotype 5.0 mm). TYPES. Holotype, male. Wafer Bay, 17-22 April 1975, col- lected by C.L. Hogue; specimen has had abdomen removed, softened, stored in microvia! on same pin as rest of specimen. Allotype, female, same data as holotype. Paratype, female, at 15-watt ultraviolet lamp. Station 6 (west bank of Rio Genio, 200 m south of river’s mouth), Wafer Bay, 27 March 1978, C.L. Hogue and S. Miller. DIAGNOSIS. Limonia (Rhipidia) hoguei resembles most closely L. (Rh.) willistoniana Alexander (originally described as costalis by Williston; for best description and figures, see Alex- ander 1970: 19, 25) from the Windward Islands (St. Vincent, Dominica) and Costa Rica. It is also similar to L. (Rh.) lu- quilloensis Alexander of Puerto Rico (Alexander, 1950: 207). The male of L. hoguei differs from both of these in having only six bipectinate flagellomeres (instead of seven) and sharply pointed gonapophyses (instead of bluntly tipped). I have not been able to compare females of these three species to deter- mine in what ways they differ. ETYMOLOGY. This species is named for its collector. Dr. Charles L. Hogue, in recognition not only of his field studies of the insects of Cocos Island but also his outstanding research on nematocerous Diptera, particularly the Blephariceridae. REMARKS. Three parasitic Acarina were attached to the ven- tral surface of the abdomen of the holotype. During the soften- ing process, these became dislodged. They are at present preserved in the vial containing the crane fly’s abdomen. Limonia (Caenoglochina) paniculata new species Figures 6-1 1 DESCRIPTION. Description based on 5 <3(3, 36 29, pinned. Head. Occiput dark gray with fine, silvery pollinosity. Eyes large in both sexes, contiguous or nearly so both on vertex and beneath rostrum on gular surface. Rostrum and maxillary palps sordid yellowish brown. Scape brown at base; pedicel and apex of scape sordid yellowish brown; flagellum grayish brown with 12 flagellomeres, each pilose, with verticils arising from near midlength; apical fiagellomere about 1.4 times as long as pen- ultimate one. Thorax. Pronotum brown. Mesonotal prescutum polished, yel- lowish brown with broad brown median stripe abruptly widened in posterior half. Scutum and scutellum brown on elevated parts, yellowish brown where more depressed; mesonotal postnotum light brown. Knobs of halteres dusky brown, stems pale yellowish brown. Pleural surfaces light yellowish brown ex- cept for indistinct pale brown spot above anterior coxa. Femora sordid yellowish brown, palest basally, darkest apically; tibiae and tarsi grayish brown; claws apparently not toothed. Wings lightly tinged with grayish brown, iridescent; stigma not much darker than ground color, its limits indistinct; diffuse brown clouding along vein R3, in much of the cell R, (es- pecially along costa) and outer cell R3. Vein Sc, joins C and Sc2 joins R, opposite approximately midlength of Rs. Cell 1st M, nearly rectangular in most specimens, basal section of M3 being in nearly transverse alignment with m cross-vein; when m is more diagonal, M, still bends rectangularly. Abdomen of Male. Terga 2-8 grayish brown, corresponding sterna paler, from sordid yellowish brown to light grayish brown. Ninth tergum yellowish brown with brown posterior and lateral borders, broadly bilobed, with several straight, black setae on posterior one-third (Fig. 6). Basistyles each bearing six tufted lobes of varying complexity, three of these on common sclerotized base joined to ventral base of basistyle and more broadly attached by membrane across deep ventral and mesal notches (Figs. 7, 8). Most ventral of these lobes about 3 times as long as wide, truncate at apex and bearing row of stiff, golden-yellow setae with recurved tips and one long, straight seta at dorsal end of row. Middle lobe of ventral group about twice as long as wide, its rounded apex covered with tuft of yellowish hairs; dorsalmost lobe blunt, with only a few hairs. From ventromesal surface of each basistyle extends a small, flattened lobe (narrower at base than at apex in some males) bearing conspicuous, fanlike array of about 20 long, stiff, yellow setae; this fan divided into more posterior group of longer, sinuous setae and more anterior group of shorter, more strongly curved setae; arrangement fixed, setae springing back, wirelike, when dry and even after boiling in water with tri- sodium phosphate and preservation in alcohol. Above “fan” a spatulate, strongly sclerotized lobe with long terminal spine curving ventrad and row of dark hairs along apical edge above base of spine. Small, subconical, setiferous lobe on mesal sur- face of basistyle close beneath dististyle. Single, large dististyle complex, generally concave on lower or inner surface, convex on upper or outer surface, wider dorsoventrally than length from base to mesal edge. Dorsal apex of dististyle a strongly sclerotized, down-curved, acutely tipped blade grooved along its mesal surface (Figs. 6, 8), ventral apex a thick, rounded, hairy lobe curved dorsolaterad; ridge of dense, black, short but thick setae along most of mesal margin. Aedeagus (Fig. 9) elon- gate but not otherwise modified, with two recurved apical points. Gonapophyses thin, compressed, sclerotized blades, Contributions in Science, Number 335 Byers: Cocos Island Crane Flies 3 4 Contributions in Science, Number 335 Byers: Cocos Island Crane Flies slightly concave on inner surface, each with upturned, mucro- nate apex. Abdomen of Female. Terga 2-8 sordid grayish brown, corre- sponding sterna 2-7 slightly paler; terga 9-10 yellowish brown except posterior end of tergum 10 slightly darker brown. Cerci and hypovalves amber brown; rounded basal portion of sternum 8 yellowish brown; bases of hypovalves dark brown in ventral aspect. Elongate setae on ventral margins, sides, and dorsum of tergum 10 (Figs. 10, 11). Eighth tergum extended laterally around base of sternum, edges meeting at ventral midline. Cerci bowed slightly apart near midlength (Fig. 1 1 ). Body length (frons to tip of abdomen), male, about 5. 0-5. 4 mm (holotype 5.4 mm); female, 5. 6-6. 4 mm (allotype 5.9 mm). Wing length, male, 5. 2-5. 9 mm (holotype 5.6 mm); female, 5.7-6. 5 mm (allotype 6.0 mm). TYPES. Holotype, male. Wafer Bay, Cocos Island, Costa Rica, collected in Malaise trap, at Station 3 (Rio Genio, 200 m south of river’s mouth), 24 March 1978, by C.L. Hogue and S. Miller. Allotype, same label data as for holotype. Para- topotypes: 25 99, 17-22 April 1975, C.L. Hogue; 1 9, 24 March 1978; 1 3, 4 99, 25 March 1978; 1 9, 26 March 1978; 3 <33, 4 99, 27 March 1978; all 1978 specimens from Malaise traps, C.L. Hogue and S. Miller. DIAGNOSIS. Limonia ( Caenoglochina ) paniculata most closely resembles L ( C .) apicata subapicata Alexander of Flor- ida (U.S.A.) and the Amazonian species L. (C.) egae (Alex- ander) (from Ega, or Teffe, Brazil) and is very similar to L. (C.) napoensis (Alexander) (from Rio Napo, an Amazonian tribu- tary, Peru) (Alexander 1921:49-50, figs. 10, 11). L. (C.) fieldi Alexander (1967:281, figs. 1, 4), from Honduras, shares with these species the ridge of thick black setae along the margin of the dististyle and the complex basistylar lobes, but it does not have the blackened blade on the dorsal apex of the dististyle. L. paniculata differs from all other species in the subgenus in the number, shapes, and arrangement of the basistylar lobes and the complexity of their setae. ETYMOLOGY. The specific name is derived from the presence of the six setiferous lobes (Latin paniculata = tufted). Limonia (Geranomyia) species Four species of the subgenus Geranomyia were identified from among the specimens available. These include L. (G.) cocoensis Alexander, L. (G.) wigginsi Alexander, L. (G.) tycoon Alexander (= pallida Williston), and an apparently undescribed species. Ernest M. May (University of Kansas) is preparing a separate report on Geranomyia. Helius brunneus new species Figures 1 2-17 DESCRIPTION. Description based on 19 33, 4 99, and one specimen without abdomen, all pinned. Head. Occiput dark brown; eyes in both sexes large, nearly contiguous, separated on vertex and ventrally by only width of one or two ommatidia. Rostrum and maxillary palps brown; ros- trum subequal in length to rest of head in both males and females. Scape and pedicel brown, somewhat depressed; ped- icel with subterminal ring of appressed, dark setae; flagellum of 14 flagellomeres, brown, with numerous short, pale hairs, and with most verticils 2-3 times length of their respective segments. Thorax. Pronotum dull brown; mesonotal prescutum slightly polished brown to light brown, some individuals (including ho- lotype) with indistinct darker brown median line and two sub- median, setiferous lines. Depression between scutal lobes yellowish brown to brown. Scutellum and postscutellum light brown. Halteres dull dark brown. Pleural surfaces slightly pol- ished brown to light brown; coxae dull brown, hind coxae with numerous dark setae on posterior surface, other coxae with few, scattered setae. Femora, tibiae, and proximal three-fourths of basitarsi brown, grading into yellowish white on apical one- fourth of basitarsi. Tarsomeres 2-4 nearly white; 5 yellowish. Middle and hind basitarsi with mesal spurlike, elongate seta just beyond end of tibia. Wings (Fig. 16) strongly tinged with grayish brown, stigma distinct but not conspicuously darker than ground color. Veins R: + 3 and R4 + 5 generally parallel for half their length beyond level of r-m, diverging near apex of wing. Cross-vein r-m short but present in some individuals, absent by contact of R4 + 5 and M]+2 in others. Cell 1st M, large, about twice as long as great- est width. Cross-vein m-cu approximately in alignment with r-m, shorter than or subequal to basal section of M, + 4. Abdomen of Male. Dark brown, terga and sterna 2-8 with mostly sparse, short hairs but longer, more numerous hairs near ^Figures 1 through 15. Limonia hoguei, L. paniculata, and Helius brunneus. Upper scale. Figures 1-9, 12, 14; lower scale. Figures 10, II, 15. Figures 1 through 5, Limonia (Rhipidia) hoguei n. sp., male holotype. Figure 1, right antenna, lateral aspect. Figure 2, ninth abdominal tergum (9t), right basistyle, dorsal and ventral (vd) dististyles, dorsal aspect. Figure 3, right gonapophysis, mesal aspect. Figure 4, left basistyle, ventrolateral aspect. Figure 5, aedeagus and left gonapophysis (gon), ventral aspect. Figures 6 through 11, Limonia ( Caenoglochina ) paniculata n. sp.; 6-9, male paratype, 10-11, female allotype. Figure 6, ninth abdominal tergum (9t), left basistyle (bs) and dististyle, dorsal aspect. Figure 7, left basistyle and dististyle, posteroventral aspect. Figure 8, left basistyle and dististyle, mesal aspect. Figure 9, aedeagus (aed) and right gonapophysis (gon), left lateral aspect (left basistyle removed). Figure 10, terminal abdominal segments, left lateral aspect. Figure 11, cerci and part of tenth tergum, dorsal aspect. Figures 12 through 15, Helius brunneus n. sp., paratypes. Figure 12, terminal abdominal segments, male, dorsal aspect (most of right side omitted). Figure 13, basal portion of inner dististyle, enlarged to show minute spines. Figure 14, left basistyle and dististyles, aedeagus (aed) and phallosome (ps), right lateral aspect (right basistyle removed). Figure 15, terminal abdominal segments, female, left lateral aspect. Contributions in Science, Number 335 Byers: Cocos Island Crane Flies 5 lateral margins. Ninth tergum (Fig. 12, 9t) with sparse margi- nal setae at either side of broadly U-shaped median emargina- tion. Outer dististyle darkly sclerotized, slender, abruptly curved near apex to form hooklike tip; no lateral spine. Inner dististyle slender, unevenly curved inward and downward with single spine at apex and group of 4-5 microscopic, blackened spines on mesal surface near midlength (Fig. 13). Basistyle not prolonged conspicuously beyond level of attachment of dis- tistyles. Phallosome bilobed, bearing two elongate, acutely tipped processes directed first downward into genital chamber, then curving upward and mesad (Fig. 14, ps). Aedeagus (Fig. 14, aed) stout, strongly curved to form nearly complete coil. Abdomen of Female. (Fig. 15). Dark brown, with sparse, fine hairs shorter on terga than on sterna. Cerci long, slender, straight in basal half, curved evenly upward in apical half. Hy- povalves slender, straight, tapering gradually to acute apex. Length of body exclusive of rostrum and antennae, male, 4.8-6. 1 mm (holotype 6.0 mm); female, 5. 2-6. 5 mm (allotype 5.2 mm). Wing length, male, 5. 8-7. 2 mm (holotype 6.7 mm); female, 5. 8-6. 3 mm (allotype 5.8 mm). TYPES. Flolotype, male. Wafer Bay, Cocos Island, Costa Rica, collected at 15-watt ultraviolet light trap, at Station 6 Figures 16 through 20. Helius brunneus and Orimarga flavescens. Scale, Figures 18, 20. Figures 16 and 17, Helius brunneus n. sp., wings, male paratypes. Figure 16, normal wing venation (wing length 7.3 mm). Figure 17, abnormal wing venation of discal area (R-radius, M-media). Figures 18 through 20, Orimarga ( Diotrepha ) flavescens n. sp., female holotype. Figure 18, terminal abdominal segments, left lateral aspect; 8s- eighth sternum, cr-cercus, hv hypovalve Figure 19, wing venation (wing length 5.1 mm). Figure 20, cerci, dorsal aspect. 6 Contributions in Science, Number 335 Byers: Cocos Island Crane Flies (Rio Genio, 200 m south of river’s mouth), 27 March 1978, by C.L. Hogue and S. Miller. Allotype, female, and 3 8 <3, 1 9 paratypes, Rio Genio, Cocos Island, 27 March 1978. Addi- tional paratypes: Wafer Bay, 17-22 April 1975 (5 88, 2 99, 1 without abdomen); Wafer Bay, at UV light, 27 March 1978 (2 88); Rio Genio, 25 March 1978 (8 88). DIAGNOSIS. Helius brunneus belongs to the albitarsis group and, on the basis of size, structure, and color, appears to resem- ble most closely H. micracanthus Alexander from the state of Sao Paulo, Brazil (Alexander 1945a) and, somewhat less closely, H. rectus Alexander from the Federal District of Brazil (Alexander 1945b). H. micracanthus is slightly smaller in body and wing length, to judge from the type specimen. In H. brun- neus, the basitarsi are less extensively white than in micra- canthus, and the halteres are dusky brown, not “dirty white.” The inner dististyle is markedly more curved than in micra- canthus. The two species have in common the strongly curved processes of the phallosome. In brunneus, the r-m cross-vein is often absent, but a larger sample of micracanthus might show that this is also true of that species. ETYMOLOGY. Although the body color of brunneus is not di- agnostic, the species takes its name from that characteristic (Latin brunneus = brownish). REMARKS. A striking venational anomaly was observed in one individual. This involved the area of the discal cell (cell 1st M2) and the radial sector (Fig. 17). Near its origin, the Rs is angular, with a short spur vein; the discal cell is not formed, and the veins in that area of the wing are thickened and atypically arranged (cf. Fig. 16). Orimarga (Diotrepha) flavescens new species Figures 18-20 DESCRIPTION. Description based on one female and one in- dividual with abdomen broken off, both pinned. Head. Occiput grayish brown with sparse, long, yellow setae; vertex and frons paler, sordid yellowish brown; rostrum and maxillary palps brown. Eyes black, large, only narrowly sepa- rated on vertex and ventrally behind rostrum. Antennal scape brown, pedicel light brown, flagellum dull yellowish, with 14 flagellomeres, most bearing verticils subequal to their length, except apical flagellomere with four nearly terminal setae twice its length. Thorax. Pronotum yellowish brown; prescutum, scutum, and scutelium dull yellowish; postnotum (postscutellum) yellowish brown. Pleural sclerites and coxae dull yellowish. Femora of middle and hind legs nearly white except for apical black band (about 8% of total femoral length). (Both front legs missing from holotype.) Tibiae white, with narrow, black apical band. Basitarsus pale grayish white, tarsomeres 2 and 3 light gray, 4 and 5 dark gray. Ratio of femur: tibia: basitarsus: tarsomeres 2: 3: 4: 5 = 85: 83: 54: 13: 5: 2: 2. Claws simple, strongly curved. Wings (Fig. 19) lightly tinged with yellowish gray, without stigmal darkening; veins yellow. Sc joins C shortly before level of midlength of Rs; Sc2 at apex of Sc, . R, curves evenly to join R2 + 3; no vein R1+2 (this vein possibly indicated by faint defle- ction in R, just before junction with R2 + 3). R4 + 5 nearly perpen- dicular to Rs and R2 + 3 at their junction. Cross-vein m-cu about three times its length before level of origin of Rs; cross-vein r-m approximately its length beyond fork of M. Halteres grayish yellow. Abdomen of Female. General coloration yellowish, with light brown, transversely wrinkled annulations at posterior ends of segments 2-4 and paler annulation at end of segment 5. Abdo- men long, slender; length of segments 2-6 about 3.5-4 times their greatest diameter. Segments 7-8 short, enlarged abruptly in diameter from 6. Terga 9 and 10 fused dorsally, partially sep- arated laterally (Fig. 18). Cerci short, not contiguous dorsally (Fig. 20), expanded laterally at base, somewhat concave ven- trolaterally, with setiferous ventral margin; apex glabrous, densely sclerotized, conspicuously upturned and sharp. Hypo- valves more than twice length of cerci, only slightly upcurved to sharp tip, strongly sclerotized in apical one-third and along dorsal and ventral margins. Body length (frons to tip of hypovalves) of female holotype about 7.6 mm. Wing length 5.1 mm. Hind femur 5.0 mm. An- tenna about 1 .2 mm. TYPES. Holotype, female. Wafer Bay, Cocos Island, Costa Rica, collected in Malaise trap, at Station 3 (Rio Genio, 200 m south of river’s mouth), 26 March 1978, by C.L. Hogue and S. Miller. Paratype (abdomen broken): Chatham Bay, 22 January 1967, I.L. Wiggins. DIAGNOSIS. Orimarga (Diotrepha) flavescens is apparently most closely related to Orimarga (D.) omissinervis Alexander, a species known only from Bolivia (Alexander 1913). Like O. fla- vescens, that species has pale legs with darkened femoral and tibial apices and lacks vein R,+2; however, it has a brown thorax, dark brown abdomen, and grayish wings and is some- what larger than flavescens. ETYMOLOGY. The specific name refers to the general colora- tion of this fly (Latin flavescens = yellowish). Gonomyia (Lipophleps) puer Alexander This species is widespread from southern United States to Mex- ico, Central America, and northern South America (Ecuador, Guyana, Peru), including the West Indies. Cocos Island re- cords: in Malaise trap. Station 3 (Rio Genio), Wafer Bay, 24 March 1978 (3 99), same but 25 March (8 99), same but 26 March (2 88, 8 99), same but 27 March (2 99); light trap, 15- watt UV, Station 4 (400 m south of mouth of Rio Genio), Wafer Bay, 24 March 1978, (2 <38, 1 9), same but Station 6 (200 m south of mouth of Rio Genio), 27 March (1 8); all col- lected by C.L. Hogue and S. Miller. On 22 January 1967, I.L. Wiggins collected 13 99 at Chatham Bay (Bahia de Chatham). Gonomyia (Paralipophleps) pleuralis (Williston) Even more widespread than Gonomyia puer, this species ranges from southeastern United States through the West Indies and Contributions in Science, Number 335 Byers: Cocos Island Crane Flies 7 Bermuda to Brazil, Bolivia, Guyana, and Peru. Cocos Island records: Wafer Bay, 17-22 April 1975, C.L. Hogue (1 2); in Malaise trap. Station 3 (Rio Genio) Wafer Bay, 24 March 1978 (1 2), same but 25 March (1 2), same but 26 March (1 <3,5 22); light trap, 1 5-watt UV, Station 2 (Rio Genio at first rapids above high tide level). Wafer Bay, 23 March 1978 (3 22), same but Station 6 (Rio Genio), 27 March (1 <5); all collected by C.L. Hogue and S. Miller. The l.L. Wiggins collection, made at Chatham Bay, 22 January 1967, contains 30 individuals (2 88, 28 22 or lacking abdomen). ACKNOWLEDGMENTS 1 am particularly indebted to Dr. Charles L. Hogue for making the recent collections available to me for study, and to him and Scott Miller for doing the field work. Their trips to Cocos Is- land were generously supported by Mr. and Mrs. Richard Steele (1978) and Mr. Timothy Doheny ( 1975). I thank also Dr. Paul H. Arnaud, Jr., of the California Academy of Sciences, San Francisco, for the loan of many of Wiggins’ specimens. For technical assistance, I thank Ernest M. May, Jon K. Gelhaus, and Chen-Wen Young. Finally, I express my gratitude to my colleague in the study of crane flies, Professor Charles P. Alex- ander, for 30 years of friendly cooperation and encouragement. My studies of Tipulidae have been generously funded by the National Science Foundation, most recently by Grant No. DEB-77-1 5868 (which, however, does not relate to tropical Tipulidae). LITERATURE CITED Alexander, C.P 1913. The neotropical Tipulidae in the Hun- garian National Museum (Diptera), II. Entomol. News 24:439-449. 1914. The crane-flies collected in Costa Rica by Dr. P.P. Calvert (Tipulidae, Diptera). Jour. New York Entomol. Soc. 22:116-124. 1921. New or little-known crane-flies from the Amazo- nian region. Proc. Acad. Natur. Sci. Philadelphia 73:39- 103. 1945a. New or little-known Tipulidae (Diptera) from Sao Paulo, Brasil, III. Papeis Avulsos Depto. Zool. 7:1-44. 1945b. Records and descriptions of Brazilian Tipulidae (Dipt.), XII. Revista Entomol. 16:210-226. 1950. Notes on the tropical American species of Tip- ulidae (Diptera), VI. The tribe Limoniini, genus Limonia: subgenera Limonia, Neolimnobia, Discobola and Rhipidia. Revista Entomol. 21:161-221. 1967. Notes on the tropical American species of Tip- ulidae (Diptera), VIII. The tribe Limoniini, genus Lim- onia, concluded; Helius, Orimarga and others; tribe Pediciini; subfamily Cylindrotominae. Studia Entomol. 10:277-352. 1970. Bredin-Archbold-Smithsonian Biological Survey of Dominica: the crane flies (Diptera: Tipulidae). Smith- sonian Contrib. Zool. 45:1-57. 1978. New or little-known neotropical Tipulidae (Dip- tera), I. Trans. Amer. Entomol. Soc. 104:1-36. Alexander, C.P, and M.M. Alexander. 1970. Family Tipulidae. A catalogue of the Diptera of the Americas south of the United States, 4. Mus. Zool., Univ. Sao Paulo, 259 pp. Hertlein, L.G. 1963. Contribution to the biogeography of Co- cos Island, including a bibliography. Proc. California Acad. Sci., Ser. 4, 32:219-289. Stewart, A. 1912. Expedition of the California Academy of Sciences to the Galapagos Islands, 1905-1906, V. Notes on the botany of Cocos Island. Proc. California Acad. Sci., Ser. 4, 1:375-404. Accepted for publication 15 December 1980. 8 Contributions in Science, Number 335 Byers: Cocos Island Crane Elies Safer las.' If-' ' Eire. , • B¥W4^:i ■wn? J liii'iHilMll 4»r . Number 336 5 February 1982 ARCOSCALPELLUM HOEK AND SOLI DOBA LANDS HOEK (CIRRIPEDIA, THORACICA) FROM THE PALEOGENE OF PACIFIC COUNTY, WASHINGTON, WITH A DESCRIPTION OF A NEW SPECIES OF ARCOSCALPELLUM Victor A. Zullo km Natural History Museum of Los Angeles County • 900 Exposition Boulevard « I-os Angeles, Californ 90007 iitrimjiMirrimmii |W| IM SERIAL PUBLICATIONS OF THE NATURAL HISTORY MUSEUM OF LOS ANGELES COUNTY The scientific publications of the Natural History Museum of Los Angeles County have been issued at irregular intervals in three major series; the articles in each series are numbered individually, and numbers run consecutively; regardless of subject matter. Contributions in Science, a miscellaneous scries of technical papers describing original research in the life and earth sciences. Science Bulletin, a miscellaneous series of monographs describing original research in the life and earth sciences. This series was discontinued in 1978 with the issue of Numbers 29 and 30; monographs are now published by the Museum in the Contributions in Science series. Science Series, long articles on natural history topics, generally written for the layman. Copies of the publications in these series are sold through the Museum Bookshop. EDITORIAL BOARD Robert Gustafson John M. Harris Robert; J. Lavenberg Camm G. Swift EDITOR Robin A. Simpson Leon G. Arnold Lawrence G. Barnes Robert. L. Bezy Kenneth E. Campbell Fred S. Truxal David P. Whistler Edward C. Wilson John W. Wright Printed at The Castle Press, Pasadena, California. ARCOSCALPELLUM HOEK AND SOLIDOBA LANES HOEK (CIRRIPEDIA, THORACICA) FROM THE PALEOGENE OF PACIFIC COUNTY, WASHINGTON, WITH A DESCRIPTION OF A NEW SPECIES OF ARCOSCALPELLUM1 Victor A. ZuIIo2 ABSTRACT. Arcosealpellum raricostatum Withers, 1953, previously known from the Italian and Cuban Eocene, and A. knapptonensis n. sp. occur in concretions in the basal beds of a chronostratigraphic equivalent of the Lincoln Creek Formation exposed along the Columbia River near Knappton, Pacific County, Washington. The middle units have yielded only fragmentary arcoscalpellid remains. The upper beds contain a spe- cies of Solidobalanus Hoek, 1913, similar to 5. ( Hesperibalanus ) sookensis (Cornwall, 1927) from the Sooke Formation (Juanian Mol- luscan Stage) of Vancouver Island, British Columbia. The molluscan faunas of the middle and upper beds are characteristic of the Matlockian and Juanian Molluscan Stages, respectively (Zemorrian benthic For- aminiferal Stage, or Oligocene). Molluscs are not known from the basal unit, but the arcoscalpellids suggest a late Eocene age (Galvinian Mol- luscan or Refugian benthic Foraminiferal Stage) based on their affinities with late Eocene European species. The two species of Arcosealpellum Hoek, 1907, are the first scaipelloid barnacles to be reported from the Pacific Coast Cenozoic. INTRODUCTION Calcareous concretions from marine sediments exposed along the north shore of the Columbia River near Knappton, Pacific County, Washington (Figure 1) contain numerous, well-pre- served, but disarticulated capitular plates of two species of Ar- coscalpellum Hoek, 1907, and the shells and a few opercular plates of an archaeobalanid resembling Solidobalanus ( Hesperi- balanus) sookensis (Cornwall, 1927). The arcoscalpellids are the first scaipelloid barnacles to be recorded from the Pacific Coast of North America. Arcosealpellum in the traditional sense is an extant, cosmopolitan genus with a fossil record extending back to the Late Cretaceous. It is represented by more than 100 extant species found primarily at bathyal and abyssal depths, and by over 50 fossil species from Upper Cretaceous and Tertiary in- shore shelf deposits. Nine fossil species have been reported from North America. As many have perceived, the traditional concep- tion of the genus Arcosealpellum includes a diversity of species whose relationships are not altogether clear. Several attempts have been made to isolate species-groups within the genus, but only Zevina ( 1 978) has proffered a comprehensive revision of the extant species. In Zevina’s classification, the species of Ar- coscalpellum, together with those of Holoscalpellum Pilsbry, 1907, are distributed among 12 genera in the subfamily Ar- Contributions in Science, Number 336, pp. 1-9 Natural History Museum of Los Angeles County, 1982 coscalpellinae Zevina. No attempt has yet been made to reevalu- ate the classification of fossil species in light of Zevina’s revision. This may not be possible for many fossil species, because generic assignment in Zevina’s scheme is dependent upon knowledge of the total armature of the capitulum, and many fossil species are known only from a few disarticulated capitular plates. Buck- eridge ( 1980) has proposed a new genus for a group of Tertiary species from Australasia characterized by heavily calcified ca- pitular plates and an absence of pits on the interior of the scutum for the placement of males. The classificatory significance of these features and their distribution in fossil and extant species outside of the Australasian region are unknown. Archaeobalanids are the oldest known balanoid barnacles, first appearing in middle Eocene rocks, and are the most com- monly encountered balanoids in Paleogene deposits. Three fossil species have been described from the Pacific Coast Tertiary: Solidobalanus ( Hesperibalanus ) cornwalli (Zullo, 1966) from the middle or upper Eocene Cowlitz Formation of Lewis County, Washington; S. (H.) sookensis from the Oligocene Sooke Forma- tion of Vancouver Island, British Columbia; and S. (H.) proinus (Woodring, 1 950) from the Pliocene of central and southern Cal- ifornia (Zullo, 1979a). The extant North Pacific species. S'. ( H .) hesperius (Pilsbry, 1916), is common in Pleistocene deposits of the Pacific Northwest. Hesperibalanus Pilsbry, 1916, was syn- onymized with Solidobalanus Hoek, 1913, by Henry and Mc- Laughlin (1967) but was reinstated as a subgenus of Sol- idobalanus by Newman and Ross ( 1 976). Of the four extant and nine extinct species presently included in Hesperibalanus, only two, S. (H ) hesperius and S. (//.) proinus, can be ascribed to this subgenus with certainty. The remainder are included presently as a matter of convenience, awaiting a much needed revision of the free-living Archaeobalaninae. The Knappton Cirripedia are significant in several respects. The presence of arcoscalpellids extends the known Paleogene 1 . Review committee for this contribution: Ellen J. Moore, William A. Newman, and Edward C. Wilson. 2. Department of Earth Sciences, The University of North Carolina at Wilmington, North Carolina 28403, and Research Associate, Inverte- brate Paleontology, Natural History Museum of Los Angeles County. ISSN 0459-0113 distribution of this group into the northeastern Pacific. The oc- currence of archaeobalanids supports previous indications that free-living members of this group are widespread in Paleogene deposits (Zullo and Baum, 1979). Perhaps most importantly, the knappton species suggest that cirripeds are useful biostratigra- phic indicators, discussion of which follows the systematic account. STRATIGRAPHY OF THE KNAPPTON EXPOSURES The knappton beds are considered chronostratigraphic equiv- alents of the Lincoln Creek Formation of Beikman and others ( 1 967) (replacement name for the Lincoln Formation of Weaver, 1912, and of subsequent authors). In the Grays Harbor basin located northeast of the knappton .locality, the Lincoln Creek Formation consists of up to 2,740 m of tuffaceous siltstone and sandstone containing scattered concretions and concretionary beds. The Lincoln Creek Formation is considered to range from the late Eocene to earliest Miocene (Rau, 1958, 1964; Armen- trout, 1975, 1 977). The lower part of the formation is correlated with the keasey Formation of Oregon and the Toutle Formation (Gries Ranch beds) of southern Lewis County, Washington. The upper part of the formation is equivalent to the Sooke Formation of Vancouver Island, British Columbia, and the upper part of the Twin River Formation of Washington (Armentrout, 1977). According to Ray Wells (personal communication, 1 980), who has mapped the region and measured and described the section, much of the knappton section is covered, including the contacts with under- and overlying units. That part of the section consid- ered correlative with the Lincoln Creek Formation consists of about 305 m of thin-bedded and laminated tuffaceous siltstone and sandy siltstone bearing occasional concretions, some of which are several meters in length. This unit is overlain by at least 1 50 m of concretionary sandstone, sandy siltstone, and silt- Figure 1. Part of the Knappton, Washington 7.5-minute quadrangle show- ing location of “Lincoln Creek” faunal units. Inset map shows location of knappton quadrangle in Pacific County. stone that is equivalent to the Miocene Astoria Formation of western Oregon. The underlying unit is not exposed. Fossils from the Knappton exposures, including those described here, were collected by James and Gail Goedert, who divided the “Lincoln Creek” part of the section into four informal faunistic units (Fig- ure 2): (1) a lower unit characterized by barnacle-bearing con- cretions and an abundance of the trace fossil Tisoa De Serres; (2) an overlying unit containing tisoans, sponges, small aturiid nautiloids, and many decapod crustacean and marine vertebrate remains; (3) a glass sponge zone; and (4) an upper unit character- ized by an abundance of invertebrate and marine vertebrate re- mains, including large aturiid nautiloids, but with few sponges and no tisoans. Frey and Cowles (1969, 1972) reported on the single, double, U-shaped, and branching burrows of Tisoa from the Knappton locality (primarily unit 1 ) and noted the presence of the decapod crustacean Callianassa knapptonensis Rathbun, 1926, based on a record in Weaver (1942). Armentrout (personal communication, 1 979) refers the molluscan fauna of unit 2 to the Echinophoria rex zone (= Matlockian Molluscan Stage, = lower Zemorrian benthic Foraminiferal Stage, = lower Oligocene), and that of unit 4 to the Echinophoria apta zone ( = Juanian Molluscan Stage, = upper Zemorrian benthic For- aminiferal Stage, = upper Oligocene). CIRRIPED BIOSTRATIGRAPHY The two identifiable arcoscalpellids are from faunal unit 1 (Nat- ural History Museum of Los Angeles County Invertebrate Paleontology, LACMIP, locality 5844). A new species of Ar- coscalpellum (?strict sense), represented by numerous carinae, scuta, and terga, is most similar to A. gassinensis (de Alessandri, 1 906) from the Eocene Calcare di Gassino on the Bussolino side of Gassino, northeast of Turin, Italy. The second species, repre- sented by partial carinae, is ascribed to A. raricostatum Withers, 1 953, also described from the Calcare di Gassino near Bussolino. Withers (1953, p. 59) considered the Calcare di Gassino to be Auversian and middle Eocene, and correlative with the Upper Bracklesham beds at Whitecliff Bay, Isle of Wight. Withers’s age determination appears to be based on the occurrence of Num- mulites variolarius (Lamarck) in both units. Herb and Hekel (1973), among others, have shown that N. variolarius first ap- pears in the uppermost middle Eocene (Biarritzian) but con- tinues throughout the upper Eocene (Bartonian/Priabonian) in many parts of Europe, including northern Italy. In addition to the current practice of considering the Auversian as basal upper Eocene, Cita (1973), in a review of the Tertiary strata of the Italian Piedmont, regarded the Calcare di Gassino as Priabo- nian. Thus, it would appear that Arcoscalpellum gassinensis and A. raricostatum are upper rather than middle Eocene species. The two species of Arcoscalpellum from the Knappton section thus suggest that the basal part of the exposed “Lincoln Creek” is upper Eocene (= Galvinian Molluscan or Refugian benthic Foraminiferal Stage). This conclusion is supported by the cir- riped faunal “facies” of the upper Eocene (Jacksonian Gulf Coastal Plain Stage) Principe Formation in the vicinity of Havana, Cuba. In addition to A. raricostatum (see systematic account), this unit contains two other species of Arcoscalpellum, Contributions in Science, Number 336 Zullo; Paleogene Cirripedia of Pacific County, Washington A. habanense Withers, 1953, and A. sanchezi (Withers, 1926), that are closely related both to A. gassinensis and the new Knappton species. It is unfortunate that the only cirriped remains so far re- covered from unit 2 (LACMIP locality 5843) are poorly preserved terga. These plates may represent a species of Ar- coscalpellum, but terga are not sufficient in distinguishing be- tween the A. gassinensis complex and the quite distinctive species of the known Oligocene fauna. Knowledge of the identity of this species might prove helpful in determining the age of unit 2. Many European geologists would place the Eocene-Oligocene boundary at the base of the Rupelian (see Withers, 1953; Van Eysinga, 1975). This shift in the boundary relegates the tradi- tional European lower Oligocene Lattorfian and Tongrian Stages to the upper Eocene. The crux of arguments favoring this change is that Lattorfian and Tongrian faunas are more similar to Barto- nian/Priabonian faunas than they are to Rupelian faunas. Rec- ognition of upper Eocene equivalents and of the Eocene- Oligocene boundary in North America is further complicated by increasing endemism of invertebrate megafaunas and by appar- ent disagreement in correlation between upper Eocene cal- careous nannofossil and planktonic foraminiferal zones. In addition, there is increasing evidence from potassium-argon, rubidium-strontium, and fission track age determinations in Eu- rope, the Caribbean, and the Atlantic and Gulf Coastal Plains that the Eocene-Oligocene boundary is closer to 33 m.y. than to the 37-m.y. date proposed by Hardenbol and Berggren (1978) (Harris and Zullo, 1980). Those faunas and faunal zones whose ages have been related to radiometric age determinations must be reevaluated in light of this evidence. Although opercular plates are preserved in some of the ar- chaeobalanid shells from unit 4 (LACMIP locality 5842), the sediment is too well indurated to permit their intact extraction. The shells are quite similar to those of Solidobalanus sookensis, FORAMINIFERAL STAGES MOLLUSCAN STAGES KNAPPTON BEDS KNAPPTON CIRRIPEDIA IIOCENE SAUCESIAN NE WPORTI AN "ASTORIA" FM. PIILARIAN ? UJ Z JUANIAN UNIT 4 LACMIP 5842 SOLIDOBALANUS AFF. SOOKENSIS UJ O o ZEMORRIAN UNIT 3 (3 _i O UNIT 2 MATLOCKIAN LACMIP 5843 UPPER EOCENE ! REFUGI AN GALVINIAN UNIT 1 LACMIP 5844 ? ARCOSCALPELLUM i KNAPPTONENSIS __A^_R AB LGjO 3J 4T HM Figure 2. Stratigraphic distribution of Knappton cirripeds and suggested correlation of the Knappton beds (partly after Armentrout, personal commu- nication, 1979). Contributions in Science, Number 336 Zulio: Paleogene Cirripedia of Pacific County, Washington 3 but the internal morphology of the single scutum examined is sufficiently different to question assignment of the Knappton specimens to that species without additional opercular plate ma- terial. It is clear that the Knappton species is unrelated either to the upper Eocene S’, cornwalli or to true Hesperibalanus. The occurrence of a species of Solidobalanus similar to 5. sookensis in unit 4 is in agreement with the Oligocene age assignment derived from molluscan data. SYSTEMATIC ACCOUNT Subclass Cirripedia Burmeister, 1834 Order Thoracica Darwin, 1854 Suborder Lepadomorpha Pilsbry, 1916 Family Scalpellidae Pilsbry, 1916 Subfamily Arcoscalpellinae Zevina, 1978 Genus Arcoscalpellum Hoek, 1907 Arcoscalpellum knapptonensis n. sp. Figures 3-7, 9-17 DIAGNOSIS. Broad, slightly bowed carina with arched tectum, faintly ridged parietes, and very narrow, inwardly turned mtra- parietes; broad, trapezoidal, longitudinally striate scutum with low, flat, apico-basal ridge and decidedly obtuse basitergal angle; tergum subtriangular, longitudinally striate, twice as long as wide, with acute apico-basal ridge and without marked distinc- tion between upper and lower carinal margins. Distinguished from A. gassinensis and A. habanense by its narrower carina without conspicuous ridges on or bordering the tectum; from A. sanchezi and A. euglyphum by its carinal parietes that do not flare outwardly; from A. choctawensis and A. toulmini by its much broader tergum and thicker capitular plates. LOCALITY. LACMIP locality 5844, “Lincoln Creek Forma- tion,” faunal unit 1, on Columbia River approximately 122 m east of boundary between sections 8 and 9, T 9 N, R 9 W, USGS 7.5-minute topographic quadrangle of Knappton (1973 ed.), near Knappton, Pacific County, Washington. MATERIAL EXAMINED. Disarticulated carinae, scuta, and terga in four concretions. Holotype (LACMIP no. 6270) and paratypes (LACMIP nos. 6266-6269, 6330-6333) are in the in- vertebrate paleontology collection of the Natural History Mu- seum of Los Angeles County. DESCRIPTION. Carina slightly bowed, broad, length about four times width; tectum moderately to strongly arched, flatten- ing towards basal margin, sometimes with incipient medial ridge; basal margin broadly V-shaped; parietes half as wide as tectum, normal to tectum, and separated from tectum by one or two inconspicuous, narrow, longitudinal ridges; parietes ornamented by up to five faint longitudinal ridges; intraparietes very narrow, separated from parietes by narrow but conspicuous ridge, and turned inward at approximately a 45° angle. Scutum strongly arched, trapezoidal, length less than twice width; apico-basal ridge flat, low, slightly curved; occludent mar- gin slightly convex; basal and lateral margins straight, nearly equal in length, their junction forming an angle greater than 90°; tergal margin concave; tergolateral margin rounded, outer sur- face ornamented by faint, irregularly placed, longitudinal striae best developed on tergal side of plate; adductor muscle pit shal- low, ill-defined. Tergum subtriangular, elongate, length about twice width, lon- gitudinally striate; apico-basal ridge well developed, acute to flat-topped, straight to gently curved, and situated less than one- third the width of the plate from the carinal margin; a second, low, curved ridge may be present extending from the apex to the scutal margin; carinal margin weakly convex, not sharply di- vided into upper and lower halves; occludent margin slightly convex, short, about eight-tenths length of scutal margin; scutal margin gently sinuous. DISCUSSION. Of the Tertiary species whose carinae are known, only the following have arched tecta: Arcoscalpellum euglyphum Withers, 1924, p. 1 1, pi. 2, figs. 1- 2); Withers (1953, p. 233, pi. 34, fig. 1 ); Buckeridge ( 1 980, p. 122, fig. 34, as a member of a newly proposed genus); lower Oligocene (Whaingaroan-Duntroonian, = Latorffian- Rupelian), New Zealand. A. gassinensis (de Alessandri, 1 906, p. 252, pi. 13, figs. 10-14, as Scalpellum michelottianum var. gassinensis ); Withers (1953, p. 212, pi. 28, figs. 1-9); upper Eocene, Calcare di Gassino, northeast of Turin, Italy. A. habanense Withers (1953, p. 218, pi. 30, fig. 1); upper Eocene, Principe Formation, near Havana, Cuba. A. hartleyi (Withers, 1936, p. 590, pi. 11, figs. 1-5); Withers (1953, p. 232, pi. 32, figs. 1 -5); (?)iower Miocene, Surma Series, Bangladesh. A. sanchezi (Withers, 1926, p. 617, pi. 26, figs, 1-7); Withers (1953, p. 215, text-fig. 84, pi. 29, figs, 1-12); upper Eocene, Principe Formation, near Havana, Cuba. Arcoscalpellum knapptonensis is most similar to A. gassinen- sis from which it differs primarily in proportions of the plates. Arcoscalpellum gassinensis has a broader carina with narrower parietes set off from the tectum by conspicuous ridges, a broader scutum that in other respects is quite similar to that of A. knapp- tonensis, and a tergum that is distinguished by its shorter occlu- dent margin and, conversely, longer scutal margin. Arcoscalpel- lum habanense is readily distinguished by the great breadth of the carina and the presence of prominent, acute, longitudinal ridges on the tectum and parietes. The carina of A. hartleyi is quite similar to that of A. knapptonensis and differs only in hav- ing a less markedly arched tectum. Arcoscalpellum habanense and A. hartleyi are known only from carinae. The carina of A. sanchezi is broader and more strongly bowed than that if A. knapptonensis, and its parietes and intraparietes flare outwards. Its scutum is narrower and has decidedly concave tergal and lateral margins. The tergum of A. sanchezi is similar to that of A. knapptonensis, except that the upper carinal margin is concave rather than straight. The carina of A. euglyphum has a strongly arched tectum, and the carinal parietes differ markedly from those of A. knapptonensis in flaring outwards and being bordered both on their inner and outer margins by prominent ridges. Weisbord (1977) described two species of Arcoscalpellum from the Paleogene of Alabama for which the carinae are not Contributions in Science, Number 336 Zullo: Paleogene Cirripedia of Pacific County, Washingtoi Figures 3 through 8. Arcoscalpellum spp. Figures 3 through 7. Arcoscalpellum knapptonensis n. sp. (3) exterior of scutum, paratype LACMIP 6266; (4) interior of scutum, paratype LACMIP 6267; (5) exterior of tergum, paratype LACMIP 6268; (6) interior of tergum, paratype LACMIP 6269; (7) exterior of Carina, holotype LACMIP 6270. Figure 8. Arcoscalpellum raricostatum Withers, exterior of carina, hypotype LACMIP 6271. Scale bars represent 5 mm. Contributions in Science, Number 336 Zullo: Paleogene Cirripedia of Pacific County, Washington 5 6 Contributions in Science, Number 336 Zullo: Paleogene Cirripedia of Pacific County, Washington known. Arcoscalpellum^! ) choctawensis Weisbord from the up- per Eocene (Jacksonian Stage) North Twistwood Creek Member of the Yazoo Clay has thin scuta and terga; the interior of the scutum is marked by an apical furrow, and the tergum is narrow and elongate. Arcoscalpellum toulmini Weisbord from the Pal- eocene (Landenian Stage) Porters Creek Formation has a nar- rower tergum than A. knapptonensis with a distinctly shorter occludent margin, and a broader scutum with a slightly acute basitergal angle and unequal tergal and lateral margins. ETYMOLOGY. Geographic, after Knappton, Washington. Arcoscalpellum raricostatum Withers, 1953 Figure 8 Arcoscalpellum raricostatum Withers (1953, p. 224, pi. 36, figs. 1-2). Arcoscalpellum aff. raricostatum n. sp. Withers ( 1 953, p. 224, pi. 36, fig. 3). LOCALITY. LACMIP locality 5844, “Lincoln Creek Forma- tion,” faunal unit 1, on Columbia River approximately 122 m east of boundary between sections 8 and 9, T 9 N, R 9 W, USGS 7.5-minute topographic quadrangle of Knappton (1973 ed.), near Knappton, Pacific County, Washington. MATERIAL EXAMINED. Two carinae in one concretion associ- ated with A. knapptonensis: one lacking apical quarter (LAC- MIP hypotype no. 627 1 ); the other lacking both apical and basal parts (LACMIP hypotype no. 6335). DESCRIPTION. The carinae are slightly bowed and have flat tecta that become gently arched towards the basal margin. The tecta are bordered by prominent ribs that bear five longitudinal ridglets. The parietes are narrow and normal to the tecta. The basal margins of the carinae are V-shaped. DISCUSSION. Withers (1953) described A. raricostatum on the basis of two carinae from the Calcare di Gassino. Withers also described a basal fragment of a carina from the Principe Formation of Cuba that he considered was similar to but specif- ically distinct from A. raricostatum. The Cuban specimen differs from the types in having a slightly arched tectum with regularly spaced growth depressions and a less V-shaped basal margin. The Knappton carinae agree with the type specimens in the pro- nounced angulation of the basal margin and in lacking growth depressions but have the basally arched tectum characteristic of the Cuban carina. The Knappton carinae, in sharing features both with the types of A. raricostatum and the Cuban carina, suggest that the observed differences merely reflect individual variation within a single, widely distributed species. The mor- phologic variation seen in these five carinae is no greater than that to be found in the extant descendant of A. raricostatum , A. micheloltianum (Seguenza, 1876) [= A. velutinum (Hoek, I 883)], whose geographic distribution includes the Atlantic, In- dian, and western Pacific Oceans. For these reasons, I assign the Cuban and Knappton carinae to A. raricostatum The presence of these carinae amid the numerous capitular plates in the concretions ascribed to A. knapptonensis raises the question as to whether some of the scuta and terga may also be referable to A. raricostatum. Only the carina is known for this species, but as indicated by Withers (1953), A. raricostatum is similar to A. micheloltianum nanum Withers, 1953, from the middle Miocene (Helvetian) of northern Italy, and to A. michel- ottianum micheloltianum from the Pliocene through Recent. If the similarities seen between the carinae of these species extend to the scuta and terga, then there are no observable scuta and terga in the concretions that approach the form seen in A. michelottianum. Secondly, the two carinae occur in a single con- cretion, and the associated terga and scuta that are identifiable in this concentration do not differ from those associated with car- inae of A. knapptonensis in other concretions. In terms of Zevina’s revision of scalpellid classification, A. raricostatum can be assigned to the restricted genus Arcoscalpel- lum on the basis of its close affinity to the type species of Ar- coscalpellum. A. michelottianum. Arcoscalpellum knapp- tonensis, on the basis of capitular plate morphology, also appears to resemble most closely species of the restricted genus Arcoscalpellum. Suborder Balanomorpha Pilsbry, 1916 Family Archaeobalanidae Newman and Ross, 1976 Subfamily Archaeobalaninae Newman and Ross, 1976 Genus Solidobalanus Hoek, 1913 Subgenus Hesperibalanus Pilsbry, 1916 Solidobalanus ( Hesperibalanus ) sp., aff. S'. ( H. ) sookensis (Cornwall, 1927) Figure 1 8 LOCALITY. LACMIP locality 5842, “Lincoln Creek Forma- tion,” faunal unit 4. on north bank of Columbia River about 1 .6 km northeast of Knappton, Pacific County, Washington, center of N '/2 of the N V2 of section 9, T 9 N, R 9 W, USGS 7.5-minute topographic quadrangle of Knappton ( 1973 ed ). MATERIAL EXAMINED. Six shells in four concretions and one broken scutum; LACMIP hypotype nos. 6334, 6336. DISCUSSION. The Knappton specimens of Solidobalanus agree with those of S. sookensis from the Sooke Formation of Figures 9 through 18. Arcoscalpellum and Solidobalanus spp Figures 9 through 17. Arcoscalpellum knapptonensis n. sp. (9) exterior of tergum, paratype LACMIP 6268, height 27 mm; (10) interior of tergum, paratype LACMIP 6269, height 24 mm; (1 1) external mold of tergum, paratype LACMIP 6330, height 24 mm; (12) interior of scutum, paratype LACMIP 6267, height 1 8 mm; (13) external mold of scutum, paratype LACMIP 6331, height 20 mm; (14) exterior of carina, holotype LACMIP6270, height 30 mm; (15) interior of carina, paratype LACMIP6332, height 27 mm;(!6)exteriorof carina, paratype LACMIP 6333, height 26 mm; (17) side view of same carina. Figure 18. Solidobalanus sp., aff. S. sookensis (Cornwall), rostral view of shell, hypotype LACMIP6334, height 16 mm. Contributions in Science, Number 336 Zullo: Paleogene Cirripedia of Pacific County, Washington 7 southern Vancouver Island in possessing a recurved carina, a large rhomboidal orifice, broad radii with oblique summits, and an irregularly plicate shell. Opercular plates are difficult to ex- tract from the indurated matrix, and only the interior of a single, broken scutum could be examined in any detail. This scutum differs from those described for S. sookensis in lacking a de- pressor muscle pit and in appearing to have a straight rather than recurved ridge bordering the adductor muscle pit. Solidobalanus sookensis and the Knappton species differ markedly from the upper Eocene Cowlitz Formation species 5. cornwalli in having broad, transparietal radii with crenulate sutural edges, rather than narrow, non-transparietal radii with smooth sutural edges, and in possessing an unusually high and thickened ridge bordering the scutal adductor muscle pit. The Pliocene species S’, proinus and the Pleistocene and extant spe- cies S. hesperius can be distinguished by their possession of a true scutal adductor ridge and of callosities and rugosities on the interior of the scutum. BARNACLES AS BIOSTRATIGRAPHIC INDICATORS Little consideration has been given to the potential of cirripeds as biostratigraphic indicators, particularly in North America. Cheetham (1963) and Weisbord (1977) recognized the value of barnacles in biostratigraphy, but only two authors have at- tempted to develop biostratigraphic zonations based on barnacle assemblages. Mellen (1973), utilizing Collins’ (1973) study of the lepadomorphs from the Upper Cretaceous of Alabama and Mississippi, was able to recognize three zones in the “Selma Chalk.” Zullo (1979b, 1980) was able to distinguish four as- semblage zones in the middle Eocene through lower Miocene formations of North Carolina. Two of these zones, the upper middle Eocene Arcoscalpellum subquadratum zone and the lower upper Eocene A. jacksonense zone, can be recognized throughout the southeastern Atlantic and eastern Gulf Coastal Plains. The Knappton arcoscalpellids demonstrate that certain barna- cles are useful in interregional and, in this case, intercontinental correlation. If it appears improbable that sessile organisms are capable of attaining rapid and widespread distribution, it is only necessary to compare the modern distribution of a related spe- cies. Arcoscalpellum michelottianum is found in the North and South Atlantic basins, the Indian Ocean, and the western Pacific between 40 and 2900 m. This species was originally described from the Astian and Plaisancian of Sicily and is considered to have evolved from A. raricostatum through the Italian Miocene taxon A. michelottianum nanum. The absence of A. michelot- tianum in the eastern Pacific may be an artifact of sampling but is probably related to the closing of Central American seaways during the time that this species was attaining its present dis- tribution. Thus, it appears that A. michelottianum achieved vir- tual worldwide distribution in a three-million-year period through dispersal of planktonic larval stages. Examples of wide- ranging species are not limited to the Lepadomorpha (“goose barnacles”), for several balanomorph species (“acorn barna- cles”), including Balanus trigonus Darwin, 1854, B. calidus Pils- bry, 1916, and B. venustus Darwin, 1 854, have developed tropico- politan or Tethyan distributions by natural means. Reconnaissance of Tertiary and particularly Paleogene marine units of the Atlantic and Gulf Coastal Plains and the Pacific Coast indicates that an array of lepadomorph and balanomorph remains has been overlooked. Systematic collection and analysis of these cirriped assemblages may provide a valuable tool in stratigraphic interpretations. ACKNOWLEDGMENTS I thank Edward C. Wilson, Natural History Museum of Los Angeles County, for providing the cirripeds collected by James and Gail Goedert from the Knappton section, for information concerning the locality, and for reviewing the manuscript; Ray Wells, U.S. Geological Survey, for permitting the use of his notes concerning the stratigraphy of the Knappton section; John M. Armentrout, Mobil Oil Corporation, for information included herein on molluscan biostratigraphy and for his review of the manuscript; William A. Newman, Scripps Institution of Oceanography, for providing the use of J.S. Buckeridge’s Ph.D. dissertation and for his review of the manuscript; and Ellen J. Moore, U.S. Geological Survey, for her review of the manuscript. LITERATURE CITED Alessandri, G. de 1906. Studi monografici sui Cirripedi fossili dTtalia. Palaeontographia Italica, Mem. Paleontol. 12: 207- 324. Armentrout, J.M. 1975. Molluscan biostratigraphy of the Lin- coln Creek Formation, southwest Washington, In Weaver, D., ed.. Future energy horizons of the Pacific Coast; Paleogene Symposium and selected technical papers. Annual Meeting of the Pacific Sections, Amer. Assoc. Petrol. Geol., Soc. Econ. Paleontol. and Mineral., Soc. Econ. Geol., Long Beach, Calif., pp. 14-48. 1 977. Cenozoic molluscan stages of Oregon and Washington. Abstracts with programs, Geol. Soc. Amer. 9(7):882— 883. Beikman, H.M., W.R. Rau, and H.C. Wagner. 1967. The Lincoln Creek Formation Grays Harbor Basin, southwestern Wash- ington. U.S. Geol. Survey Bull. 1244-1:11-114. Buckeridge, J.S. 1980. The fossil barnacles (Cirripedia: Thor- acica) of New Zealand and Australia. Ph D. dissertation, Univ. Auckland, New Zealand, 431 p. Cheetham, A.H. 1963. Gooseneck barnacles in the Gulf Coast Tertiary. Jour. Paleontol. 37:393-400. Cita, M B. 1973. Bacino Terziario del Piemonte. In Desio, A., ed., Geologia dell’ Italia. Unione Tipografico-Editrice Torin- ese, Torino, pp. 455-456. Collins, J.S. 1973. Cirripedes from the Upper Cretaceous of Alabama and Mississippi, eastern Gulf region, U.S. A., I. Pal- aeontology. Bull. British Mus. (Nat. Hist.) 23:351-380. Cornwall, I.E. 1927. Fossil Cirripedia from the upper Oligocene Sooke Formation of Vancouver Island, B.C. Univ. California Publ. Bull. Dept, of Geol. Sci. 1 6(9):399— 408. 8 Contributions in Science, Number 336 Zullo: Paleogene Cirripedia of Pacific County, Washington Darwin, C. 1 854. A monograph on the sub-class Cirripedia, with figures of all the species. The Balanidae, The Verrucidae, etc. Ray Society, London, 684 p. Frey, R.W., and J.G. Cowles. 1969. New observations on Tisoa , a trace fossil from the Lincoln Creek Formation (mid-Tertiary) of Washington. The Compass 47:1-22. 1972. The trace fossil Tisoa in Washington and Oregon. The Ore Bin 34(7): 113-119. Hardenbol, J., and W.A. Berggren. 1978. A new Paleogene nu- merical time scale. In Cohee, G. V., and others, eds.. The geologic time scale. Amer. Assoc. Petrol. Geol., Studies in Geol. 6:213-234. Harris, W.B., and V.A. Zullo. 1 980. Rb-Sr glauconite isochron of the Eocene Castle Hayne Limestone, North Carolina. Geol. Soc. Amer. Bull. 91, pt. 1: 587-592. Flenry, D.P., and PA. McLaughlin. 1967. A revision of the sub- genus Solidobalanus Hoek (Cirripedia Thoracica) including a description of a new species with complemental males. Crusta- ceana 1 2( 1 ):43— 58. Herb, R., and H. Hekel. 1973. Biostratigraphy, variability and facies relations of some upper Eocene Nummulites from northern Italy. Eclogae geol. Helv. 66(2):4 19-445. Hoek, P.P.C. 1 883. Report on the Cirripedia collected by H.M.S. Challenger during the years 1873-1876. Rept. Sci. Res. Voy- age H.M.S. Challenger , Zool. 8(25): 1-1 69. 1 907. The Cirripedia of the S76ogtf-Expedition, A. Cirripedia Pedunculata. S76oga-Expeditie 31a, E. J. Brill, Leyden, p. 1 128. 1913. The Cirripedia of the S76oga-Expedition, B. Cirripedia Sessilia. S76oga-Expeditie 31b, E. J. Brill, Leyden, p. 129-147. Mellen, F.F. 1973. Cirripedes from the Upper Cretaceous of Alabama and Mississippi, eastern Gulf region, U S. A., II. Ge- ology. Bull. British Mus. (Nat. Hist.) 34:381-388. Newman, W.A., and A. Ross. 1 976. Revision of the balanomorph barnacles; including a catalog of the species. San Diego Soc. Nat. Hist. Mem. 9:1-108. Pilsbry, H.A. 1907. The barnacles (Cirripedia) contained in the collections of the U. S. National Museum. Bull. U. S. Natl. Mus. 60: 1-22. 1916. The sessile barnacles (Cirripedia) con- tained in the collections of the U. S. National Museum; includ- ing a monograph of the American species. Bull S. Natl. Mus. 93: 1-366. Rathbun, M.J. 1926. The fossil stalk-eyed Crustacea of the Pa- cific slope of North America. Bull. U. S. Natl. Mus. 138: 1- 155. Rau, W.W. 1958. Stratigraphy and foraminiferal zonation in some of the Tertiary rocks of southwestern Washington. U.S. Geol. Survey Oil and Gas Investigations Chart OC-57, 2 sheets. 1964. Foraminifera from the northern Olympic Peninsula, Washington. U.S. Geol. Survey Prof. Paper 374- G:G1-G33. Seguenza, G. 1876. Richerchi paleontologiche intorno ai Cir- ripedi terziari della provincia de Messina. Con appendice in- torno ai Cirripedi viventi nel Mediterraneo e sui fossili terziari dellTtalia Meridionale. Parte II. Atti Accad. Pontianiana 10: 267-481. Van Eysinga, F.W.B. 1975. Geologic time table, 3rd ed. Elsevier Scient. Publ. Co., Amsterdam, 1 sheet. Weaver, C.E. 1912. A preliminary report on the Tertiary paleon- tology of western Washington. Washington Geol. Survey Bull. 15:1-80. 1942. Paleontology of the marine Tertiary for- mations of Oregon and Washington. Washington Univ. Publ. in Geol. 5:1-790. Weisbord, N.E. 1977. Some Paleocene and Eocene barnacles (Cirripedia) of Alabama. Bull. Amer. Paleontol. 72:143-146. Withers, T.H. 1924. The fossil cirripedes of New Zealand. Bull. Geol. Survey of New Zealand 10: i-vi, 1-48. 1926. Scalpellum sanchezi sp. n., a cirripede from the lower Miocene of Cuba. Ann. Mag. Nat. Hist. 18:616-621. 1936. Miocene cirripedes (Scalpellum and Tessarelasma n. g.) from India. Ann. Mag. Nat. Hist. 1 8:589- 593. 1953. Catalogue of fossil Cirripedia, vol. III. Tertiary. British Mus. (Nat. Hist.), London, xv -I- 396 p. Woodring, W. P. 1950. In Woodring, W.P, and M.N. Bramlette. Geology and paleontology of the Santa Maria District, Cal- ifornia. U. S. Geol. Survey Prof. Paper 222: 1-185. Zevina, G.B. 1978. A new system of the Family Scalpellidae Pilsbry (Cirripedia, Thoracica). 2. Subfamilies Arcoscalpel- linae and Meroscalpellinae. Zoologichesky Zhurnal Akade- mia Nauk SSSR 57(9): 1 343-1 352 (Russian with English abstract). Zullo, V.A. 1966. A new species of Balanus (Cirripedia: Thor- acica) from the late Eocene Cowlitz Formation of southern Washington, U.S. A. Crustaceana 1 1(2): 198-204. Zullo, V.A. 1979a. Thoracican Cirripedia of the lower Pliocene Pancho Rico Formation, Salinas Valley, Monterey County, California. Contributions in Science 303:1-13. 1979b. Biostratigraphy of Eocene through Miocene Cirripedia, North Carolina Coastal Plain. In Baum, G.R., W.B. Harris, and V.A. Zullo, eds.. Structural and strat- igraphic framework for the Coastal Plain of North Carolina. Field trip guidebook, Carolina Geol. Soc. 1979: 73-85 1980. Biostratigraphy of middle and upper Eocene barnacles (Cirripedia), Gulf and southeastern Atlan- tic Coastal Plains. Abstracts with programs, Geol. Soc. Amer. 12:555. Zullo, V.A., and G.R. Baum. 1979. Paleogene barnacles from the Coastal Plain of North Carolina (Cirripedia, Thoracica). Southeastern Geol. 20(4):229-246. Received 25 August 1980; accepted for publication 21 July 1981. Contributions in Science, Number 336 Zullo: Paleogene Cirripedia of Pacific County, Washington INSTRUCTIONS FOR AUTHORS The Natural History Museum of Los Angeles County publishes papers reporting the resuits of original research in the life and earth sciences in its Contributions in Science series. Individual Contributions are issued at irregular intervals and range in size from 8-page papers to lengthy monographs. Manuscripts submitted for publication must undergo internal and outside review; priority is given to manuscripts reporting the results of studies involving the collections of the Museum. Manuscripts should be prepared in accordance with the requirements outlined below and submitted to the appro- priate section of the Museum. Authors must adhere to the Internationa! Codes of Botanical and Zoological Nomenclature. In addition, authors are urged to comply with all recommendations of these codes. A definitive and differential diagnosis is required for any newly proposed taxon. Authors proposing new taxa must indicate that all primary types have been deposited in an appropriate scientific institution, citing the institution by name and providing the relevant identifying details. A Spanish summary is required for all manuscripts dealing with Latin American subjects. Sum- maries in other languages are not required but are strongly recommended where appropriate. PREPARATION OF MANUSCRIPT Type manuscript double-spaced, including title page, abstract, text., acknowledgments, references, tables, and figure titles. A paper must total a minimum of 8 printed pages (as a guideline, 3.5 pages of draft equals about 1 page of final printed copy). Each manuscript component should begin on a new page, in this sequence: title page; abstract; Spanish summary, if required; text; acknowledgments; literature cited; tables, each table— complete with title and footnotes — on a separate page; and figure titles. All pages should be numbered consecutively. Submit original illustrations. Line drawings should be prepared for 50 percent reduction, and should not exceed 14 by 17.5 inches (35 by 43 cm) in size; photographs are usually reproduced at full size and should not exceed 7 by 8.75 inches ( 1 7 by 22 cm, full page width) or 3.25 by 4.5 inches (8 by 1 1 cm, single column). SUBMISSION OF MANUSCRIPT Mail three copies of the manuscript in a heavy paper envelope, enclosing the manuscript copies and figures in cardboard, if necessary, to prevent bending of figures during mail handling. Manuscripts should be accompanied by a covering letter from the author who will be responsible for correspondence regarding the manuscript. The covering letter should contain a statement that the manuscript has been seen and approved by all authors. The letter should give any additional informa- tion that may be helpful to reviewers, such as information on prior publication of any part of the manuscript, and whether or not the author(s) will be willing to meet the cost of reproducing color illustrations. Include copies of any permissions needed to reproduce published material. Manuscripts will be reviewed for possible publication with the understanding that they are being submitted to one journal or series at a time and have not been published, simultaneously submitted, or already accepted lor publication elsewhere. This does not preclude consideration of a manuscript that has been rejected by another publication or of a complete report that follows publication of preliminary findings elsewhere. Copies of any possibly duplicative material should be submitted with the manuscript that is being sent for consideration, as should any peer reviews done before the manuscript was completed. ADDITIONAL GUIDELINES Additional guidelines for authors are available upon request from the Museum Publications Offu . Q 11 L52X NH Number 337 5 February 1982 CIENCE WOLFCAMPIAN RUGOSE AND TABULATE CORALS (COELENTERATA: ANTHOZOA) FROM THE LOWER PERMIAN MCCLOUD LIMESTONE OF NORTHERN CALIFORNI A Edward C. Wilson mm Natural History Museum of Los Angeles County • 900 Exposition Boulevard « Los Angeles, California 90007 —— —I— !■■■!! INI ■■ III !!■! Illlll III III ll II I llllll llll I IIIIIIIIIU ill III SERIAL PUBLICATIONS OF THE NATURAL HISTORY MUSEUM OF LOS ANGELES C OUNTY The scientific publications of the Natural History Museum of Los Angeles County have been issued at irregular intervals in three major series; (he articles in each series are numbered individually, and numbers run consecutively, regardless of subject matter. Contributions in Science, a miscellaneous series of technical papers describing original research in the life and earth sciences. Science Bulletin, a miscellaneous series of monographs describing original research in the life and earth sciences. This series was discontinued in 1978 with the issue of Numbers 29 and 30; monographs are now published by the Museum in the Contributions in Science series. • Science Series, long articles on natural history topics, generally written for the layman. Copies of the publications in these series are sold through the Museum Bookshop. For instructions for authors, refer to Contribution No. 336 or contact the Museum Publications Office. Leon G. Arnold Lawrence G. Barnes Robert L. Bezy Kenneth E. Campbell EDITORIAL BOARD Robert Gustafson John M. Harris Robert J. Lavenberg Camm C. Swift EDITOR Robin A. Simpson Fred S. Truxal David P. Whistler Edward C. Wilson John W. Wright Printed at The Castle Press, Pasadena, Californ WOLFCAMPIAN RUGOSE AND TABULATE CORALS (COELENTERATA: ANTHOZOA) FROM THE LOWER PERMIAN MCCLOUD LIMESTONE OF NORTHERN CALIFORNIA Edward C. Wilson Contributions in Science, Number 337 Natural History Museum of Los Angeles County 5 February 1982 ISSN 0459-0113 Natural History Museum of Los Angeles C ounty 900 Exposition Boulevard Los Angeles, California 90007 CONTENTS ABSTRACT I INTRODUCTION 1 PREVIOUS WORK 1 REGIONAL GEOLOGY 3 STRUCTURE AND RELATIONSHIPS 3 STRATIGRAPHY, AGE, AND CORRELATION 5 Relationships of Sections in the McCloud Limestone 5 Bayha Section 5 Potter Ridge Section 5 Hirz Mountain Section 8 McCloud Point Section 8 Section East of McCloud Bridge 8 Section West of McCloud Bridge 8 Occurrence and Comparison of Related Permian Coral Faunas 8 PALEOECOLOGY 9 SYSTEMATIC MATERIALS AND METHODS 10 SYSTEM ATICS 10 Order Rugosa Milne Edwards and Haime, 1 850 10 Solitary Rugosa 10 Genus Aulophyllum Milne Edwards and Haime, 1 850 10 Aulophyllum (?) sp 10 Genus Clisiophyllum Dana, 1 846 10 Clisiophyllum gabbi Meek, 1864 10 Clisiophyllum oweni n. sp 11 Cyathopsid (?) 19 Genus Gshelia Stuckenberg, 1888 19 Gshelia americana n. sp 19 Genus Heterocaninia Yabe and Hayasaka, 1 920 23 Heterocaninia langenheimi n. sp 23 Heterocaninia (?) sp 25 Fasciculate Rugosa 25 Genus Durhamina Wilson and Langenheim, 1 962 25 Durhamina sublaeve ( Meek, 1864) 25 Genus Heritschioides Yabe, 1950 29 Heritschioides carneyi n. sp 29 Heritschioides coogani n. sp 33 Heritschioides durhami n. sp 33 Heritschioides gavini n. sp 34 Heritschioides hammani n. sp 34 Heritschioides merriami n. sp 35 Heritschioides rowetti n. sp 35 Heritschioides skinneri n. sp 41 Heritschioides smithi n. sp 41 Heritschioides stevensi n. sp 45 Heritschioides wexoi n. sp 45 Heritschioides wildei n. sp 47 Heritschioides (?) californiense Meek, 1 864 47 Genus Mccloudius n. gen 49 Mccloudius fluvius n. sp 49 Genus Siphonodendron M’Coy, 1 849 51 iii Siphonodendron hongi n. sp 51 Genus Yatsengia Huang, 1932 51 Yatsengia fletcheri n. sp 51 Yatsengia kenneyi n. sp 52 Yatsengia scheetzi n. sp 52 Cerioid Rugosa 53 Genus Bass i us n. gen 53 Bassius mccloudensis n. sp 57 Genus Dillerium n. gen 57 Dillerium potterensis n. sp 57 Genus Kleopatrina McCutcheon and Wilson, 1 963 59 Subgenus Kleopatrina McCutcheon and Wilson, 1963 59 Kleopatrina (Kleopatrina) raubae n. sp 59 Subgenus Porfirievella Minato and Kato, 1965 59 Kleopatrina (Porfirievella) mckibbinae n. sp 59 Kleopatrina (Porfirievella) peggyae n. sp 62 Kleopatrina ( Porfirievella ) whitneyi n. sp 63 Kleopatrina (Porfirievella) zulloi n. sp 63 Genus Langenheimia n. gen 65 Langenheimia klamathensis n. sp 65 Genus Petalaxis Milne Edwards and Haime, 1 852 65 Petalaxis allisonae n. sp 65 Petalaxis besti n. sp 67 Petalaxis kennedyi n. sp 69 Petalaxis pecki n. sp 69 Petalaxis sutherlandi n. sp 69 Genus Traskina n. gen 73 Traskina shastensis n. sp 73 Cerioid-Astreoid Rugosa 73 Genus Arachnastraea Yabe and Hayasaka, 1916 73 Arachnastraea fergusoni n. sp 73 Arachnastraea fryi n. sp 75 Order Tabulata Milne Edwards and Haime, 1 850 75 Genus Bayhaium Langenheim and McCutcheon, 1959 75 Bayhaium merriamorum Langenheim and McCutcheon, 1959 75 Bayhaium virginiae n. sp 77 Genus Enigmalites Tchudinova, 1 975 77 Enigmalites roberti n. sp 77 Genus Michelinia de koninck, 1841 79 Michelinia nelsoni n. sp 79 Genus Neomultithecopora Lin, 1 963 79 Neomultithecopora sandoi n. sp 79 Genus Syringopora Goldfuss, 1 826 83 Syringopora mccutcheonae Wilson and Langenheim, 1 962 83 Syringopora multattenuata McChesney, 1 860 83 LOCALITIES 83 UCMP Localities 83 LACMIP Localities 86 ACKNOWLEDGMENTS 86 LITERATURE CITED 87 iv WOLFCAMPIAN RUGOSE AND TABULATE CORALS (COELENTERATA: ANTHOZOA) FROM THE LOWER PERMIAN MCCLOUD LIMESTONE OF NORTHERN CALIFORNIA' Edward C. Wilson* 2 ABSTRACT. Six sections measured across the McCloud Limestone, Shasta County, northern California, are Lower Permian (Wolfcampian Series), represent a composite thickness of at least 5,500 feet (about 1,700 meters), and contain rich invertebrate faunas including at least 49 species of rugose and tabulate corals. Forty-two species in 17 genera of rugose corals are described, including 36 new species and 5 new genera. Seven species in five genera of tabulate corals are described, including four new species. The general generic composition is characteristic of the Durhaminid Coral Province known from rocks in the Ural Mountains, Novaya Zemlya, Spitzbergen, Arctic North America, and western North America as far south as southern California. Some genera appear to be endemic to western North America. The corals did not form reefs but were randomly spaced on calcareous bioclastic substrates in shallow waters of the Corddleran eugeosyncline near the Permian paleoequator in an area temporarily lacking much volcanic sedimentation. INTRODUCTION The primary purpose of this study is to determine the diversity and stratigraphic distribution of the stony corals of the McCloud Limestone, Shasta County, California. Six sections were mea- sured and intensively sampled, and some additional collections were made at localities nearby the measured sections from the southernmost outcrop of the formation to about 20 miles (about 32 km) north along its general northeast trending strike (Fig. 1 ). In the remaining approximately 14 miles (about 22.5 km) of outcrop north of the measured sections, inspections were made on High Mountain and Bald Mountain, but the few corals col- lected were very poorly preserved. The section at Tombstone Mountain, which contains the highest Permian (Leonardian) rocks in the formation, is under study. Fossils were selected for collecting in the following manners: ( 1 ) all corals, where feasible, were collected; (2) fusulinids were searched for near the exposed bases and tops of the sections, but within the sections they generally were collected only where abundant; (3) brachiopods were collected at localities where they appeared to be well preserved; (4) bryozoans were collected where abundant; (5) mollusks (bivalves, gastropods, cephalo- pods, rostroconchs) were collected where well preserved; (6) other conspicuous fossils were collected if well preserved, readily attainable, and of potential stratigraphic significance. The re- sults are a relatively detailed sampling of the fossil corals of the southern part of the formation and a reconnaissance collection of other fossils. The specimens are deposited in the University of California Museum of Paleontology at Berkeley and the Natural History Museum of Los Angeles County, Invertebrate Paleontol- ogy Section. Field work was pursued during the following intervals: 9-10 March, 19 March-4 April, 19-22 June, 7-1 1 September 1963; 1 9-20 April, 26-29 April 1964; 29 June 1967; 1 9-28 September 1968; 21-22 May 1970; 26 August-6 September 1973; 1-8 Oc- tober 1976; 21-29 June 1979; 1-15 July 1981. PREVIOUS WORK Pioneer work on older rocks in Shasta County was begun by Trask (1853:15), who later found (Trask, 1855:50) a “Car- boniferous” limestone there that cropped out from “the Pitt River in a northerly distance for about thirty or forty miles, forming a portion of the canyon of McCloud’s Fork. . . .” He was followed by the California Geological Survey under J.D. Whitney, who collected from the limestone on 20 and 21 Septem- ber 1862 (Brewer, 1930:323). From this collection, Meek ( 1 864) described fusulinids, corals, brachiopods, and a gastropod, all reportedly Carboniferous. Whitney (1865:326) estimated that the formation was 1 ,000 feet thick and conformable to the forma- tions above and below. Fairbanks ( 1893:35) named the formation McCloud Lime- stone. In a differing opinion. Turner ( 1 894:230) considered it to be part of the Calaveras Formation. The classical study of the Redding Quadrangle by Diller (1906) fixed the name McCloud Limestone and assigned the formation to the Pennsylvanian. Hinds (1932:273) first placed the McCloud Limestone in the Permian. Using fusulinids, Wheeler ( 1 933) confirmed this deter- mination and later reported (Wheeler, 1935) fusulinids in “seven distinct faunal horizons” from the southern part of the forma- tion. Thompson and Wheeler (1946) described fusulinids from 1 Review committee for this contribution: John M. Harris, William J. Sando, and Calvin H Stevens. 2. Section of Invertebrate Paleontology, Natural History Museum of Los Angeles County, Los Angeles, California 90007. Contributions in Science, Number 337, pp. 1-90 Natural History Museum of Los Angeles County, 1982 ISSN 0459-01 13 ^ENTO 2 Contributions in Science, Number 337 Wilson: Permian Corals of California the McCloud Limestone and Nosoni Formation as far north as Potter Ridge, concluding that the former formation was Wolf- campian and Leonardian and the latter was Guadalupian in age. Fusulinid paleontology and biostratigraphy were continued by Skinner and Wilde ( 1 965), who showed that the formation repre- sented an aggregate thickness in excess of 7,000 feet and zoned it for its entire outcrop length into eight complexly distributed zones, ranging from possible Upper Pennsylvanian into Lower Permian. Wilde (1971 : 364 ) later determined that the lowest zone is Permian. Additional reports on fusulinids, corals, brachiopods, and mol- lusks from the formation were published by Hayasaka ( 1 936:64), Langenheim and McCutcheon (1959), Watkins (1973, 1974, 1975), and Wilson (1967a and b, 1970, 1971, 1980:89). Other information about the formation has continued to accu- mulate. Coogan ( 1 960), Skinner and Wilde ( 1 965), and Watkins (1973) pointed out that fossils at the top of the underlying Baird Formation at Bollibokka Mountain are Pennsylvanian or Per- mian, rather than Mississippian as at the type locality. Working farther south, Albers and Robertson (1961) reported that the McCloud Limestone is separated from the overlying Nosoni For- mation by the largest fault in the area, the McCloud fault, along which quartz diorite is intruded. Lanphere, Irwin, and Hotz (1968) cited an age of 246 m.y. for the quartz diorite, indicating that the intrusion was a Permian event rather than a Jurassic one as previously believed. Evans (1977) reported on the economic geology of the southernmost McCloud Limestone. Study of the petrography of the McCloud Limestone indi- cated to Demirmen and Harbaugh (1965) that it is formed largely of bioclastic sediments that originated in clear waters of a eugeosyncline, somehow protected from the volcanic sediments typical of the rest of the section. In summary, although considerable work has been done on other aspects of the McCloud Limestone, thorough paleontologi- cal study has been limited to the fusulinids, although minor work has been done with the corals, brachiopods, and mollusks. REGIONAL GEOLOGY Rocks older than Silurian or younger than Triassic have not been positively identified among the eastward dipping, north-south trending, eugeosynclinal strata that rise south of Shasta Lake from beneath disconformably overlying Cretaceous and younger rocks and disappear northwards beneath the volcanic flows of the southern flanks of Mount Shasta. At least half of the entire rock column is composed of layered volcanic rocks. Upper Paleozoic rocks represent parts of the Mississippian, Pennsylvanian, and Permian. They are known in ascending order as the Bragdon Formation, Baird Formation, McCloud Lime- stone, Nosoni Formation, and Dekkas Formation. Various divi- sions (members, groups) have been proposed based on lithology, and some zones and zonules have been erected based on fossils. The Bragdon Formation has Mississippian fossils. The Baird For- mation contains Mississippian, Pennsylvanian, and possibly Per- mian fossils. The overlying three formations all are Permian. The Pit stock, a Permian intrusive, follows the Baird-McCloud contact for many miles, locally intrudes faults in parts of these and other formations, and is followed by the McCloud River for much of its course south from Bollibokka Mountain In the Pot- ter Ridge area, an intrusion of mafic quartz diorite cuts the Pit stock in places. STRUCTURE AND RELATIONSHIPS The Paleozoic and Lower Mesozoic rocks of the Shasta Lake area form a slightly sigmoid north-south trending homocline, which dips eastward at angles varying from 45 to 60 . Coogan (1960:245) reported synclines and anticlines in the Bollibokka Mountain region. An additional small anticline is present in this area on the west bank of the McCloud River near the high-water line north of the McCloud River Bridge and south of Nawtawa- ket Creek. Albers and Robertson (1961, map) showed a north- south trending anticline in the block of McCloud Limestone north of Marble Creek. For its entire length, the McCloud Lime- stone is broken by transverse faults into discrete blocks. Numer- ous high-angle faults occur within some of these blocks. There has been disagreement about the nature of the contact of the McCloud Limestone with the underlying and overlying formations at some localities. Although the scope of this report is primarily the coral paleontology of the formation, the following observations and conjectures may improve understanding of the structural and stratigraphic relationships. 1 . On the eastern bank of the McCloud River directly north of the McCloud River bridge, the contact between the McCloud Limestone and the underlying Baird Formation consistsof purple tulTaceous sediments, including conglomerate, and interbedded limestone, and seems to be gradational and sedimentary. The gradational rocks here were identified (Coogan, 1960:250; Skin- ner and Wilde, 1965:1 1) as the Baird Formation but are higher (Pennsylvanian or Permian) than at the Baird type locality ( Mississippian). 2. A high-angle fault separates the McCloud Limestone and the Nosoni Formation along the east side of the ridge crest due east of the McCloud River bridge on Bollibokka Mountain and continues south for at least a mile. Diller (1906, map 3, top structure section) apparently recognized this fault, but some later workers considered the contact to be conformable. 3. About a mile north of the McCloud River bridge on the east bank of the river opposite Wittawaket Creek, Coogan ( 1 960:249) described a depositional contact between the McCloud Lime- stone and the overlying Nosoni Formation. This is in a different fault block than the McCloud Bridge section of the present study. I inspected Coogan’s contact and recognized that the lithology of the apparent top of the McCloud Limestone there Figure 1. McCloud Limestone exposures in Shasta and Siskiyou Counties, northern California, showing locations of measured stratigraphic sections: ( 1 ) Bay ha section, (2) Potter Ridge section, (3) Hirz Mountain section, (4) McCloud Point section, (5) McCloud Bridge East section, (6) McCloud Bridge West section. Contributions in Science, Number 337 Wilson: Permian Corals of California 3 4 Figure 2. Columnar sections of the McCloud Limestone showing locations of coral collections. differs from that at the top of the McCloud Bridge section, but I feel that the contact is worth further investigation. 4. At Hirz Mountain, an apparently gradational contact be- tween the McCloud Limestone and the underlying Baird Forma- tion is exposed along the road a few meters above the beginning of the steep upgrade that leads abruptly from the nearby horizon- tal access road to the summit of the mountain. The gradational contact has limestone beds that contain fusulinids. This contact also was recognized by Watkins (1973:1755), who erected the Hirz Mountain Limestone Member of the Baird Formation for nearby limestone lenses below the Baird-McCloud contact and considered it to be equivalent to fusulinid zone A, now known to be Wolfcampian. If this determination is correct, then the Penn- sylvanian-Permian boundary here lies in the upper Baird Forma- tion. It also implies that the fault block along the road should be in fusulinid zone A and therefore must be in a different fault block from the one measured at Hirz Mountain by Skinner and Wilde (1965), who assigned their section at Hirz Mountain wholly to fusulinid zone D. My Hirz Mountain section is not in the same fault block as the road and has zone D fusulinids. 5. Both the upper and lower contacts of the McCloud Lime- stone on the ridge between Marble and Potter Creeks are with an intrusion of mafic quartz diorite that probably rose along high- angle faults as emphasized by Albers and Robertson ( 1 96 1:41 ). 6. In the Bayha section of the present study, both the upper and lower contacts of the McCloud Limestone are covered. A fine- grained shale was dug out here within 2 feet (0.6 meters) of the upper contact. No evidence of a conglomerate was seen. The lower contact, though covered by talus and deep soil, has small springs in places, suggesting the presence of an underlying fault. 7. A pebble conglomerate is present stratigraphically below definite Nosoni Formation rocks on the south side of the summit of the first knoll due south of the Gray Rocks. This hill is greatly faulted, and whether the conglomerate lies within the Nosoni Formation or at the base of the formation is not easily demon- strated because rocks below the conglomerate are not exposed. Fusulinids within the clasts of this conglomerate are referable to species in the McCloud Limestone. 8. Albers and Robertson (1961:57) recognized part of an ex- tensive fault separating the McCloud Limestone and the Nosoni Formation between the Gray Rocks and the Curl Creek areas, named it the McCloud Fault, and described it as dipping west at a moderately steep angle, with left lateral and reverse movement, amount of displacement unknown. The full extent of this fault should be mapped to determine if it had a role in causing the present odd distribution of only older rocks in the southern part of the McCloud Limestone and only younger ones in the northern part. STRATIGRAPHY, AGE, AND CORRELATION Relationships of Sections in the McCloud Limestone By 1935, the discrete fault blocks of the southernmost 9 or so miles (about 14.5 km) of the McCloud Limestone were known to correlate biostratigraphically. It was believed that the same cor- relation eventually would be obtained everywhere in the forma- tion when its approximately 30-mile (about 48-km) length had been investigated This expectation was shown to have been sim- plistic in 1965 when Skinner and Wilde published their mono- graph on the fusulinid faunas and biostratigraphy of the forma- tion for its entire outcrop area. They stated ( 1 965: 1 2) that “the maximum thickness observed in any one place ... is about 2,400 feet, but if the maximum observed thickness of the various [fusulinid] faunal zones were present in a single section the ag- gregate would be in excess of 7,000 feet.” Correlation of the fault blocks proved to be complex. Of their 23 sections measured across the formation, 7 contain only a single zone, 12 have two zones, 3 have three zones, and 1 has four zones. No section has more than four of the eight zones. The validity of this zonation has been substantiated by its recognition elsewhere in western North America. Permian corals of western North America are not yet known thoroughly enough to be used more than locally with any as- surance as index fossils. In general, however, the results obtained with the corals in the present study support the fusulinid zona- tion of Skinner and Wilde (1965) in the limited sense that the Potter Ridge and Bayha sections can be correlated using corals, as they can be with fusulinids, whereas the four more northerly sections, which could not be correlated by corals, were shown by Skinner and Wilde (1965) to be at various lower stratigraphic positions in the formation. Six sections were measured in the McCloud Limestone with Jacob’s staff or tape and Brunton pocket transit (Fig. 2). The tables of Mandlebaum and Sanford ( 1 952) were used to compute the thickness of the taped sections. Ranges of coral genera in the formation were tabularized (Fig. 3), and occurrences of coral species were organized into a checklist (Fig. 4). Comments about the sections that relate to their stratigraphy, age, and correlation follow. Bayha Section. The lower beds of my Bayha section contain very abundant specimens of the Wolfcampian index fusulinid Pseudoschwagerina. Four species of corals that occur in the Bayha section also were found in the Potter Ridge section. Seven species of corals found in the Bayha section were not recognized in the Potter Ridge section. Three species of corals described by Meek (1864) were not found by me in the Bayha section or elsewhere, although they reportedly were collected from that area. The large archaeogastropod Omphalotrochus whitneyi (Meek, 1864) occurs in the Bayha section, and the types were collected from there or nearby. It has a wide stratigraphic range in the McCloud Limestone and is the type species for the genus, now recognized as an index fossil for the Upper Paleozoic in North America, South America, Europe, and Asia. Near the top of the Bayha section are beds with abundant specimens of a biscuit-shaped blue-green alga. These are overlain with beds of particularly abundant and large pelmatozoan columnals that are not present elsewhere in the section. This sequence of algae and columnals also occurs near the top of the Potter Ridge section. The occurrences of coral species common to both sections and the sequential regularity of most of them and other fossils in each section confirm the correlation of the two sections with fusulinids by Skinner and Wilde (1965). Potter Ridge Section. This section has been correlated with the Bayha section by Skinner and Wilde (1965) with fusulinids. I have described 15 species of corals from there, 4 of which also Contributions in Science, Number 337 Wilson: Permian Corals of California 5 WOLFCAMPI AN FUSULINID ZONES ENIGMALITES CLISIOPHYLLUM c D HERITSC HIOIDES NEOMULTITHECOPORA AULOPHY LLUM l?l HETEROCANINIA YATSENGI A SYRINGOPORA KLEOPATRINA I PORFIRIEVELLAI MICHELI NIA ARACHNASTRAEA MCCLOUDIUS r SIPHONODENDRON BASSIUS KLEOPATRINA I KLEOPATRINA I DILLERIUM TRASKINA L ANGEN HE I M I A DURHAM I NA * * (Stratigraphic position uncertain but within Zone E or F). Figure 3. Stratigraphic ranges of coral genera in the McCloud Limestone Wolfcampian Series fusulinid zones of Skinner and Wilde (1965). 6 CHECKLIST OF MCCLOUD LIMESTONE WOLFCAMPIAN SERIES STONY CORALS Section W. of McCloud River Bridge & nearby Section E. of McCloud River Bridge & nearby [ McCloud Pt. section Hirz Mt. section & nearby Potter Ridge section Bayha sectic >n Southern outcrop | Meek (1864) specimens C/3 uj H _l SPECIES o O _i 1 dO«T .'V'Ol.l (OcolxHjmvl S.unifri u't Ha Irfornin . ( Ce rlif in H Vrous P I,ATK, I . KiM.!,--;- d-i’. .! .Young. Contributions in Science, Number 337 Wilson: Permian Corals of California 13 Figure 6. Clisiophyllum gabbi Meek, 1 864. Fusulinids in Parts a-d and f are Pseudoschwagerina robusta ( Meek, 1 864). Part a, MCZ paralectotypes, 14 MCZ Meek collection block no. 23, complete calyx figured by Meek (1864, pi. 1, fig la), xl Part b, MCZ lectotype (lower right) and paralectotypes, MCZ Meek collection block no. 19, x0.8. Part c, MCZ lectotype (lower left arrow) and paralectotypes, MCZ Meek collection block no. 19, obverse of Part b xO 8 Part d, MCZ paralectotype, MCZ Meek collection unnumbered corallum, figured by Meek ( 1 864, pi. 1 , fig. 1 ), missing apical part presumed lost xl Part e, MCZ paralectotype, MCZ Meek collection no. 24, x0.5. Part f, MCZ paralectotype, MCZ Meek collection block no. 1, xl. Figure 7. Clisiophyllum gabbi Meek, 1 864. Fusulinids are Pseudoschwagerina robusta (Meek, 1 864). All parts x3. Parts a and b, MCZ lectotype I rom MCZ Meek collection block no. 19, transverse and longitudinal sections. Parts c through f, UCMP hypotype 37 1 71 , serial transverse sections through one 15 corallum. Part g, UCMP hypotype 37172, longitudinal section. Part h, LACM1P hypotype 6344, longitudinal section. Figure 8. Clisiophyllum gabbi Meek, 1864. Fusulinids are Pseudoschwagerina robusta (Meek, 1864). All parts x3„ Parts a and b, UCMP hypotype 16 371 74, serial transverse sections through one corallum. Part c, UCMP hypotype 37173, longitudinal section. Part d, UCMP hypotype 37175, longitudinal section. 'yji \ . 'x A Figure 9. Aulophyllum (?) sp Clisiophyllum oweni n. sp and cyathopsid (?). Parts a and b. Aulophyllum (?) sp. Part a, LAC Ml P hypotype 6342, transverse section, xl .5. Part b, LACMIP hypotype 6343, longitudinal section, x3. 17 Parts c through e. C. oweni n. sp., all parts x3. Parts c and d, LACMIP holotype 6345, transverse (c) and longitudinal (d) sections. Part e, LACMIP paratype 6346, transverse section. Part f. Cyathopsid (?), UCMP hypotype 37176, transverse section, x 1 . 5 Figure tO. Gshelia americana n. sp„ LACMIP holotype 6347. Parts a and b, serial transverse sections, x3. Parts c through e, serial transverse sections, 18 xl.5. Part f, longitudinal section, xl.5. more major septa (28 at 8 mm, 39 at 18 to 22 mm), longer minor septa, and a less symmetrical axial structure. Etymology. The species is named for Mr. Phillip G. Owen. Cyathopsid (?) Figure 9f Description. Corallites large, solitary, exterior unknown; adoles- cent stages unknown; adult corallite diameters 6 to 7 cm; septa of 2 orders, 76 to 90 each; major septa 14 to 21 mm long, thin in dissepimentarium, somewhat dilate in cardinal quadrants in tab- ularium, thin in counter quadrants, cardinal septum short, 10 to 16 mm long, and dilate; minor septa short, 3 to 6 mm long, thin, confined to periphery of dissepimentarium; dissepimentarium wide, 6 to 1 1 mm; dissepiments in 11 to 20 ranks, pseudo- herringbone between minor and major septa, herringbone or complete (many straight, some concentric) between major septa; narrow stereozone on axial row of dissepiments; corallite wall thin, generally not preserved. Documentation. UCMP hypotypes 37 1 76-37 178 Three coralla from UCMP localities D-841 (hypotype 37178) and D-842 (hy- potypes 37 1 76-37 1 77) were studied. Discussion. The preservation of these specimens is poor due to partial crushing, apparent preburial abrasion of most of the ex- ternal walls and adjacent dissepimentaria, and loss of structures due to styolite resorption. The corals are not uncommon but difficult to collect. It is important to record the existence of these large solitary corals because they are associated with both cerioid and fasciculate compound rugose corals at these localities. The description is based upon transverse sections of ephebic stages only. It was not possible to make longitudinal sections of the coralla available. However inadequate the preservation, it seems obvious to me that these specimens are referable to the Cyathopsidae because of their resemblances to Bothrophyllum Trautschold, 1879, Timania Stuckenberg, 1 875, and Gshelia. No North American Permian cyathopsid has been described with such a great number (90) of major septa, and therefore compari- sons with the few other recorded species are unnecessary. Genus Gshelia Stuckenberg, 1888 Gshelia americana n. sp. Figures lOa-f; 1 la-e; 12a-d Diagnosis. A species of Gshelia characterized by the combina- tion of very large corallites, numerous major septa, and minor septa confined to the dissepimentarium. It has larger corallites and shorter minor septa than G. rouilleri. It has shorter minor septa than G. (?) lonsdalei. It has longer major septa and shorter minor septa than G. rouilleri var. breviseptata. It has a more complex axial structure than G. (?) nikitini. External Description. Corallites solitary, cylindrical beyond short apices with angles of 30-35°, straight, irregularly curved, or with abrupt 90° or smaller angle bends, large diameters to 8.1 cm, lengths to 47 cm (incomplete) and more (holotype length 3 1 cm); epitheca with closely spaced, fine growth lines and widely spaced annual (?) constrictions; calices deep, sides sloping steep- ly inwards and downwards, cardinal fossula prominent. Transverse Section Description, Holotype. Corallite circular to sub- circular, 5.4 cm maximum diameter; septa of 2 orders, 50 to 55 each at maturity, all extending to corallite wall; major septa thin in dissepimentarium, highly dilate in cardinal quadrants in tab- ularium in adolescent and early adult stages, slightly dilate else- where in tabularium, 14 to 19 mm long in adult stage, cardinal septum shortened, 1 0 to 12 mm long in adult stage; minor septa thin, short, confined to peripheral U of dissepimentarium in adult stage, not crossing dissepimentarium, not seen at diameters less than 30 mm; major septa at corallite diameters 27, 35, 47, and 50 mm number 48, 51, 55, and 52 (actual decrease), respectively; microstructure of dilate septa shows fibers at right angle to sin- uous, dark, centerline; cardinal fossula open, with 3 to 5 tabulae arched around axial end in some sections; axial structure present only in adolescent stage, clisiophy lloid at 20-mm corallite diame- ter with slightly dilate medial plate 2.5 mm long and 6 discon- tinuous septal lamellae united by 5 concentric tabellae (axial structure absent at corallite diameters 30 mm and greater, leav- ing open tabularium); dissepimentarium wide, width 3 to 9 mm in adult stages, not observed at corallite diameters less than 20 mm; dissepiments in 10 to 17 ranks in adult stage, regular pe- ripherally, herringbone axially; stereozone on axial row of dis- sepiments in cardinal quadrants; corallite wall very thin, width 0. 1 to 0.2 mm, silicified. Transverse Section Description, Paratype Variation. Paratype 37183: major septa at corallite diameter 6 mm (smallest diame- ter sectioned in this species) number 20, all thin, withdrawn from center, without axial structure; major septa at corallite diameter 1 8 mm number 40, all dilate, with axial structure; major septa at corallite diameter 45 mm number 51, cardinals dilate, counters thin, axial structure absent. Paratype 37184: major septa at corallite diameter 17 mm number 32, all dilate except 1 counter quadrant on apical side of section, all dilate on calicular side of section. Paratype 6348: major septa at corallite diameter 37 mm num- ber 49, cardinal quadrant septa dilate, counter quadrant septa thin, cardinal septum short, major septa at corallite diameter 60 mm number 64, cardinal quadrant septa slightly dilate, counter quadrant septa thin, cardinal septum short. This specimen shows middle to late adult stage transition by thinning cardinal quad- rant septa. Paratype 6355 (largest specimen): major septa at corallite di- ameters 78 mm and 81 mm (sections 95 mm apart) number 60 in each, cardinal septum short, 16 mm long, dilate, cardinal quad- rant septa slightly dilate, those neighboring fossula more so, counter quadrant septa thin, major septa length 28 mm total, 9 to 1 3 mm in tabularium, not meeting; minor septa short, 4 to 6 mm long, not crossing dissepimentarium; dissepimentarium regular, wide, 13 to 19 mm width, dissepiments as in holotype; tab- ularium 41 mm wide, open in center; wall thin, silicified. Longitudinal Section Description, Holotype. Dissepimentarium in adult stage (35 mm below calyx, corallite diameter 48 mm) of 1 0 to 17 ranks of steeply dipping cystose dissepiments, smaller at periphery (presumably where minor septa present), larger ax- ially; tabulae in same section mostly flat with downturned edges. Contributions in Science, Number 337 Wilson: Permian Corals of California 19 Figure 1 1. Gshelia americana n. sp. All parts x 1.5. Parts a through c, LACMIP paratype 6348, serial transverse sections (a, b) and longitudinal section 20 (c). Parts d and e, LACMIP paratype 6349, serial transverse sections. Figure 12. Gshelia americana n. sp. All parts xl .5. Parts a and d, UCMP paratype 37180, transverse (a) and longitudinal (d) sections. Part b, LACMIP paratype 6355, transverse section. Part c, UCMP paratype 37182, transverse section. 21 Us 5! t 22 Figure 13. Heterocaninia langenheimi n. sp., UCMP holotype 37195, serial transverse sections (Part a, x3. Parts b e, xl.5). 9 to 1 2 per cm; tabellae at peripheral edges of some tabulae; wall as in transverse section; longitudinal section of adolescent stage not observed. Documentation. L ACM IP holotype 6347, LAC M IP paratypes 6348-6357, UCMP paratypes 37179-37194. Twenty-three thin sections and numerous polished sections from 27 coralla from LACMIP locality 1 133 (holotype 6347, paratypes 6348-6357), UCMP localities D-800 (paratypes 37183-37184, 37189), D-819 (paratypes 37185-37187), D-821 (paratype 37179), D-822 (paratype 37188), D-826 (paratypes 37180-37181, 37 1 90), D-83 1 (paratypes 37191 -37 1 93), and D-832 (paratypes 37 1 82, 37 1 94) were studied. Discussion. This species has all the characters of the genus: earliest stages without axial structures, followed by short interval with axial structures, and adult stages lacking axial structures (Dobrolyubova, 1940:71-81). The McCloud Limestone occur- rence is the first record of the genus for North America in the Permian. Schouppe (1961 :368) stated that Gshelia occurs in the “Unterkarbon . . . Nordamerica” but I have not been able to verify this. Hill (1956:292) listed the range as “M. Carb.-U. Carb., USSR-Spitz.[bergen],” but it also occurs in the Lower Permian of Russia. Gshelia americana is comparable to the type species, G. rou- illeri Stuckenberg, 1888, from the Upper Carboniferous of the Moscow Basin, in size, number of septa, development of the axial structure, and some minor features. G. americana has, however, minor septa that are confined to the periphery of the dissepimen- tarium, whereas G. rouilleri has minor septa in many coralla that cross the dissepimentarium and may extend slightly into the tabularium. Schouppe (1961:367) stated that there are four species of Gshelia in Spitzbergen: G. calophylloides (Holtedahl, 1913), G. nikitini (Stuckenberg, 1905), G. ruprechti ( Stuckenberg, 1895), and G. lonsdalei (Keyserling, 1854). These were described and figured by Heritsch (1939:37-57) under the genus Si- phonophyllia Scouler, 1844. Only G. nikitini was shown (Heritsch, 1 939:54, pi. 2, fig. 3) to have a juvenile axial struc- ture, and that was a simple columella without the axial lamellae present in the type species of the genus. Although the adult stages resemble those of Gshelia, there are other genera so simi- lar that I believe the Spitzbergen species’ generic assignment to be in doubt until the juvenile stages are figured adequately. If, however, these four species should turn out to belong in Gshelia, the following differences can be pointed out from G. americana. G. (?) nikitini has a simpler axial structure. G. (?) calophylloides appears similar in the adult stages but the juve- nile stages are unknown, and Heritsch (1939) figured no longitu- dinal sections so these cannot be compared. G. (?) ruprechti, as figured by Heritsch (1939, pi . 2, figs. 6, 7; pi . 1 1 , fig. 8) does not exhibit many characters of the genus, is probably not Gshelia, and has minor septa that cross the dissepimentarium. G. (?) lons- dalei, as figured by Heritsch (1939, pi. 3, fig. 2; pi. 19, fig. 15) has long minor septa that cross the dissepimentarium and extend into the tabularium. G. rouilleri var. breviseptata Dobrolyubova and Kabakovich, 1948 (p. 21, pi. 11, figs. 1-9; pi. 12, figs. 1-3; pi. 13, figs. 1-2), from the Upper Carboniferous of the Moscow Basin, has longer minor septa and shorter major septa than G. americana. G. elliptica Chi, 1931 (p. 17, pi. 3, figs. 8a-c), does not belong in this genus and is not comparable to G. americana. Both Gshelia (?) sp. of Easton and Melendres ( 1964:413, figs. 1 a 1 c), from a clast in a Miocene conglomerate in the Philippine Islands, and Gshelia cf. calophylloides of Schouppe (19o 1:368, fig. 4), from the Upper Carboniferous of Yugoslavia, are known from specimens inadequately preserved for close comparison with other species. Etymology. The species is named for North America. Genus Heterocaninia Yabe and Hayasaka, 1920 Heterocaninia langenheimi n. sp. Tigures 1 3a-e; 1 4a-d Diagnosis. A species of Heterocaninia characterized by the combination of very large corallites, a modest number of septa, and well-developed minor septa in the dissepimentarium. It has much larger corallites, fewer septa, and better developed minor septa than H tholusitabulata. External Description. Corallites solitary, large, trochoid becom- ing cylindrical, maximum observed length 36 cm, maximum ob- served diameter 7.6 cm; coral 1 i te wall transversely lined and wrinkled, larger wrinkles (rugae) about 1 to 2 cm apart, fine transverse lines number about 30 in 1 cm vertically along corallite. Transverse Section Description. Corallites roughly circular, diam- eters to 7.6 cm (generally between 6 and 7 cm); major septa well developed, dilate in cardinal quadrants in tabularium through- out ontogeny, thin in dissepimentarium, thinner in counter quad- rants in tabularium but somewhat dilated there in juvenile stages; major septa in holotype number 36 at 23 mm, 43 at 35 mm, 52 at 54 mm, 52 at 57 mm, 57 at 68 mm; septa number 74 in largest (74-mm) paratype, UCM P 37 1 96; cardinal septum short; septa number 24 in youngest observed (diameter 1 1 mm) para- type, 6359; minor septa well developed, at 1 1 -mm corallite diam- eter crossing dissepimentarium (2 ranks) and few entering tabularium as nubs on inner wall, in adult stages confined to outer '/2 to % of dissepimentarium; axial structure large, variable, clisiophy lloid in juvenile stages, clisiophylloid or aulophylloid in adult stages, occupying most of tabularium, regular to irregular, with relatively few septal lamellae abutting straight to sinuous medial plate and connected by axial tabellae; dissepimentarium well developed, as much as 1.5 to 2 cm wide in adult stages; dissepiments highly variable, herringbone, pseudoherringbone, concentric, angulo-concentric, and (in largest adult stages) with some small lateral cystose ones alongside some septa; false wall developed axially on row of dissepiments bordering tabularium in some corallites, thickest in cardinal quadrants and juvenile to early adult stages; corallite wall very thin, 0.1 to 0. 1 5 mm thick, silicified. Longitudinal Section Description. Dissepimentarium of 12 to 23 ranks of steeply to very steeply dipping cystose to elongate, small to large dissepiments, generally more cystose peripherally, more Contributions in Science, Number 337 Wilson: Permian Corals of California 23 24 Contributions in Science, Number 337 Wilson: Permian Corals of California elongate axially; tabellae of 2 series: axial and periaxial; axial tabellae gently domed inwards and upwards towards and around medial plate and/or septal lamellae, about 15 in I cm; periaxial tabellae in I or 2 series, horizontal, sagging, or sloping inwards and upwards generally to axial tabellae. Documentation. UCMP holotype 37195, UCMP paratypes 37196-37202, 37283, L ACM I P paratypes 6358-6373. Nine thin sections and numerous polished sections from 24 coralla from UCMP localities D-819 (paratypes 37200-37202), D-822 (paratypes 37196, 37198-37199, 37283), D-826 (paratype 37197), and D-83I (holotype 37195) and LACMIP localities 1132 (paratypes 6366-6370), 1133 (paratypes 6359-6365, 6371-6373), and 4458 (paratype 6358) were studied. Discussion. Genera of many large solitary Upper Paleozoic cor- als are badly in need of reinvestigation and redefinition. This coral does not fit readily into any of them. I have chosen to place it in Heterocaninia because the type species, H. tholusitabulata Yabe and Hayasaka, 1920, from the l.ower Carboniferous of China, has distinctly dilated septa in the cardinal quadrants and a similar appearing axial structure, whereas other genera of the Auiophyllidae and Cyathopsidae with somewhat similar charac- ters do not have such pronounced cardinal quadrant septal dila- tion. Hill ( 1 956:290) stated that Heterocaninia has “minor septa absent,” but Cotton (1973:95) reported “. . . minors very short and in weathered specimens may not be observed.” H. latt- genheimi has minor septa that do not cross the dissepimentarium in postjuvenile stages, and some specimens have parts of the coral that contained these weathered away. The genus has not been heretofore recorded in North Ameri- can Permian rocks. The size and other characters clearly dis- tinguish H. langenheimi from any previously described Permian coral. The specimens occur in growth positions in “log jam” abun- dance in beds with Gshelia americana. The shape of the cylindri- cal corallum is twisted in many specimens, even as much as 360 in very long specimens, and many specimens exhibit sharp angles (as much as 90 ), suggesting that they may have grown upwards until top heavy, fallen over, and then resumed upward growth. Calices of most specimens face towards the upper bedding plane surfaces. Internal structures, such as the cardinal fossula, are randomly oriented to the external curvature of the corallites, which is in contrast to observations made of some other solitary corals. Etymology. The species is named for Dr. R.L. Langenheim, Jr. Heterocaninia (?) sp. Figures 1 4e— f Transverse Section Description. Corallite solitary, circular to sub- circular, maximum observed diameter 29 mm (dissepimentar- ium missing); major septa number 38 at 20-mm corallite diame- ter, 41 at 22-mrn corallite diameter, all lanceolate and highly dilate, 6 to 7 mm long in tabularium, many touching axial struc- ture and continuous with septal lamellae; cardinal septum short- ened, about 1 mm long in tabularium; cardinal fossula closed; minor septa not observed; dissepimentarium not present due to eroded surface of corallite; axial structure large, filling entire tabularium, composed of highly sinuous medial plate (?) and septal lamellae, many formed from attenuate ends of septa, joined by tabellae; corallite wall not observed. Documentation. UCMP hypotype 37203. Two thin sections and six polished sections from one corallum from UCMP locality D-86! were studied. Fragments of other corallites were present in the rock. Discussion. This coral occurs with Clisiophyllum gabbi Meek, 1864, which has fewer and much thinner septa and a better defined axial structure. The extreme dilation of all the septa and the nature of the axial structure clearly distinguish Hetero- caninia (?) sp. from other species found in the McCloud Lime- stone. It somewhat resembles ? Clisiophyllum sp. B of Rowetl, 1 969 (p. 58, pi. 7, fig. 3) from the Lower Permian of Alaska, but Rowett’s specimen has a greater number of septa and an open cardinal fossula. The generic referral is questionable since the dissepimen- tarium was not preserved and no longitudinal section was ob- tainable. Probably the coral represents a late neanic stage. Fasciculate Rugosa Genus Durhamina Wilson and Langenheim, 1962 Durhamina sublaeve ( Meek, 1 864) Figures 5:4, 4a, 4b; 1 5a-c; 1 6a-e Lithostrotion mamillare (?), Castlenau, (sp.). Meek, 1864:5 ( pars ), pi. 1 , figs. 4, 4a, 4b. L. sublaeve Meek, 1 864:6 (pars), pi. I , figs. 4, 4a, 4b. Lithostrotion mamillare, var . sublaevis Meek, 1864, legend pi. 1 , figs. 4, 4a, 4b. Lithostrotion ? Meek, 1864:7 (not figured). Not Lithostrotion mamillare? of Meek, 1 864, pi. 1 , figs. 3, 3a. External Description. Corallum phaceloid, maximum observed diameter 15 cm; epitheca with distinct, closely spaced rugae; calyx not observed. Transverse Section Description. Corallites circular, diameter 10 to 1 2 mm, touching to as much as 20 mm distant; septa of 2 orders, 1 9 to 23 each, thin, straight, or slightly sinuous; major septa 2.5 to 4 mm long, generally all but 1 or 2 withdrawn from axial structure; minor septa 0.2 to 1 .0 mm long, well developed; dis- sepimentarium regular, 1 to 3 mm wide; dissepiments concentric, angulo-concentric, straight, or herringbone; axial structure aulophylloid where most fully developed, with very sinuous me- dial plate (not everywhere distinguishable from axial tabellae) and an indeterminable number of highly sinuous septal lamellae, generally withdrawn from septa except in some corallites 1 Figure 14. Heterocaninia langenheimi n. sp. and Heterocaninia (?) sp. Parts a through d. H langenheimi n. sp , all parts xl .5 Part a, UCMP holotype 37195, longitudinal section. Parts b through d, UCMP paratype 37 1 96, serial transverse sections (b, c) and longitudinal section (d). Parts e and f. Heterocaninia (?) sp., UCMP hypotype 37203, serial transverse sections, x3. Contributions in Science, Number 337 Wilson: Permian Corals of California 25 Figure 15. Durhamina sublaeve ( Meek, I 864), Heritschioides (?) californiense (Meek, I 864), and Kleopatrina ( Porfirievella) whitneyi n. sp. Parts a through c. D. sublaeve, all parts xl . Part a, MCZ holotype, MCZ Meek collection corallum no. 1 7, figured by Meek ( 1 864, pi. I , fig. 4). Part b, MCZ hypotype, MCZ Meek collection corallum no. 16, labeled " Lithostrotion ” by Meek. Parte, MCZ hypotype, MCZ Meek collection corallum no. 14, labeled " Lilhostrotion ” by Meek. Parts d and e. H. ('.’) californiense. Part d, MCZ lectotype (arrow), paralectotypes, MCZ Meek collection block no. 13, x0.5. Part e, lectotype, detail of Part d, figured by Meek ( 1 864, pi I , fig. 2b), x I Part f. K\P.) whitneyi n. sp., MCZ holotype, MCZ Meek collection corallum no. 15, figured by Meek ( 1 864, pi. I , fig. 3), xl .5. 26 Contributions in Science, Number 337 Wilson: Permian Corals of California Wilson: Permian Corals of California 27 Contributions in Science, Number 337 Figure 16. Durhamina sublaeve (Meek, 1864). All parts x3. Parts a and b, MCZ holotype, MCZ Meek collection corallum no. 17, transverse (a) and 28 longitudinal (b) sections. Part c, MCZ hypotype, MCZ Meek collection corallum no. 1 4, transverse section. Parts d and e, MCZ hypotype, MCZ Meek collection corallum no. 16, longitudinal (d) and transverse (e) sections. (counter?) or 2; medial plate attached to 1 septum present in some corallites without septal lamellae; corallite wall very thin, 0. 1 to 0.2 mm wide. Longitudinal Section Description. Dissepimentarium of 1 to 4 ranks of various sized (none highly inflated) dissepiments; tab- ellae of 2 loosely defined ranks, axial and periaxial, both inclined inwards and upwards to medial plate; tabulae tent-shaped where ranks of tabellae break down. Documentation. Meek’s ( I 864, pi. 1 , figs. 4, 4a, 4b) figured speci- men is recognized here as the holotype in accordance with ICZN article 73(a): “If a nominal species is based on a single specimen, that specimen is the ‘holotype.’ ” The specimen is no. 17 in the MCZ collection and the only one labeled Lithostrotion sublaeve. I prepared two thin sections from it. MCZ specimens nos. 1 4 and 16 are referred here to Durhamina sublaeve. I prepared one thin section of the former and two of the latter. Nos. 1 4 and 1 6 were labeled “ Lithostrotion ?” and apparently were used by Meek for his description (Meek, 1 864:7) under that designation. They are fasciculate, as Meek reported, although a cerioid cor- allum was figured erroneously by him (Meek, 1 864, pi. 1 , figs. 3, 3a) with this name. Discussion. The type locality, as with the other Meek (1864) McCloud Limestone specimens, is presumed to be somewhere in the formation south of the Pit River. Since I did not encounter additional specimens of the species in my field work and since diagnostic fusulinids were not found in the matrix remaining on Meek’s coralla, the exact stratigraphic position is uncertain. In this area, the McCloud Limestone is within fusulinid zone E and zone F. Rowett (1969:43) published a key for eight species of Dur- hamina. D. sublaeve compares closely only with D. cordillerensis (Easton, I960) from the Permian of east-central Nevada and southern California and with D. alaskaensis Rowett, 1969, from the Permian of Alaska. All three species have “axial tabellae not highly cystose at maturity” and “comparatively small corallites with few speta.” D. cordillerensis. type species of the genus, is similar to D. sublaeve, but the former has a more complex axial structure where most fully developed, with more major septa touching it and more lonsdaleoid dissepiments. D. alaskaensis has somewhat larger corallites and a greater number of septa than D. sublaeve. Genus Heritschioides Yabe, 1950 Heritschioides carneyi n. sp. Figures 1 8a-c Diagnosis. A species of Heritschioides characterized by the combination of large corallites, a large number of septa, a simple axial structure, and minor septa generally confined to the dissepi- mentarium. It has larger corallites than H. buttensis, H. sum- mitensis, H. parvum. and H. stevensi. It has shorter minor septa than H. hammani, H. merriami, H. wexoi. H. columbicum. H coogani, H. durhami, H. rowetti, H. washburni, H. wildei, H gavini, and H. smithi. It has a simpler and more symmetrical axial structure than H. hillae, H. moormanensis, H ochocoensis. and H. skinneri. External Description. Corallum phaceloid, hemispheroidai, max- imum observed diameter 18 cm; corallites subparallel, closely spaced to touching; epitheca and calyx not observed. Transverse Section Description. Corallites circular to subcircular, diameter 1 2 to 18 mm, closely spaced, touching to as much as 1 I mm apart; septa of 2 orders, 25 to 29 each, straight to broadly sinuous, generally thin in dissepimentarium (0.1 to 0.2 mm wide), dilate in tabularium (about 0.3 mm at base) becoming attenuate axially; major septa withdrawn from axial structure (except counter? septum in some), 4 to 6 mm long, except short- ened cardinal septum 3 to 5 mm long; minor septa well devel- oped, 1 .3 to 2.4 mm long, generally confined to peripheral part of dissepimentarium, some crossing dissepimentarium, few enter- ing tabularium as short spines; fossula small, open, formed by shortened cardinal septum; dissepimentarium regular or her- ringbone; stereozone developed on some dissepiments separating dissepimentarium from tabularium; axial structure clisiophy 1- loid, generally simple and asymmetrical, relatively small, ellipti- cal to subcircular, dimensions 1 by 3 mm to 3.5 by 4.5 mm; medial plate well developed, straight to sinuous, dilate, width 0.2 to 0.25 mm, length 2 to 4 mm; 2 to 10 sinuously and irregularly developed septal lamellae present; axial tabellae generally few, connecting septal lamellae in somewhat asymmetrical spider- web pattern; corallite wall about 0.2 to 0.3 mm wide. Longitudinal Section Description. Dissepimentarium of I to 5 (generally 3 to 4) steeply dipping ranks of large and small cystose dissepiments; tabellae of 2 zones, axial and periaxial; axial tab- ellae small, cystose, dipping inwards and upwards to medial plate; periaxial tabellae variously developed, about 15 per cm, generally I or 2 ranks sloping inwards and upwards to axial tabellae, but outer ones may be variously inclined to flat; tabulae present where axial structure consists only of medial plate. Documentation. L ACM IP holotype 6374, LACMIP paratype 6375. Six thin sections and 65 polished sections from two coralla from LACMIP locality 4457 were studied. Discussion. H. carneyi differs from other species of the genus that have similar sized corallites and similar numbers of septa by possessing a variable axial structure that generally is extremely simple with few axial lamellae and by possessing minor septa that generally are confined to the dissepimentarium and appear only as spines in the few instances that they reach the tab- ularium. These relationships are shown graphically for this and other species in the genus on Figure 17, which informally and artificially combines species of Heritschioides based on four readily observable characters. H. carneyi may be the oldest species of the genus recognized thus far. All other species in the McCloud Limestone are strat- igraphically higher than H carneyi. as are the species described from Permian rocks in Oregon, Nevada, and most probably Alaska. The exact stratigraphic position of the type species, H. columbicum Smith, 1935, is uncertain however (Wilson, 1980:90). Oekentorp et al. (1978:387) figured a coral from the Upper Permian of Oman and referred it to H. columbicum. It should be examined again for characters of the genus. Wilson (1980) con- sidered reports of Heritschioides from rocks outside western North America to be doubtful. Watkins ( 1973: 1 761 ) cited C.H. Stevens as having identified a coral from the Hirz Mountain Limestone member of the Baird Contributions in Science, Number 337 Wilson: Permian Corals of California 29 30 Contributions in Science, Number 337 Wilson: Permian Corals of California Figure 17. Informal combinations of Heritschioides species based on corallite diameters, numbers of major septa, lengths of minor septa, and symmetry of axial structures. Figure 18. Heritschioides carneyi n. sp. and H coogani n. sp. All parts x3. Parts a through c. H. carneyi n. sp., LACMIP holotype 6374, transverse (a, b) and longitudinal (c) sections. Part d. H. coogani n. sp., LACMIP paratype 6379, longitudinal section. Figure 19. H eritschioides coogani n. sp. All parts x3. Parts a through c, LACM I P holotype 637 6, transverse (a, b) and longitudinal (c) sections. Part d, 32 LACMIP paratype 6379, transverse section. Formation on llirz Mountain as Heritschioides sp. This is below the McCloud Limestone and Pennsylvanian or Permian in age (Watkins, 1 973: 1 743). I have not seen this coral. Etymology. The species is named lor Mr. William Alderman Carney. Heritschioides coogani n. sp. Figures 18d, 1 9a d Diagnosis. A species of Heritschioides characterized by the combination of relatively long and thin septa, a symmetrical and complex axial structure, large corallites, and a large number of major septa. It has thinner minor septa and a more symmetrical axial structure than H. columbicum. It has smaller corallites, a more regular axial plate, and shorter minor septa than H dur- hami. It has larger corallites, more numerous septa, and a more complex axial structure than H. washburni. It has more and better developed septal lamellae than H. wildei. It has thinner and more numerous septa, shorter minor septa, and more numer- ous and less sinuous septal lamellae than //. woodi. External Description. Corallum phaceloid, hemispheroidal, max- imum observed diameter 1 2 cm, height 1 0 cm; corallites parallel to subparallel, very closely spaced (some nearly cerioid); epi- theca and calyx not observed. Transverse Section Description. Corallites circular to subcircular (some angular in part where corallite nearly cerioid), closely spaced, diameter 1 5 to 19 mm; septa of 2 orders, 25 to 33 each, straight to slightly sinuous, thin in dissepimentarium (about 0. 1 mm wide), dilate in tabularium (0.2 to 0.3 mm at base) becoming attenuate axially; major septa touching or slightly withdrawn from axial structure, 6 to 10 mm long (generally 6 to 8), except shortened cardinal septum 4 to 6 mm long; minor septa well developed, 2 to 5.5 mm long, crossing dissepimentarium, many entering tabularium as nubs or dilate septa as much as 1 mm long; fossula small, open, formed by shortened cardinal septum, 1 to 1.5 mm deep; dissepimentarium regular, with dissepiments concentric, angulo-concentric, pseudoherringbone, or concave outwards; stereozone developed on some dissepiments separating dissepimentarium from tabularium; axial structure clisiophyl- loid, large, complex, symmetrical, generally circular, diameters 3 to 5 mm; medial plate well developed, 3 to 5 mm long, thin or dilate, straight to sinuous, connected to counter septum in some corallites; septal lamellae numerous, 1 3 to 26, rather evenly de- veloped opposite one another in most complex axial structures; axial tabellae in 3 to 6 ranks, connecting septal lamellae in sym- metrical spider-web pattern in best developed axial structures; corallite wall about 0.2 mm wide, rarely fused in closely spaced corallites to form joint cerioid-type wall. Longitudinal Section Description. Dissepimentarium wide, of 4 to 6 ranks of steeply dipping cystose dissepiments (transverse sec- tions indicate more ranks in some areas of corallites); tabellae of 2 well-defined zones, axial and periaxial; axial tabellae small to large, cystose to straight, dipping inwards and upwards to medial plate in 1 to 6 ranks; periaxial tabellae in 1 to 3 ranks, straight to gently domed, dipping inwards and upwards to axial tabellae. Documentation. LACMIP holotype 6376, LACMIP paratypes 6377-6379. Six thin sections and 54 polished sections from four coralla from LACMI P locality 1119 were studied. Discussion. H . coogani is comparable to six other species, as can be seen in Figure I 7. //. wildei n. sp. is somewhat similar but has far fewer septal lamellae that are less regularly developed on each side of the axial plate. H durhami n. sp. has larger cor- allites, longer minor septa, and a less regular axial plate. H. rowetti n. sp. has much smaller corallites and fewer septa. //. columbicum (Smith, 1 935), the type species from British Colum- bia, has longer minor septa that are more dilate and a somewhat less symmetrical axial structure. //. washburni ( Merriam, 1 942), from the Permian Coyote Butte Formation of Oregon, has more septa, larger corallites, and a somewhat smaller and simpler axial structure. //. woodi Wilson and Langenheim, 1962, from the Permian (Leonardian) Arcturus Formation of eastern Nevada, has longer minor septa, fewer and more sinuous septal lamellae, and somewhat fewer septa. Etymology. The species is named for Dr. Alan H. Coogan. Heritschioides durhami n. sp. Figures 20a-b Diagnosis. A species of Heritschioides that has larger corallite diameters than any other species in the genus. External Description. Corallum phaceloid and fragmentary, maximum observed diameter 1 I cm; epitheca highly silicified, weathered; calyx not observed. Transverse Section Description. Corallites circular, diameters very large, 1 7 to 34 mm, touching to as much as 7 mm apart; septa of 2 orders, 26 to 29 each, straight to somewhat sinuous, thin in dis- sepimentarium (about 0 I mm), dilate in tabularium (about 0.24 to 0.4 mm); major septa very slightly withdrawn from axial struc- ture (some bent axially near axial structure), 5 to 8 mm long, except shortened cardinal septa (1.7 to 3.2 mm long); minor septa very well developed, 1.7 to 3.2 mm long, extending across dissepimentarium and well into tabularium (about 0.5 mm); fos- sula small, open; dissepimentarium regular, 2 to 4 mm wide; dissepiments generally angulo-concentric, less generally her- ringbone, concentric, or straight; stereozone on axial row of dis- sepiments absent or very inconspicuous; axial structure clisiophy I loid, rather symmetrical, large, circular, diameter 5 to 8 mm; medial plate inconspicuous, not dilate; septal lamellae thin, variously developed, 1 8 to 3 1 , sinuous, curved, or straight, not all traversing axial structure to medial plate; axial tabellae very numerous, generally straight between septal lamellae; cor- allite wall about 0. 1 to 0. 1 5 mm wide, with septa inserted in V- shaped depressions. Longitudinal Section Description. Dissepimentarium wide, of 4 to 10 ranks of steeply dipping, various sized dissepiments; tabellae of 2 zones, axial and periaxial; axial tabellae of various sizes, cystose to elongate, dipping steeply inwards and upwards to me- dial plate in 5 to 6 ranks on each side; periaxial tabellae cystose to straight, sloping inwards and upwards to axial tabellae in 2 to 3 ranks. Documentation. LACM I P holotype 6380. Two thin sections and 13 polished sections from one corallum from LACMIP locality 1 1 25 were studied. Discussion. The corallite diameters of H. durhami are much Contributions in Science, Number 337 Wilson: Permian Corals of California 33 larger than those of any other species yet recorded for the genus. A 20-mm diameter is large for any species of Heritschioides , but one corallite in this corallum is 34 mm in diameter and the others are 20 mm or more. This alone should clearly distinguish it from other species. In addition, the combination of a large number of septa (as many as 29), the great length of the minor septa, the complex and regularly shaped large axial structure with its thin medial plate distinguishes it from H. wildei, which occurs in the same unit and has a similar number of septa, H. skinneri, which is much smaller and has much shorter minor septa, and other spe- cies with a similarly large number of septa. Etymology. The species is named for Dr. J. Wyatt Durham. Heritschioides gavini n. sp. Figures 20c-d, 21 a-c Diagnosis. A species of Heritschioides characterized by the combination of large corallites, a large number of septa, long minor septa, and an asymmetrical axial structure. It has larger corallites, more septa, and longer minor septa than H. sniithi. External Description. Corallum phaceloid, hemispheroidal, max- imum observed diameter 27 cm; corallites subparallel, touching to as much as 1 .5 cm apart; calyx 1 9 mm deep in corallite 24 mm in diameter, with large central axial boss; epitheca not observed. Transverse Section Description. Corallites circular to subcircular, diamters 21 to 27 mm, generally closely spaced or touching; septa of 2 orders, 34 to 36 each, straight to somewhat sinuous, generally thin in dissepimentarium (about 0.2 mm wide), rarely dilate in dissepimentarium (about 0.4 mm wide), dilate in tab- ularium (0.2 to 0.4 mm wide); major septa slightly withdrawn or touching axial structure, 6 to 8 mm long, except shortened cardi- nal septum 3.5 to 4 mm long; minor septa well developed, 3.5 to 5 mm long, crossing dissepimentarium and entering tabularium 0.5 to 2 mm; fossula small, open (rarely closed by neighboring septa), formed by shortened cardinal septum; dissepimentarium regular (rarely lonsdaleoid where 1 or 2 septa fail), generally 3 to 4 mm wide; dissepiments pseudoherringbone, concentric, an- gulo-concentric, or straight; stereozone separating tabularium from dissepimentarium; axial structure clisiophylloid, large, di- ameter 5 to 8 mm, circular to subcircular, complex, generally asymmetrical (rarely symmetrical), with sinuous medial plate (straight in neanic stages), numerous (15 to 30) septal lamellae crossing several ranks of axial tabellae (rarely all structures coated with stereoplasm); corallite wall about 0.2 mm wide. Longitudinal Section Description. Dissepimentarium of 3 to 9 (generally 4 to 5) ranks of steeply dipping small globose and few elongate dissepiments; tabellae of 2 zones, axial and periaxial; axial tabellae of 4 to 5 ranks, steeply dipping inwards and up- wards to axial tabellae, also some flat, horizontal, or sloping in- wards and downwards (rare “clinotabellae”). Documentation. LACMIP holotype 6381, LACMIP paratypes 6382-6388. Seven thin sections and 78 polished sections from eight coralla from LACMIP locality 1285 were studied. Discussion. H. gavini has corallites among the largest in diame- ter in the genus. H durhami also has large corallites ( 1 7- to 34- mm diameters) but fewer septa (26 to 29) as well as other dif- ferences, including an essentially symmetrical axial structure. H. sniithi has corallites much smaller in diameters (15 to 21 mm) and fewer septa (26 to 32), although it also has an essentially asymmetrical axial structure. //. washburni (Merriam, 1942:375, pi. 55, figs. 2, 5, 7, 9, 11, 13) would seem similar according to the description, but examination of the type speci- mens (USN M 1 43427- 143431) confirms that they are smaller (largest corallite diameter about 20 mm), although the numbers of major septa (33 to 35) are close, and the axial structure on some specimens is rather asymmetrical but less so than in H. gavini. The species is not closely comparable with any other species. Preservation of the corallites is not good. The beds have been partly metamorphosed, and there is some crushing in most cor- allites. Omphalotrochus sp. was barely discernible as a “ghost” in the same unit. Cerioid rugose corals also were present but impossible to collect. Etymology. The species is named for Mr. Peter T. Gavin. Heritschioides hammani n. sp. Figures 22a-e Diagnosis. A species of Heritschioides characterized by the combination of relatively small corallites, a simple axial struc- ture, long minor septa, and a small number of major septa. It has smaller corallites and a simpler axial structure than H merriami and H. wexoi and straighter septal ends than H wexoi. External Description. Corallum phaceloid, maximum observed diameter 10 cm; corallites parallel, closely spaced; epitheca and calyx not observed. Transverse Section Description. Corallites circular, diameter 1 0 to 1 2 mm, touching or as much as 5 mm apart; septa of 2 orders, 1 9 to 23 each, dilate in tabularium (many also dilate in dissepimen- tarium), straight to slightly sinuous; major septa 2.5 to 4.5 mm long, except cardinal septum about 2 mm long, generally slightly withdrawn from axial structure, except counter septum, which may touch axial structure in some; minor septa well developed, 0.5 to 1 .5 mm long, generally extending slightly into tabularium as spines; fossula small, open, about 1 mm deep; dissepimen- tarium regular, 0.5 to 1.0 mm wide; dissepiments concentric or pseudoherringbone; false wall stereozone present; axial structure clisiophylloid, elongate, with only 1 or 2 septal lamellae; medial plate well developed, nearly straight, as much as 4 mm long, connected in some corallites to counter septum; axial tabellae number about 2 or 3 ranks on each side of medial plate, rather symmetrically developed; corallite wall 0. 1 to 0.2 mm wide. Longitudinal Section Description. Dissepimentarium narrow, of 1 to 3 (rare) ranks of small, steeply dipping dissepiments; tabellae of 2 zones, axial and periaxial; axial tabellae of 1 to 3 ranks, elongate, steeply inclined inwards and upwards to medial plate; periaxial tabellae generally of 1 or 2 ranks, horizontal or gently inclined inwards and upwards to axial tabellae. Documentation. UCMP holotype 37204. Five thin sections and 43 polished sections from one corallum from UCMP locality D-828 were studied. Discussion. This corallum is stratigraphically the highest colo- nial rugose coral found in the McCloud Bridge section. It has been so highly altered that many of the corallites are not pre- 34 Contributions in Science, Number 337 Wilson: Permian Corals of California served in detail, and most of the limestone matrix has been al- tered to masses of radiating black crystals that color entire beds black like some limestone beds of the Hirz Mountain section. A few corallites are well enough preserved to observe the structure. The simplicity of the axial structure, the protrusion of the minor septa into the tabularium as spines, the size, and the num- ber of septa readily distinguish H. hammani from other species, except H. merriami n sp. and H. wexoi n. sp. H. hammani is differentiated in the discussions following the descriptions of these. Etymology. The species is named for Mr. Howard Hamman. Heritschioides merriami n. sp. Figures 23a-c Diagnosis. A species of Heritschioides characterized by the combination of moderately large corallites, a moderately com- plex and symmetrical axial structure, relatively few major septa, long minor septa, and relatively straight major septal ends. 1 1 has larger corallites and a more complex axial structure than H hammani. It has straighten septal ends and a somewhat more symmetrical and complex axial structure than H. wexoi. External Description. Corallum phaceloid, hemispheroidal, max- imum observed height 20 cm, width 1 1 cm; corallites parallel, closely spaced, many touching; epitheca poorly preserved, show- ing some transverse wrinklings; calyx not observed. Transverse Section Description. Corallites circular to subcircular, diameter 1 3 to 15 mm, very closely spaced; septa of 2 orders, 22 to 23 each, straight to somewhat sinuous, thin in dissepimen- tarium (about 0.1 mm wide), dilate in tabularium (about 0.2 mm at base), thinning axially; major septa 3.5 to 6 mm long, touching or slightly withdrawn from axial structure, except cardinal sep- tum 2.5 to 3 mm long; minor septa well developed, 2 to 3.5 mm long, crossing dissepimentarium and entering tabularium as much as 0.5 mm; fossula small, open, about 1 mm deep; dissepi- mentarium regular, dissepiments angulo-concentric, straight, pseudoherringbone; thin stereozone formed on some dissepi- ments at tabularium interface; axial structure clisiophylloid, symmetrical, occupying most of tabularium beyond major septa, elliptical to circular, dimensions 3.5 by 3.5 mm, 2 by 3 mm, 2.5 by 4 mm; medial plate well developed, straight to somewhat sinuous, thin to slightly dilate, length 2.0 to 2.5 mm; septal lamellae number 8 to 14, generally straight to somewhat sinuous, rather regularly developed, many opposite one another on each side of medial plate; corallite wall about 0.1 mm wide. Longitudinal Section Description. Dissepimentarium of 1 to 6 steeply dipping ranks of large and small cystose dissepiments; tabellae of 2 zones, axial and periaxial; axial tabellae small, cystose, dipping inwards and upwards to medial plate; periaxial tabellae number about 17 in 1 vertical cm, highly varied in in- clination, flat, gently domed, generally dipping inwards and up- wards to axial tabellae, generally of 1 rank. Documentation. LACMIP holotype 6389. Three thin sections and 15 polished sections from one corallum from LACMIP lo- cality 1 120 were studied. Discussion. H. merriami has corallite diameters generally greater than 10 mm, a maximum number of major septa fewer than 25, minor septa that extend into the tabularium, and a symmetrical axial structure. As can be seen from Figure 17, it therefore is comparable only with H hammani and H. wexoi. H hammani has a much smaller corallite diameter (10 to 12 mm) and a much simpler axial structure than H. merriami. It is dis- tinguished from H. wexoi in the discussion after that species description. Etymology. The species is named for Dr. Charles W. Merriam. Heritschioides rowetti n. sp. Figures 23d, 24a-c Diagnosis. A species of Heritschioides characterized by the combination of relatively small corallites, relatively few septa, long minor septa, and a symmetrical axial structure. It has smaller corallites and fewer septa than H. columbicum, H coogani. H. durhami, H. washburni. H. wildei, and H. woodi. External Description. Corallum phaceloid, large, exact shape not seen, maximum size of largest fragment 33 cm long, 5 cm wide; corallites closely spaced or touching in some coralla, as much as 2 cm apart in others; epitheca with low, rounded, transverse growth wrinklings, irregularly spaced 1 to 3 mm apart; calyx not observed. Transverse Section Description. Corallites circular to subcircular, diameter 9 to 1 3 mm, touching to as much as 7 mm apart; septa of 2 orders, 17 to 24 each, generally somewhat dilate in tab- ularium, straight to slightly sinuous; major septa 3 to 5.5 mm long, except cardinal septum 2.4 to 3.2 mm long, generally very close to or touching axial structure, rare counter septa connected with medial plate; minor septa well developed, 1 to 1 .5 mm long, crossing dissepimentarium and generally extending into tab- ularium as dilate spines; fossula small, open, about 2 mm deep; dissepimentarium regular, 1 .3 to 2. 1 mm wide; dissepiments con- centric or herringbone (very rarely lonsdaleoid); false wall ster- eozone present; axial structure clisiophylloid, circular to elongate, generally symmetrical appearing, with 4 to 1 1 septal lamellae; medial plate generally well developed, slightly dilate (rarely thin), straight to somewhat sinuous, 2 to 4 mm long, connected in some corallites to counter septum; axial tabellae in 2 to 4 ranks on each side of medial plate, outer rank dilate in some corallites; corallite wall 0. 1 to 0.2 mm wide. Longitudinal Section Description. Dissepimentarium moderately wide, of 1 to 3 ranks of mixed small and large, steeply dipping dissepiments; tabellae of 2 zones, axial and periaxial; axial tab- ellae of 2 to 5 ranks, elongate, steeply inclined inwards and up- wards to medial plate, not uniform in size; periaxial tabellae generally of 1 or 2 ranks, horizontal or gently inclined inwards and upwards to axial tabellae or sagging or gently inclined out- wards and upwards to dissepimentarium. Documentation. LACMIP holotype 6390, LACMIP paratypes 6391-6399. Four thin sections and 57 polished sections from ten coralla from LACMIP locality 1 1 25 were studied Discussion. H. rowetti superficially resembles H parvum Ste- vens, 1967, and H. buttensis Stevens, 1967, both from the Leonardian Arcturus Formation of White Pine County, Nevada. It differs from the former species by having longer minor septa and from the latter species by having a narrower dissepimen- Contributions in Science, Number 337 Wilson: Permian Corals of California 35 Figure 20. Heritschioides durhami n. sp. and H. gavini n. sp. All parts x3. 36 Parts a and b. H durhami n. sp., LACM1P holotype 6380, transverse (a) and longitudinal (b) sections. Parts c and d. H. gavini n. sp., LACM1 P holotype 6381, transverse sections. Figure 21. Heritschioides gavini n. sp. All parts x3. Part a, LACMIP paratype 6383, transverse section. Parts b and c, LACMIP holotype 6381, longitudinal sections. Figure 22. Heritschioides hammani n. sp„ UCMP holotype 37204, transverse sections (a-c) and transverse and longitudinal sections (d, e). All parts 38 x3. Figure 23. Heritschioides merriami n. sp. and H rowetti n. sp. All parts x3. Parts a through c. H merriami n. sp., LACMIP holotype 6389, transverse (a) and longitudinal (b, c) sections. Part d. H. rowetti n. sp., LACMIP paratype 6391, transverse section. 39 Figure 24. Heritschioides rowetti n. sp. All parts x3. Parts a and b, LACMIP holotype 6390, transverse (a) and longitudinal (b) sections. Part c, 40 LACMIP paratype 6391, longitudinal section. tarium, a smaller axial structure with less sinuous axial lamellae, and a better defined axial plate. There is, however, considerable variation in the corallites of H rowetli in every corallum exam- ined. H. rowetti has smaller coral I ite diameters than other spe- cies in its group as shown in Figure 1 7. Etymology. The species is named for Dr. Charles L. Rowett. Heritschioides skinneri n. sp. Figures 25a-c Diagnosis. A species of Heritschioides characterized by the combination of a moderately asymmetrical axial structure with a relatively well-defined medial plate, a relatively narrow dissepi- mentarium, large corallites, and a large number of relatively short septa. 1 1 has a smaller, less complex, and more symmetrical axial structure than H. hillae. It has a more symmetrical and a more complex axial structure and a narrower dissepimentarium than H. ochocoensis. It has shorter septa and a more symmetrical axial structure with a better defined medial plate than //. moormanensis. External Description. Corallum phaceloid, maximum observed diameter 1 1 cm; corallites parallel, closely spaced to 3 cm apart; epitheca and calyx not observed. Transverse Section Description. Corallites circular to subcircular, diameter 1 to 1 4 mm, touching to as much as 5 mm apart; septa of 2 orders, 26 to 30 each, some dilate in tabularium, straight to slightly sinuous; major septa 3.5 to 6 mm long, except cardinal septum 2.5 to 4.5 mm long, generally close to but slightly with- drawn from axial structure; minor septa thin, 0.5 to 1.5 mm long, generally confined to dissepimentarium, rarely appearing in tab- ularium as short spines; fossula small, open, about 0.5 mm deep; dissepimentarium regular, ! to 2 mm wide; dissepiments con- centric or pseudoherringbone; false wall stereozone generally not present, rarely slightly developed; axial structure clisiophylloid, large, asymmetrical, with numerous septal lamellae; medial plate variously developed, highly sinuous to nearly straight, slightly dilate to not dilate, 1.5 to 4 mm long; axial tabellae abundant in many corallites, generally asymmetrically devel- oped in spider-web pattern; corallite wall about 0.1 mm wide. Longitudinal Section Description. Dissepimentarium narrow, of 1 to 3 ranks of steeply dipping, small dissepiments; tabellae of 2 zones, axial and periaxial; axial tabellae of 2 to 3 ranks, elongate, steeply inclined inwards and upwards to sinuous medial plate, periaxial tabellae of 1 to 3 ranks, gently inclined inwards and upwards to axial tabellae. Documentation. UCMP holotype 37205. Three thin sections and 32 polished sections from one corallum from UCMP locality D-804 were studied. Discussion. H. skinneri is similar in superficial appearance to the type species of the genus, H. columbicum from British Co- lumbia, Canada. In the latter species, however, the minor septa generally extend into the tabularium, whereas in H. skinneri they are generally confined to the dissepimentarium. H. hilae Wilson and Langenheim, 1962, has a larger and more asymmetrical axial structure than H. skinneri. H. ochoeoenis (Merriam, 1942), from the Coyote Butte Formation of Crook County, Oregon, has a simpler axial structure and wider dissepimentarium. H. moor- manenis Stevens, 1967, from the Arcturus Formation of White Pine County, Nevada, has septa that generally touch the axial structure and a much more asymmetrical axial structure in which the medial plate is so sinuous as to be generally unrecog- nizable. The other nominal species are not closely comparable. Etymology. The species is named for Dr. John W. Skinner. Heritschioides smithi n„ sp. Figures 25d, 26a-c Diagnosis. A species of Heritschoides characterized by the com- bination of relatively short minor septa, a relatively small and asymmetrical or simple axial structure, relatively small cor- allites, and relatively few septa. It has shorter minor septa, a smaller axial structure, somewhat smaller corallites, and some- what fewer septa than H. gavini. External Description. Corallum phaceloid, hemispheroidal, max- imum observed diameter 1 0 cm; corallites parallel to subparallel, very closely spaced or touching; epitheca and calyx not observed. Transverse Section Description. Corallites circular to subcircular, diameter 1 5 to 2 1 mm, touching to as much as 4 mm apart; septa of 2 orders, 26 to 32 each, straight to slightly sinuous, thin in dissepimentarium (0.1 to 0.2 mm wide), dilate in tabularium (0.2 to 0.3 mm wide); major septa 6.5 to 7.5 mm long, except shortened cardinal septum 4.5 to 5 mm long; minor septa well developed, 3 to 4 mm long, generally extending across dissepi- mentarium into tabularium as spines, fossula small, open; dis- sepimentarium regular, wide; dissepiments, pseudoherringbone, straight, concentric, or angulo-concentric; false wall stereozone present; axial structure clisiophylloid, variable, simple or asym- metrical where more complex (about 2.5 by 3.5 mm in latter); medial plate slightly dilate (0.1 mm), 3 to 4 mm long, sinuous, connected in some corallites to counter septum; septal lamellae irregularly developed, generally short, 2 to 6 per corallite, not touching septa; corallite wall aboout 0.2 mm wide. Longitudinal Section Description. Dissepimentarium 2.5 to 4 mm wide, composed of 2 to 6 ranks of steeply dipping, small and cystose or large and elongate dissepiments; tabellae of 2 poorly differentiated zones: axial and periaxial; axial tabellae tent- shaped, in 1 to 3 ranks; periaxial tabellae horizontal or gently sloping inwards and upwards to axial tabellae (rarely sloping inwards and down), in I to 2 ranks, 1 5 to 20 tabellae per cm. Documentation. UCMP holotype 37206, FACMIP paratype 6400. Six thin sections and 46 polished sections from the holo- type corallum from UCMP locality D-840 and two thin sections and 24 polished sections from the paratype corallum from LACM1P locality 1 1 24 were studied. Discussion. The only other species of Heritschioides with large corallites (more than 1 0 mm diameter), a large number of major septa (more than 25), minor septa that generally extend into the tabularium, and an asymmetrical (some corallites) axial struc- ture is the type species, H. columbicum from British Columbia. H. columbicum. however, has a large, well-defined axial structure with numerous septal lamellae and axial tabellae. H. smithi has a curiously deformed looking axial structure that is composed in many corrallites of simply a twisted medial plate with a few, very short, asymmetrically placed septal lamellae in the simplest forms. Some corallites of H. smithi develop a somewhat more Contributions in Science, Number 337 Wilson: Permian Corals of California 41 Figure 25. Heritschioides skinner/ n. sp. and H. smithi n. sp. All parts x3. 42 Parts a through e. H. skinneri n. sp., UCMP holotype 37205, transverse (a, b) and longitudinal (c) sections. Part d. H. smithi n. sp., UCMP holotype 37206, transverse section. / Figure 26. Heritschioides smithi n. sp. All figures x3. Part a, UCMP holotype 37206, longitudinal section. Parts b and c, LACMIP paratype 6400, longitudinal (b) and transverse (c) sections. 43 Figure 27. Heritschioides stevensi n. sp. and H. wexoi n. sp. All parts x3. 44 Parts a and b. H. stevensi n. sp., UCMP holotype 37207, transverse (a) and longitudinal (b) sections. Part c. H. wexoi n. sp., LACM IP holotype 6401 , transverse section. complex axial structure with longer septal lamellae joined by a few axial tabellae, but these arc much more asymmetrical than the axial structure of H. columbicum. H. smithi's differences with H gavini are discussed following the description of that species. The holotype corallum is somewhat recystallized, and the structures are more apparent when studied in polished than in thin section. This has made them dillicult to photograph satis- factorily, but the paratype slides photograph better. Etymology. The species is named for D. Stanley Smith. Heritschioides stevensi n. sp. Figures 27a-b Diagnosis. A species of Heritschioides characterized by the combination of small corallites, short minor septa, a small num- ber of major septa, a relatively small and simple axial structure, and a relatively narrow dissepimentarium. It has a smaller and simpler axial structure and a somewhat narrower dissepimen- tarium than H. parvum. External Description. Corallum phaceloid, hemispheroidal, max- imum observed diameter 20 cm; epitheca and calyx not observed. Transverse Section Description. Corallites circular to subcircular, diameter 8 to 1 1 mm, touching or as much as 1 5 mm apart; septa of 2 orders, 1 8 to 22 each, straight or slightly sinuous, thin throughout; major septa 2.0 to 3.5 mm long, except cardinal septum 2.3 to 3.0 mm long, withdrawn from axial structure ex- cept counter in some corallites continuous with medial plate; minor septa short, 0.5 to 1.0 mm long, absent in parts of some corallites, generally confined to tabularium; fossula small; dis- sepimentarium regular, 0.5 to 1.9 mm wide; dissepiments con- centric, angulo-concentric, herringbone, or pseudoherringbonc; axial structure clisiophylloid, asymmetrical, very simple, elon- gate, 1.5 to 2.5 mm long, about 1 .0 mm wide, composed of medial plate, septal lamellae, and axial tabellae; medial plate thin, straight to somewhat sinuous, attached to counter septum in some corallites; septal lamellae short, irregularly developed, 1 to 4 in number, corallite wall about 0. 1 mm wide. Longitudinal Section Description. Dissepimentarium of 1 to 4 ranks of dissepiments, various sizes and various inclinations; elongate dissepiments generally gently inclined, small cystose dissepiments generally steeply inclined; tabellae of 2 poorly de- fined ranks, axial and periaxial; axial tabellae of 1 to 2 ranks, steeply inclined inwards and upwards to medial plate; periaxial tabellae striaght to slightly doomed, more gently inclined in- wards and upwards to periaxial tabellae. Documentaiton. UCMP holotype 37207. Two thin sections and 34 polished sections from one large corallum from UCMP lo- cality D-862 were studied. Discussion. Only one other species, H. parvum Stevens, 1976, from the Arcfurus Formation of White Pine County, Nevada, is comparable to H. stevensi in corallite diameter, numbers of septa, length of minor septa, and simplicity of the axial structure. //. parvum may be distinguished readily because it has dilate major septa in the tabularium, a more complex axial structure where it is most highly developed (more axial tabellae), and minor septa better developed and more consistently present throughout the corallites, a few of which even extend into the tabularium. Etymology. The species is named for Dr. Calvin FI. Stevens. Heritschioides wexoi n. sp. Figures 27c, 28a-e Diagnosis. A species of Heritschioides characterized by the combination of a large number of septa, long minor septa, moder- atley large corallites, a moderately complex axial structure, and axial ends of many septa twisted in vortex fashion. It has larger corallites and a more complex axial structure than H hammani. It has more twisted septal ends and a somewhat less symmetrical and less complex axial structure than H. merriami. External Description. Corallum phaceloid, maximum diameter observed 1 5 1 cm, height 46 cm; corallites parallel to subparallel, closely spaced; epitheca with faint rugae; calyx with axial boss, not well preserved Transverse Section Description. Corallites circular, diameter 1 0 to 1 6 mm, touching or as much as 6 mm apart; septa of 2 orders, I 8 to 25 each, dilate in tabularium, major septa 3.5 to 4.5 mm long, except cardinal septum about 3 to 3.5 mm long (some cardinal septa not shortened), straight to sinuous, axial ends attenuate in many, twisted in axial vortex fashion to axial structure, with- drawn and not curved in others; minor septa well developed, 1 .25 to 2 mm long, generally crossing dissepimentarium, many ex- tending into tabularium as dilate spines 0.25 to 0.5 mm long; fossula small, open, inconspicuous or absent in some; dissepimen- tarium regular, 1 to 1.5 mm wide, rarely to 3 mm wide; dissepi- ments concentric, angulo-concentric, or pseudoherringbone; false wall stereozonc present; axial strucutre clisiophylloid, highly varialbc, ranging from medial plate with I or 2 septal lamellae to complex symmetrical or asymmetrical form, at- tached to major septa in some, unattached in others; medial plate generally sinuous, slightly dilate, I to 3.5 mm long; axial tabellae in to 3 ranks on each side of medial plate, variably developed, in some combined with septal tabellae forming “spider web"; cor- allite wall about 0.2 mm wide. Longitudinal Section Description. Dissepimentarium moderately wide, of 2 to 5 ranks of small, steeply dipping dissepiments; tabellae of 2 poorly defined zones, axial and periaxial taballae of 3 to 4 ranks, tent-shaped, sloping inwards and upwards to medial plate; periaxial tabellae of 1 to 2 ranks, horizontal (few) or in- clined inwards and upwards to axial tabellae. Documentation. LAC Mil-* holotype 6401, I. ACM IP paratypes 6402-641 I . Eight thin sections and 86 polished sections from 1 1 coralla from LACMIP locality 5316 were studied. Discussion. H. wexoi resembles only H. hammani and H. mer- riami.These types have corallite diameters generally greater than 1 0 mm, 25 or fewer major septa, and minor septa that generally extend into the tabularium, a combination of characters that distinguishes these from other species of the genus. H. wexoi has widely variable axial structures, some of which are attached to the twisted axial ends of major sept. H merriami differs in having asymmetrical, complex axial structure and no twisted septal ends. H hammani has smaller corallites (10-12 Contributions in Science, Number 337 Wilson: Permian Corals of California 45 Figure 28. Heritschioides wexoi n. sp. All parts x3. Parts a and c. LACM1P holotype 6401 , longitudinal sections. Parts b, d, and e, LACMIP paratype 46 6402, longitudinal (b, e) and transverse (d) sections. mm), a simpler axial structure, and lacks the twisted septa of H. wexoi. Etymology. The species is named for Mr. John B. Wexo. Heritschioides wildei n. sp. Figures 29a-e Diagnosis. A species of Heritschioides characterized by the combination of large corallites, numerous septa, relatively short and thin minor septa, and a symmetrical axial structure with a well-defined medial plate. It has shorter and somewhat thinner minor septa and a better defined medial plate than H colum- bicum. It has a broader dissepimentarium and generally more dilate major septa in the tabularium than H coogani It has generally smaller corallites, a somewhat narrower dissepimen- tarium, and a more symmetrical axial structure with a better defined medial plate than H. durhami. It has larger corallites and more septa than H. rowetti. It has fewer septa and a more sym- metrical axial structure than H washburni. It has shorter and thinner minor septa, fewer herringbone dissepiments, and a somewhat shorter medial plate than H. woodi. External Description. Corrallum phaceloid, hemispheroidal, maximum observed diameter 12 cm; epitheca badly preserved, with some transverse wrinklings; calyx deep with walls sloping nearly vertically inwards to depth of about 1 .2 cm; calicular floor crossed by prominent septa; axial boss central, large, inversely conical, diameter about 4 mm, height about 4 mm. Transverse Section Description. Corallites circular, diameter 14 to 1 8 mm, touching to as much as 5 mm apart; septa of 2 orders, 24 to 30 each, straight to somewhat sinuous, thin in dissepimen- tarium (about 0.1 mm), dilate in tabularium (about 0.2 to 0.3 mm); major septa generally slightly withdrawn from axial struc- ture, 4.5 to 6.9 mm long, except shortened cardinal septa gener- ally 0.5 to ! .0 mm shorter; minor septa well developed, 1 .5 to 3.5 mm long, generally extending across dissepimentarium into tab- ularium about 0.5 mm, dilate in tabularium; fossula small, open, 0.5 to 1.0 mm deep; dissepimentarium regular, 1.5 to 2.5 mm wide; dissepiments herringbone near corallite wall, otherwise concentric, angulo-concentric, or straight; stereozone on row of dissepiments separating dissepimentarium from tabularium thin; axial structure clisiophylloid, large, circular to subcircular, diameter 3.5 to 5.0 mm; medial plate well developed, dilate and nearly straight to thin and rather sinuous, about 1.5 to 4.0 mm long, 0. 1 to 0.2 mm wide where dilate; septal lamellae thin, vari- ously developed, 1 1 to 1 9, sinuous to straight, unequal number on each side of medial plate; axial tabellae numerous, intercepting septal lamellae at near right angles, straight, curved, or slightly sinuous (rare); corallite wall about 0.2 mm wide. Longitudinal Section Description. Dissepimentarium wide, of 4 to 10 ranks of rather steeply dipping, various sized dissepiments; tabellae of 2 zones, axial and periaxial; axial tabellae of various sizes, cystose, dipping steeply inwards and upwards to medial plate in 5 to 8 ranks on each side; periaxial tabellae generally sloping gently inwards and upwards to axial tabellae in 1 to 4 ranks. Documentation. UCMP holotype 37208, LACMIP paratypes 6412-6413. Five thin sections and 42 polished sections from three coralla from LACMIP locality 1 125 (paratypes) and UCMP locality D-852 (holotype) were studied. Discussion. H. wildei occurs in the same beds as H rowetti, but the latter has much smaller corallites and fewer septa. H. wildei is closest to H woodi Wilson and Langenheim, 1962, from the Leonardian Arcturus Formation of White Pine County, Nevada, but the latter species has longer and more dilate minor septa in the tabularium, a dissepimentarium composed of a greater num- ber of herringbone dissepiments, and an axial structure that is generally more symmetrical in appearance with a straighter me- dial plate. H washburni (Merriam, 1942) from the Permian Coyote Butte Formation of Crook County, Oregon, has a greater number of septa, generally less well-developed minor septa, and a generally less symmetrical axial structure. H. columbicum (Smith, 1 935), from British Columbia has longer minor septa. H. coogani has a narrower dissepimentarium. Etymology. The species is named for Dr. Garner L. Wilde. Heritschioides (?) californiense Meek, 1 864 Figures 5:2b; I 5d-e; 33b-d; 33e (?) Lithostrotion? Californiense Meek, 1864, p. 6, pi. I, fig. 2b (and (?) 2, 2a, 2c). External Description. Corallum dendroid (?), maximum ob- served diameter 22 cm; epitheca apparently smooth except for regularly spaced rugae; calyx deep with walls sloping steeply downwards and inwards to relatively horizontal floor; axial boss prominent, central; septa present in calyx except on axial boss. Transverse Section Description. Corallites generally circular, di- ameter 1 7 to 22 mm, touching or short distances apart; septa of 2 orders, 26 to 34 each, dilate, somewhat thicker in tabularium; major septa 4 to 9 mm long, most somewhat withdrawn from axial structure, few continuous with septal lamellae; minor septa 2 to 4 mm long, extending into tabularium in smaller corallites, confined to dissepimentarium in larger corallites; dissepimen- tarium regular, as much as 5 mm wide; dissepiments concentric, angulo-concentric, pseudoherringbone, herringbone; false wall stereozone poorly developed; axial structure aulophylloid, very irregular, composed of a loosely arranged group of elements in which the medial plate (if present), septal lamellae, and periaxial tabellae not readily distinguished; corallite wall 0.1 to 0.3 mm wide. Longitudinal Section Description. Dissepimentarium of 3 to 5 ranks (transverse section shows possible 12) of various size dis- sepiments (small and cystose to large and highly elongate), steep- ly dipping; tabellae of 2 rather irregular zones; axial tabellae slope inwards and upwards to septal lamellae (and medial plate?), apparently more steeply inclined in smaller corallites; periaxial tabellae horizontal or sloping gently inwards and up- wards (rarely the reverse) to axial tabellae. Documentation. Three thin sections, one clearly showing a mother and daughter corallite, were prepared and studied from the corallum in the MCZ collection numbered I 3. This corallum was figured by Meek ( 1 864, pi. 1 , fig. 2b) and is here designated the lectotype. A label glued to the block reading “No. 13 Contributions in Science, Number 337 Wilson: Permian Corals of California 47 Figure 29. Heritschioides wildei n. sp. All parts x3. Parts a through c, UCMP holotype 37208, transverse (a) and longitudinal (b, c) sections. Parts d and 48 e, LACMIP paratype 6412, transverse (d) and longitudinal (e) sections. Lithostrotion Californiense, Meek” presumably is written in Meek’s hand. A dozen or so other corallites on this block may be the same taxon but were not sectioned. The type specimens presumably were collected in the McCloud Limestone somewhere south of the Pit River. The ab- sence of Pseudoschwagerina robusta in the exposed matrix sug- gests that they came from higher in the section than the basal 50 feet, where this fusulinid is very abundant. The corals therefore were collected from either fusulinid zones E or F. Discussion. A transverse section of one coral 1 ite in the type corallum suggests the presence of both cardinal and counter open tabular fossulae, but these may be the results of folds in the axial ends of the mesenteries that caused septa to be formed that were doubled back on themselves at the tips. A transverse section of another corallite in the same corallum suggests the presence of an open tabular cardinal fossula. A small transverse section of a single corallite (Fig. 33e) from 210 feet above the base of the Potter Ridge section (UCMP loc. D-836) is very similar in many respects to this species and does have an open tabular fossula and other features questionably like those of Heritschioides I did not collect specimens of this species from the Bayha section (pre- sumably the type area) and, until additional coralla are avail- able, the generic assignment is problematic. This species is unlike any other nominal species that I have examined. Genus Mccloudius n. gen. Diagnosis. Mccloudius is a phaceloid coral with two orders of septa, a fossula formed by a shortened cardinal septum, a clisiophylloid axial structure, a well-developed regular dissepi- mentarium that becomes peripherally lonsdaleoid in late ephebic stage, and tabellae of two ranks, both of which slope inwards and upwards. Type Species. McCloudius fluvius n. sp. Description. As for the type and only known species. Discussion. Mccloudius somewhat resembles Lonsdaleia M’Coy, 1849, but the latter has an almost wholly lonsdaleoid dissepimentarium, no fossula, and is a Lower and Middle Car- boniferous genus. Lonsdaleoides Heritsch, 1936, was reported originally from the Lower Permian of the Carnic Alps and subse- quently from the Upper Pennsylvanian or Lower Permian of Japan and Spain. Its peripheral, sporadically developed series of lonsdaleoid dissepiments is similar to that of Mccloudius, but Lonsdaleoides has no fossula and its axial structure is much thickened and unlike that of Mccloudius. The lonsdaleoid dis- sepimentarium of Mccloudius distinguishes it from Herit- schioides Yabe, 1950, which has a regular dissepimentarium. Other genera are not closely comparable to Mccloudius. Etymology. The genus is named for the McCloud River. Mccloudius fluvius n. sp. Figures 30a-f Diagnosis. Because the genus Mccloudius is erected as mono- typic, the diagnoses of the genus and of the type species are identical. External Description. Corallum phaceloid; size, corallite spac- ing, epitheca, calyx not observed. Transverse Section Description. Corallites circular to subcircular, diameters 1 3 to 20 mm, not closely spaced, some touching, others as much as 1 0 mm apart; septa of 2 orders, 25 to 33 each, straight to slightly sinuous, generally thin in dissepimentarium (about 0.7 mm wide), dilate in tabularium (about 0.3 at base) becoming attenuate axially; major septa generally withdrawn from axial structure (few touching), 4 to 6 mm long, except shortened cardi- nal septum 3 to 3.5 mm long, extending into tabularium 3 to 4 mm; minor septa well developed in most corallites, poorly in some, generally entering tabularium as spines 0.5 to 1.0 mm long, where present in dissepimentarium, total lengths are 1 .5 to 3 mm; fossula small, open, formed by shortened cardinal septum, possible depressed tabellae; dissepimentarium variable with on- togeny; ( I ) regular in most corallites with diameters less than 14 to 16 mm, dissepiments herringbone, pseudoherringbone, straight, concentric, angulo-concentric, (2) lonsdaleoid pe- ripherally in corallites with diameters from about 1 5 to 20 mm, width 1.5 to 3.8 mm, not lonsdaleoid at thinnest widths, pe- ripheral 2 to 3 mm lonsdaleoid at 2.5- to 3.8-mm widths, axial 0.2 to 1 .2 mm regular; lonsdaleoid dissepiments axially convex, vari- able sized, 2 to 6 rows, not naotic; stereozone separating tab- ularium from dissepimentarium in larger, some smaller, corallites; axial structure clisiophylloid, variable, subcircular, di- ameter 2 to 6.5 mm, complex, symmetrical to asymmetrical; medial plate 1 to 3 mm long, straight to sinuous, combined with 5 to I 2 septal lamellae (some spines in larger corallites), numerous tabellae; corallite wall 0.2 to 0.4 mm thick, with pronounced V- shaped insertions for septal bases except where lonsdaleoid. Longitudinal Section Description. Dissepimentarium of 1 to 10 rows (5 to 1 0 in larger corallites) steeply dipping ranks of dissepi- ments varying in shape from small and cystose to large and elon- gate; tabellae of 2 zones, axial and periaxial; axial tabellae elongate, dipping inwards and upwards to medial plate, 4 to 5 ranks; periaxial tabellae various developed, 1 0 to 15 per cm, 1 to 3 ranks sloping inwards and upwards to axial tabellae. Documentation. LACMIP holotype 6414, LACMIP paratype 6415, UCMP paratype 37209. Eight thin sections and 39 pol- ished sections from three coralla from LACMIP locality 1 1 14 (holotype 6414, paratype 641 5) and UCMP locality D-836 (par- atype 37209) were studied. Discussion. The lower beds of the McCloud Limestone on Potter Ridge have abundant fasciculate corals, but they generally are highly metamorphosed by the massive intrusion at the base, which also intrudes the lower beds in many places. It was not possible to determine if Mccloudius fluvius is present in the beds below the type locality. Unfortunately also, even though the cor- allites clearly branch, it was not possible to determine the precise size or nature of the coralla because many of them were broken before burial. However, they clearly are fasciculate. There is no other species known to me with which this coral can be compared. Contributions in Science, Number 337 Wilson: Permian Corals of California 49 Figure 30. Mccloudius fluvius n. gen., n. sp. All parts x3. Parts a through c, LACM1P holotype 6414. transverse (a, b) and longitudinal (c) sections. 50 Parts d and e, LACMIP paratype 6415, serial transverse sections. Part f, UCMP paratype 37209, longitudinal section Etymology. Fluvius. the species name, is Latin and means river. Genus Siphonodendron M’Coy, 1849 Siphonodendron hongi n. sp. Figures 3 i a-b Diagnosis. A species of Siphonodendron characterized by the combination of large coral I ites, relatively few major septa, very rare dissepiments, and a columella attached to one septum. It has larger corallites, fewer major septa, far fewer dissepiments, and a columella attached to fewer septa than S. paueiradialis. External Description. Corallum phaceloid, external shape not observed, maximum fragmentary diameter 4.5 cm; corallites cyl- indrical to subcylindrical, parallel, touching to as much as 4 mm apart, many elongated laterally to remain in contact with neigh- boring corallite; calyx, epitheca not observed. Transverse Section Description. Corallites circular to subcircular, 5.5 to 7 mm wide, touching to 4 mm apart (generally closer), some attached to nearby corallites by extensions of wall about 5 mm long, 7 mm wide (no internal connections between cor- allites); septa of 2 orders, 14 to 17 each, straight to somewhat curved, thin; major septa 1 to 2 mm long, thin, withdrawn from columella except 1 (counter?) attached; minor septa rudimen- tary spines, 0.2 to 0.5 mm long; dissepimentarium absent or pre- sent as 1 row; axial structure a columella, simple, elongate, from 0.8 mm long by 0.3 mm wide to I . I mm long by 0.6 mm wide, attached to 1 septum (counter); corallite wail 0. 1 to 0.3 mm wide. Longitudinal Section Description. Columella central; tabulae flat, gently sagging, or gently inllated, inclined inwards and upwards to columella, outer ends straight, rarely slightly downturned, 1 9 per cm; tabellae absent. Documentation. UCMP holotype 37210. Two thin sections and 27 polished sections from one corallum from UCMP locality B-4837 were studied. Discussion. Lithodendron paueiradialis M’Coy, 1844, the type species for Siphonodendron, from the Lower Carboniferous (Mississippian) of Ireland (the type locality), also has been re- ported from England, Scotland, Russia, and China (Hill, 1940; 170). Hill (1940:165) synonymized Siphonodendron with Lithostrotion Fleming, 1 828, which has a cerioid type species. 1 prefer to use Siphonodendron rather than to call this species a “fasciculate Lithostrotion .” The type species of Siphonodendron has corallites 4 to 5.5 mm in diameter, 18 to 22 major septa, a dissepimentarium, and a columella attached generally to both the counter and cardinal septa. These features clearly distinguish it from S. hongi. S. pauciradiaius and S. hongi are, however, remarkably similar at the generic level, both having the lateral attachments of some corallites to others seen in cross section. Unfortunately, the holo- type of 5'. hongi occurred in rock near an intrusion and is highly silicified, with well-developed quartz crystals in some vuggy places, so that features of all the corallites are not preserved. I know of no other Permian species with which S. hongi can be compared. This single corallum appears to be the first record of Siphonodendron in the Permian. Etymology. The species is named for Mr. Chang Ki Hong. Genus Yatsengia Huang, 1932 Yatsengia fletcheri n. sp. Figures 3 1 c-d Diagnosis. A species of Yatsengia characterized by the combina- tion of large corallites, a large number of septa, and a relatively broad and continuous disepimentarium. It has larger corallites and more major septa than Y. kenneyi n. sp. It has larger cor- allites and a broader, more continuous dissepimentarium than Y. scheetzi n. sp. It has larger corallites and more septa than species of the genus described from Asia. External Description. Corallum phaceloid, maximum observed diameter 5 cm; corallites subparallel, closely spaced; epitheca and calyx not observed. Transverse Section Description. Corallites circular to subcircular, diameter 1 2 to 15 mm, closely spaced, touching to as much as 3 mm apart; septa of 2 orders, 24 to 27 each, thin in dissepimen- tarium, dilate in tabularium; major septa confluent with septal lamellae, 4 5 to 5.5 mm long, about 0.5 mm wide at periphery of tabularium; minor septa short, 0.7 to 1.5 mm long, confined to dissepimentarium except axial ends dilate and join laterally with major septa to form false wall; dissepimentarium regular, lacking in parts of some corallites; dissepiments straight, concentric, an- gulo-concentric, pseudoherringbone; axial structure aulophyl- loid, large; septal lamellae sinuous, confluent with major septa, generally thin, rarely dilate, number equal to or fewer than num- ber of major septa; false wall stereozone as much as 1 mm wide, formed of dilate axial ends of minor septa and dilate peripheral ends ot major septa in tabularium, with ends of minor septa either exposed axially to tabularium or covered by extensions of major septa; corallite wall about 0.1 mm wide, with bases of major and minor septa fitted into V-shaped grooves; microstruc- ture of septa fibrous, with libers at right angles to straight or somewhat sinuous central line. Longitudinal Section Description. Dissepimentarium of 1 to 3 ranks of steeply dipping, various sized, noninflated dissepiments; tabellae in 2 zones; axial tabellae of I to 3 ranks, elongate, steeply inclined inwards and upwards to septal lamella (?) or medial plate (?); periaxial tabellae in 1 to 3 ranks, gently inclined in- wards and upwards to axial tabellae; calyx about 8 mm deep, with narrow pointed axial boss about 5 mm high. Documentation. UCMP holotype 3721 1. Two thin sections and 13 polished sections from one corallum from UCMP locality D-858 were studied. Discussion. This species and the following two are referred to Yatsengia on the basis of comparisons with the figures and de- scription of the type species, asiatica (Huang, 1932:56, pi. 5, fig. 1 ) and other discussions and figures of this and other species by Yoh (in Yoh and Huang, 1932:32, pi. 9, figs. 3a-3c), Minato Contributions in Science, Number 337 Wilson: Permian Corals of California 51 (1955: 117-120, pi. 18, figs. 7-10; pi. 29, figs. 1-4; pi. 37, fig. 10; pi. 41, fig. 3; pi. 43, fig. 3; text-figs. 9A-9M), and Fontaine (1961: 1 38- 1 43, pi. 8, figs. 3, 4; pi. 9, fig. 4-6; pi. 10, figs. 1 , 2; pi. 32, figs. 3-5; pi. 33, figs. 3, 4). These authors erected species from China (the type locality), Japan, and Cambodia. Fontaine (1961:140) also considered the genus to be present in Iran and Turkey. All occurrences seem to be in Permian rocks, ranging from Lower to Upper. Yatsengia has not heretofore been recorded from North Amer- ica. My reference of three McCloud Limestone species to it is based on their similarity to the Asiatic species in having axial structures with most septa joined to septal lamellae, apparent lack of medial plates in most axial structures, narrowness or impersistency of the dissepimentaria, shortness of the minor septa, general presence of an excentric false wall, and lack of fossulae or shortened cardinal septa. It has not been possible to compare the microstructure closely. Y.fletcheri has a larger corallite diameter ( 1 2 to 15 mm) than the Asiatic species of Yatsengia (4 to 8.7 mm) and has more septa than any of them (24 to 27 compared with 13 to 20). The dif- ferences with Y. kenneyi n. sp. are discussed under that species, which is apparently lower in the section although still Wolfcam- pian. V.' scheetzi n. sp., also from the McCloud Limestone, is apparently higher in the section, although also still Wolfcampian, and has smaller corallite diameters (7 to I 1 mm) and a much less continuous dissepimentarium than Y. fietcheri. Etymology. The species is named for Mr. William Fletcher. Yatsengia kenneyi n. sp. Figures 3 le; 32a-b Diagnosis. A species of Yatsengia characterized by the combina- tion of relatively small corallites, a relatively small number of major septa, and a relatively continuous dissepimentarium. It has smaller corallites and more major septa than Y fietcheri It has somewhat smaller corallites, somewhat fewer septa, and a more continuous dissepimentarium than Y. scheetzi n. sp. It has larger corallites and more septa than species of the genus de- scribed from Asia. External Description. Corallum phaceloid, maximum observed diameter 1 0 cm; corallites generally parallel, closely spaced; epi- theca rather smooth with widely spaced rugae; calyx not observed. Transverse Section Description. Corallites circular to subcircular, diameter 7.5 to 9.5 mm, closely spaced, touching to as much as 8 mm apart; septa of 2 orders, 17 to 20 each, thin in dissepimen- tarium, slightly dilate in tabularium; major septa 2.0 to 2.5 mm long, generally confluent with septal lamellae; minor septa short, 0.5 to 0.7 mm long, generally confined to dissepimentarium, rarely extending slightly into tabularium as short nubs, axial ends coalescing with lateral extensions of major septa at dissepi- mentarium-tabularium interface to form prominent false wall stereozone where dissepimentarium present; dissepimentarium regular, absent in parts of some corallites; dissepiments angulo- concentric, concentric, pseudoherringbone, straight; axial struc- ture apparently aulophylloid, large, with irregular septal lamellae generally connected with major septa (if true medial plate is present, it is too sinuous to be discernible); false wall stereozone as much as 0.7 mm wide, formed of expanded axial ends of minor septa and bases of major septa in tabularium; corallite wall about 0.3 mm wide, with bases of major and minor septa fitted into V-shaped grooves; microstructure of septa fi- brous, with fibers at right angles to straight or slightly sinuous central line. Longitudinal Section Description. Dissepimentarium of 1 to 3 ranks of steeply dipping, various sized, noninflated dissepiments (transverse sections show that it is not present in parts of some corallites, but this not observed in longitudinal sections avail- able); tabellae of 2 zones, boundary irregular in places; axial tabellae of 1 to 2 ranks, various sizes, steeply inclined inwards and upwards to 1 or more thin septal lamellae (medial plate?); periaxial tabellae large or small, in 1 to 3 ranks, inclined inwards and upwards to axial tabellae. Documentation. UCMP holotype 37212. Three thin sections and 36 polished sections from one corallum from UCMP locality D-805 were studied. Discussion. Y. kenneyi differs from the type and other Asiatic species in its generally greater corallite diameters (7.5 to 9.5 mm compared to 4 to 8.7 mm), generally greater number of septa ( 1 7 to 20 major septa compared to 1 3 to 20), and other characters. It may be distinguished from the two other new species from the McCloud Limestone by its smaller corallite diameters (7.5 to 9.5 mm compared to 1 2 to 1 5 mm for Y. fletcheri and 7 to 1 1 mm for Y sheetzi) and fewer number of major septa (17 to 20 compared to 24 to 27 for Y. fietcheri and 1 8 to 26 for Y. scheetzi). The dissepimentarium is more discontinuous in Y. scheetzi than in Y. kenneyi. Etymology. The species is named for Mr. Paul Kenney. Yatsengia scheetzi n. sp. Figures 32c-g; 33a Diagnosis. A species of Yatsengia characterized by the combina- tion of moderately large corallites, a large number of major septa, and a very discontinuous dissepimentarium. It has larger corallites, more major septa, and a less continuous dissepimen- tarium than Y. kenneyi. It has smaller corallites and a less contin- uous dissepimentarium than Y. fietcheri. It has larger corallites and more septa than species of the genus described from Asia. External Description. Corallum phaceloid, maximum observed diameter 10 cm; corallites subparallel to parallel, closely spaced; calyx and epitheca not observed. Transverse Section Description. Corallites circular to subcircular, diameter 7 to 1 1 mm, closely spaced, touching to 3 mm apart; septa of 2 orders, 18 to 26 each, lanceolate throughout; major septa 2.5 to 3.5 mm long, about 0.5 mm wide at base where lanceolate, many confluent with septal lamellae in largest cor- allites, cardinal and/or counter septa only touch axial structure in smaller corallites; minor septa 0.5 to 0.8 mm long, width where lanceolate as in major septa; dissepimentarium very discon- tinuous, present only in parts of some corallites; dissepiments straight or pseudoherringbone; axial structure generally aulo- phylloid, some clisiophylloid or dibunophylloid; medial plate where present thin to slightly dilate, straight to very sinuous; 52 Contributions in Science, Number 337 Wilson: Permian Corals of California septal lamellae number I to 1 0, straight to very sinuous, attached to major septa in some large corallites, generally not in smaller corallites; false wall stereozone thin or not present due to discon- tinuity of dissepimentarium; corailite wail denticulate, with septa set in V-shaped grooves; microstructure of septa fibrous, with fibers at right angles to straight or somewhat sinuous central line. Longitudinal Section Description. Dissepimentarium discon- tinuous, may be absent for as much as 1 cm; where present, of 1 to 2 ranks of small, steeply dipping dissepiments, numbering about 6 in 0.5 cm; tabellae of 2 zones; axial tabellae of 1 to 2 ranks, steeply to gently inclined inwards and upwards to medial plate (where present); periaxial tabellae of 2 to 3 ranks, more gently inclined inwards and upwards to axial tabellae, various sizes; corailite wall as much as 0.7 mm wide. Documentation. UCMP holotype 37213, UCMP paratypes 37214-3721 8, LACM I P paratype 6416. Seven thin sections and numerous polished sections from seven coralla from UCMP lo- calities D-834 (paratype 37218), D-836 (holotype 37213, para- types 37214, 37216), and D-838 (paratypes 37215, 37217) and LACMIP locality 1114 (paratype 6416) were studied. Discussion. The nature of the axial structure in this species, which ranges from aulophy lloid through cl isiophy 1 loid and di- bunophylloid, is troublesome, but all types are present in some coralla and therefore demonstrate that the variation is natural. Two figures of the holotype of Yatsengia asiatica Huang, 1932 (pi. 5, figs, la, 1 b), seemingly show a similar variation. Y. scheetzi is larger than the Asiatic species and has a greater number of septa. It has somewhat larger corallites than Y. ken- neyi(l to 1 I mm compared to 7.5 to 9.5 mm) and has more septa (18 to 26 major septa compared to 17 to 20). It has smaller corallites than Y fletcheri, which ranges from 12 to 15 mm in diameter, and has generally fewer septa ( Y. fletcheri has 24 to 27). The dissepimentarium of Y. scheetzi is more discontinuous and narrower than those of the other two species of Yatsengia from the McCloud Limestone. Etymology. The species is named for Mr. James Scheetz. Cerioid Rugosa Genus Bassius n. gen. Diagnosis. Bassius is a cerioid rugose coral with two orders of septa, clisiophylloid axial structure, wide lonsdaleoid dissepi- mentarium, thin wall, tabellae of two ranks — axial and periaxial, the former generally clustered dissepiment-like around the co1- umella, the latter generally flat, domed, or slightly sagging, h' , i- zontal or sloping inwards and upwards, rarely sloping inwards and downwards. Type Species. Bassius mccloudensis n. sp. Description. As for the type and only species firmly referred to the genus. Discussion. Kleopatrina ( Porfirievella ) is similar in some charac- ters to Bassius, but the latter has a wide and more consistently developed lonsdaleoid dissepimentarium and a much thinner corailite wall. The separation of Kleopatrina into two subgenera is largely a subjective matter based on the presence of more or fewer septa that cross the dissepimentarium to the corailite wall. Even the most lonsdaleoid species of Kleopatrina, such as A. (Porfirievella) arcturusensis Stevens, 1967, from the Lower Per- mian of eastern Nevada, have “. . . in large corallites generally less than one-half of the major septa extend to the theca” (Ste- vens, 1967:427). Ipciphyllum Hudson, 1958, shares some characters with Bassius, but the former is not lonsdaleoid. WentzellophyUum Hudson, 1958, is cerioid, clisiophylloid, and lonsdaleoid, but it has tertiary septa, which Bassius lacks. IVentzelella Grabau in Huang, 1932, has both teritary septa and a mostly regular dis- sepimentarium, unlike Bassius, which has neither. Ipciphyllum, WentzellophyUum, and Wentzelella are Permian and from the classic Tethys area. Only the former genus has been reported in North America ( Ipciphyllum tschernyschewi Minato, I960, from the Arctic Islands of the Northwest Territories, Canada). Lonsdaleia M’Coy, 1849, is a Mississippian coral, the type species of which is fasciculate, although Hill (1956:306) would include cerioid coralla in it as well. I consider fasciculate and cerioid coralla to be generic characters and on this basis alone differentiate Lonsdaleia from Bassius, but in addition they are dissimilar in many other ways. Lonsdaleia is restricted to the Carboniferous by most paleontologists. Hudson ( 1 958: 1 82) referred a specimen from the Permian of northern Iraq to 'Lonsdaleia' chaoi Huang var. The species of Huang (in Yoh and Huang, 1 932:35), from the Permian of China, has an aulophylloid axial structure, and the specimens of Hudson are clisiophylloid (Hudson, 1958:183), which are generic dif- ferences. Minato and Rato ( 1 965a:2 1 1 ) referred Huang’s species to Wentzelophyllum and noted that it has tertiary septa. Both these species were reported by Hudson ( 1958) and Minato and Kato ( 1 965b) to have “clinotabulae” (clinotabcllae), a name pro- posed by Hudson (1958:177) for periaxial tabellae that slope downwards and inwards to the axial tabellae and form a deep groove around the axial boss, as shown in Hudson’s ( 1958, pi 33, figs. 4a, 4b) reference to his figure of Waagenophyllum indicum ( Waagen and Wentzel, 1 886) as having examples of these struc- tures. Occasional tabellae sloping downwards and inwards in a coral in which they generally slope upwards and inwards do not form the deep groove of Waagenophyllum Hayasaka, 1925, and, in my opinion, should not be considered clinotabcllae or heavily weighed systematically. I think that the specimens referred to by Hudson ( 1 958: 1 82) as 'Lonsdaleia' chaoi Huang var. do not have clinotabellae, since his figures (Hudson, 1958, pi. 32, fig. 1 and text-figs. 3c-3d) show no such groove. Minato and Kato ( 1 965a:2 11) erected Wentzelophyllum (?) gelikhanense for the 'Lonsdaleia' chaoi Huang var. of Hudson from Iraq and described tertiary septa “locally found along walls as mere short septal ridges.” I see no certain evidence of tertiary septa in Hudson’s ( 1 958, pi. 32, fig. 2) best transverse section of the holotype. If tertiary septa are lacking, then perhaps W. (?) gelikhanense should be referred to Bassius. The thin sections of the holotype should be examined. Minato and Kato ( 1 965a: 2 1 3 ) named Wentzelophyllum (?) tabasense for a coral from the Permian of Iran. They noted (Minato and Kato, 1 965a:21 5) that it “. . . lacks any trace of the tertiary septa . . . typical to the genus Wentzelophyllum." Their figures (Minato and Kato, 1965a, pi. 20) show transverse and Contributions in Science, Number 337 Wilson: Permian Corals of California 53 Figure 31. Siphonodendron hongi n. sp., Yatsengia fletchen n. sp., and V. kenneyi n. sp. All parts x3. 54 Parts a and b. 5. hongi n. sp., UCMP holotype 37210, transverse section (a; note Pseudoschwagerina at center) and longitudinal section (b). Parts c and d. Y. fletchen n. sp., UCMP holotype 37211, transverse (c) and longitudinal (d) sections. Part e. Y. kenneyi n. sp., UCMP holotype 3721 2, transverse section. Figure 32. Yatsengia kenneyi n sp. and Y. scheetzi n. sp. All parts x3. Parts a and b. Y. kenneyi n. sp., UCMP holotype 37212, transverse (a) and longitudinal (b) sections. 55 Parts c through g. Y scheetzi n. sp. Parts c and d, UCMP holotype 37213, transverse (c) and longitudinal (d) sections. Parts e through g, UCMP paratype 37214, transverse (e) and longitudinal (f, g) sections. 1 1 / Y>S| ¥ CmL&L/ Kir V -r' ^\^W' 56 Contributions in Science, Number 337 Wilson: Permian Corals of California longitudinal sections with characters similar to Bassius. The lon- gitudinal sections lack what I consider to be clinotabellae, al- though Minalo and Kato considered them to be present. This holotype also should be examined for possible inclusion in Bassius. If the two species from Iraq and Iran should prove to be refer- able to Bassius, then this would be a rare instance of true Tethys massive corals occurring both there and in North America. Etymology. The genus is named for Mr. John S.P. Bass. Bassius mccloudensis n. sp. Figures 34a-c Diagnosis. Because the genus Bassius is erected as monotypic, the diagnoses of the genus and of the type species are identical. External Description. Corallum cerioid, hemispheroidal, max- imum observed diameter 4.5 cm; calyx not observed. Transverse Section Description. Corailites polygonal, 5 to 7 sided, 1 1 to 1 7 mm wide at greatest diameter; septa of 2 orders; major septa 19 to 25 each, 1 .5 to 3 mm long, extending into dissepimen- tarium 0.5 to 1 .0 mm, thin in dissepimentarium (where present), represented in places by septal spines on dissepiments, slightly dilate in tabularium, withdrawn from axial structure; minor septa commonly absent in adult corailites or represented by a few septal spines in outer ranks of dissepiments and on wall, absent in tabularium; axial structure clisiophylloid, oval, 2 to 3 mm long, with well-defined axial plate having 5 to 6 short septal lamellae opposite each other on each side, connected by axial tabellae; dissepimentarium highly lonsdaleoid, 3 to 4 mm wide; dissepi- ments highly lonsdaleoid in peripheral ranks, concave axially, variously sized, with some axial spines, concentric, angulo-con- centric, or pseudoherringbone near axial part where major septa present; corallite wall thin and smooth or slightly denticulate, entire, 0. 1 to 0.2 mm wide. Longitudinal Section Description. Dissepimentarium of 2 to 5 ranks of small cystose or larger elongate dissepiments, steeply dipping, about 20 per cm; tabellae of 2 series, axial and periaxial; axial tabellae clustered about columella dissepiment-like or slop- ing steeply inwards and upwards, 1 to 3 ranks each side; periaxial tabellae flat, domed, or slightly sagging, generally sloping in- wards and upwards to axial tabellae, rarely sloping inwards and downwards, about 20 per cm. Documentation. UCMPholotype 37219. Three thin sections and 19 polished sections from one corallum from UCMP locality D-l 8 1 were studied. Discussion. The only two species known to me with which this species might be compared are from the Permian of Iran and Iraq and have been discussed above under “Remarks” for the genus. It is uncertain if they may be referred to Bassius, but if they should be, then the most obvious difference would be that B. mccloudensis has no minor septa in the tabularium whereas both of the other species have well-developed minor septa there. Etymology. The species is named for the McCloud River. Census Dillerium n. geo. Diagnosis. Dillerium is a cerioid coral with two orders of septa, a styliform columella, a regular dissepimentarium, and tabulae that slope inwards and upwards. Type Species. Dillerium potterensis n. sp. Description. As for the type species since the genus is described as monotypic. Discussion. Dillerium may be distinguished readily from Kleopatrina McCutcheon and Wilson, 1963, by the former’s greatly thickened columella and possession of tabulae, whereas the latter has a medial plate, variously abutted by septal lamel- lae, and two well-defined zones of tabellae. It can be differenti- ated from other somewhat similar appearing cerioid corals by the lack of three orders of septa, the lack of two or three zones of tabellae (no clinotabellae), the lack of a fossula, the lack of a clisiophylloid, aulophylloid, or dibunophylloid axial structure, the lack of an incomplete wall, and the possession, as described in the diagnosis, of complete walls, two orders of septa, a styliform columella, a regular dissepimentarium, and tabulae. Etymology. The genus is named for Mr. Joseph S. Diller. Dillerium potterensis n. sp. Figures 35a-c Diagnosis. Because the genus Dillerium is erected as mono- typic, the diagnoses of the genus and of the type species are identical. External Description. Corallum cerioid and hemispheroidal; maximum observed diameter 5.5 cm; other external features not preserved. Transverse Section Description. Corailites 4 to 7 sided, 14 to 16 mm wide at greatest diameter; septa of 2 orders, 13 to 17 each, straight to somewhat sinuous; major septa generally withdrawn from axial structure, 0.5 to 1.5 mm long, generally dilate in tabularium (to 0. 1 5 mm thick), thin in dissepimentarium; minor septa 0.3 to 0.5 mm long, confined to dissepimentarium or ex- tending into tabularium as short spines, thin throughout; dissepi- mentarium regular, narrow, 0.5 mm wide or less; dissepiments concentric, angulo-concentric, straight, herringbone (rare), or pseudoherringbone (rare); axial structure a highly dilate col- umella (styliform), round, oval, lenticular in shape, measuring 1 .0 by 1 .0 mm, 1 .0 by 1 .5 mm, to 0.5 by 1 .5 mm, with dark line in center representing axial plate and, in some, 1 to 3 short dark Figure 33. Yatsengia scheetzi n. sp. and Heritschioides (?) californiense (Meek, I 864). All parts x3. Part a. Y. scheetzi n. sp., LACMIP paratype 6416, transverse section. Parts b through d. H. (?) californiense, MCZ lectotype, MCZ Meek collection block no. 1 3, transverse and longitudinal sections ( b), transverse section (c), and longitudinal section (d). Part e. (?) H californiense. UCMP hypotype 37350, transverse section. Contributions in Science, Number 337 Wilson: Permian Corals of California 57 58 Figure 34. Bassius mccloudensis n. gen., n. sp., UCMP holotype 37219, transverse (a, b) and longitudinal (c) sections. All parts x3. lines representing axial lamellae, not connected to septa; cor- allite wail 0. 1 5 to 0.3 mm wide, with dark line in center, V-shaped depressions for septal bases. Longitudinal Section Description. Dissepimentarium of 1 to 2 stee- ply dipping ranks of small cystose dissepiments; tabulae slope steeply inwards and upwards to columella, generally about 20 per cm. Documentation. UCMP holotypc 37220. Two thin sections and 17 polished sections from one corallum from UCMP locality D-843 were studied. Discussion. Although the preservation of the corallum on which this new genus and species are based is not very good (somewhat metamorphosed, somewhat crushed), the characters clearly can be made out with some patience. Dillerium potterensis is not close to any other species that I have seen, and the genus is erected as monotypic. Etymology. The species is named after Potter Ridge. Genus Kleopatrina McCutcheon and Wilson, 1963 Subgenus Kleopatrina McCutcheon and Wilson, 1963 Kleopatrina (Kleopatrina) raubae n. sp. Figures 35d-e Diagnosis. A species of Kleopatrina ( Kleopatrina ) charac- terized by the combination of large corallites, abundant septa, a complex axial structure, and thin septa in the tabularium. It has larger corallites, more septa, and a more complex axial structure than K. (K.) ftatateeta and K. (K.) wilsoni. It has larger corallites and/or more septa than all the species of the subgenus described from Russia except K. (K.) magnifica. It has a more complex axial structure and septa that are thinner in the tabularium than K. (K.) magnifica. External Description. Corallum cerioid, hemispheroidal, max- imum observed diameter 8 cm (probably much greater in un- collected specimens); calyx not observed; corallites parallel. Transverse Section Description. Corallites 4 to 7 sided, 9 to 14 mm wide at greatest diameter; septa of 2 orders, 16 to 23 each, straight to slightly sinuous, rarely lonsdaleoid; major septa gener- ally withdrawn from axial structure, touching in some corallites, 2.5 to 5.5 mm long, slightly dilate in tabularium or thin through- out; minor septa poorly developed (absent in parts of some cor- allites), 1 .0 to 2.0 mm long, generally confined to outer V2 to % of dissepimentarium, but extending into tabularium in places as nubs or spines 0.5 mm long; dissepimentarium generally regular, width 1 to 5 mm, generally about 3 mm; dissepiments concentric, herringbone, pseudoherringbone, or lonsdaleoid (uncommon); axial structure clisiophylloid, circular to subcircular, 1 .5 to 2 mm diameter, formed of straight to sinuous, thin or slightly thickened medial plate about 1.5 to 2 mm long with 2 to 8 radiating, straight to sinuous septal lamellae connected by 1 to 2 axial tabellae, generally not touched by axial ends of septa, rarely touched by 1 or more; corallite wall 0. 1 5 to 0.2 mm wide. Longitudinal Section Description. Dissepimentarium of 4 to 7 stee- ply to gently dipping ranks of mixed large elongate and large and small cystose dissepiments; tabellae of 2 zones, axial and periax- ial; periaxial tabellae flat, gently domed, with straight peripheral edges, rarely small and cystose (sloping in and up to periaxial tabellae where axial, sloping in and up to dissepiments where peripheral), 20 to 30 per cm; axial tabellae in 1 to 3 ranks, large to small, elongate (some cystose), steeply sloping inwards and up to medial plate, 20 to 32 per cm. Documentation. LACM1P holotype 6417, LACMIP paratypes 6418-6419, UCMP paratypes 37221-37222. Eleven thin sec- tions and 64 polished sections from five coralla from UCMP locality D-842 (paratypes 37221-37222) and LACMIP locality 1122 (holotype 6417, paratypes 64 1 8-6419) were studied. Discussion. K. (K.) raubae can be distinguished from the other McCloud Limestone species of Kleopatrina described herein by the absence or relative rarity of lonsdaleoid dissepimentaria. The other new species have predominately lonsdaleoid dissepimen- taria, which relegate them to the subgenus Porfirievella Minato and Kato, 1 965b:7 1 , type species Wentzelella grandis Dobro- lyubova, 1941 (in Soshkina, Dobrolyubova, and Porfiriev, 1 941 : 1 97, 264, pi. 52, figs, la lb). Minato and Kato ( 1 968:363) suggested later that their name be changed to Uralnevadaphyl- lum to avoid confusion with Porfirieviella Ivanovsky, 1 963, which has a one letter difference. ICZN Article 56(a) (International Commission on Zoological Nomenclature, 1964) states that “even if the difference between two genus-group names is due to only one letter, these two names are not to be considered hom- onyms,” therefore the subgeneric name Porfirievella of Minato and Kato cannot be invalidated on the grounds given by them and must stand. Previously, only two North American species have been re- ferred to the nominal subgenus: the type species, K. ftatateeta (McCutcheon and Wilson, 1961), from the Lower Permian of east central Nevada, and K wilsoni Minato and Kato, 1965b, from the Lower Permian of southeast Nevada. Both have smaller corallites, fewer septa, and less complex axial structures than K. raubae. Minato and Kato ( 1965b:69) based K. wilsoni on a para- type of K. ftatateeta but rejected a longitudinal section (Mc- Cutcheon and Wilson, 1961, pi. 123, fig. 5) from the same cor- allum (their holotype), considering it to belong to K. ftatateeta. which illustrates the difficulties of achieving objective systemat- ics when working with variable corals. K raubae has greater corallite diameters and/or more abun- dant septa than the Russian species of K. (Kleopatrina). except for K. (K.) magnifica (Porfiriev, 1941, in Soshkina, Dobro- lyubova, and Porfiriev: 1 99, 265, pi. 53, figs, la-lc; pi. 54, figs, la-lb) from the Lower Permian of the Ural Mountains. The latter species, however, has a less complex axial structure and septa that are thickened in the tabularium, which distinguish it from K. (K.) raubae. Etymology. The species is named for Mrs. Lenor Raub Subgenus Porfirievella Minato and Kato, 1965b Kleopatrina (Porfirievella) mckibbinae n. sp. Figures 36c-f Diagnosis. A species of Kleopatrina ( Porfirievella ) charac- terized by the combination of a generally simple axial structure Contributions in Science, Number 337 Wilson: Permian Corals of California 59 Figure 35. Dillerium potlerensis n. gen., n. sp. and Kleopatrina (Kleopatrina) raubae n. sp. 60 Parts a through c. D. polterensis n. gen., n. sp., UCMP holotype 37220, transverse section (a, x3), longitudinal section (b, x3), and transverse section (detail of a, x7.5). Parts d and e. K. (K.) raubae n. sp., LACMIP holotype 6417, transverse (d) and longitudinal (e) sections. Both parts x3. Figure 36. Kleopatrina (Porfirievella) peggyae n. sp. and A (Pj mckibbinae n. sp. All parts x3. Parts a and b. K. (P) peggyae n. sp., UCMP holotype 37223, transverse (a) and longitudinal (b) sections. 61 Parts c through f. K. (P.) mckibbinae n. sp. Parts c through e, UCMP holotype 37224, transverse (c) and longitudinal (d, e) sections. Part f, LACM1P paratype 6420, transverse section touched by many septa and a moderately lonsdaleoid dissepi- mentarium. It has more septa touching the axial structure and a less ionsdaleoid dissepimentarium than K. (P.j peggyae n. sp. and K. (P.) whitneyi n. sp. It has a less complex axial structure touched by fewer septa than K. (P.j illipahensis. It has a less complex axial structure touched by more septa than K. (P.) arcturusensis. It has a smaller and less complex axial structure than K. (P.j nevadensis. It has an axial structure touched by more septa than K. (P.j grandis, K. (P.) radiala. and K. (P) stylidophylloides. External Description. Corallum cerioid, hemispheroidal to sub- hemispheroidal, maximum diameter 13 cm (represented by col- lected fragments, many from much larger coralla), calyx not observed. Transverse Section Description. Corallites 4 to 7 sided, 6 to 1 1 mm wide at greatest diameter; septa of 2 orders, 1 2 to 20 each (gener- ally 15 to 17), straight to somewhat sinuous, absent or repre- sented by spines on dissepiments where lonsdaleoid; major septa nearly all touching axial structure in some corallites, fewer touching in others, some touching in nearly all corallites, 1.5 to 3.5 mm long, generally thin throughout except few somewhat dilate to 0.2 mm thick in tabularium; minor septa poorly devel- oped, absent in parts of many corallites, 0.5 to 1.5 mm long, rarely crossing dissepimentarium; dissepimentarium width 0.5 to 3.0 mm, generally about 2.0 mm; dissepiments concentric, her- ringbone, pseudoherringbone, lonsdaleoid (common); axial structure variable, clisiophy 1 loid (connected to counter and car- dinal septum in many and appearing dibunophylloid), formed of thin to thickened medial plate 1 to 2 mm long, rarely dilate to as much as 0.6 mm, generally simple with 1 to 7 irregularly spaced, short septal lamellae, some without septal lamellae, some with septal lamellae joined by axial tabellae into complex structure, many of which are touched by attenuate septal ends; corallite wall 0. 1 5 to 0.3 mm wide; corallites rarely sciophy lloid. Longitudinal Section Description. Dissepimentarium of 1 to 6 (generally 2 to 4) ranks of large elongate and small and large cystose dissepiments; tabellae poorly zoned in most corallites, not zoned in some; periaxial tabellae fiat, domed, horizontal to gently sloping inwards and upwards to axial tabellae, 1 8 to 30 per cm; axial tabellae steeply sloping inwards and upwards to col- umella, 1 8 to 20 per cm; tabulae tent-shaped or domed, about 20 per cm. Documentation. UCMP holotype 37224, UCMP paratypes 37225-37229, LACMIP paratypes 6420-6423. Twenty-three thin sections and 183 polished sections from 10 coralla from UCMP localities A-7096 (holotype 37224, paratypes 37225- 37226, 37228-37229) and D-842 (paratype 37227) and LACMIP localities 1 123 (paratypes 6420-6421) and 1 124 (par- atypes 6422-6423) were studied. Discussion. The numerous coralla of this species examined show it to be variable in the complexity of the axial structure as seen in transverse and longitudinal sections and the consequent zoning, or lack of zoning, in the tabularium as seen in longitudinal sec- tion. The holotype has a relatively complex axial structure for the species. A particularly characteristic feature of the species is the common attachment of both the counter and cardinal septa to the axial structure, giving it a di bu nophy I loid-1 ike character. The other three species of Kleopatrina (Porfirievella) de- scribed herein from the McCloud Limestone are comparable in corallite sizes and septal numbers. The two species from lower in the formation both have much more highly lonsdaleoid dissepi- mentaria as seen in transverse section and lack the number of septa touching the axial structure. K. (P.j whitneyi n. sp. — holo- type figured by Meek (1864, pi. 1, figs. 3, 3a) — is somewhat similar in the dissepimentarium, but the axial structure does not show any counter-cardinal septa attachments. Meek’s specimen may have come from the Bayha area, but its high degree of silicification is unlike those collected in the Bayha section in this study. The species of Kleopatrina ( Porfirievella ) from elsewhere show a marked similarity in corallite diameters and septal numbers. However, they differ from K. (P) mckibbinae in the following manners: K. (P.j illipahensis (Easton, 1960) from the Lower Per- mian Pequop Formation of east-central Nevada has a much more complex axial structure that is touched by almost every septum. K. (P.j arcturusensis Stevens, 1967, from the Lower Permian Arcturus Formation of east-central Nevada also has a more com- plex axial structure and septa that are more withdrawn from it. K (P.) nevadensis Stevens, 1967, from the Lower Permian Arcturus (?) Formation of east-central Nevada has a much larger and far more complex axial structure. K. (P.j grandis (Dobrolyubova, 1941) from the Lower Permian of the Ural Mountains of Russia, the type species of the subgenus, has fewer septa touching the axial structure. K. (P.j stylidophylloides (Dobrolyubova, 1941 ) and K. (P.j stylidophylloides var. radiata. (Porfiriev, 1941), also from the Permian of the Ural Mountains, likewise have axial structures that are more withdrawn from the axial ends of the septa. Etymology. The species is named for Mrs. Jean McKibbin. Kleopatrina (Porfirievella) peggyae n. sp. Figures 36a-b Diagnosis. A species of Kleopatrina (Porfirievella) charac- terized by the combination of a simple axial structure touched by few septa and a highly lonsdaleoid dissepimentarium. It has a simpler axial structure touched by fewer septa than K. (P.j zulloi n. sp. It has a more lonsdaleoid dissepimentarium than K. (P) whitneyi n. sp. and K (P.j mckibbinae. It has a simpler axial structure than K. (P.j illipahensis. K. (P.j arcturusensis. and K. (P.j nevadensis. It has thinner septa and a more lonsdaleoid dissepi- mentarium than K. (P.j grandis. It has more septa and a more lonsdaleoid dissepimentarium than K. (P.j stylidophylloides var. radiata. External Description. Corallum cerioid, maximum observed di- ameter 10 cm; no well-preserved calices or wall observed. Transverse Section Description. Corallites polygonal, 5 to 7 sided, 7 to 1 1 mm wide at greatest diameter; septa of 2 orders; major septa number 17 to 18, 2 to 3 mm in length, extending into tabularium 1 to 1.5 mm, thin in dissepimentarium, thin or slightly dilate in tabularium; minor septa absent or represented by septal spines on wall, dissepiments, and (rare) false wall. 62 Contributions in Science, Number 337 Wilson: Permian Corals of California fewer than major septa; axial structure variable: columella or medial plate alone, dilate or thin, straight or slightly sinuous, or simplified clisiophy I loid with 1 to 4 short septal lamellae; dissepi- mentarium lonsdaleoid; corallite wall denticulate, 0.3 to 0.5 mm wide. Longitudinal Section Description. Dissepimentarium of 1 to 4 ranks of inflated dissepiments; dissepiments number about 8 in I cm; tabeilae of 2 series; axial tabellae small, 1 to 3 ranks on each side of columella, steeply dipping; periaxial tabellae generally horizontal or sloping inwards and upwards to axial tabellae; col- umella sinuous. Documentation. UCMP holotype 37223. Two thin sections and 21 polished sections from one corallum from UCMP locality D-827 were studied. Discussion. K. (P) peggyae has thinner walls, a thinner axial plate, fewer septa touching the axial structure, and generally a simpler axial structure than K. (P.) zulloi n. sp. from the Hirz Mountain section. It is more highly lonsdaleoid and differs in other characters from K. (P) whitneyi n. sp. from south of the Pit River and K. (P) mckibbinae from the Potter Ridge section. K. (P) illipahensis, K. (P.) arcturusensis, and K (P) nevadensis from eastern Nevada have more complex axial structures than K. (P.) peggyae. The Russian species of the subgenus are not very similar to this species. K. (P) grandis, type species for the subgenus, has major septa dilated in the tabularium and is somewhat less lonsdaleoid. K. (P.) stylidophylloides var. radiata has somewhat fewer septa and a less lonsdaleoid dissepimentarium. Etymology. The species is named for Mrs. Peggy McCain. Kleopatrina (Porfirievella) whitneyi n. sp. Figures 5:3, 3a; 15f; 37a-b Lithostrotion mamillare? Meek, 1864, plate 1, figs. 3, 3a. Diagnosis. A species of Kleopatrina (Porfirievella) charac- terized by the combination of a simple axial structure touched by few septa and a moderately lonsdaleoid dissepimentarium. It has an axial structure touched by fewer septa and a less lonsdaleoid dissepimentarium than K. (P) zulloi n. sp. It has a less variable axial structure and a less lonsdaleoid dissepimentarium than K. (P) peggyae. It has an axial structure touched by fewer septa than K. (P.) mckibbinae. It has a simpler axial structure than K. (P) arcturusensis. K. (P.) illipahensis. and K. (P) nevadensis. External Description. Corallum cerioid, maximum observed di- ameter 3 cm; calices and epitheca not observed. Transverse Section Description. Corallites polygonal, 4 to 7 sided, 6 to 1 1 mm wide at greatest diameter; septa of 2 orders; major septa number 17 to 20, total length 2 to 5 mm, extending into dissepimentarium 0.5 to 1 .5 mm; generally withdrawn from axial structure; minor septa same number or fewer than majors, ex- tending into tabularium about 0.3 mm in some corallites, absent in tabularium of others, represented in places only by septal spines on dissepiments; axial structure clisiophy lloid, with sin- uous medial plate 1.5 to 2 mm long; septal lamellae absent or maximum of 6, short, loosely connected by axial tabellae; dis- sepiments highly lonsdaleoid in parts of some corallites, not lons- daleoid in others but straight, angulo-concentric, pseudoherring- bone, herringbone; dissepimentarium about 4 mm wide at great- est width; wall rather sinuous, not denticulate, about 0.2 mm wide. Longitudinal Section Description. Dissepiments of I to 6 ranks of steeply dipping, small or inflated dissepiments; tabellae of 2 zones, axial and periaxial; axial tabellae large, of 1 rank, sloping upwards to the columella, rarely continuous with periaxial tab- ellae to form tabulae; periaxial tabellae horizontal in places, gen- erally sloping inwards and upwards gently to axial tabellae, about 16 per cm; columella sinuous. Documentation. Two thin sections and several polished sections from the holotype corallum, MCZ Meek collection specimen no. 15, were studied. This specimen was labeled “No 15 Lithostro- tion mamillare Castlenau ? sp.” and is not the “ Lithostrotion ?’’ of Meek (1864:7), which he indicated is figured on his plate 1, figures 3, 3a. The specimen labeled “ Lithostrotion ?" is a fasciculate coral and was not figured by Meek. The type locality is uncertain but probably is somewhere in the McCloud Limestone south of the Pit River at one of the localities collected by the Whitney Survey in September 1 862. Other cor- al la of this species were not collected by me, and the high degree of silicification suggests that the specimen did not come from my Bayha section, where few corals are so highly silicified. Unfortu- nately, no fusulinids are associated with the holotype, but the formation in this area is within the fusulinid zones E and F. Cerioid corals are common in places in zone F of the Bayha section. Discussion. K. (P) whitneyi can be distinguished from the other three new species of the subgenus from the McCloud Limestone by the number of septa in contact with the axial structures and the degree of lonsdaleoid development of the dissepimentaria as well as other features. K. (P.) zulloi n. sp. from the Hirz Mountain section has many more septa that touch the axial structure and is much more lonsdaleoid. K. (P) peggyae from Bollibokka Moun- tain has a much more lonsdaleoid dissepimentarium and a differ- ent kind of axial structure. K. (P) mckibbinae from the Bayha and Potter Ridge sections has many more septa that touch the axial structure. K. (P) illipahensis, K. (P.) arcturusensis, and K. (P) nevadensis, all from the Lower Permian of Nevada, have more complex axial structures and other differences. Etymology. The species is named for Mr. Josiah Dwight Whitney. Kleopatrina (Porfirievella) zulloi n. sp. Figures 37c-e Diagnosis. A species of Kleopatrina ( Porfirievella ) charac- terized by the combination of a moderately simple axial struc- ture touched by many septa and a highly lonsdaleoid dissepimen- tarium. It has longer septa, a more complex axial structure touched by more septa, and more inflated dissepiments than K. (P.) peggyae. It has a more lonsdaleoid dissepimentarium than K. (P) mckibbinae and K. (P) whitneyi. It has a simpler axial struc- ture than K. (P) arcturusensis. K. (P) illipahensis, and K. (P) nevadensis. External Description. Corallum cerioid, maximum observed di- Contributions in Science, Number 337 Wilson: Permian Corals of California 63 Figure 37. Kleopairina (Porfirievella) whitneyi n. sp., K. (P.) zulloi n. sp., and Langenheimia klamathensis n. gen., n. sp. All parts x3. 64 Parts a and b. K. (Pj whitneyi n. sp., MCZ holotype, MCZ Meek collection corallum no. 1 5, transverse (a) and longitudinal (b) sections. Parts c through e. K. (P) zulloi n. sp., UCMP holotype 37230. longitudinal (c, e) and transverse (d) sections. Part f. L. klamathensis n. gen., n. sp., UCMP holotype 3723 1 , longitudinal section. ameter 15 cm, calyx 2 to 3 mm deep, with steeply sloping walls, rather flat floor, tall central axial boss, which may be attached to one side of calyx, presumably by counter septum. Transverse Section Description. Corallites polygonal, 5 to 7 sided, 6 to 9 mm wide at greatest diameter; septa of 2 orders, I 5 to 19 each, thin throughout, or very slightly dilate in tabularium; ma- jor septa 1 to 2 mm long, extending into tabularium 0.5 to 0.8 mm; minor septa generally confined to dissepimentarium, about 0.3 mm long; axial structure clisiophy 1 loid, with well-defined me- dial plate, 1 to 4 septal lamellae, and tabellae (juvenile axial structure a columella attached to counter septum); dissepiments lonsdaleoid in many places, elsewhere concentric and her- ringbone; corallite wall denticulate, with or without septal spines where lonsdaleoid, 0.2 to 0.3 mm wide. Longitudinal Section Description. Dissepimentarium of 1 to 3 (rare) ranks of steeply dipping, generally inflated dissepiments; tabellae of 2 series; axial tabellae dissepiment-like against col- umella; periaxial tabellae generally flat or sagging, rarely sloping inwards and upwards to the axial tabellae; columella somewhat sinuous. Documentation. UCMP holotype 37230. Four thin sections and 44 polished sections from one large corallum from UCMP lo- cality D-855 were studied. Discussion. K. (P.) zulloi is similar in some respects to K. (P) peggyae from Bollibokka Mountain. The latter, however, has a simpler axial structure from which the septa are somewhat more withdrawn and generally less inflated dissepiments. K. (P) whit- nevi and K. (P.) mckibbinae, both from higher in the McCloud Limestone, have less highly developed lonsdaleoid dissepimen- taria as well as other characters that distinguish them from K. (P) zulloi. K. (P. ) zulloi may readily be distinguished from the three spe- cies of the subgenus from Nevada, K. (P.) illipahensis, K. (P.) arcturusensis, and K. (P) nevadensis by its simpler axial structure. Etymology. The species is named for Dr. Victor A. Zullo. Genus Langenheimia n. gen. Diagnosis. Langenheimia is a cerioid rugose coral with large corallites, no fossula, septa of two orders, well-defined aulophyl- loid axial structure, thin and complete wall, tabellae of two ranks — axial and periaxial, the former sloping steeply inwards and upwards, the latter variously inclined but generally gently inwards and upwards to the axial tabellae. Type Species. Langenheimia klamathensis n. sp. Description. As for the type and only known species. Discussion. Langenheimia has some characters in common with other Permian cerioid rugose corals, but they have the following, among other, important characters that distinguish them from Langenheimia. Wentzelella Grabau in Huang, 1932, Wentzelloides Yabe and Minato, 1944, and Wentzelophyllum Hudson, 1 958, all have tertiary septa. Yokoyamaella Minato and Kato, 1965a, has thick mural septa. Ipciphyllum Hudson, 1958, and Kleopalrina McCutcheon and Wilson, 1963, both have clisiophylloid axial structures. Parawentzelella Fontaine, 1961, also has a clisiophylloid axial structure as well as the controver- sial “canals” between walls of corallites. Etymology. The genus is named for Dr. Ralph L. Langenheim, Jr. Langenheimia klamathensis n. sp. Figures 37f; 38a-b Diagnosis. Because the genus Langenheimia is erected as mono- typic, the diagnoses of the genus and of the type species are identical. External Description. Corallum cerioid, hemispheroidal; max- imum observed diameter 13 cm; calyx and epitheca not seen. Transverse Section Description. Corallites polygonal, 5 to 7 sided, 1 5 to 27 mm wide at greatest diameters; septa of 2 orders, 23 to 27 each; major septa 3.5 to 8.5 mm long, extending into tab- ularium 3 to 3.5 mm, thin in dissepimentarium, dilate and lan- ceolate in tabularium, maximum width about 0.5 mm near base in tabularium; minor septa thin, 2 to 3.5 mm long, generally confined to dissepimentarium, rarely penetrating slightly into tabularium; axial structure aulophylloid, of numerous irregular cystose tabellae, elongate, 4 to 5 mm long, 2 to 2.5 mm wide, not touched by septa; dissepiments variously angulo-concentric, con- centric, herringbone, or pseudoherringbone, with irregular small dissepiments near bases of some septa; corallite wall entire, thin, about 0. 1 3 mm wide. Longitudinal Section Description. Dissepimentarium wide, of 5 to 10 ranks of small, cystose to large, elongate, steeply dipping dissepiments; tabellae of 2 series, axial and periaxial; axial tab- ellae small and cystose to larger and elongate, in 3 to 8 ranks each side, steeply inclined inwards and upwards; periaxial tabellae straight, domed, or sagging, in 1 to 3 ranks, variously inclined gently inwards and upwards (generally), horizontal (uncom- mon), or inwards and downwards (uncommon) to axial tabellae. Documentation. UCMP holotype 37231, UCMP paratype 37232. Three thin sections and 30 polished sections from two coralla from UCMP locality D-845 were studied. Discussion. This coral is the only species that presently seems to be referable to the genus. It does not appear to be comparable to any other described Upper Paleozoic coral. A poorly preserved corallum from UCMP locality A-7096 in the Bayha section of the McCloud Limestone may belong to the same genus and spe- cies, but detailed observation of morphological characters is im- possible due to metamorphism and styolitic resorption of parts of the corallite. I have interpreted the longtudinal section as showing two se- ries of tabellae, each consisting of several ranks. The tabellae of the periaxial series are so variously inclined that there might be considered to be three series of tabellae in some parts of the section, but this is not a consistent feature. Etymology. The species is named for the Klamath Mountains. Genus Petalaxis Milne Edwards and Haime, 1852 Petalaxis allisonae n. sp. Figures 39a-b Diagnosis. A species of Petalaxis characterized by the combina- tion of small corallites, a moderately large number of major Contributions in Science, Number 337 Wilson: Permian Corals of California 65 Figure 38. Langenheimia klamathensis n. gen., n. sp. Both parts x3. Part a, UCMP holotype 3723 1 , transverse section. Part b, UCMP paratype 37_3~, 66 transverse section. septa, a very dilate columella, and a moderately broad dissepi- mentarium with few septal spines. It has smaller coraliites, fewer major septa, and a somewhat more dilate columella than P hesti n. sp. It has smaller coraliites and a somewhat less dilate col- umella than P. kennedyi n. sp. It has smaller coraliites, a more dilate columella, and a narrower dissepimentarium than P pecki n. sp. It has somewhat smaller coraliites, a more dilate columella, and a broader dissepimentarium with fewer septal spines than P. Sutherland i n. sp. It has a more dilate columella than P hrokawi, P. mokomokensis, P. dilatata, and P. occidentalis. External Description. Corallum cerioid, maximum observed di- ameter 5 cm; calyx 3 to 4 mm deep, with steeply sloping walls, small central axial boss. Transverse Section Description. Coraliites polygonal, 5 to 7 sided, 6 to 7.5 mm wide at greatest diameter; septa of 2 orders, 1 3 to 16 each, fewer minor septa where lonsdaleoid, thin in dissepimen- tarium, slightly dilate in tabularium; major septa I to 2 mm long, extending into tabularium about 1 mm where dissepimentarium lonsdaleoid, generally absent or represented by septal spines on dissepiments; minor septa fewer than major septa in most cor- allites, extending slightly into tabularium in places; axial struc- ture a columella, generally lenticular, with smooth edges, attached to counter septum, rarely to cardinal septum also, touched by other septa in some coraliites; dissepiments lons- daleoid, generally inflated, many bearing septal spines, forming false wall separating dissepimentarium and tabularium; corallite wall denticulate, about 0.2 mm wide, generally without septal spines where dissepiments highly inflated. Longitudinal Section Description. Dissepimentarium ot 1 to 3 ranks of steeply dipping, somewhat inflated dissepiments; tab- ulae generally flat and horizontal, some sloping gently inwards and upwards to columella, some sagging, in places broken up into tabellae, with outer series mostly flat and horizontal, inner series gently inclined inwards and upwards or even tent-shaped (rare); columella slightly sinuous, 0.2 to 0.4 mm wide. Documentation. Holotype UCMP 37233. Two thin sections and 10 polished sections from one corallum from UCMP locality D-803 were studied. Discussion. Petalaxis presently is distinguished by the posses- sion of predominately flat and horizontal tabulae from Acro- cyathus d’Orbigny, 1849 (a senior synonym of Lithostrotionella Yabe and Hayasaka, 1915, according to Easton, 1973), which generally has tent-shaped tabulae (Sutherland, 1977). Although there is considerable variation in the shape and inclination of tabulae in the McCloud Limestone species of Petalaxis, many are flat and horizontal, but few are tent-shaped. Some tabulae of the McCloud specimens even sag deeply, which is unlike the type species of either Petalaxis or Acrocyathus. Four species from the Permian of North America may be re- ferred to Petalaxis: P. hrokawi (Wilson and Langenheim, 1962), Nevada; P dilatata (Easton, 1960), Nevada; P. mokomokensis (Easton, 1960), Nevada; P occidentalis (Merriam, 1942), Oregon. In addition, unfigured paratypes nos. 6 and 7 (USNM nos. 174371 and 174373) of Lithostrotionella americana Hayasaka, 1 936, likely are from the McCloud Limestone exposures south of the Pit River and might be referrable to Petalaxis. I have not seen them. The holotype of L. americana is from the Mississip- pi of Kentucky, and it is improbable that the California Per- mian paratypes are placed correctly in the same species under current concepts. P. allisonae does not closely resemble any of the four Permian species listed above, all of which occur in somewhat younger rocks. It is superficially similar, in some characters, to P hesti n. sp. and P kennedyi n sp. from higher in the formation at Hirz Mountain and Potter Ridge, respectively. These two species, however, have larger coraliites, more highly denticulate walls, and generally more septa, which are more withdrawn from the columella. P allisonae is the lowest record for the genus in the McCloud Limestone. Etymology. The species is named for Dr. Carol W. Allison. Petalaxis besti n. sp. Figures 39c-f Diagnosis. A species of Petalaxis characterized by the combina- tion of large coraliites, a great number of major septa, a lenticular columella, and a very broad dissepimentarium with few septal spines. By the large number of septa alone, it can be distinguished from P allisonae, P. kennedyi n. sp., P pecki n. sp., P sutherlandi n. sp., P hrokawi, P mokomokensis, P dilatata, and P. occidentalis. External Description. Corallum cerioid, maximum observed di- ameter 8 cm; external features not preserved. Transverse Section Description. Coraliites polygonal, 5 to 7 sided, 7 to 1 1 mm wide; septa of 2 orders, 1 7 to 21 each, thin in dissepi- mentarium and tabularium; major septa 1 to 2 mm long, extend- ing into tabularium 1 to 1.5 mm; minor septa generally represented by spines on inner wall, dissepiments, and outer wall; axial structure a columella, roughly lenticular, with sinuous edges and, in places, I or 2 septal lamellae, generally attached to counter septum, not touched by other septa; dissepimentarium lonsdaleoid except in immature coraliites, with large dissepi- ments, some with septal spines; corallite wall highly denticulate, about 0.5 mm wide. Longitudinal Section Description. Dissepimentarium of 1 (rarely 2) ranks of steeply dipping, inflated dissepiments; tabulae varia- bly inclined: flat and horizontal to sinuous and sloping down- wards and inwards to columella, but generally flat and horizontal near columella; columella only slightly sinuous, 0.3 to 0.9 mm wide. Documentation. UCMP holotype 37234, UCMP paratypes 37235-37237. Four thin sections and 54 polished sections from four coralla from UCMP localities D-857 (three paratypes) and D-858 (holotype) were studied. Discussion. This species can be distinguished readily from oth- ers in the genus from the McCloud Limestone by its greater number of septa, as well as other features. P. occidentalis , from Contributions in Science, Number 337 Wilson: Permian Corals of California 67 s^V” Figure 39. Petalaxis allisonae n. sp., P. besti n. sp., and P kennedyi n. sp. All parts x3. 68 Parts a and b. P allisonae n. sp., UCMP holotype 37233, transverse (a) and longitudinal (b) sections. Parts c through f. P besti n. sp., UCMP holotype 37234, longitudinal (c-e) and transverse (f) sections. Parts g through i. P. kennedyi n. sp., UCMP holotype 37238, transverse (g, h) and longitudinal (i) sections. the Permian of Oregon, has smaller corallites and fewer septa than P. besti. Etymology. The species is named for Dr. R. V. Best. Petal axis kennedyi n. sp. Figures 39g-i Diagnosis. A species of Petalaxis characterized by the combina- tion of large corallites, a moderately great number of major septa, a highly dilate (most circular in transverse section) col- umella, and a moderately broad dissepimentarium with few sep- tal spines. It has somewhat larger corallites and a somewhat more dilate columella than P. allisonae. It has a much more dilate columella than P besti, P. pecki n. sp., P sutherlandi n. sp., P brokawi, P mokomokensis, P. dilatala, and P. occidentalis. External Description. Corallum cerioid, maximum observed di- ameter 1 3 cm; calices and wall not observed. Transverse Section Description. Corallites polygonal, 5 to 7 sided, 4 to 1 0 mm wide at greatest diameter; septa of 2 orders, 1 4 to 17 each, thin throughout or slightly dilate in tabularium; major septa 1.5 to 2.5 mm long, extending into tabularium 0.2 to 0.4 mm; axial structure a dilate, lenticular columella, generally smooth edged, attached to counter septum (rarely to cardinal septum also); dissepiments lonsdaleoid, large, some with septal spines; corallite wall denticulate, with long septal spines in places, 0.3 to 0.6 mm wide. Longitudinal Section Description. Dissepimentarium of 1 (rarely 2-3) ranks of steeply dipping, very highly inflated dissepiments; tabulae generally flat and horizontal or sloping downwards and inwards, turning up slightly at columella, rarely inwards and upwards; columella straight, about 0.3 mm wide. Documentation. UCMP holotype 37238, UCMP paratypes 37239-37241, LACMIP paratype 6424. Five thin sections and 76 polished sections from five coralla from UCMP localities D-836 (holotype 37238), D-837 (paratypes 37239-37240), and D-841 (paratype 37241 ) and LACMIP locality I 1 1 6 (paratype 6424) were studied. Discussion. P kennedyi has similar corallite diameters and num- bers of septa to P allisonae, but the latter has much longer septa and many more ranks of dissepiments. P besti n. sp. has a greater number of septa, a thicker wall, and a narrower dissepimen- tarium than P. kennedyi. The other two McCloud Limestone new species of this genus also are from Potter Ridge, but they have much more dilate columellae than P kennedyi. The species is dissimilar in many ways to the other species already known from the Permian of Oregon and Nevada. Etymology. The species is named for Dr. George L. Kennedy. Petalaxis peeki n. sp. Figures 40a-e Diagnosis. A species of Petalaxis characterized by the combina- tion of large corallites, a moderately great number of major septa, a lenticular columella, and a narrow dissepimentarium with many septal spines. It has a thinner columella than P. al- lisonae and P. kennedyi. It has a smaller number of septa and a narrower dissepimentarium with more septal spines than P. besti. It has a somewhat greater number of septa and a broader dissepi- mentarium with fewer septal spines than P. sutherlandi n. sp. External Description. Corallum cerioid, maximum observed di- ameter 8 cm; calices and wall not observed. Transverse Section Description. Corallites polygonal, 5 to 7 sided, 5.5 to 10 mm wide at greatest diameter; septa of 2 orders, 12 to 17 each, thin, generally absent in dissepimentaria of adult corallites; major septa 0.5 to I 0 mm long; minor septa well developed, about 0.2 mm long; axial structure a columella (stereocolumella) somewhat lenticular, smooth-edged, max- imum observed size about 0.6 mm wide, 1.0 mm long, attached to counter septum, rarely to other septa; dissepiments lons- daleoid, inflated, some with septal spines; corallite wall denticu- late, generally without septal spines in adult corallites, 0.2 to 0.3 mm wide. Longitudinal Section Description. Dissepimentarium of 1 to 3 ranks of elongate, steeply dipping, inflated dissepiments; tabulae slope steeply downwards and inwards to columella, becoming flat near junction or turning up slightly at columella, breaking down into tabellae in places, with outer series like clinotabellae; col- umella essentially straight, 0.5 to 0.9 mm wide. Documentation. UCMP holotype 37242, UCMP paratypes 37243-37245. Seven thin sections and 40 polished sections from four coralla from UCMP localities D-842 (holotype 37242, para- type 37243) and D-843 (paratypes 37244-37245) were studied. Discussion. P pecki has a thicker columella (stereocolumella) than any other species of the genus from the McCloud Limestone except P. sutherlandi n. sp. The latter species may be dis- tinguished by its thicker columella, generally thinner walls, septa, and dissepiments, and more open dissepimentarium as seen in transverse section. P occidentalis from the Permian Coyote Butte Formation of Oregon is similar to P pecki in cor- allite size, septal number, dilation of columella, and inclination of tabulae/tabellae, but it has a thicker wall, septa closer to the columella, and a false wall in contact with the corallite wall in more places. P. pecki is not similar to other Permian North Amer- ican species of the genus. Etymology. The species is named for Mr. Joseph H. Peck, Jr. Petalaxis sutherlandi n. sp. Figures 40f- i Diagnosis. A species of Petalaxis characterized by the combina- tion of moderately large corallites, a moderately great number of major septa, a lenticular columella, and a narrow dissepimen- tarium with many septal spines. It has somewhat larger cor- allites, a thinner columella, and a narrower dissepimentarium with more septal spines than P. allisonae. It has somewhat smaller corallites, fewer septa, and a narrower dissepimentarium with more septal spines than P. besti It has a thinner columella and a narrower dissepimentarium with more septal spines than P. kennedyi. It has somewhat fewer septa and a narrower dissepi- mentarium with more septal spines than P. pecki. It has a nar- rower dissepimentarium and a generally more lenticular colum- ella than P brokawi, P. mokomokensis, P. dilatata, and P. occidentalis. External Description. Corallum cerioid, maximum observed di- ameter 8 cm; external features not preserved. Transverse Section Description. Corallites polygonal, 5 to 7 sided. Contributions in Science, Number 337 Wilson: Permian Corals of California 69 Figure 40. Petalaxis pecki n. sp. and P Sutherland i n. sp. All parts x3. Parts a through e. P pecki n. sp. Parts a through c, UCM P holotype 37242, transverse (a) and longitudinal (b, c) sections. Parts d and e.UCMP paratype 37244, transverse (d) and longitudinal (e) sections. Parts f through i. P Sutherland i n. sp. Parts f and i, UC M P holotype 37246. transverse ( f) and longitudinal ( i ) sections. Parts g and h, UC M P paratype 37247, transverse (g) and longitudinal (h) sections. 70 Contributions in Science, Number 337 Wilson: Permian Corals of California c jfi ' b V^' .., .<0^1 >•< -5.M WW $0*1 ’ l>.' • '*$!!»? J X*rV* ml Contributions in Science, Number 337 Wilson: Permian Corals of California 71 mm Figure 41. Traskina shastensis n. gen., n. sp. and Arachnastraea fryi n. sp. All parts x3. 72 Parts a through c. T. shastensis n. gen., n. sp. Parts a and b, UCMP holotype 37253, transverse (a) and longitudinal (b) sections. Part c, UCM paratype 37254, transverse section. Parts d and e. A. fryi n. sp., UCMP holotype 37256, transverse (d) and longitudinal (e) sections. 4 to 9 mm wide at greatest diameter; septa of 2 orders, with most coral I ites having fewer minor than major septa; major septa I 2 to 15 in number, 0.4 to 1.0 mm long, generally present only in tabularium, in places represented by septal spines on corallite wall or dissepiments, thin throughout; minor septa short septal spines in tabularium, absent or reduced in number in some cor- allites, in places represented by septal spines on corallite wall and dissepiments; axial structure a columella (stereocolumella), round or somewhat oval, generally smooth edged, maximum ob- served size about 0.7 mm wide, 1.5 mm long, attached to counter septum, very rarely in contact with other septa; dissepiments very large, so few in each corallite that dissepimentarium ap- pears nearly “empty”; corallite wall thin, sinuous, apparently vestigially denticulate, about 0.2 mm wide, with septal spines in few places. Longitudinal Section Description. Dissepimentarium of 1 row of inflated, very steeply inclined dissepiments, 8 to 9 per cm; tab- ulae generally straight, sloping gently downwards and inwards toward columella, rarely sigmoid and sloping steeply down (clinotabellae-l ike); columella straight or slightly sinuous, max- imum observed diameter 0.9 mm. Documentation. UCMP holotype 37246, UCMP paratypes 37247-37252. Four thin sections and 76 polished sections from seven coralla from UCMP locality D-845 were studied. Discussion. This species probably is related to P pecki, which occurs lower in the same section. Differences are discussed under P. pecki. P. sutherlandi is not otherwise similar to other species of Petalaxis from the Permian of North America. Etymology. The species is named for Dr. Patrick K. Suther- land. Genus Traskina n. gen. Diagnosis. Traskina is a sciophylloid cerioid rugose coral with two orders of intermittently developed septa, simplified clisio- phylloid axial structure of straight to sinuous medial plate and few sinuous septal lamellae, highly lonsdaleoid dissepimen- tarium, complete denticulate wall, tabulae sloping inwards and upwards to medial plate, rarely separated into irregular axial and periaxial tabeilae with similar inclination. Type Species. Traskina shastensis n. sp. Description. As for the type and only known species. Discussion. Traskina is superficially similar to Kleopatrina ( Por - firievella) Minato and Kato, 1965b, but differs by having only vestigial septa (where developed at all), well-developed tabulae, and, in places, poorly zoned tabeilae whereas the latter taxon has strongly developed septa and two well-defined zones of tabeilae. The axial structures of Traskina are a medial plate and septal lamellae, both of which are generally sinuous (even zigzag in places), whereas the same structures of Kleopatrina are more nearly straight. Finally, all the skeletal elements of Traskina are the thin sciophylloid kind whereas those of Kleopatrina are robust. Traskina has vestigial septa and thin skeletal elements similar to Sciophyllum Flarker and McLaren, 1950, known from the Upper Mississippian (?) near the Yukon-Alaska boundary, the Lower Pennsylvanian of Japan, and the Lower Permian of east- central Nevada (Wilson and Langenheim, 1962:514) and possi- bly the Pennsylvanian and Lower Permian of the Ural Moun- tains, Russia (as Thysanophyllum aseptatum Dobrolyubova, 1936). Flowever, Traskina has an axial structure and Scio- phyllum has none. Etymology. The genus is named for Dr. John B. Trask. Traskina shastensis n. sp. Figures 41 a-c Diagnosis. Because the genus Traskina is erected as monotypic, the diagnoses of the genus and of the type species are iden- tical. External Description. Corallum cerioid, maximum observed di- ameter 10 cm; calices and epitheca not seen. Transverse Section Description. Corallites polygonal, 5 to 7 sided, 5.5 to 8 mm wide at greatest diameter; septa of 2 orders; major septa number 13 to 15 where well developed (rare), generally numbering fewer than 10, conspicuously absent in parts of most corallites, total length 0.4 to 1.0 mm, extending into tabularium 0.2 to 0.5 mm, generally present as spines on corallite wall and false wall; minor septa rare, present as septal spines, absent in tabularium; axial structure clisiophylloid, of straight (uncom- mon), sinuous, or zigzag, thin or somewhat dilate medial plate crossed by 1 to 4 sinuous or zigzag septal lamellae, with medial plate rarely attached to counter septum; dissepiments highly lonsdaleoid, not discernible in many corallites, so that dissepi- mentarium appears nearly structureless; corallite wall 0. 1 to 0.2 mm thick, straight or somewhat denticulate, with septa originat- ing from V-shaped depressions. Longitudinal Section Description. Dissepimentarium of 1 to 2 ranks of highly inflated, steeply dipping, cystose dissepiments; tabulae straight or slightly domed, with straight or slightly down- turned peripheral ends, sloping gently inwards and upwards to columella, 18 to 20 per cm, rarely broken down into tabeilae; columella straight to somewhat sinuous. Documentation. UCMP holotype 37253, UCMP paratype 37254. Five thin sections and 42 polished sections from two cor- alla from UCMP locality D-843 were studied. Discussion. There does not seem to be any described species that should be placed in Traskina with the type species. Etymology. The species is named for Shasta Lake. Cerioid-Astreoid Rugosa Genus Arachnastraea Yabe and Hayasaka, 1916 Arachnastraea fergusoni n. sp. Figures 42a-e Diagnosis. A species of Arachnastraea characterized by the combination of very large corallites, well-developed and rela- tively long minor septa, and a complex axial structure. It has larger corallites, better developed minor septa, and a more com- plex axial structure than A. fryi n. sp., A. molli. and A. manchurica. External Description. Corallum cerioid-astreoid; maximum ob- served diameter 1 1 .5 cm; calices deep with prominent axial boss; external features not well preserved. Contributions in Science, Number 337 Wilson: Permian Corals of California 73 74 Figure 42. Arachnastraea fergusoni n. sp. All parts x3. Parts a through c, LAC MI P holotype 6425, transverse (a) and longitudinal (b, c) sections. Parts d and e, UCMP paratype 37255, transverse sections. Transverse Section Description. Corallites polygonal, 5 to 7 sided, 1 0 to 16 mm wide at greatest diameter; septa of 2 orders, 1 6 to 1 8 each; major septa 2.5 to 5 mm long, extending into tabularium 0.5 to 1.5 mm, generally slightly dilate in tabularium, most touching axial structure but not confluent with axial tabellae; minor septa well developed, extending across dissepimentarium and penetrating tabularium as dilate spines; axial structure arachnoid, circular, 2 to 3.5 mm wide, generally symmetrical, with well-developed medial plate and septal number equal to or greater than number of septa, tightly connected by tabellae; dis- sepimentarium regular; dissepiments straight, angulo-con- centric, or pseudoherringbone; corallite wall cerioid- astreoid, very discontinuous, about 0.3 mm wide. Longitudinal Section Description. Dissepimentarium of 3 to 6 ranks, various sized cystoid or elongate, gently dipping dissepi- ments, about 22 per cm; tabellae of 2 well-defined ranks, axial and periaxial; axial tabellae elongate, of 2 to 4 steeply dipping ranks, 38 to 40 per cm; periaxial tabellae straight, slightly domed with downturned edges, gently dipping, 20 to 28 per cm; wall about 0.2 mm wide where present. Documentation. LACMIP holotype 6425, UCMP paratype 37255. Five thin sections and 29 polished sections from two cor- alla from LACMIP locality 1 125 (holotype 6425) and UCMP locality D-852 (paratype 37255) were studied. Discussion. The UCMP paratype of A. fergusoni is a small cor- allum (4 cm diameter) and has smaller corallites (diameter 6 to 1 0 mm) and slightly fewer major septa ( 1 3 to 1 6) than the holotype. 1 consider it to be a juvenile corallum, although the possibilities exist that it may belong to another species or repre- sent variation within the species. Both coralla were closely associ- ated in the same bed. A. fryi n. sp. from the Potter Ridge section of the McCloud Limestone has smaller corallites, a more continuous wall, and a less complex axial structure than A. fergusoni. A. molli (Stuck- enberg, 1888) from the Middle Carboniferous of Russia and A. manchurica Yabe and Hayasaka, 1916 (the type species), from the Lower Permian (?) of Manchuria have smaller corallites (di- ameter 6 to 8 mm), simpler axial structures, septa more continu- ous with the axial lamellae, and less well-developed minor septa. This is the first record for the genus in the Americas. Etymology. The species is named for Mr. Eugene Ferguson. Arachnastraea fryi n. sp. Figures 41 d-e Diagnosis. A species of Arachnastraea characterized by the combination of moderately large corallites, well-developed mi- nor septa, and a moderately complex axial structure. It has smaller corallites, somewhat shorter minor septa, and a some- what less complex axial structure than A. fergusoni. It has larger corallites, a more complex axial structure, and better developed minor septa than A. molli and A. manchurica. External Description. Corallum cerioid-astreoid, maximum ob- served diameter 12 cm; calyx not observed. Transverse Section Description. Corallites polygonal, 5 to 7 sided, 6.5 to 10 mm wide at greatest diameter; septa of 2 orders, 1 5 to 17 each; major septa 2.5 to 3.5 mm long, extending into tab- ularium 0.5 to 1 .0 mm, thin throughout or slightly dilate in tab- ularium, generally connected with septal lamellae of axial structure; minor septa well developed, extending across dissepi- mentarium and barely penetrating tabularium; axial structure arachnoid, with thin, well-developed medial plate, septal lamellae generally equal in number to and confluent with major septa; dissepimentarium regular; dissepiments straight, con- centric, angulo-concentric, or pseudoherringbone; corallite wall cerioid-astreoid, about 0.2 mm thick. Eongitiduinal Section Description. Dissepimentarium of 4 to 6 ranks of gently inclined dissepiments, about 33 per cm; tabellae of 2 ranks, axial and periaxial; axial tabellae sloping either steep- ly or gently inwards and upwards to columella; periaxial tabellae sloping gently inwards and upwards to axial tabellae; columella slightly sinuous; wall clearly shows interruptions. Documentation. UCMP holotype 37256. Two thin sections and 12 polished sections from one corallum from UCMP locality D-837 were studied. Discussion. In several features, Arachnastraea fryi appears to be close to A. molli from the Middle Carboniferous of Rus- sia, as figured by Hill (1956, figs. 194, 3a, 3d) and Soshkina, Dobrolyubova, and Kabakovich ( 1 962:343, figs. 1 06a, 1 06b). All illustrations show remarkably similar axial structures, partially dilate septa, and denticulate walls, where present. The specimen figured by Hill shows corallites with as many as 21 major septa and apparently no minor septa, whereas that of Soshkina, Dobrolyubova , and Kabakovich has well-developed minor septa and major septa numbering about 1 4. They may represent differ- ent species. Because it has fewer septa and well-developed minor septa, A. fryi is not specifically close to the specimen of A. molli figured by Hill. The specimen figured by Soshkina, Dobrolyubova, and Kabakovich has somewhat fewer septa than A. fryi. but their A. molli has septa that are dilate in the dissepimentarium of some corallites. A. fryi has no dilate septa in the dissepimentarium and its dissepimentarium is much wider than that shown in the Soshkina, Dobrolyubova, and Kabakovich illustrations. Differences with A. fergusoni n. sp. also from the McCloud Limestone, are listed in the discussion of that species. Etymology. The species is named for Dr. Wayne L. Fry. Order Tabulata Milne Edwards and Haime, 1850 Genus Bayhaium Langenheim and McCutcheon, 1959 Bayhaium merriamorum Langenheim and McCutcheon, 1959 Figures 43a-g Bayhaium merriamorum Langenheim and McCutcheon, 1959: 100, pi. 19, figs. 1-6. Bayhaium merriamorum Langenheim and McCutcheon. Lafuste, 1963:1 127, text-figs. 1-3. Documentation. UCMP hypotypes 37257-37260. Five thin sec- tions and 14 polished sections from one corallum from UCMP Contributions in Science, Number 337 Wilson: Permian Corals of California 75 Figure 43. Bayhaium merriamorum Langenheim and McCutcheon, 1959. All parts x3. Parts a and b, UCMP hypotype 37257, transverse (a) and 76 longitudinal (b) sections. Part c, UCMP hypotype 37258, transverse section Parts d and e, UCMP hypotype 37259, transverse (d) and longitudinal (e) sections. Parts f and g, UCMP hypotype 37260, transverse (f) and longitudinal (g) sections. locality D-181 (hypotype 37257), four thin sections and four polished sections from one corallum from UCMP locality D-838 (hypotype 37258), and four thin sections and 21 polished sec- tions from two coralla from UCMP locality D-839 (hypotypes 37259-37260) were studied. Discussion. Since a very active search was made by me for this coral and 1 found that it occupied narrow stratigraphic ranges in the Bayha and Potter Ridge sections, I conclude that these occur- rences represent a biostratigraphic correlation between the sec- tions and probably the intervening type locality. Bayhaium virginiae n. sp. Figures 44a-f Diagnosis. A species of Bayhaium characterized by the com- bination of thick walls in the immature regions and the rarity of septa and tabulae. It has thicker walls in the immature region and far fewer septa and tabulae than B merriamorum. It has far fewer septa and tabulae than B. vallum. External Description. Corallum cerioid-fasciculate, generally hemispherical, maximum observed diameter 4 cm, outer cor- allites cylindrical, closely spaced. Transverse Section Description. In immature portion, corallites cerioid, generally 1 to 1 .5 mm in diameter ( 1 is 2 mm), generally without septa ( 1 has 3 short septa), with sclerenchyme-thickened walls (generally about 0.3 mm thick on each side of separating dark line, 1 thickened Multithecopora-Wke to 0.9-mm-thick wall and constricted 0.3-mm thick tabularium), with rare infun- dibuliform tabulae about 0. 1 6 mm thick, no apparent connecting tubules. In mature portion, corallites 1 to 1 .5 mm diameter, 0.6 to 1 .0 mm apart, with walls 0.3 to 0.5 mm thick. Longitudinal Section Description. In mature portion, as above, with randomly clumped tabulae, 2 to 4 per mm, generally con- cave, rarely tabellae or infundibuliform; connecting tubules tun- nel-like, rare, randomly scattered, with 0.4- to 0.5-mm wide openings with tabulae passing through; calices as much as 6 mm from top to first tabulae below. Documentation. UCMP holotype 37261, UCMP paratype 37262. Six thin sections and nine polished sections from two coralla from UCMP locality D-852 were studied. Discussion. Bayhaium merriamorum. the type species of the genus, is known only from the type locality and nearby areas in the McCloud Limestone, which are stratigraphically higher in the formation and south of the locality where B virginiae n. sp. was collected. The latter species has much thicker walls in the immature region, far fewer septa where developed at all, and far fewer tabulae than B. merriamorum. B vallum Hoare, 1964, from the Wolfcampian Sunflower Formation of Elko County, Ne- vada, is more similar to B. merriamorum than to B. virginiae. The latter species can be distinguished readily from B vallum by many of the same features that distinguish it from B. mer- riamorum. especially the lack of, or poor development of, septa and tabulae. However, both B. virginiae and B. vallum have thicker walls in the immature parts of the corallum than has B merriamorum. This is the third species of Bayhaium erected. The genus has been reported only from the Lower Permian (Wolfcampian) Cor- dilleran eugeosynclinal rocks of California and Nevada. Etymology. The species is named for Virginia A. McCutcheon Langenheim. Genus Enigmalites Tchudinova, 1975a Enigmalites roherti n. sp. Figures 44g-h; 45a-c Diagnosis. A species of Enigmalites characterized by the com- bination of large corallites, poorly developed septal spines and connecting processes, and relatively abundant dissepiments and tabulae. It has fewer septal spines, fewer connecting processes, and more dissepiments and tabulae than E. lectus. It has larger corallites than E. tschernyschewi, E. lautus, and E. largus. External Description. Corallum basally reptant, distally erect, phaceloid, irregularly hemispheroidal, maximum observed height 9 cm, width 9 cm (both incomplete); epitheca with coarse rugae; connecting tubules rare. Internal Description. Corallites 2.2 to 2.5 mm in diameter, touch- ing to as much as 1 corallite diameter distant; wall 2-layered, 0.2 to 0.4 mm thick, epitheca microstructure not observed, scle- renchyme layer fibrous, with fibers radially oriented, at right angles to epitheca in transverse and longitudinal sections, not bundled into trabeculae; dissepiments numerous, 1 to 3 ranks, steeply dipping in longitudinal section, circling the interior of the wall in transverse section, leaving central or slightly off-centered tabularium; tabularium open or rarely crossed by flat or slightly domed tabulae; septal spines short, conical, irregularly devel- oped, not present in all corallites, maximum number of 1 4 where fully developed (rare), generally only a few present on wall, on sclerenchyme, and on dissepiments (lonsdaleoid-like), pointing gently inwards and upwards in longitudinal sections; connecting tubules and mural pores very rare. Documentation. UCMP holotype 37263, UCMP paratypes 37264-37265. Six thin sections and 57 polished sections from three coralla from UCMP locality A-7101 were studied. Discussion. Tchudinova ( 1 975a: 13) based this genus on a new species and a species of Stuckenberg (1895:19) and later ( 1975b:429) placed two other new species in it. These specimens were collected in the Ural Mountains. Tchudinova’s three species are Lower Permian and Stuckenberg’s species is Upper Carboniferous. The genus is characterized by the combination of an epitheca lined by a thin layer of sclerenchyme, incomplete tabulae taking the form of dissepiments and leaving a tabularium-like central area that is occupied in places by tabulae, septal spines, and rare connecting processes. Probably some species presently placed in Syringopora will be assigned by other workers to Enigmalites. E. roberti is the first report of the genus outside of Russia. E. roberti is close to E. lectus Tchudinova, 1975a, the type species, in corallite size and thickness of the wall. E. roberti has fewer and less well developed septal spines, many present on the dissepiments, rarer connecting processes, more abundant dis- sepiments, and more tabulae than E. lectus. Both E. roberti and E. lectus have larger corallites than E. tschernyschewi (Stucken- Contributions in Science, Number 337 Wiison: Permian Corals of California 77 78 Contributions in Science, Number 337 Wilson: Permian Corals of California berg, 1895), E. lautus Tchudinova, 1975b, and E. largus Tchudinova, 1 975b. Etymology. The species is named for Mr. Robert B Spangenberg. Genus Michelinia de Koninck, 1 841 Michelinia nelsoni n. sp. Figures 45d -h Diagnosis. A species of Michelinia characterized by the com- bination of moderately large corallites, lack of mural pores, and presence of generally complete tabulae. It has smaller corallites than M. hranneri, M. harkeri, and M. referta. It has larger cor- allites than M. eugeneae. M. subcylindrica, and M. tenuicula. It has no mural pores such as those present in M. exilimura, M. latebrosa, M. scopulosa, and M. spissata. External Description. Corallum cerioid, small, maximum ob- served diameter 2.1 cm, roughly hemispherical, covered ven- trally by wrinkled corailite wall; calices 3.5 to 4.5 mm wide, 2 to 5 mm deep, with flat floors and smooth, slightly flaring walls. Transverse Section Description. Corallites polygonal, 5 to 7 sided, 3.5 to 4.5 mm wide; mural pores not observed, apparently absent; combined thickness of 2 corailite walls 0.3 to 0.5 mm Longitudinal Section Description. Tabulae generally complete, ir- regularly spaced, much thinner than corailite walls, about 0.06 mm thick, 0.5 to 1.5 mm apart, generally flat and horizontal, some slightly sagging or domed, some at angle across corailite. Documentation. UCMP holotype 37266, UCMP paratypes 37267-37269, LACMIP paratypes 6426-6429. Six thin sections and 16 polished sections from five coralla and three etched cor- alla from UCMP locality D-852 (holotype 37266, paratypes 37267-37269) and LACMIP locality 1125 (paratypes 6426- 6429) were studied. Discussion. No species of Michelinia or its closely related and perhaps synonymous genus Pleurodictyum Goldfuss, 1829, has been described previously from the Permian of North America (Sando, 1 974, 1 980a; Wilson, 1 974), although Rowett ( 1 969: 15) reported Michelinia sp. in the Permian of Alaska, but did not describe it in that paper as he stated he had. Of the 1 1 species described from the Pennsylvanian of North America (Sando, 1974, 1980a), only four have corailite diameters comparable to those of M. nelsoni. These are M. exilimura Mather, 1915, from the Morrow Formation near Choteau, Oklahoma, M. latebrosa Moore and Jeffords, 1945, from the Otterville Limestone near Ardmore, Oklahoma, and elsewhere in Texas, northwestern Arkansas, and northeastern Oklahoma, M. scopulosa Moore and Jeffords, 1945, from the Lower Pennsylvanian of northeastern Oklahoma and northwestern Arkansas, and M. spissata Moore and Jeffords, 1945, from the Brentwood Limestone near Brent- wood, Arkansas. M nelsoni apparently lacks mural pores, whereas all of the above Pennsylvanian species have them vari- ously developed from “not abundant” to “very numerous.” In addition, the tabulae of M. spissata are mostly incomplete, whereas those of M. nelsoni are mostly complete. Etymology. The species is for Dr. Samuel J. Nelson. Genus Neomultithecopora Lin, 1963 Neomultithecopora sandoi n. sp. Figures 46a-f; 47a-f Diagnosis. A species of Neomultithecopora characterized by the combination of relatively large corallites, clumped tabulae, and rare connecting tubules or mural pores. It has larger cor- allites than N. syringoporoides, N. simplex. N. uralica, and N. repens. It has smaller corallites than N. berkhi. External Description. Corallum basally reptant, distal ly erect, phaceloid, irregularly hemispheroidal, maximum observed height 9 cm, width 7 cm (both incomplete); epitheca with fine and coarse rugae. Internal Description. Corallites 1.8 to 2.6 mm diameter (gener- ally about 2.0 mm), touching to as much as 2 corailite diameters distant; tabulae distally concave, convex, or elongate and cyst- like, clumped in groups of as many as 6 in 2-mm vertical space, generally only I shape in a clump but all shapes may occur in a single corailite, present where sclerenchyme is thin or thick, ab- sent in most parts, some crossing through openings connecting corallites; corailite wall distally thin, elsewhere thickened by sclerenchyme to as much as 1 1 mm thick, leaving tabularium as narrow as 0.1 mm; septal spines regularly superposed and jux- taposed grid-like in longitudinal and horizontal rows, conical, 0. 1 to 0.2 mm long, about 0. 1 mm diameter at base, dark brown color contrasting with lighter sclerenchyme, in transverse section pres- ent on corailite walls and in some on inner layers of sclerenchyme (as many as 30 on former, 15 on latter), in longitudinal section pointing inwards and gently upwards, about 0.1 mm apart, in tangential longitudinal section circular cross sections of spines in 2 or more vertical rows, about 4 or 5 in 1 mm vertically, spines in parallel rows regularly opposite one another; septal spines not present in every corailite or everywhere in corallites that have them; connecting processes scattered, rare, becoming mural pores between touching corallites or, more rarely, connecting tu- bules as long as 2.5 mm, with diameters to 1 . 1 mm between more distant corallites; sclerenchyme composed of minute, elongate fibers at right angles to corailite wall in both transverse and longitudinal sections, not bundled into trabeculae. Documentation. UCMP holotype 37270, UCMP paratypes 37271-37279, LACMIP paratypes 6430-6438. Eighteen thin sections and 202 polished sections from 19 coralla from UCMP Figure 44. Bayhaium virginiae n sp. and Engimalites roberti n. sp. All parts x3. Parts a through f. B virginiae n. sp. Parts a through d, UCMP holotype 37261, transverse (a-c) and longitudinal (d) sections. Parts e and f, UCMP paratype 37262, transverse and longitudinal sections. Parts g and h. E roberti n. sp., UCMP holotype 37263, transverse sections. Contributions in Science, Number 337 Wilson: Permian Corals of California 79 Figure 45. Engimalites roberti n. sp. and Michelinia nelsoni n. sp. All parts x3. 80 Parts a through c. E. roberti n. sp. Part a, UCMP holotype 37263, longitudinal section. Parts b and c, UCMP paratype 37264, transverse (b) and longitudinal (c) sections. Ar'x/fiP Parts d through h. M. nelsoni n. sp. Parts d through g, UCMP holotype 37266, transverse (d, g) and longitudinal (e, f) sections. Part h, LAC. paratype 6426, longitudinal section. Figure 46. Neomultithecopora sandoi n. sp. All parts x3. Parts a through c, UCMP holotype 37270, transverse (a, b)and longitudinal (c) sections. Parts d through f, UCMP paratype 37275, transverse (d) and longitudinal (e, f) sections. 81 Figure 47. Neomultithecopora sandoi n sp All parts x3 Parts a and b, UCMP paratype 37277, transverse (a) and longitudinal (b) sections Parts c and 82 d, UCMP paratype 37278, transverse (c) and longitudinal (d) sections. Parts e and f, LACM IP paratype 6438, transverse and longitudinal sections (e) and longitudinal section (f). localities D-80I (paratypes 37271-37273), D-803 (holotype 37270), D-822 (paratype 37274), D-831 (paratypes 37275- 37276), D-832 (paratype 37277), and D-858 (paratypes 37278- 37279) and LACMIP localities 1125 (paratypes 6430-6431), 1 1 32 (paratypes 6432-6433), 1 1 33 (paratypes 6434-6437), and 4457 (paratype 6438) were studied. Discussion. This is the first report of Neomultithecopora in North America and the first record of it in the Permian. The type species, N. syringoporoides Lin, 1963, is from the Lower Car- boniferous of south China. Other species occur in the Lower Carboniferous, and questionably Upper Carboniferous, of the Ural Mountains and Novaya Zemlya of Russia (Kachanov, 1967:24; Tchudinova, 1 97 5a: 17). This genus is separated from Multithecopora Yoh, 1927, chiefly on the basis of the presence of very symmetrically ar- ranged septal spines in the former and perhaps also by micro- structural differences of the sclerenchyme. Multithecopora seems to have sclerenchyme arranged in concentric layers, whereas Neomultithecopora has sclerenchyme composed of radi- ally arranged fibers. The only other tabulate coral with highly thickened walls from the McCloud Limestone, Bayhaium Lan- genheim and McCutcheon, 1959, also has sclerenchyme com- posed of calcite fibers (Lafuste, 1 963), but they differ from those of Neomultithecopora in being bundled into trabeculae. Some of the coraila of N. sandoi have wholly or partially si 1- icified walls. The silicified parts are a pale blue, whereas the septal spines are brown. The septal spines arc most obvious when observed in reflected light but not always apparent when exam- ined in thin section by transmitted light. N. sandoi has greater corallite diameters than N. syr- ingoporoides from the Lower Carboniferous of south China, N. simplex Kachanov, 1967, from the Lower Carboniferous of the Ural Mountains, and N. repens (Stuckenberg, 1895) from the Upper Carboniferous of the Ural Mountains. The latter species is here placed in the genus based on the report of Tchudinova ( 1 975a: 1 7) that it has septal spines. N. berkhi (Gorskiy, 1951), from the Lower Carboniferous of Novaya Zemlya and the Ural Mountains (Kachanov, 1967:29) has greater corallite diameters than N. sandoi. N. uralica Kachanov, 1967, from the Lower Carboniferous of the Ural Mountains and Novaya Zemlya, has abundant connect- ing tubules, whereas N. sandoi has very few. Etymology. The species is named for Dr. William J. Sando. Genus Syringopora Goldfuss, 1826 Syringopora mccutcheonae Wilson and Langenheim, 1962 Figures 48a-b Syringopora mccutcheonae Wilson and Langenheim, 1962:515, pi. 89, figs. 1 1-13. S. mccutcheonae Wilson and Langenheim. Langenheim and Langenheim, 1965:235. Documentation. UCMP hypotype 37280. Five thin sections and 35 polished sections from one corallum from UCMP locality D-842 were studied. Discussion. This species is geographically widespread in Lower Permian strata of eastern Nevada (Wilson and Langenheim, 1962; Langenheim and Langenheim, 1965). The McCloud Limestone corallum agrees very closely with the Nevada speci- mens in corallite diameter, absence of septa and septal spines, and general occurrences of connecting processes at the same levels in the coraila. Apparent differential silici fication in some corallites gives the impression that sclerenchyme-embedded septa are present. Syringopora multattenuata McChesney, 1 860 Figures 48c-f Syringopora multattenuata McChesney, 1860:75. Syringopora multattenuata McChesney, 1867:2, pi. 2, fig. 4. Syringopora multattenuata McChesney. McCutcheon, 1961:1014, pi. 121, figs. 1-8. Documentation. UCMP hypotypes 37281 -37282. Six thin sec- tions and 30 polished sections from one corallum from UCMP locality D-825 (hypotype 37281) and two thin sections and 26 polished sections from one corallum from UCMP locality D-853 (hypotype 37282) were studied. Discussion. This species was thoroughly reviewed by McCutcheon (1961), who summarized the geographic occur- rence as “widespread in Pennsylvanian rocks of the United States ... in the Permian of Spitzbergen ... in the Wolfcampian of the Arrow Canyon Range, Clark County, Nevada.” The McCloud Limestone specimens agree closely with the neotype and hypotypes of McCutcheon (1961), except that the corallite diameters of the McCloud specimens, while generally about 2 mm, do range as high as 2.5 and 2.8 mm in a few corallites, which is somewhat greater than the 2.2-mm maximum cited by McCutcheon ( 1961 .1014). Sando (1965:32) commented on the differences and sim- ilarities of 5. multattenuata and S. occidentalis Meek, 1877, from the Pennsylvanian Weber Sandstone or Permian Park City Formation of Utah, and concluded that “. . . in Syringopora multattenuata the axial tube is more clearly defined and the incomplete tabulae are more numerous, more variable in size, and generally more inflated than in S. occidentalis." LOCALITIES The following localities are entered in the UCMP and LACMIP locality registers. They all are located in the McCloud Lime- stone. Abbreviations also are used for United States Geological Survey (USGS), township (T), range (R), north (N), south (S), east (E), and west (W). References to fusulinid zones are those of Skinner and Wilde ( 1 965), with age determinations amended by Wilde (197 1:364). UCMP Localities A-7096. Bayha section (lower part is the Tunnel Inn section of Skinner and Wilde, 1965). It is on the ridge most likely collected by the California Geological Survey in 1 862. The traverse begins at the lowest exposure near the base of the W face of the ridge with summit marked “1693” in the NE !4 of section 1 5, T 33 N, Contributions in Science, Number 337 Wilson: Permian Corals of California 83 Figure 48. Syringopora mccutcheonae Wilson and Langenheim, 1962, and S. multattenuata McChesney, 1859. All parts x3. 84 Parts a and b. S. mccutcheonae, UCMP hypotype 37280. transverse (a) and longitudinal (b) sections. Parts c through f. 5. multattenuata. Parts c and d, UCMP hypotype 37281, transverse (c) and longitudinal (d) sections. Parts e and f, UCMP hypotype 37282, transverse (e) and longitudinal (f) sections. R 4 W, as shown on the USGS 7.5-minute topographic quad- rangle of Project City, California (1957), continues due E over the summit of the ridge, down the E dip slope into cover, offsets slightly S and continues up a smaller limestone hill, down again into cover and ends at the second of two 2-foot (0.6-meter) thick limestone units, which crop out of deep soil in the saddle shown in the NW Vi of section 14, T 33 N, R 4 W, of the same map. Thicknesses recorded for this section include sills of quartz di- orite found in it. A-7096 is 500 feet ( 1 52.4 meters) above the base of the section and has abundant cerioid corals, small solitary corals, and rare, large pelmatozoan columnals. Fusulinid zone F. A-7101. W bank of the McCloud River at river level and N of the McCloud River bridge as shown in the NE 'A of section 3 1 , T 35 N, R 3 W, of USGS 15-minute topographic quadrangle of Bollibokka Mountain, California (1957). Silty limestone with some coquinoid beds of echinoid plates and spines, syringoporid corals, fusulinids. Same as LACMIP locality 1123. Fusulinid zone F. B-4837. This locality is in the NE 1 A of the SW ! A of section 35, T 33 N, R 4 W, as shown on the USGS 7.5-minute topographic quandrangle of Project City, California (1957). It is a nearly covered, small limestone outcrop N of a small tributary that flows into the East Fork of Stillwater Creek. A single corallum of a fasciculate coral was collected here by Mr. Chang Ki Elong of Korea in 1962 and brought to me at UCMP for identification. The matrix on the coral has a specimen of the fusulinid Pseudo- schwagerina robusta (Meek, 1864). I examined the locality but found no other fossils. Fusulinid zone E. D-181. Bayha section, 415 feet ( 1 26.4 meters) above the base. Cerioid tabulate corals, cerioid rugose corals. Fusulinid zone F. D-800. Lowest black limestone unit on spur SE of the forest service building 1 A mile S of the McCloud River bridge on Gil- man Road as shown in the SE 1 A of section 32, T 35 N, R 3 W, of the USGS 15-minute topographic quadrangle of Bollibokka Mountain, California (1957). Large solitary rugose corals, Omphalotrochus, Husledia Hall and Clarke, 1893, fusulinids, fenestellid bryozoans. Fusulinid zone A. D-801. Estimated 300 feet (91.4 meters) stratigraphically above D-800. Syringoporid corals to 2 feet (0.6 meters) in diame- ter, large solitary rugose corals, Omphalotrochus. Fusulind zone B. D-803. Float collection made an estimated 700 feet (213.4 meters) stratigraphically above D-801 . Large solitary rugose cor- als, cerioid rugose corals, syringoporid corals. Fusulinid zone B or C. D-804. An estimated 150 feet (45.7 meters) stratigraphically above D-803 on spine of spur. Fasciculate rugose corals. Fusulinid zone C. D-805. Due E of D-804 on spine of spur an estimated 300 feet (9 1 .4 meters) stratigraphically below summit of main N-S ridge. Fasciculate rugose corals. Fusulinid zone B. D-819. McCloud Bridge East section. Probably same as Skin- ner and Wilde’s (1965) McCloud Guard Station section. This traverse begins at the lowest exposures of the formation on the first spur S of the E end of the McCloud River bridge as shown in the NE Vi of the SE 'A of section 3 1 , T 35 N, R 3 W, of USGS 1 5- minute topographic quadrangle of Bollibokka Mountain, Cal- ifornia (1957), and continues E over the top of the high ridge, stopping at cover in the saddle at the east end of the dip slope as shown in the NE 'A of the SW 'A of section 32, T 35 N, R 3 W, same map. D-819 is 1 140 feet (347.5 meters) stratigraphically above the base of the section. Many silicified fossils, large soli- tary corals, Conocardium langenheimi, Omphalotrochus, fusulinids. Same as LACMIP locality 4458. Fusulinid zone A. D-821. Wide bench below sheer cliff 410 feet (125 meters) stratigraphically above D-8 1 9. Fusulinids, solitary rugose corals, syringoporid corals, brachiopods, bryozoans, bivalves. Same as LACMIP locality 1 132. Fusulinid zone B. D-822. Bench 34 feet (11 meters) stratigraphically above D-831. Large solitary rugose corals, syringoporid corals. Fusulinid zone B. D-825. Beds 310 feet (94.5 meters) stratigraphically above D-822. Syringoporid corals. Fusulinid zone C. 0-826. Bench 130 feet (40 meters) stratigraphically above D-825. Large solitary rugose corals. Fusulinid zone C. 0-827. Beds 20 feet (6.1 meters) stratigraphically above D-826. Fusulinids and cerioid rugose corals. Fusulinid zone C. 0-828. Topographic summit of main ridge of section, 670 feet (204 meters) above D-827. Fusulinids, fasciculate rugose corals. Fusulinid zone C. 0-831. Approximately 200 yards ( 1 83 meters) S of D-8 1 9 and about 15 feet (4.6 meters) stratigraphically higher. Fusulinids, abundant large solitary rugose corals, abundant syringoporid corals, Conocardium langenheimi. Same as LACMIP locality 1 1 33 Fusulinid zone A. 0-832. Approximately 1 00 yards (91.4 meters) S of D-83 1 and an estimated 10 feet (3 meters) stratigraphically higher. Syr- ingoporid corals and large solitary rugose corals. Fusulinid zone A. D-834. Potter Ridge section. The traverse begins at the lowest exposure of the formation on the W end of the ridge between Potter and Marble Creeks as shown in the SE 'A of section 23, T 33 N, R 4 W, USGS 15-minute topographic quadrangle of La- moine, California ( 1 957), and continues up the ridge crest to the base of the intrusive overlying the last limestone unit in the sad- dle E of the nob on the ridge as shown in the N W ]A of the SW ]A of section 24, T 33 N, R 4 W, same map. Thicknesses recorded for this section include sills of quartz diorite. D-834 is 130 feet (39.6 meters) above the base of the section. Fasciculate rugose corals. Fusulinid zone E. D-836. Beds 80 feet (24.4 meters) stratigraphically above D-834. Fasciculate and large solitary rugose corals. Same as LACMIP locality 1114. Fusulinid zone E. D-837. Beds 25 feet (7.6 meters) stratigraphically above D-836. Fusulinids, cerioid and cerioid-astreoid rugose corals. Fusulinid zone E. D-838. Beds 65 feet (19.8 meters) stratigraphically above D-837. Fasciculate rugose corals, cerioid syringoporid corals, Omphalotrochus. Fusulinid zone E. D-839. Beds 10 to 35 feet (3 to 10.8 meters) stratigraphically above D-838. Cerioid syringoporid corals. Fusulinid zone F. D-840. Beds 230 to 340 feet (70 to 103.6 meters) strat- Contributions in Science, Number 337 Wilson: Permian Corals of California 86 igraphically above D-839. Fasciculate rugose corals. Fusulinid zone F. D-841. Beds 285 feet (86.8 meters) stratigra phi ically above D-840. Fusulinids, solitary and cerioid rugose corals. Fusulinid zone F. D-842. Beds 320 feet (97.5 meters) stratigraphically above D-84 1 . Large solitary rugose corals with and without axial struc- tures, cerioid rugose corals, syringoporid corals, large beller- ophontid gastropods. Same as LACMIP locality 1116. Fusulinid zone F. D-843. Beds 20 feet (6.1 meters) stratigraphically above D-842. Fusulinids, large solitary and cerioid rugose corals, gas- tropods. Fusulinid zone F. D-845. Beds 105 feet (31.9 meters) stratigraphically above D-843. Large cerioid rugose corals. Fusulinid zone F. D-852. Summit of H irz Mountain, dip slope E of the foot of the lookout tower as shown in the SE Va of section 7, T 35 N, R 3 W, of the USGS 15-minute topographic quadrangle of Bollibokka Mountain, California (1957). Very rich fauna including fusulin- ids, syringoporid corals, solitary, fasciculate, cerioid-astreoid, and cerioid rugose corals, brachiopods, and gastropods. Same as LACMI P locality 1 1 25. Probably fusulinid zone D. D-853. Hirz Mountain section. The traverse begins at the lowest exposure of the formation of the NW face of the N end of the ridge that forms the summit of Hirz Mountain, as shown in the SE Va of section 7, T 35 N, R 3 W, of the USGS 15-minute topographic quadrangle of Bollibokka Mountain, California (1957) and continues E over the N end of the summit ridge and down the E face into cover as shown in the SE Vi of section 7 and the SW Va of section 8, T 35 N, R 3 W, of the same map. D-853 is 125 feet (37.1 meters) stratigraphically above the base. Syr- ingoporid corals. Fusulinid zone D. D-855. Beds 420 feet (128 meters) stratigraphically above D-853. Fusulinids, cerioid and fasciculate rugose corals, brachiopods. Fusulinid zone D. D-857. Beds 240 feet (73 meters) stratigraphically above D-855. Fusulinids, cerioid and fasciculate rugose corals. Fusulinid zone D. D-858. Beds 10 feet (3 meters) stratigraphically above D-857. Cerioid and fasciculate rugose corals, syringoporid corals. Fusulinid zone D. 0-861. Basal 50 feet ( 1 5.2 meters) of the Bayha section. Abun- dant fusulinids including Pseudoschwagerina robusta (Meek, 1864), solitary and fasciculate rugose corals, brachiopods, Omphalotrochus whitneyi (Meek). Some of Meek’s ( 1 864) speci- mens were collected from this stratigraphic interval, but not necessarily from this locality. See A-7096 for map description. Same as LACMIP locality 1119. Fusulinid zone E. 0-862. Beds 245 feet (74.7 meters) stratigraphically above D-861. Fusulinids, fasciculate rugose corals. Fusulinid zone F. LACMIP Localities 1114. See UCMP locality D-836. 1116. See UCMP locality D-842. 1119. See UCMP locality D-861. 1120. Beds in the Bayha section, 380 feet (1 15.8 meters) strat- igraphically above base. Fasciculate rugose corals. Fusulinid zone F. 1121. Beds 45 feet ( 1 3.7 meters) stratigraphically above 1 120. Large fasciculate corals. Fusulinid zone F. 1122. Beds 45 feet ( 13.7 meters) stratigraphically above 1121. Cerioid and fasciculate rugose corals. Fusulinid zone F. 1 123. See UCMP locality A-7096. Fusulinid zone F. 1124. Beds 10 feet (3.0 meters) stratigraphically above 1123. Fasciculate and cerioid rugose corals. Fusulinid zone F. 1125. See UCMP locality D-852. 1132. See UCMP locality D-821. 1133. See UCMP locality D-831. 1285. McCloud Point section. The traverse begins on the E bank of the McCloud River Arm of Shasta Lake at the first limestone exposure on Point McCloud as shown in the NW Va of the SW Va of section 5, T 35 N, R 3 W, of the USGS 1 5-minute topographic quadrangle of Bollibokka Mountain, California ( 1 957), and ends at the highest exposure of the formation W of and about the same elevation as Gilman Road. This is the upper part of the McCloud Point section of Skinner and Wilde (1965). In this section, only this locality contained corals. It is 600 feet (182.5 meters) stratigraphically above the base of the section. Fasciculate and cerioid rugose corals. Impossible to collect the cerioid corals. Fusulinid zone G. 4457. Basal 1 00 feet (30.5 meters) of the McCloud Bridge West section (see 53 1 2 for description), along strike on the SW side of the hill from the line of section and topographically above Gil- man Road. Fasciculate rugose corals. Fusulinid zone A. 4458. See UCMP locality D-819. 5312. McCloud Bridge West section. The traverse begins at the lowest exposure of the formation on the SW slope of the hill at the W end of the McCloud River Bridge and NW of Gilman Road and runs NE to the crest of the hill and down the NE side into cover as shown on the NE Va of section 3 1 , T 35 N, R 3 W, of the USGS 15-minute topographic quadrangle of Lamoine, Cal- ifornia ( 1 957), and the SE Va of section 30, T 35 N, R 3 W, of the USGS 15-minute topographic quadrangle of Bollibokka Moun- tain, California (1957). Locality 5312 is 260 feet (79.2 meters) stratigraphically above the base of the section. Solitary rugose corals, brachiopods, Omphalotrochus (?). Fusulinid zone A. 5316. Beds 220 feet (67 meters) stratigraphically above 5312. Enormous fasciculate rugose coralla (largest 5 feet in diameter), Omphalotrochus. Fusulinid zone A. 5318. Beds 75 feet (22.9 meters) stratigraphically above 5316, at the summit of the hill. Large solitary rugose corals, brachiopods. Fusulinid zone A. 5319. Beds 1 to 25 feet (0.3 to 7.6 meters) stratigraphically above 5318, at the top of the dip slope. Large solitary rugose corals. Fusulinid zone A. ACKNOWLEDGMENTS I am grateful to R.L. Langenheim, Jr., for suggesting this study. J. Wyatt Durham supervised much of the early stages of the work and pointed out many things that had escaped my atten- tion. W.H. Easton kindly discussed characters of some species of Durhamina with me. V.A.M. Langenheim and R.R. Snelling critically read the manuscript and made numerous helpful sug- 86 Contributions in Science, Number 337 Wilson: Permian Corals of California gestions. The following men were field assistants on various trips: W.A. Carney, Jr., William Fletcher. P.T. Gavin, H E Hamnian, Paul Kenney, PG. Owen, Bradford Riney, James Scheetz, and R.B. Spangenberg. Chang Ki Hong donated part of a corallum collected by him to the University of California Museum of Pa- leontology for me to study. John B. Wexo financed transportation, subsistence, and two field assistants for one field trip. The De- partment of Paleontology, University of California, Berkeley, provided transportation for several field trips and subsistence for 6 days work. The University of California Committee on Re- search awarded me a Student Research Grant for 1 0 days subsis- tence in the field for myself and one field assistant. The Natural History Museum of Los Angeles County provided my salary, transportation, some field assistance, and some subsistence for several field trips. The Harvard Museum of Comparative Zool- ogy loaned their fossils from the McCloud Limestone described by Meek ( 1 864). Officials of the Shasta-Trinity National Forest issued permits to collect the McCloud Limestone in areas where it crops out on their land. John Alford and the AWA Land Com- pany permitted access to the Bayha section. The McCloud River Club allowed use of their roads to reach a section near Cold Springs on a spur of Bald Mountain. The Calaveras Cement Company let me examine their Gray Rocks quarry. All prepara- tion of fossils, including the numerous thin sections, were done by me using facilities at the University of California Museum of Paleontology, the Natural History Museum of San Diego, and the Natural History Museum of Los Angeles County. Photo- graphs are by Mr. Richard Meier. LITERATURE CITED Albers, J.P., and J.F. Robertson. 1961. Geology and ore deposits of east Shasta copper-zinc district, Shasta County, California. United States Geological Survey, Professional Paper 338:1 107. Brewer, W.H. 1930. Up and down California in 1 860- 1 864, the journal of William H. Brewer, Professor of Agriculture in the Sheffield Scientific School from 1864 to 1903, ed. F.P. Far- quhar. Yale University Press, New Haven. 601 pp. Chi, Y.S. 1931. Weiningian (Middle Carboniferous) corals of China. Palaeontologia Sinica, ser. B, 12(5): 1-70. Coogan, A.H. 1960. Stratigraphy and paleontology of the Per- mian Nosoni and Dekkas Formations (Bollibokka Group). University of California, Publications in Geological Sciences 36( 5 ):243— 3 1 6. Cooper, G.A. 1957. Permian brachiopods from central Oregon. Smithsonian Miscellaneous Collections 1 34( 1 2): I -79. Cotton, Geoffrey. 1 973. The rugose coral genera. Elsevier Scien- tific Publishing Company, Amsterdam. 358 pp. Dana, J.D. 1846. Genera of fossil corals of the family Cyatho- phyllidae. American Journal of Science and Arts, ser. 2, 1:178-189. Demirmen, Ferruh, and J.W. Harbaugh. 1965. Petrography and origin of Permian McCloud Limestone of northern California. Journal of Sedimentary Petrology 35( 1 ): 1 36-1 54. Diller, J.S. 1906 Redding folio, California. United States Geo- logical Survey, Geologic Atlas of the United States 138:1-14. Dobrolyubova, T.A. 1936 The corals of the Upper Carbonifer- ous of the western slope of the Middle Urals and their strat- igraphic importance. Vsesoiuzyni nauchno-issledo-vatelskii. Institut mineral’nogo Syria. Trudy 103:1-68. 1940. The Rugosa corals of the Upper Car- boniferous of the Moscow basin. Moscow, Travaux dc Plnstitut Paleontologique 9(3): 1 -88. Dobrolyubova, T.A., and N.V. Kabakovich. 1948. Some repre- sentatives of the Rugosa in the Middle and Upper Carbonifer- ous of the Moscow Basin (in Russian). Izdatelstvo Akademiia Nauk SSR, Trudy Paleontologicheskovo Instituta 1 4( 2 ): 1 37. d’Orbigny, Alcide. 1849. Note sur des polypiers fossiles. Victor Masson, Paris. I 2 pp. Dunbar, C.O., and J.W. Skinner. 1936. Schwagerina versus Pseudoschwagerina and Paraschwagerina. Journal of Pa- leontology 1 0( 1 ):83— 9 1 . Easton, W.H 1 960. Permian corals from Nevada and California. Journal of Paleontology 34(3 ):570— 583. 1973. On the tetracorals Acrocyathus and Lithostrotionella and their septal morphology. Journal of Pal- eontology 47( 1 ): 1 2 1 - 1 35. Easton, W.H., and M.M. Melendres. 1964. A Paleozoic coral from Mindoro. Journal of Paleontology 38(2):4 1 2—4 1 4. Evans, J.R. 1977. Economic geology of McCloud limestone. South Gray Rocks area, Shasta County, California. California Geology 30(6): 123-132. Fairbanks, H.W. 1892. The pre-Cretaceous age of the meta- morphic rocks of the California Coast Ranges. American Ge- ologist 9: 153 166. 1893. Geology and mineralogy of Shasta County. California State Mining Bureau, Annual Report I 1:24-53. Fleming, John. 1 828. History of British animals. Edinburgh, 565 pp. Fontaine, Henri. 1961. Les Madreporaires Paleozoiques du Viet- Nam, du Laos et du Cambodge. Archives Geologiques du Viet-Nam 5:1 -276. Ginsburg, R N. 1960. Ancient analogues of Recent stromato- lites. Report of the 21st International Geological Congress, Copenhagen 22:26-35. Goldtuss, G.A. 1826-1833. Petrefacta Germaniae. Arnz, Dus- seldorf. 252 pp. Gorskiy, LI. 1 95 1 . Carboniferous and Permian corals of Novaya Zemlya. Trudy NIIG Arktiki 32:1-12. Hall, James, and J.M. Clarke. 1 893. An introduction to the study of the genera of Paleozoic Brachiopoda. New York Geological Survey, Paleontology 8(1): 1-367. Harker, Peter, and D.J. McLaren. 1950. Sciophyllum, a new rugose coral from the Canadian Arctic. Canada, Geological Survey, Bulletin 15:29-34. Hayasaka, Ichiro. 1925. On the fauna of the anthracolithic lime- stone of Omi-Mura in the western part of Echigo. Tohoku Imperial University Science Reports 8(1 ): 1 83. 1936. On some North American species of Lithostrotionella. Taihoku Imperial University, Faculty of Sci- ence, Memoir 1 2(5):47— 73. Contributions in Science, Number 337 Wilson: Permian Corals of California 87 Heritsch, Franz. 1936. Korallen der Moskauer-, Gishcl- und Schwagerinen-stufe der Karnischen Alpen. Palaentographica 83 (A): 99-162. 1939. Die Korallen des Jungpalaozoikums von Spitzbergen. Arkiv for Zoologi 3 1 A( 1 6): 1 -138. Hill, Dorothy. 1940. A monograph on the Carboniferous rugose corals of Scotland, Part 3, pp. 1 15-204. Palaeontographical Society, London. 1956. Rugosa. Vol. F, Coelenterata, pp. 233- 245, 252-324 in R.C. Moore, ed.. Treatise on invertebrate paleontology. University of Kansas Press, Lawrence. Hinds, N.E.A. 1932. Diastrophic epochs in the southern Klamath Mountains. Geological Society of America, Bulletin 43:273-274. Hoare, R.D. 1964. Permian corals from northern Nevada. Jour- nal of Paleontology 38(3):496-504 1966. New name for Cornwallia Hoare, 1964, and a new species of Bayhaium from northern Nevada. Jour- nal of Paleontology 40( 1 ): 1 48-1 50. Holtedahl, Olaf. 1913. Zur Kenntnis der Karbonablagerungen des westlichen Spitzbergens. II Allgemeine stratigraphische und tektonische Beobachtungen. Norway, Skrifter udgivne af Videnskabsselskabet i Kristiania 23:1-91. Huang, T.K. 1932. Permian corals of southern China. Palaeon- tologia Sinica, ser. B, 8(2 ): 1 163. Hudson, R.G.S. 1958. Permian corals from northern Iraq. Pa- laeontology 1(3): I 74—192. International Commission on Zoological Nomenclature. 1964. International code of zoological nomenclature, ed. comm, chairman N.R. Stoll International Trust for Zoological No- menclature, London. 176 pp. Irelan, William, Jr. 1891. Preliminary mineralogical and geo- logical map of the state of California. Britton and Rey, San Francisco. 1 p. Ivanovsky, A.B. 1963. Ordovician and Silurian Rugosa of the Siberian Platform. Institute of Geology and Geophysics, Sibe- rian Branch, Akademia Nauk, 160 pp. Kachanov, Y. I. 1967. New data on corals of the genus Neo- multithecopora from the Lower Carboniferous of Novaya Zemlya and the Urals. Paleontological Journal 1(3 ):23 30. Keyserling, Alexander von. 1854. Palaontologische Bemerkungen. In A.G. Schrenk, Reise nach dem Nordosten des Europaischen Russlands ... 2. Dorpat. Koninck, L.G de. 1841-1844. Description des animaux fossiles qui se trouvent dans le terrain Carbonifere de Belgique. Liege, 650 pp. Lafuste, J.G. 1963. Note on the microstructure of the Permian tabulate coral Bayhaium. Journal of Paleontology 37(5): 1 127-1 131. Langenheim, R.L., Jr., and V.A. McCutcheon. 1959. Bayhaium merriamorum, a new Permian tabulate coral from California. Journal of Paleontology 33( 1 ):99-102. Langenheim, V.A.M., and R.L. Langenheim, Jr. 1965. The Bird Spring Group, Chesterian through Wolfcampian, at Arrow Canyon, Arrow Canyon Range, Clark County, Nevada. Trans- actions Illinois Academy of Science 58(4):225-240. Lanphere, M.A., W.P Irwin, and P.E. Hotz. 1 968. Isotopic age of the Nevadan orogeny and older plutonic and metamorphic events in the Klamath Mountains, California. Geological Soci- ety of America, Bulletin 79:1027-1052. Lewis, H.P 1927. On Auloclisia. a new coral genus from the Carboniferous Limestone. Proceedings of the Geological Soci- ety of Yorkshire 2 1 :29-46. Lin, Pao-Yu. 1963. Some Carboniferous and Permian Tabulata of south China. Acta Palaeontologica Sinica 1 1 (4):579— 596. Mandlebaum, Hugo, and J.T. Sanford. 1952. Table for comput- ing thickness of strata measured in a traverse or encountered in a bore hole. Bulletin of the Geological Society of America 63:765-776. Mather, K.F. 1915. The fauna of the Morrow Group of Arkansas and Oklahoma. Bulletin of the Scientific Laboratories of Deni- son University 1 8( 3 ):59— 284. McChesney, J.H. 1860. Descriptions of new species of fossils from the Paleozoic rocks of the western states. Ext. Transac- tions, Chicago Academy of Science 1:1-76. 1867. Descriptions of fossils from Paleozoic rocks of the western states, with illustrations. Chicago Acad- emy of Sciences, Transactions 1 : 1 -57. McCutcheon, V.A. 1961 . Redescription of Syringopora multat- tenuata McChesney. Journal of Paleontology 35( 5): 1 0 1 4— 1016. McCutcheon, V.A., and E.C. Wilson. 1961. Ptolemaia. a new colonial rugose coral from the Early Permian of eastern Ne- vada and western Russia. Journal of Paleontology 35(5): 1020- 1028. : 1963. Kleopatrina. new name for Ptolemaia McCutcheon and Wilson. Journal of Paleontology 37(1):299. M'Coy, Frederick. 1 844. A synopsis of the characteristic Car- boniferous Limestone fossils of Ireland. Dublin, 207 pp. 1 849. On some new genera and species of Pa- laeozoic corals and foraminifera. Annals and Magazine of Natural History, ser. 2, 3(1 3): I -20, 3(14): 1 19-136. Meek, F.B. 1864. Descriptions of the Carboniferous fossils. In Meek, F.B., and W.M. Gabb, Palaeontology of California 1 :3- 16. Caxton Press, Philadelphia. 1877. Part I. Palaeontology. In Report of the geological exploration of the 40th parallel . . ., Professional Papers of the Engineer Department, U.S. Army 18:1-197. Merriam, C.W. 1942. Carboniferous and Permian corals from central Oregon. Journal of Paleontology 1 6( 3):372— 38 1 . Milne Edwards, Henri, and Jules Haime. 1 850. A monograph of the British fossil corals. First part. Introduction; corals from the Tertiary and Cretaceous formations. Palaeontographical Society, London. 85 pp. 1852. A monograph of the British fossil corals. Third part. Corals from the Permian Formation and the Moun- tain Limestone. Palaeontographical Society, London. 63 pp. Minato, Masao. 1 955. Japanese Carboniferous and Permian cor- als. Hokkaido University, Contribution from the Department of Geology and Mineralogy. 540:1-202. 1 960. Eine Permische koralle von Konig Oscars- land im nordwestlichen Ellesmereland. Stockholm Contribu- tions in Geology 6(3):25— 36. Minato, Masao, and Makoto Kato. 1965a. Waagenophyllidae. Hokkaido University, Journal of the Faculty of Science, ser. 4, Geology and Mineralogy 1 2(3-4): 1-241 . 88 Contributions in Science, Number 337 Wilson: Permian Corals of California 1965b. Durhaminidae (tetracoral). Hokkaido University, Journal of the Faculty of Science, ser. 4, Geology and Mineralogy 1 3( 1 ): 1 1-86. 1968. Ural nevadaphy llum, a new subgeneric name for Porfirievella Minatoand Kato, 1965. Palaeontologi- cal Society of Japan, n.s.. Transactions and Proceedings 72:363. Moore, R.C., and R.M. Jeffords. 1945. Description of Lower Pennsylvanian corals from Texas and adjacent states. Univer- sity of Texas Publication 440 1 :77-208. Moore, R.C., Dorothy Hill, and .1 W. Wells. 1956. Glossary of morphological terms applied to corals. Vol. F, Coelenterata, pp. 245-251 in R.C. Moore, ed.. Treatise on invertebrate paeiontology. University of Kansas Press, Lawrence. Nicholson, H.A., and H. Thomson. 1876. Descriptions of some new or imperfectly understood forms of Palaeozoic corals. Pro- ceedings of the Royal Society of Edinburgh 9( 94 ) : 1 49— 1 50. Oekentorp, Klemens, C. Montenat, and Henri Fontaine. 1978. Eine kleine Korallenfauna aus dem unteren Operperm von Saiq, Oman ( Arabische Halbinsel). Neues Jahrbuch fur Geol- ogie und Palaontologie, Abhandlungen 1 55(3):374-397. Rowett, C.L. 1969. Upper Palaeozoic stratigraphy and corals from the east-central Alaska Range, Alaska. Arctic Institute of North America, Technical Paper 23:1-120. . 1975. Provinciality of Late Paleozoic inverte- brates of North and South America and a modified interconti- nental reconstruction. Pacific Geology 10:79-94. 1977. Provinciality of Late Paleozoic corals of North and South America and a modified pre-drift reconstruc- tion. France, Bureau de Recherches Geologiques et Minieres, Publication 89: 1 90- 1 95. Sando, W.J. 1965. Revision of some Paleozoic coral species from the western United States. United States Geological Survey, Professional Paper 503E: 1 -38. 1974. Checklist of North American Late Pal- eozoic coral species (Coelenterata, Anthozoa). United States Geological Survey, Bulletin 1387:1-36. 1 980a. Supplement to checklist of North Amer- ican Late Paleozoic coral species (Coelenterata, Anthozoa). Journal of Paleontology 54(3):632-633. 1980b. The paleoecology of Mississippian cor- als in the western conterminous United States. Acta Palaeon- tologica Polonica 25(3-4):6 1 9-63 1 . Schouppe, Alexander. 1961. Beschreibung der korallenfauna. In Ramovs, Anton and Alexander von Scouppe, Karbon und Perm in Vitanje-Gebirge-Ostkarawanken-NW-Jugoslawien. Neues Jahrbuch furGeologie und Palaontologie, Monatshefte 7:354-374. Scouler, J. 1844. Siphonophyllia cylindrica. In McCoy, F., A synopsis of the characters of the Carboniferous limestone fos- sils of Ireland. Dublin. 1 87 pp. Skinner, J.W., and G.L. Wilde. 1965. Permian biostratigraphy and fusulinid faunas of the Shasta Lake area, northern Cal- ifornia. University of Kansas, Paleontological Contributions, Protozoa. 6:1-98. 1966. Permian fusulinids from Pacific north- west and Alaska. University of Kansas, Paleontological Contri- butions 4:1-64. Smith, .I P 1894. Metamorphic series of Shasta County, Califor- nia. Journal of Geology 2:588-612. Smith, Stanley. 1 935. Two anthracolithic corals from British Co- lumbia and related species from the Tethys. Journal of Paleon- tology 9( 1 ):30-42. Soshkina, E.D., T.A. Dobrolyubova, and N B. Kabakovich. 1962. Podklass Tetracoralla. InOsnovy Paleontologii, ed. B.S. Sokolov, pp. 236-356. Izdatel’stvo Akademii Nauk SSSR, Moscow. Soshkina, E.D., T.A. Dobrolyubova, and G.S. Porfiriev. 1941 The Permian rugose corals of the European part of the USSR. Vol. 5 pt. 3, fasc. 1 , pp. I -304, in Paleontology of USSR. Moscow: Academy of Sciences of USSR Press. Stevens, C.H. 1967. Leonardian (Permian) compound corals of Nevada. Journal of Paleontology 4 1 (2):423 43 1 . 1975. Occurrence and dispersal of boreal mas- sive Rugosa in the Early Permian. Pacific Geology 10:33-42. 1977. Permian depositional provinces and tec- tonics, western United States. Pages 1 I 3-1 35 in J.H. Stewart, ed.. Paleozoic paleogeography of the western United States. Society of Economic Paleontologists and Mineralogists, Los Angeles. Stuckenberg, A. A. 1888. Anthozoen und bryozoen des oberen mittelrussischen Kohlenkalks. St. Petersburg, Memoires du Comite Geologique 5(4): 1 -54. 1895. Korallen und Bryozoen der Steinkohienablegerungen des Ural und des Timan. St. Pe- tersburg, Memoires du Comite Geologique 1 0( 3): 1 —244. 1905. Die Fauna der oberkarbonischen Suite des WolgaDurchbruches bei Samara. St. Petersburg, Memoires du Comite Geologique 23: 1 3-23. Sutherland, PK. 1977. Analysis of the Middle Carboniferous rugose coral genus Petalaxis and its stratigraphic significance. Memories du Bureau de Recherches Geologiques et Mineres 89:185-189. Tasch, Paul. 1957. Fauna and paleoecology of the Pennsylvanian dry shale in Kansas. Geological Society of America, Memoir 67(2):365— 406. Tchudinova, 1.1. 1975a. Revision of the Syringopora described by A. A. Stuckenberg. Paleontological Journal 9( 1 ):7-19. 1975b. Permian syringoporids of the middle Urals. Paleontological Journal 9(4):427 433. Thompson, M.L. 1964. Fusulinacea. In Treatise on invertebrate paleontology, ed. R.C. Moore, (C) Protista 2( 1 ):358-436. Thompson, M L., and H.E. Wheeler. 1946. Permian fusulinids of northern California. Geological Society of America, Memoir 17:21-36. Trask, J.B 1853. Prof. John B. Trask’s report on the geology of the Sierra Nevada or California Range. California State Sen- ate Document 59:1-30. I 855. Report on the geology of the coast moun- tains; embracing their agricultural resources and mineral pro- ductions. Also, portions of the middle and northern mining districts. California State Senate Document 14:1-95. Trautschold, H. 1879. Die Kalkbruche von Mjatschkowa. So- ciete des Naturalistes de Moscou, Nouveaux Memoires 14:1 — 82. Contributions in Science, Number 337 Wilson: Permian Corals of California 89 Turner, H.W. 1894. Geological notes on the Sierra Nevada. American Geologist I 3:228-249, 297-3 1 6. Waagen, Wilheim, and Joseph Wentzel. 1886. Salt Range Fos- sils, Part 6, Productus Limestone Fossils, Coelenterata. Pa- laeontologia Indica ser. 13, 1:835-924. Watkins, Rodney. 1973. Carboniferous faunal associations and stratigraphy, Shasta County, northern California. American Association of Petroleum Geologists, Bulletin 57(9): 1743- 1764. 1974. Carboniferous brachiopods from north- ern California. Journal of Paleontology 48( 2):304— 325. 1 975. Paleoecology of some Carboniferous Pec- tinacea. Lethaia 8:1 25- 131. 1979. Carboniferous rocks of the eastern Klamath Mountains, California. United States Geological Survey, Professional Paper 1 1 IOCC:33-38. Wells, J.W. 1957. Corals. Geological Society of America, Memoir 67(2):773-782. Wheeler, H E. 1 933. Fusulinids of the McCloud and Nosoni For- mations of north California. Geological Society of America, Bulletin 44:2 1 8. 1935. The fauna and correlation of the Mc- Cloud Limestone of northern California. Stanford University, Bulletin ser. 6, 18:83-85. Whitney, J.D. I 865. Geological Survey of California. Geology 1 . Report of progress and synopsis of the field-work, from 1 860 to 1864. Caxton Press, Philadelphia. 498 pp. Wilde, G.L. 1971. Phylogeny of Pseudofusulinella and its bear- ing on Early Permian stratigraphy. Smithsonian Contributions to Paleobiology 3:363-379. Wilson, E.C. 1 967a. Primary type-material of the Early Permian archaeogastropod Omphalotrochus whitneyi (Meek, 1864). Journal of Paleontology 4 1 (5): 1 1 1 3- 1 1 1 5. 1967b Fusulina gracilis Meek, 1864 (For- aminiferida): Proposed invalidation of neotype-material, des- ignation as type-species of Eoparafusulina Coogan, I960, and related actions. Bulletin of Zoological Nomenclature 24(4):234-236. 1970. Conocardium langenheimi sp. n. (Mol- lusca: Bivalvia) in the Lower Permian Series of the McCloud Limestone, northern California. Natural History Museum of Los Angeles County, Contributions in Science 184:1-14 1971. Corals and other significant fossils from the Upper Paleozoic McCloud Limestone of northern Califor- nia. Dissertation Abstracts International 31(7):4233B. 1974. Bibliographic index of North American Permian rugose and tabulate coral species. Journal of Paleon- tology 48(3):598-606. 1980. Redescription of type specimens of the Permian rugose coral Waagenophyllum columbicum Smith, 1935, type species of Heritschioides Yabe, 1950. Journal of Paleontology 54( I ):85— 92. Wilson, E.C., and R.L. Langenheim, Jr. 1962. Rugose and tabu- late corals from Permian rocks in the Ely Quadrangle, White Pine County, Nevada. Journal of Paleontology 36(3):495-520. Yabe, Hisakatsu. 1950. Permian corals resembling Waagenophyllum and Corwenia. Proceedings of the Japan Academy 26(3):74-79. Yabe, Hisakatsu, and Ichiro Hayasaka. 1915-1916. Palaeozoic corals from Japan, Korea, and China. Geological Society of Japan, Journal 22(261 ):55-70, 22(263):79-92, 22(264):93- 109, 22(265):127-142, 23(27 1 ):57-83. 1920. Geographical research in China, 1911 — 1916, Atlas of fossils. Tokyo Geographical Society, Tokyo. 57 pp. Yabe, Hisakatsu, and Masao Minato. 1944. Wentzelloides maiyaensis Yabe et Minato, gen. et sp. nov. aus dem Perm des Sud-Kitakami-Gebirges. Japanese Journal of Geology and Geography 9:141-142. Yancey, T.E. 1975. Permian marine biotic provinces in North America. Journal of Paleontology 49(4):758— 766. 1976. Permian positions of the northern hemi- sphere continents as determined from marine biotic provinces. Pages 239-247 in Jane Gray and A.J. Boucot, eds.. Historical biogeography, plate tectonics, and the changing environment. Oregon State University Press, Corvallis. Yoh, S.S. 1 927. On a new genus of syringoporoid coral from the Carboniferous of Chihli and Fentien Provinces. Bulletin of the Geological Society of China 5:291-293. Yoh, S.S., and T.K. Huang. 1 932. The coral fauna of the Chihsia Limestone of the lower Yangtze Valley. Palaeontologia Sinica, ser. B, 8(1): I -52. Submitted 1 April 1981. Accepted for publication 2 October 1981. 90 Contributions in Science, Number 337 Contributions in Science, Number 337 !i m SERIAL PUBLICATIONS OF THE NATURAL HISTORY MUSEUM OF LOS ANGELES COUNTY The scientific publications of the Natural History Museum of Los Angeles County have been issued at irregular intervals in three major series; the articles in each series are numbered individually, and numbers run consecutively, regardless of subject matter. • Contributions in Science, a miscellaneous series of technical papers describing original ;g$;ii:ip . I'-- . j . ..V'i:. .C’,.': r;rj jy&t-/*'- m research in the life and earth sciences. ® Science Bulletin, a miscellaneous series of monographs describing original research in the life and earth sciences. This series was discontinued in 1978 with the issue of Numbers 29 and 30; monographs are now published by the Museum in the Contributions in Science series. ® Science Series, long articles on natural history topics, generally written for the layman. Copies of the publications in these series are sold through the Museum Bookshop, For instructions for authors, refer to Contribution No. 336 or contact the Museum Publications Office. EDITORIAL BOARD Till Leon G. Arnold Lawrence G. Barnes Robert L. Bezy Kenneth E. Campbell Robert. Gustafson John M. Harris Robert J. Lavenberg Camm C. Swift Fred S. Truxal David P Whistler Edward C. Wilson John W. Wright EDITOR Robin A. Simpson IS Printed at The Castle Press, Pasadena, California. l REVISED STATUS OF NET-WINGED MIDGES OF THE GENUS BIBIOCEPHA LA IN NORTH AMERICA BASED ON A STUDY OF QUANTITATIVE VARIATION IN THE MALES (DIPTERA: BLEPHARICERIDAE)1 Charles L. Hogue2 ABSTRACT. Samples of male Bibiocephala from North America were morphometrically analyzed to determine the presence or absence of sub- sets of specimens of possible taxonomic significance. Both univariate and multivariate (principal components analysis, unweighted pair group clus- ter and block cluster analysis) techniques were used. Some subsets were recognized, but none was believed to have specific or subspecific signifi- cance. A single, moderately variable species exists, B grandis Osten Sacken, 1 874. Bibiocephala kelloggi Garrett, 1922, B griseus Curran, 1923, and B nigripes Alexander, 1965, are synonyms for B. grandis. The adult, pupa, and larva of B grandis are redescribcd, and the distribution and biology of the species are discussed. INTRODUCTION In North America, four nominate species have been referred to the genus Bibiocephala Osten Sacken, 1874. The identities of these have never been properly defined, and they have been vari- ously treated in the fragmentary literature dealing with the Blephariceridae. Alexander (1965) gave the only general ac- count of the genus; he recognized two valid and one questionably distinct species. After accumulating all available specimens of this genus from museum and other collections, and finding my- self unable to separate the described species, using the charac- ters prescribed by their original authors, I conducted an analysis de novo to determine the presence or absence of valid entities, unbiased by previous separations. My first attempts to reduce the sample by inspective (often called “intuitive” or “classical”) taxonomic procedure were to- tally without success, the specimens possessing an apparent high degree of overall similarity and continuous variation in virtually all characters. Indeed, few characters with consistent and mea- surable variation could be found at all. Suspecting that quantita- tive characters might exist by which groups of specimens could be defined, I subjected the material to a regimen of statistical procedures that seemed to me most likely to reveal any such groups that might lie hidden to the purblind eye. My choice of procedures was also partly determined by availability of com- puter facilities and programs (see METHODS). Only adult male specimens were available in sufficient num- bers for analysis. Females and immatures were too damaged and few in number to give reliable quantitative results, and in no case Contributions in Science, Number 338, pp. 1-16 Natural History Museum of Los Angeles County, 1982 were the latter associated with adults; identities, therefore, could not be correlated with decisions made pertaining to the other stages. These stages were measured and univariate statistics cal- culated for descriptive purposes, however. My conclusions, having been based only on the one stage, must be considered provisional. However, in the interest of providing a basis for review works now in progress, mainly “Flies of the Nearctic Region” (ed., G.C.D. Griffiths, University of Alberta), I am publishing here a report on my studies of the best material available. ANALYSIS OPERATIONAL BASIS Because past taxonomists concerned with North American Bibiocephala have recognized more than one species, my task has been to determine whether or not there are subsets of speci- mens among the material conforming to pheno-(morpho-)species criteria. These criteria are: 1 . Low level of absolute variation in many characters. 2. Variation of most characters unimodal and normal. 3. Significantly higher degree of overall similarity within sub- sets than between subsets. 4. Suites of associated states (of biologically independent characters) present. Of course, these elements in the definition of phenospecies are relative, and final decisions of their taxonomic and nomen- clatorial application rest with the taxonomist. This section is an explanation of my application and represents only one approach to the problem of species resolution. MATERIALS A total of 786 specimens was available for this study, 526 imma- tures and 260 adults. From the latter, 1 60 were measured and are 1 . Review committee for this contribution: George W. Byers, Julian P Donahue, and F. Christian Thompson. 2. Entomology Section, Natural History Museum of Los Angeles County, Los Angeles, California 90007 ISSN 0459-01 13 my “entire” sample; 30 of these were partly dissected for mor- phometric analysis and constitute the “subsample” referred to in the section on METHODS et seq. A complete accounting of these specimens and their data is given under SPECIMENS EXAMINED. TERMINOLOGY In the following discussions and descriptions, terminology essen- tially follows that of my other papers on the Blephariceridae (particularly Hogue, 1973, 1978). Abbreviations for taxonomic characters are explained in the following section, those for speci- men repositories are as follows: American Museum of Natural History (AMNH), Academy of Natural Sciences of Phila- delphia (ANSP), British Museum (Natural History) (BMNH), Brigham Young University, Bean Life Science Museum (BYU), California Academy of Sciences (CAS), California Insect Sur- vey Collection, University of California at Berkeley (CIS), Ca- nadian National Collection, Biosystematics Research Institute, Agriculture Canada (CNC), C.P Alexander collection, Amherst, Massachusetts (CPA), Cornell University (CU), Cen- tral Washington State University (CWS), Illinois Natural His- tory Survey (INHS), Johns Hopkins University (JH), Kansas State University (KS), University of Kansas (KU), Natural His- tory Museum of Los Angeles County (LACM), Museum of Comparative Zoology, Harvard University (MCZ), Oregon State University (OS), Ohio State University (OSU), Royal On- tario Museum ( ROM), San Jose State University (SJS), Univer- sity of Alberta (UAT), University of California at Davis (UCD), University of Idaho (UI), U.S. National Museum of Natural History (USNM), University of Wyoming (UW), and Wash- ington State University (WSU). Statistical notation follows that in general use plus special terms from Sneath and Sokal ( 1 973). TAXONOMIC CHARACTERS Microscopic examination of both external and genital structures of male Bibiocephala revealed very few characters with variation adequate for taxonomic purposes. In fact, the vast majority of those characters usually found useful in species definition in the Blephariceridae were simply invariant by any reasonable mea- sure. This logically included characters whose variation was, for practical purposes, indeterminable by reason of their subtlety (integumental coloration), irresoluteness (number of setae on the apex of the outer dististyle), or imperfection (body length). The characters finally chosen are the following: Continuous variates 1. Wing length. (WING) 2- 1 0. Lengths of leg segments (femur, tibia, and basitarsus only; the shortness of the remaining tarsal segments precludes accurate measurement). (FEM 1, TIB 1, TAR 1, FEM 2, etc.) 1 1 . Length of outer dististyle of genitalia. (OTDSYL) 1 2. Length of fused basistyles of genitalia. (BASYL) Meristic variates 13. Number of median setae on proctiger of genitalia. (PRCTSM) 1 4. Number of median setae on ninth tergite. (TERSM) 15. Presence or absence of pilosity on upper eye. (EYHAIR) No qualitative characters were used. Because the configura- tions of genital structures are generally very useful for species discrimination in the Blephariceridae, I made a careful compari- Figures 1 through 4. Bibiocephala grandis. Outer dististyle of male genitalia. Rotated views to illustrate varied appearance depending on aspect. Figure 1, dorsal; Figure 2, dorsomesal; Figure 3, near mesal; Figure 4, full mesal. 2 Contributions in Science, Number 338 Hogue: Revised Status of Bibiocephala son of these in both male and female specimens and concluded that none had consistent, appreciable states of variation. At first impression, several features seemed to exhibit shape differences; but upon careful stereoscopic inspection, I saw that they were no more than outline changes produced by different viewing angles. The genitalia are complex, and it is therefore difficult to mount them on slides with all parts in a standard position. Thus, with the compound miscroscope, the same element may look different in plane perspective, depending upon very slight rotations of posi- tion relative to other elements. Only by viewing on all sides can the true shape be determined. Examples of this problem are the entire dististyle and the basal lobe of the inner dististyle. As the former is rotated ( Figs. 1 -4), its appearance changes drastically. The rounded, flat apex of the latter lobe may appear pyramidal if viewed from the oblique position usually assumed in dorsoventral whole mounts of the genitalia. METHODS When direct inspection of the samples did not reveal subsets of specimens that met my phenospecies criteria, I sorted them mathematically and applied a variety of reductions and tests. These I chose on the basis of parsimony and the pragmatic avail- ability locally of data processing systems. The analyses were carried out with the aid of packaged computer programs avail- able at the Computer Center of the University of Southern Cal- ifornia in Los Angeles and included “Biomedical Computer Programs, P-Series” (Dixon, 1977). Where applicable in the fol- lowing discussion, the exact programs used from the latter are cited by their system call codes (e.g., “BMDP4M”). Appropriate direct measures of criteria 1 and 2 are frequency distributions, bivariate correlations (to reveal possible ratio char- acters and to partially assess character interrelations), and the standard univariate descriptions of variation (standard devia- tion, coefficient of variation, Chi-square test for goodness of fit to normalcy, skewness and kurtosis, normal probability, and fre- quency plots). Interrelations of continuously varying characters were also read from a principal components analysis (PCA) ap- plied to them (BMDP4M). Criterion 3 was “tested” by means of an unweighted pair group cluster analysis (UPGMA) based on the average distance algorithm (BMDP2M). Distance measure was the sum of squares (Euclidian distance), both between individuals and be- tween ascending clusters of individuals. Two analyses were run, one on the raw data and a second on PCA scores of the 12 continuous mensural characters plus the raw meristic data. The PCA scores were obtained from factoring of a variance- covariance matrix; both these scores and the raw meristic values were standarized by scaling before the distances were calcu- lated. This second procedure follows the example of Agafitei and Selander (1980), who point out its advantages over the first in limiting the effect of a general size factor to a single character and in removing the effect of redundant or correlated characters by transforming them into retrolinear combinations of the origi- nals. To evaluate clusters, I followed the recommendation of Sneath and Sokal (1973:304) that differences between branch points of a dendrogram ought to be more than several standard errors of the similarity coefficient used in order to be significant. Some idea of outliers and groupings of specimens can be de- rived also from the ordination content of PCA. This derives from bivariate plots of the first three components and plots of eigen- vectors of each character, which I also completed. Finally, using what seemed to be the four best uncorrelated and disjunctively variant characters, as determined from the cal- culations described above (i.e., WING, EYHA1R, PRCTSM, and TERSM ), I processed ranked raw data for each according to a block clustering procedure (BMDP3M). To maximize the pos- sibility of recognizing significant subsets and to reflect the trend toward bimodality in their frequency distributions, I ranked the characters into only two classes each. One of the characters was size and was represented by wing length, which- -from its domi- nance of the first principal component — seems to be the best single measure of this character. This method of clustering by modal blocks identifies and groups specimens that have similar patterns for subsets of ranked character states. It seems to me to simulate most closely the methodology of (inspective) taxonomy in which one seeks suites of character states that vary dis- junctively in parallel between taxa and thus indicate fundamen- tal (probably genetic) discontinuities characteristic of phenetic and phyletic species. The method is discussed at length by Har- tigan ( 1 976). The univariate statistics and plots were made from the entire sample; a subsample of 30, representing complete specimens from as many distinct localities as possible, was used for the PCA and cluster analysis. The specimens in the subsample are identi- fied under SPECIMENS EXAMINED by the italic number appearing between the parentheses also enclosing the repository abbreviation. RESULTS Univariate statistics for the data are given in Table 1 . Frequency distributions for all the characters fail the Chi-square goodness- of-fit test for normality at a highly significant level; only wing length has a distribution even approaching normal but is other- wise asymmetrical as are the curves for all the characters. Sec- ondary grouping of class intervals for all characters produces generally flat-topped curves with mild to moderate skewness to the right (most) or left. Several curves have elongate positive tails for which a few very large specimens from an Oregon popu- lation are responsible. Primary curves drawn to the original fre- quencies appear multimodal, although the sample size is not large enought to permit reliable assessment of the detailed form of the distribution. A few characters have slight bimodal trends and were used in the two-state block cluster analyses as described below. Coefficients of variation range from 9.0 to 14.2. The dendrograms obtained by the two UPGMA cluster analy- ses are reproduced in Figures 5 and 6. Since the subsample came from a large geographic area and a moderate coefficient of varia- tion was observed for the continuous characters, I felt justified in choosing fairly high multiples of the standard error of the mean of the distances between specimens (Euclidian distance, dj , ) to judge significance. With results from the raw data, these values are 10; with scaled data from PCA scores plus raw meristics, 6. Contributions in Science, Number 338 Hogue: Revised Status of Bibiocephala 3 A B H — C — I h-DH < o < X > < LU I— Z X) — -T z o o z < 3 o X < O — — 0.968 1 . 205 1 .260 1.359 1 .424 1 . 445 1.613 1.752 1 .782 1.833 1.839 1.938 2.036 2. 154 2.190 2.284 2.369 2.372 2.308 2.355 2.685 2.794 2.867 3.225 3.242 5.072 5.100 5- 192 8.068 ca CM < od < 3 < cc o _j o CD nO =H < ID MD O >- OO CM < 3 < CO 00 < -2. 5. 5 5.5 ■2. 5.2 0.5 ■ 5.2 < 2 < < < 2: < o 2: LA < < tn < I — cd < oa < 3 < LD ^ CO LlJ ZD — J ^ >- < ■ — LA vD h- < CD f— Z cc — LU □= CD CO —I < < 3 (A N LA 4- CNl O JZ < CD O DC >- OA CO -J- CNJ — ■ — o CD CD < Z Z OZ — 002 : _l o o o ^ >- CJ ^ < LU CJ h- 2 CO < CO < h- o oa vO OO OO nO CsJ cnJ O n: < CD n: < < LU CJ \— 2 CO < CO < < CJ I— 2 CO < CO LA -d- OD O (NJ CNl CA O CD LU QC O CNJ O OD — C\j CNJ — CNJ Figure 6. Dendrogram representing results of unweighted pair group cluster analysis of Bibliocephala subsample, based on principal components scores of continuous variables. Specimen numbers and general localities at top; amalgamation distance measures at left. Italicized values to right of stems indicate multiples of standard error of similarity coefficient determining the branch. Significance level is 6 with SE of the mean of dji = 1.1 5. Contributions in Science, Number 338 Hogue: Revised Status of Bibiocephala 5 Table I. Univariate statistics of Bibiocephala sample, males. Characters N Mini- mum Maxi- mum 1 nter- quartile Range Mean (Mode) SD Coelf Var. Goodness of Fit Skewness - SE Kurtosis ^ SE WING 102 7.0 n.i 8. 3-9.5 8.9 0.80 9.0 0.76 1.28 -0.40 FEM 1 98 2.5 3.9 2.9-3. 3 3.1 0.29 9.3 <0.20 0.87 0.43 TIB 1 97 2.6 4.3 3. 1-3.5 3.3 0.33 10.1 <0.20 1.76 0.30 TAR 1 97 1.16 2.20 1.62-1.86 1.74 0.18 10.5 0.54 -0.42 0.37 FEM 2 103 2.8 4.8 3. 4-3. 9 3.7 0.35 9.6 <0.20 1.76 2.02 TIB 2 103 2.7 4.5 3. 2-3. 6 3.4 0 36 10.6 <0.20 2.63 1.66 TAR 2 102 0.82 1.34 1 .00- 1.18 1.09 0.1 1 10.4 <0.20 0.83 - 1.41 FEM 3 99 4.3 6.9 5. 2-5. 9 5.5 0.53 9.5 <0.20 0.48 -0.30 TIB 3 97 3.7 6.0 4. 6-5. 4 5.0 0.49 9.9 0.33 -0.68 -1.01 TAR 3 92 1.60 2.60 1.99-2.24 2.10 0.20 9.4 0.32 -0.54 0.22 OTDSYL 40 0.45 0.71 0.53-0.58 0.56 0.05 8.5 <0.20 2.33 2.43 BASYL 101 1.20 2.75 1.70-2.02 1.91 0.27 14.2 <0.20 3.82 2.74 PRCTSM 40 2 1 1 5-8 (6) 2.17 — <0.20 — — TERSM 39 14 42 O cr (30) 5.67 — <0.20 — — EYHAIR 102 - + — — — — — — — Using this arbitrary standard, the former yielded four significant clusters with more than one individual (Fig. 5, groups A-D). Three of these (A-C, plus specimen 1 I ) form a significant group of a higher order distinct from the fourth (D). The latter dendro- gram has only two significant clusters (Fig. 6, A-B). Cophenetic correlation coefficients were obtained between re- spective dendrograms and their similarity matrices (raw' data: 0.82; PCA scores: 0.78), but a poor match was found between the two dendrograms (0.59). The total number of data points used in the block clustering analysis was 120; from these, 12 blocks (single and multiple) were developed (Fig. 7). The ratio of data points to blocks recog- nized is a measure of the significance of the blocks (subsets). In the present instance, the ratio was 0.10, evidence of a “good” analysis according to BMDP3M (Dixon, 1977:647). The PCA yielded six isolates (Table 2, Fig. 8), only one of which (composed of Oregon specimens 21 and 22) is really dis- tinct. All scores are fairly close to the centroid of the third ordi- nate, but there are vague linear trends along all axes, especially along the second. There is also a better concordance of groups and general localities than in either of the cluster analyses. CONCLUSIONS The reasonable conclusion of inspective analysis of the visually uniform sample — that no taxonomically recognizable subsets are present — is confirmed by metric analysis. All univariate characteristics are consistent with those ex- pected theoretically from a homogeneous sample, i.e., a sample with many internal minor subsets. This would be expected from a single species distributed over a wide geographic area with multi- ple populations and small morphological (especially size) differences. Although a few subsets were formed and measured as signifi- cant in the dendrograms from UPGMA clustering, and thus would appear to meet criterion 3 for phenospecies, they are prob- ably not significant. My greatest reason for this conclusion is the lack of concordance between the two cluster analyses (rC[C2 = 0.59). With the exception of the New Mexico specimens (distinc- tive character state of EYHAIR) and those from Oregon (large size), there are no concordant groups even of only two specimens. I regard the second analysis (based on PCA scores) as more indicative than the first because of the removal of the effects of correlation. Here, the “staircase” form of the dendrogram is highly reflective of homogeneity, i.e., nearly equal small units. Moreover, the dominating effect of size from many characters (leg segment lengths) is reduced to a single character. Although a vague cline in size is evident from smaller, extreme northern (Yukon) specimens to larger, extreme southern (New Mexico, Oregon) specimens, a single taxonomic entity is indicated for the sample. The specimens from New Mexico, with pilosity present on the upper eye division, plotted together in both dendrograms and are phenospecies. However, I choose not to give them taxonomic rank at present on the basis of this single character; more mate- rial is needed. Note that they (19, 20) do not segregate in the PCA analysis in which EYHAIR was not included (Fig. 8). The conclusion of homogeneity derives also from assessment of the block cluster analysis. The pattern of clustering, like that in the second UPGMA analysis, indicates homogeneity; many small groups of mixed character states appear. There is no con- cordance with the results of block and UPGMA clustering, ex- cept in the recognition again of the distinctiveness of the New Mexico and Oregon specimens. The distinctiveness of the latter is borne out by their outlying position in the bivariate plots of the first three principal components. Homogeneity of the remainder of the sample is confirmed by this ordination method; points for all other specimens are more or less evenly clustered around the centroids with a slight linear trend along each axis (clinal varia- 6 Contributions in Science, Number 338 Hogue: Revised Status of Bibiocephala tion). (EYHAIR, not being a continuous variate, was not in- cluded in the PCA.) The ultimate conclusion derivable from the results of all the statistical procedures applied to the 1 5 characters of the present sample is that there are no subsets of specimens clearly and consistently definable as phenospecies. Accordingly, I am mak- ing a taxonomic decision to consider all specimens as represent- ing a single species and so recognize them nomenclaturally (see TAXONOMY). Certainly I would urge dipterists to collect Bibiocephala assiduously so that this conclusion can be tested in the future on a much larger sample of both males and females, and on immatures also if reared or closely associated with adults. TAXONOMY I am foregoing a consideration of the status of Amika Kitakami, 1950, relative to Bibiocephala and am treating these genera as separate until their relationships can be clarified. Thus the fol- lowing description for the one species of the latter may be consid- ered the same as for the genus. Reference to the tribe Blepharicerini is as provisionally de- fined by Zwick (1977:24, and in personal communication). Bibiocephala Osten Sacken SYNONYMY Bibiocephala Osten Sacken, 1874:564-566. Type species Bibiocephala grandis Osten Sacken, 1 874, by monotypy. Bibionus Curran, 1923:266-267. Type species Bibionus griseus Curran, 1923, by original designation. EXSYNONYMY3 (North America only) Bibiocephala comstocki Kellogg, 1903:192, and of Aldrich, 1905:171; Kellogg, 1907:13; Williston, 1908:152; Cole and Lovett, 1921:229; Walley, 1927:1 16; Wirth and Stone, 1956:389; Hennig, 1968:42 [Agathon comstocki (Kellogg)]. Bibiocephala doanei, of Kellogg, 1903:194-195; Aldrich, 1905:171; Kellogg, 1907:13; Williston, 1908:152 [Agathon doanei (Kellogg)]. Bibiocephala elegantulus von Roeder, 1890:230, 232, and of Kellogg, 1903:193; 1907:13; Aldrich, 1905:171; Williston, 1908:152 [Agathon elegantulus (von Roeder)]. Bibiocephala snowi Kellogg, 1903:211. Nomen nudum [Agathon elegantulus (von Roeder)] Bibiocephala sp. B Johannsen, 1934:52, pi. 23, fig. 206 [Philorus californicus Hogue]. Bibiocephala sp. D Johannsen, 1934:52, pi. 23, fig. 204 [Dioptopsis sequoiarum (Alexander)]. Bibiocephala grandis Osten Sacken SYNONYMY Bibiocephala grandis Osten Sacken, 1874:566. Fig. wing. Type locality: USA, “Colorado Mountains” (exact locality un- 3. 1 am introducing this new term here to apply to listings of nominate species originally or at some time combined with a genus under consid- eration but newly recognized in other combinations. Figure 7. Block diagram showing results of block cluster analysis of Bibiocephala subsample. Specimens with similar character states are grouped. Character codes are as follows: EYHAIR absent (i ), present (2); PRCTSM (number of setae) 2-7 (1), 8-11 (2); WING (length in mm) 7. 0-9.0 (1), 9.1-1 1 .1 (2); TERSM (number of setae) 14-25 (1), 26- 43(2). Contributions in Science, Number 338 Hogue: Revised Status of Bibiocephala known but stated to be between 8,000-10,000 feet elevation). LECTOTYPE male, and four PARALECTOTYPE males, by present designation, from original syntype series of five males, all MCZ Bibiocephala kelloggi Garrett, 1922:91. Type locality: CAN- ADA, British Columbia, Cranbrook (?, possibly arrived on train from east). Holotype female (erroneously cited as male in origi- nal description), CPA. NEW SYNONYMY. Bibionus griseus Curran, 1 923: 268-269. Type locality: CAN- ADA, Alberta, Nordegg. Holotype male, CNC. NEW SYNONYMY. Bibiocephala nigripes Alexander, 1965:2-3. Type locality: USA, Idaho, Elmore County, Sawtooth State Forest, near Feath- erville, 4,900 feet. Holotype male, CPA. NEW SYNONYMY. Bibiocephala grandis, of Loew, 1877:95; Osten Sacken, 1877:194, 1891:409, 1895:161; Kellogg, 1903, 1907:13; Aldrich, 1905:171; Williston, 1908:152; Coquillett, 1910:514; Cole and Lovett, 1 921 :229; Walley, 1927:1 15;Wirthand Stone, 1956:389; Figure 8. Ordination plot of scores of first three components resulting from principal components analysis of Bibiocephala continuous variates. Sequence of specimens and approximate localities at top. 8 Contributions in Science, Number 338 Hogue: Revised Status of Bibiocephala Alexander, 1958:820, 1963:50, 1965; Stone, 1965:99; Cole, 1969:74. Bibiocephala grisea, of Stone, 1965:99; Cole, 1969:74. Bibiocephala griseus, of Walley, 1927:115; Alexander, 1958:820, 1963:50, syn. kellcggi. Bibiocephala kelloggi, of Walley, 1927:116; Alexander, 1958:820, 1963:50; Stone, 1965:99; Cole, 1969:74. Bibiocephala nigripes , of Hogue, 1973:9. descriptions Adult (Figs. 9-16) Rather large, robust, well-pigmented Blephariceridae. Coloration. Integument generally well sclerotized, gray-brown, legs tending to be yellow-orange basad, darker areas of body with silvery pollinose opalescence. Wing membrane of male infuscate, of female hyaline. Chaetotaxy. Thorax, abdominal tergite I (laterally), head scle- rites, and lower eye division pilose. Size. Wing length: Range, male (N = 102), 7.0 to 11.1 mm (interquartile range, 8.3 to 9.5 mm), mean = 8.9 mm; female (N = 54), 8.4 to 1 2.3 mm (9.2 to 10.9 mm), mean = 10.2 mm. Head. Structure: Generally as found in family. Antennal fossae approximate (slightly closer in male), mesofrons very narrow, slightly convex. Male (Fig. 11). Sclerites: Suprafrons obliterated dorsad by contiguous eyes, ventral portion triangular, flat; no suprafrontal carina. Parietal sclerite reduced, laterally oblique ventrad. Ocel- lar lobes joined to form a sessile tubercle, lenses inserted on lobes obliquely. Eyes: Upper division well differentiated from lower, about 3 times larger in area than lower; callus oculi very narrow to obsolete laterad. Proboscis: Short, length of free portion dis- tinctly less than head width Mandibles absent. Palpus: 5-seg- mented, length proportions 3:5:5:4:6. Female (Fig. 1 2). Sclerites: Suprafrons narrow, interocular dis- tance equal to 5 ommatidial diameters. Eyes: Upper and lower divisions about equal, former only one-fifth larger in area than latter; callus oculi present. Proboscis: Mandibles present. Palpus: 5-segmented, length proportions 3:5:6:4:7. Otherwise similar to male. Antenna (Fig. 9). Short, length approximately 0.7 times head width; 1 5-segmented (intersegmental membrane between first two flagellomeres often incomplete; these each slightly longer than remaining flagellomeres); scape oval, pedicel elongate, flagellomeres 3-13 all (including apical) of equal length, ovoid, appearing somewhat contracted in life. Thorax and Abdomen. Shape of sclerites typical for subfamily. Wing (Fig. 10). Elongate, subtriangular, broad (broadest across basal third), jugum more expansive than in other genera of Blepharicerini. Venation: Vein R2 + 3 present, long, straight, joining R, near its extreme; M3 + 4 (false crossvein M-Cu) present. Dorsal mac- rotrichia on veins R5 and M3 complete, on M, + 2 and M4 apical only; ventral macrotrichia complete on vein R4 Legs. Forefemur strongly upcurved, slightly incrassate. Re- maining segments straight and simple; tarsomeres 5, all un- modified. Tibial spurs absent from foreleg, a pair each on mid- and hindlegs. Claws simple, similar, and not enlarged. Table 2. First three principal components of the covariance Bibiocephala grandis specimens, for 1 2 continuous variates. matrix of 30 Components Character 1 2 3 WING 0.90 0.27 0.06 FEM ! 0.29 -0.10 0.01 TIB 1 0.36 -0.07 -0.02 TAR 1 0.17 -0.07 0.08 FEM 2 0.41 -0.05 — 0.0 1 TIB 2 0.41 -0.02 -0.05 TAR 2 0 10 -0.04 0.04 FEM 3 0.61 -0.10 -0.06 TIB 3 0.53 -0.17 -0.02 TAR 3 0.18 -0 1 1 0.12 OTDSYL 0.04 0.02 — 0.01 BASYL 0.23 0.08 -0.09 Eigenvalues 2 14 0 16 0.04 Percent total variance 89 6 2 Male Genitalia (Figs. 13-15). Structure: Segment VIII not spe- cially modified; IXth tergite elongate with apicomesal region heavily sclerotized and densely trichiate; lobes of IXth tergite prominent, subtriangular. Basistyle proportionately much larger than in other Blepharicerini, 3.4 times length of outer dististyle, the anterior basistylar extreme projecting below and overriding Vllth sternite. Outer dististyle quadrate, incurved, the apex bi- furcate shallowly, the dorsal lobe heavily setate, the inner weakly so; dorsal margin with weakly developed median and basal lobes. Inner dististyle complex, a curved spatulate plate with a basal, short, flat capitate process (appears triangular in some specimens in dorsal view) and a long, apicodorsal, pannuceate,4 digitiform process, this with a subapicolateral swelling. Parameres and penis filaments very long, former only about twice thickness of latter, hypopygium recurved basad, leading to inverted phal- lobase. Phallosomal vesica small, nonflanged. Female Genitalia (Fig. 16). Structure: VUIth sternite lobe broadly rounded. Oviscapt reduced in overall size, basal portion ex- panded and convex, lobes not well defined. Dorsolateral portions of IXth tergum developed into a large, separate lobe, larger than cercus, which is reduced. Spermathecae three in number, large, pyriform, with tortuous necks. Chaetotaxy: Bristles of VUIth sternite lobes large and numer- ous. Apex of IXth tergite expansion densely setate, distal margin with row of very long, heavy, incurved setae. Base of oviscapt with central setal group (median bristle series). Pupa (Figs. 17-19) Large, robust. Outline shape hemielliptical (length/width pro- portions of male 2.2, of female 2.1); body deep, cross-sectional 4. Finely wrinkled, like fingerprint ridges. Contributions in Science, Number 338 Hogue: Revised Status of Bibiocephala Figures 9 through 12. Bibiocephala grandis, adult male and female. Structure and terminology used defined in text. Figure 9, antenna; Figure 10, wing (male); Figure 1 1, head of male, frontal view; Figure 12, head of female, frontal view. Figures 13 through 16. Bibiocephala grandis, genitalia. Structure and terminology used defined in text. Figure 13, external structures, male, dorsal aspect; Figure 14, ninth tergite, male, ventral view, showing setal groups; Figure 15, internal phallic complex, male, dorsal aspect; Figure 16, female genitalia, ventral aspect. 10 Contributions in Science, Number 338 Hogue: Revised Status of Bibiocephala Contributions in Science, Number 338 Hogue: Revised Status of Bibiocephala 1 1 Figures 17 through 22. Bibiocephala grandis, pupa and fourth-instar larva. Figure 17, denticulae of pupal integument; Figure 18, pupa, anterior aspect; Figure 19, pupa, general features (dorsal left, ventral right); Figure 20, larva, general features (dorsal left, ventral right); Figure 21, detailed structure of typical segment, dorsal; Figure 22, dorsal pseudopod, anterior view. 12 Contributions in Science, Number 338 Hogue: Revised Status of Bibiocephala shape semicircular to subtriangular. Pleural margins overlap venter laterally. Coloration. Dorsum and overlapping lateral margins of same black, ventral membranes white. Size. Body length: Range, male (N = 23) 7.9 to 9.7 mm, mean = 8.8 mm; female (N = 24) 8.65 to 10.5 mm, mean - 9.6 mm. Width: Range, male (N - 23) 3.35 to 4.5 mm, mean - 3.95 mm; female (N - 24) 3.75 to 5.0 mm, mean = 4.5 mm. Male about 90 percent size (length + width averages) of female. Structure. Cuticle: Dorsum of metascuta! (mesally only), scutellar, and abdominal sclerites, and of subbranchial (with mesal extension) and underfolded portions of branchial sclerite, finely and densely denticulate; denticulae (Fig. 17) recurved (giving surface a rough texture, which may be felt by running the tip of the finger anterad over the dorsum). Branchiae: Large, porrect, incurving and converging; posteri- ormost plates contiguous or overlapping in some specimens; all plates projecting appreciably beyond plane of anterior margin. Plates four in number, subparallel (slightly divergent distad), and subequal in size, rigidity, and thickness; individually broadly seculate5 in shape. Dorsal sclerites: Abdominal tergites and scutellum broad, the former appearing to reach lateral margin in dorsal view because of steep pleurae; anteromesal region of frontal sclerite of male with a pair of low, irregular bosses forming a shallow median sulcus, this sclerite evenly rounded in female. Ventral sclerites: Antennal case short, extending only to base of wing case, apex recurved. Vestigial gills, with 4 to 6 filaments each, present on abdominal sternites III-V. Pattern of extension of leg case apices same in both sexes, tip of hindleg most ex- tended, foreleg less, midleg least extended. Fourth-Instar Larva (Figs. 20-22) Large, robust. Coloration. Dorsum generally well pigmented, black; head and sclerites densely sclerotized, shiny. Venter pale, light-brownish to white. Size. Body length: Range of randomly selected specimens (N = 1 22) 5.3 to I 3.7 mm, mean = 9.6 mm; of mature specimens only (i.e., pupal branchiae visible) (N = 38) 7.8 to 13.7 mm, mean - 10.5 mm. Head capsule width: Range 1 .4 to 2.2 mm, mean = 1 .8 mm. Structure. Body divisions: Anterior division spherical in mature specimens, subtriangular to quadrate in younger individuals. Terminal division with well-difFerentiated segment VH-IX; pos- terolateral lobe of segment VII well developed, bearing large dorsal pseudopod. Intercalary segments developed A sharp sub- lateral indentation in dorsal body wall running length of abdo- men ( I— V I ) setting off dorsal pseudopod on a broad lateral lobe. Head: Small relative to body, anterior portion constricted, lat- eralia expanded, fused, leaving only a shallow sulcus, which closes completely in many (especially older) specimens. Antenna with two segments, basal segment slightly thicker (expanding slightly distad) but shorter than distal segment. Integumentary modifications: Dorsum of abdominal segments 5. Sickle-shaped. I -VII usually with antero- and posterotergal plates, these small, circular or elongate, anterior slightly larger and more oval than posterior, sometimes raised into a small tubercle; segment VIII with a single pair of very small plates; all plates evanescent in some specimens. A small tergopleural tubercle occasionally also present on segments I — VI. Dorsal pseudopods (Fig. 22) well de- veloped on segments I- VI I, these of equal size on 1 1- VI, slightly smaller on I and VII; shape antennuoconate, apex truncate, sharply angled posteriorward at basal third, a strong, sharp, erect spine arising just basad of angle. Pseudopods: Normal, that of segment I more elongate and smaller than others, which are of equal size and shape except that of VI, which is slightly smaller. Gills: Lateral gills composed of 6 digitiform filaments, arising from contiguous bases and arranged in a plane, hemirosette pat- tern. Number of filaments and overall size equal on all segments. Anal gills normal. Chaefotaxy (Figs. 20-22). Dorsum and lateral portions of venter set with dense vestiture of short secondary coniform to claviform sensilla with interspersed setiform types (more so laterad); the former also present on apicodorsal portions of dorsal pseudopods, anteroventra! margins of same densely fringed with numerous long, setiform sensilla. Distinctive primary sensilla as follows: subtergal 1-VI about in line with outer edges of dorsal plates, slightly closer to posterior plate, that on VI I very small and close to posterotergal plate, all setiform; tergopleural M-T members in contiguous pairs, all setiform; dorsopseudopodal I— V 1 1 arising from blunt apices of dorsopseudopods, heavy, short setiform. Ter- minal setae of single row of 20-40 large setiforms fringing termi- nus of segments VII I— IX. BIONOMICS Very little is known about the biology of Bibiocephala grcmdis. It is a typical blepharicerid in its general life cycle and ecological requirements, developing in swift mountain streams and leading a riparian adult life. The adults are active during the summer, most collections indicating June to August as the major flight period, although there are later (September) and earlier (April and even March) records. The well-developed mandibles of the females indicate a preda- tory habit; the males, lacking mandibles, probably subsist on liquid food sources (plant nectar and the like). The species is large by family standards, and the females probably can take fairly good-sized prey. The adaptive function of the long pile present over much of the body of both sexes is not known. The curious habit of adults to be attracted to bright silvery objects, possibly mistaking them for the water surface, has been observed by one collector of this species (in this case, “silver metal tanks,” Gittins and Barr, Idaho, Slate Creek, 1963). The winter is presumably spent in the egg stage or as very young larvae. Older larvae are encountered throughout the spring and summer and typically are taken from small to me- dium-sized rocks in riffles and other completely submerged situa- tions. They apparently are not prone to live on exposed vertical faces of large boulders and fixed rock faces as are the larvae of other blepharicerids. I have found them most frequently on the undersides of stones in gravelly stream beds. Larvae are frequently found with very heavy coatings of di- Contributions in Science, Number 338 Hogue: Revised Status of Bibiocephala 13 atoms and algae adhering to the dorsum. This may afford them a degree of protective camouflage or may simply be an accident of their lethargic existence. DISTRIBUTION Bibiocephala grcmdis occupies suitable habitats in all of the Pa- cific and Rocky Mountain states of western North America and the Canadian territories and provinces of Yukon, Alberta, British Columbia, and Saskatchewan. It is not known presently from Alaska or Arizona but is very likely to occur there also. It enters California, Nevada, and New Mexico only in their northern por- tions, and it is very doubtful if it extends its range into northern Mexico (see Fig. 23). SPECIMENS EXAMINED ADULTS CANADA. ALBERTA. Edmonton, date and collector unknown ( 9 : UAT). Jasper: 29 August 1932, R.H. Handford ( 3 : CNC — 3); Maligne Lake, 1 1 August 1926 ( 9 : CNC); Athabaska Falls, 21 August 1932, E. Hearle ( 23 A <5 , 2 9 9: CNC— 17). Athabaska River, Old Fort Point, 31 August 1932, E. Hearle (4 <3 <3, 4 9 9: CNC— 16). BRITISH COLUMBIA. Glenora, March 1941, Wickham (9: USNM). Seton Lake, 2 August 1933, J. McDunnough ( 3 , 9 : CNC). Lillooet, Seton Lake, 31 May 1926, J. McDunnough ( 9 : CNC). SASKATCHEWAN. Fenton, 27 May 1948, J.R. Vockeroth (9 <3 3 , 3 9 9 : CNC — II). North Saskatchewan River, Fort a la Corne, 4 June 1948, J.R. Vockeroth ( <3 : CNC- 10). Nipawin, 2 June 1948, J.R. Vocke- roth (11 <3 <3: CNC). Weldon, 27 May 1948, J.R. Vockeroth ( 1 3 <5 <3 : CNC— 7, 8. 9. 12). YUKON TERRITORY. Dawson, 2, 4, 29, 3 1 August 1962, R. Leech (3 <3 <3 , 2 9 9 : CNC — 13. 14). Mile 87, Dempster Highway, 4-8 August 1 973, G. and D. Wood (<3, 2 9 9: CNC— 15). UNITED STATES. CALIFORNIA, Shasta County. Pit River Dam No. 5, 14 May 1952, H P Chan- dler ( 9 : CAS). Siskiyou County. Dunsmuir, date unknown, Wickham ( 9 : USNM). COLORADO, Clear Creek County. Ex- act locality unknown, 9,000 feet, 8 August 1946, R.H. Painter ( 9 : KS). Grand County. Cameron Pass, 1 1 August 1959, H. Knutson ( 9 : KS). Poudre River Canyon, 28 June 1931, J. Not- tingham ( 9 : KU). Gunnison County. Gothic, 9,500 feet, 15 July 1934, C.P. Alexander ( 9 : CPA). Larimer County. West slope Loveland Pass, 9,850 feet, 8 August 1961, C.H. Mann ( 3 : CNC — 6). “N” Eggers, Roosevelt National Forest, 6,500 feet, 2 July 1 936, H. Spieth (5 <3 <3 : AMNH). Weld County. Platte Can- yon, 23 May 1901, H.G. Dyar and A.N. Caudell ( <3 : USNM). IDAHO, Idaho County. Slate Creek Ranger Station, attracted to silver metal tanks (part), 26-27 June 1963, A.R. Gittins and W.F. Barr (50 5 <5, 20 9 9: CAS, SJS— 23. UI— 25). Latah County. Moscow Mountain, 4 June 1910, J. Aldrich ( <3 : USNM). Lewis County Lawyers Canyon, 26 June 1963, A.R. Gittins (2 9 9: UI). Nez Perce County. Central Grade, 27 April 1960, A. R. Gittins ( <3 : UI). Lewiston, 28 March 1953, C. Terhaar ( <3 : UI); 9 June 1923, A. Melander ( 9 : USNM). Spald- ing, 24 April 1949, M. James ( 3 : USU). Owyhee County. Mur- phy Hot Springs, 20 June 1965, W.F. Barr ( 3 : UI). Shoshone County. Red Ives Ranger Station, 14 July 1969, C. Hornig ( 3 : UI). MONTANA, Gallatin County. Gallatin River, 40 mi S. Belgrade, Malaise trap, 31 July 1971, H. Telford (13: WSU — 30). Upper Gallatin Canyon, 7,000, 8,100 feet, 10, 20 July 1928, J. McDunnough (7 33: AMNH, CNC — 4. 5). Yellowstone, 20 July 1929, E. Hearle ( 9 : CNC). Malvin County. Madison Creek, 16 July 1 948, R. Hays and C. Bishopp (5 3 3 , 3 9 9: USNM). Park County. Livingston, 14 July 1903, A. Melander (2 33, 9 : USNM). NEVADA, Nye County. Currant River, Van Burch, 18 June 1936, H. Spieth ( 3 : AMNH — 1). NEW MEX- ICO, San Miguel County. Pecos: 9 June 1903, M. Grabham (10 3 3 : USNM, BMNH); no date, T.D.A. Cockerell (3 3 3: LACM — 19, 20). OREGON, Hood River County. Hood River, 2, 21, 26 June 1917, 15 June 1919, F. Cole (6 3 3 , 7 9 9 : OS— 2/, 22). UTAH, Cache County. Logan, 16 July 1933, T. Thatcher ( 3 : KU); 14 July, 1 August 1954, T. Miura (3 3 3, 9 : KU); 4 Sep- tember 1950, G. Bohart ( 3,2 9 9: UCD — 24). Smithfield, 1 1 July 1937, C. Smith and F. Harmston (3 3 3: KU — 18. USNM — 26). Duchesne County. Duchesne River and highway 35, west of Hanna, 12 July 1976, R. Baumann and V. Tipton (2 3 3, 9: BYU — 2). Summit County. Weber River, below Smith-Morehouse Creek, 5 August 1975, Sakamoto et al. ( 9 : LACM). WASHINGTON, Asotin County. Grand Ronde River near Anatone, 28 April 1963, B.F. Finnigan (2 33, 9 : WSU — 29). Asotin, 22 April 1 923, A.L. Melander ( 9 : USNM). Clarks- ton, 1 April 1907, A.L. Melander ( 3 : USNM — 27). Jefferson County. Olympic Hot Springs, 22 July 1938, E.C. VanDyke ( 9 : CAS). Pierce County Longmire, Mount Rainier, 3,000 feet, 4 June 1934, Bryant ( 9 : CAS). Whitman County. Almota, 12 April 1941, L.J. Lipavsky ( 9 : WSU). Pullman, date unknown, A.L. Melander (2 33: USNM). Wawawai, 24 April 1 909, A.L. Melander ( 3 : USNM); 7 April 1954, M.T. James and J. Quist ( 9 : WSU). Yakima County. Chinook Pass, 29 July 1949, R.H. Beamer( 9 : KU). WYOMING, Yellowstone National Park. Gal- latin Station, Yellowstone Park, 28 July 1923, A. Melander ( 3 : USNM — 28). Firehole River, 7,600 feet, Yellowstone Park, 21 July 1928, J. McDunnough ( 3 : CNC). Yellowstone Park, date and collector unknown (2 3 3, 3 9 9: CU). NO DATA ( 9 : WSU). IMMATURES CANADA. ALBERTA. Fiddle Creek, Jasper National Park, 19 June 1962, G B Wiggins ( 1 larva: ROM). BRITISH COLUM- BIA. Kicking Horse River, Yoho National Park, 26 April 1964, D.M. Wood (9 larvae: CNC). Thompson’s Bridge, Fraser Can- yon, August 1967, D. Craig (12 larvae: UAT). Mule Deer Provin- cial Park, August 1967, D. Craig (9 larvae: UAT). UNITED STATES. CALIFORNIA, Nevada County. S Fork Yuba River, 6 mi NW Nevada City, highway 49, 25 June 1973, D.R. Givens and S.D. Smith (8 larvae: CWS). COLORADO, Archuleta County. Rio Blanco-Navajo River, 1974, H. Kennedy (19 larvae: LACM). Boulder County. Left Hand Creek, near Altona, 18 February 1943, T.H. Frison and H.G. Rodeck (1 larva: USNM). Gunnison County. Copper Creek, Gothic, 27 July 1967, T.C. Emmel (36 pupae: LACM). East River, Gothic, 9,500 feet, 28 June 1967, T.C. Emmel ( 1 larva: LACM). Marble, 4 September 1966, J.T. Polhemus (34 larvae: LACM). Larimer County. Big Thompson River, Estes Park, 18 February 1943, T.H. Frison and H.G. Rodeck (2 larvae: INHS). Mineral County. Rio Grande River, Marshall Camp, 16 June 1963, D.M. Wood (3 larvae, 18 14 Contributions in Science, Number 338 Hogue: Revised Status of Bibiocephala pupae: CNC). Rio Blanco County. S Fork White River, 10 mi S Buford, 4 July 1 968, J.T, Polhemus (2 larvae, 20 pupae: LACM). Routt County. Elk River, 3 mi S Clark, 18 March 1968, B.R. Oblad (1 larva: LACM). Summit County. Stream along US 6, 8 mi W Loveland Pass, 23 August 1968, J. Emmel and O. Shields (21arvae: LACM). IDAHO, Custer County. HerdCreek, 1 1 mi S Clayton, 1 1 May 1965, S.D. Smith and E.R. Logan (2 larvae: UI). Idaho County. French Creek, 19 mi E Riggins, 25 March 1966, S.D. Smith (1 1 larvae: UI). Crooked Fork Creek, 4 mi NE Figure 23. Collection localities of Bibiocephala grandis in western North America. Powell Ranger Station, 30 March 1967, F.H. Everest (5 larvae: UI). Lemhi County. Bear Valley Creek, 12 mi SW Lemhi, “7-5-66,” H R. Gibson (2 larvae: UI). OREGON, Wallowa County. 6 mi E Lostine, Forest Service Road S-202 at national forest boundary, 24 March 1973, D.L. Peck (1 larva: CWS). Unknown abluent to Wallowa Lake, 24 April 1977, Zack and RafTa (8 larvae: WSU). UTAH, Cache County. Logan River, 25 September 1952, collector unknown (6 larvae: USNM). Du- chesne County. Duchesne River, highway 35 west of Hanna, 12 July 1976, R. Baumann and V. Tipton (3 pupae: BYU). Rock Creek near mouth of Fall Creek, 8 October 1933, Rasmussen (2 larvae: INHS). East fork of Rock Creek, 7 October 1933, Rasmussen (3 larvae: INHS). Salt Lake County. Big Cotton- wood Creek, 2 mi above power plant, approximately 6,000 feet, 25 February 1968, R and D. Koss (20 larvae: JH). Big Cotton- wood Creek, 5 mi W Brighton, The Spruces, 3 June 1976, C.L. Hogue (63 larvae: LACM). Millcreek, 3 February, 10, 30 April 1966, R.W. Baumann (7 larvae: LACM). Mill Creek, 6 mi E canyon mouth, “The Elbow,” 3 June 1976, C.L. Hogue (27 lar- vae: LACM). Millcreek Canyon Creek, 16 December 1965, R.W. Baumann (1 larva: LACM). Summit County. Bear River, near Utah-Wyoming state line, 14 May 1970, R.B. Golightly (4 larvae: LACM). Mid fork Provo River, Slate Gorge, 3 June 1976, C.L. Hogue (10 larvae: LACM). Uintah County. White River, Bonanza, 17 September, 8 November 1975, 25 April 1976, E. Hornig ( 1 1 larvae: BYU). Utah County. American Fork Canyon Creek, 7, 15 April 1966, R.W. Baumann and B.R. Oblad (5 larvae: LACM). North Fork Creek, below Sundance ski area, 5 August 1976, D.K. Sakaguchi (3 larvae, 4 pupae: BYU). WASH- INGTON, Asotin County. Asotin Creek, 14 April 1 935, collector unknown (1 larva: WSU). Jefferson County. Hoh River, Hoh River Campground, Olympic National Park, 30 June-1 July 1969, ROM field party (18 pupae: ROM). Skagit County. Mar- ble Creek, 10 mi E Marblemount, 30 June 1974, D.L. Peck (1 larva: CWS). Snohomish ? County. Pilchuck River, 12 May 1935, collector unknown (2 pupae: WSU). Whatcom County. Goodell Creek at Newhalem, highway 20, 1 July 1974, D.L. Peck ( I larva: CWS). WYOMING, Albany County. Centennial, Wyoming Summer Camp, 30 June 1936, collector unknown (1 larva, 5 pupae: USNM). Libby Creek along highway 1 30, below Snowy Range Lodge, Medicine Bow Mountains, 22 August 1968, J. Emmel and O. Shields (19 larvae: LACM). Upper and Lower Libby Creek, Snowy Range Mountains, 23 April, 3 Sep- tember, 2 October 1955, 14-15, 29 April, 23 June 1956, PH. Freytag et al. (22 larvae, I pupa: OSU, UW). Nash Fork Creek, Snowy Range Mountains, 14 August 1955, P. and R. Freytag ( 1 larva: OSU). North fork Little Laramie River, 30 April 1955, P. Freytag et al. ( 1 larva: UW). Carbon County. South Brush Creek, 54 mi W Laramie, South Brush Creek Campground, Route 1 30, 1 1 September 1 966, G. Wiggins et al. (78 larvae: ROM). Yellow- stone National Park. Gibson River near Norris, Geyser Basin, 23 August 1953, R. Coleman (3 larvae: USNM). ACKNOWLEDGMENTS I am greatly indebted to Drs. Richard Friesen, James D. Smith, and Nancy J. Agafitei for their patient guidance and instruction in multivariate statistical methods and machine computing. Contributions in Science, Number 338 Hogue: Revised Status of Bibiocephala which made the mathematical approach to my problem possible. A tremendous amount of the physical handling of material and computation was ably accomplished by Charlotte Low and Betty Birdsall; Figures 1-4 and 14 were rendered by Carol Madle. For the use of specimens loaned and donated, I wish to thank the following individuals: C.P. Alexander, Amherst, Mas- sachusetts; R.T. Allen, INF1S; Paul H. Arnaud, Jr., CAS; Rich- ard W. Baumann, Bean Life Science Museum, BYU; George W. Byers, KU; Douglas C. Craig, UAT; J.G. Edwards, SJS; Paul Freeman, BMNH; Paul Freytag and Charles A. Triplehorn, OSU; Don Fronk, UW; Harold Grant (deceased), ANSP; Maurice James, Richard Zack, and William Turner, WSU; Rich- ard Koss, JH; J.D. Lattin, OS; Norman Marston, KS; A.T. McClay (deceased), UCD; L.L. Pechuman, CU; B.V. Peterson, CNC, Biosystematics Research Institute, Agriculture Canada; Jerry A. Powell, CIS, University of California at Berkeley; S.D. Smith, CWS; Alan Stone and F. Christian Thompson, Agri- cultural Research Service, U.S. Department of Agriculture, USNM; Margaret Thayer, MCZ, Harvard University; Richard Westcott and William F. Barr, UI; Glenn B. Wiggins, ROM; and Pedro Wygodzinsky, AMNH. My thanks are also extended to my associates Julian P. Dona- hue and Roy R. Snelling for criticism of the manuscript. Finally, I am dedicating this work to Dr. Charles P. Alexander, indefatigable dipterist of the century, whose continual interest in my studies with the Blephariceridae has given me immeasurable encouragement. LITERATURE CITED Agafitei, N.J., and R.B. Selander. 1980. First instar larvae of the Vittata Group of the genus Epicauta (Coleoptera: Meloidae). Kansas Entomological Society, Journal 53:1-26. Aldrich, J.M. 1 905. A catalogue of North American Diptera (or two-winged flies). Smithsonian Miscellaneous Collections 46(1 444): 1-680. Alexander, C.P. 1958. Geographical distribution of the net- winged midges. Tenth International Congress of Entomology, Proceedings 1 ( 1956):81 3—828. 1963. Family Blepharoceridae. Pages 39-71 in Guide to the Insects of Connecticut, Part VI. The Diptera or true Hies of Connecticut, Fascicle 8 (part). Connecticut State Geo- logical and Natural History Survey, Bulletin 93. 1965. A new net-winged midge from Idaho (Blepharoceridae, Diptera). Great Basin Naturalist 25:1-3. Cole, F.R. 1969. The flies of western North America. University of California Press, Berkeley. Cole, F.R., and A.L. Lovett. 1921. An annotated list of the Dip- tera (flies) of Oregon. California Academy of Sciences, Pro- ceedings Ser. 4, 11:1 97-344. Coquillett, D.W. 1910. The type-species of the North American genera of Diptera. United States National Museum, Proceed- ings 33:499-647 Curran, C.H. 1923. Studies in Canadian Diptera II. — The gen- era of the family Blepharoceridae. Canadian Entomologist 55:266-269. Dixon, W.J., series ed. 1977. BMDP-77. Biomedical computer programs, P-Series. University of California Press, Berkeley. Garrett, C.B.D. 1922. Two new Blepharoceridae (Diptera). In- secutor Inscitiae Menstruus 9: 1 1 9-132. Hartigan, J.A. 1976. Modal blocks in dentition of west coast mammals. Systematic Zoology 25:149-160. Hennig, W. 1968. Die larvenformen der Dipteren. Part 2. Aca- demie Verlag, Berlin. Hogue, C.L. 1 973. The net-winged midges or Blephariceridae of California. California Insect Survey, Bulletin 15:1-83. 1 978. The net-winged midges of eastern North Amer- ica, with notes on new taxonomic characters in the family Blephariceridae (Diptera). Natural History Museum of Los Angeles County, Contributions in Science 29 1 : 1 -4 1 . Johannsen, O.A. 1934. Aquatic Diptera. Part I. Nemocera, ex- clusive of Chironomidae and Ceratopogonidae. Cornell Uni- versity Agricultural Experiment Station, Memoirs 164:1-71, 24 pis. Kellogg, V.L. 1 903. The net-winged midges (Blepharoceridae) of North America. California Academy of Sciences, Proceedings Ser. 3, 3:187-233. 1907. Diptera. Fam. Blepharoceridae. Genera Insec- torum 56:1-15, 2 pi. Loew, H. 1877. Revision der Blepharoceridae. Zeitschrift fur Entomologische Breslau, Neu Folge 6:54-98. Osten Sacken, C.R. 1874. Report on the Diptera collected by Lieut. W.L. Carpenter in Colorado during the summer of 1873. United States Geological and Geographical Survey of the Territories, Annual Report [7] :56 1 — 566. 1877. Western Diptera: Descriptions of new genera and species of Diptera from the region west of the Mississippi and especially from California. United States Geological and Geographical Survey of the Territories, Bulletin 3:189-354. 1891. Synopsis of the described genera and species of the Blepharoceridae. Berliner Entomologische Zeitschrift 36:407-41 1. - . 1 895. Contributions to the study of the Liponeuridae Loew (Blepharoceridae Loew, olim). Berliner Entomologische Zeitschrift 40:148-169. Von Roeder, V. 1890. Zwei neue nordamerikanische Dipteren. Wiener Entomologische Zeitung 9:230-232. Sneath, PH. A., and R.R. Sokal. 1973. Numerical taxonomy. The principles and practice of numerical classification. W.H. Freeman and Co., San Francisco. Stone, A. 1965. Family Blephariceridae. Pages 99-100 in A. Stone, C.W. Sabrosky, W.W. Wirth, R.H. Foote, and J.R. Coolson, A catalog of the Diptera of America north of Mexico. United States Department of Agriculture, Washington, D C. Walley, G.S. 1927. Review of the Canadian species of the dip- terous family Blephariceridae. Canadian Entomologist 59:112-116. Williston, S.W. 1908 (Ed. 3). Manual of North American Dip- tera. James T. Hathaway, New Haven, Connecticut. Wirth, W.W., and A. Stone. 1956. Family Blephariceridae. Pages 389-390 in R.L. Usinger, ed., Aquatic insects of California. University of California Press, Berkeley Zwick, P. 1977. Australian Blephariceridae. Australian Journal Zoology, Suppl. Ser. 46:1-121. Submitted 9 March 1981. Accepted for publication 16 October 1981. 16 Contributions in Science, Number 338 Hogue: Revised Status of Bibiocephala |N' CONTRIBUTIONS IN SCIENCE ill; REVISION OF THE EASTERN NORTH PACIFIC ANTHIIN BASSES (PISCES: SERRANIDAE) John E. Fitch KliSliSPi^ : I • #■ “ :■ : : l! h >i SERIAL PUBLICATIONS OF THE NATURAL HISTORY MUSEUM OF LOS ANGELES COUNTY The scientific publications of the Natural History Museum of Los Angeles County have been issued at irregular intervals in three major series; the articles in each series are numbered individually, and numbers run consecutively, regardless of subject matter. • Contributions in Science, a miscellaneous series of technical papers describing original research in the life and earth sciences. • Science Bulletin, a miscellaneous series of monographs describing original research in the life and earth sciences. This series was discontinued in 1978 with the issue of Numbers 29 and 30; monographs are now published by the Museum in the Contributions in Science series, • Science Series, long articles on natural history topics, generally written for the layman. Copies of the publications in these series are sold through the Museum Bookshop. For instructions for authors, refer to Contribution No. 336 or contact the Museum Publications Office. Leon G. Arnold Lawrence G. Barnes Robert L. Bezy Kenneth E. Campbell Robert Gustafson John M. Harris Robert J. Lavenberg Camra C. Swift EDITORIAL BOARD Fred S. Truxal David P. Whistler Edward C. Wilson John W. Wright EDITOR Robin A. Simpson HI I® REVISION OF THE EASTERN NORTH PACIFIC ANTHIIN BASSES (PISCES: SERRANIDAE)' John E. Fitch* 2 ABSTRACT: Four species of anthiin basses inhabit eastern North Pacific waters: Pronotogrammus multifasciatus Gill, 1863; P eos Gilbert, 1890; Hemanthias peruanus Steindachner, 1874; and H signifer (Gar- man, I 899). Anthias gordensis Wade, 1946, and Holanthias sechurae Barton, 1947, are relegated to the synonymy of Pronotogrammus multi- fasciatus. Pronotogrammus Gill, 1863, has priority over Holanthias Gunther, 1 868. P. eos should be generically reassigned. Centristhmus Garrnan, 1899, is relegated to the synonymy of He- manthias Steindachner, 1874. H. delsolari is a nomen nudum that has been applied to specimens of H. signifer in Peruvian waters (Chirichigno, 1974). Two of the four species occur in Californian waters: Pronotogrammus multifasciatus (Hobson, 1975, as Anthias gordensis ), and Hemanthias signifer (misidentified as H. peruanus by Wales [1932]). An identification key is presented; the key utilizes facial scalation, pored lateral-line scale counts, eye diameter, finray counts, and other easily observed characters for differentiating the four eastern North Pa- cific anthiins. Otoliths (sagittae) of these species are illustrated because of their occurrence as remnants of prey in food habit studies. INTRODUCTION An attempt to identify an anthiin that had been caught around 3 February 1977 in 1 92 meters of water off San Onofre, California, revealed considerable taxonomic confusion and contradiction among accounts of eastern North Pacific Anthiinae. Although an abundance of literature, including original descriptions (Gill, 1 863; Steindachner, 1 874; Gilbert, 1890; Garrnan, 1899; Wade, 1946; and Barton, 1947), was available for anthiins in the eastern North Pacific, few faunal reports and identification keys in- cluded more than one or two species (Jordan and Evermann, 1896; Hiyama, 1937; Flildebrand, 1946; Hildebrand and Barton, 1949; Berdegue, 1956; Chirichigno, 1974; Walford, 1974; Miller and Lea, 1976; Ramirez and Gonzales, 1976; and Thomson, Findley, and Kerstitch, 1979). In addition, characters given in these publications often were incorrect or inadequate, making it impossible to compare the same features on different species. As a result, numerous anthiins in collections at the California Acad- emy of Sciences (CAS, or SU-CAS), Natural History Museum of Los Angeles County (LACM), Scripps Institution of Oceanography (SIO), and University of California at Los An- geles (UCLA) were misidentified. Obviously, a critical review of eastern North Pacific anthiins was sorely needed. During the past several years, all eastern Pacific anthiins in the CAS, Cabrillo Marine Museum (CMM), LACM, SIO and Contributions in Science, Number 339, pp. 1-8 Natural History Museum of Los Angeles County, 1982 UCLA collections were examined, as was selected material in the University of Costa Rica (UCR) collection. Proportional measurements in thousandths of standard length (SL), were made on 147 of these specimens representing the known size ranges for the four anthiins that inhabit eastern North Pacific waters (Table 1 ). The data were used to construct the following key and species accounts. Only primary syn- onymies have been included in the species accounts. Otoliths (sagittae) of the four species are illustrated because these frequently occur in digestive tracts and scats of various predators, and might be found in coastal Indian camp sites or in fossil deposits. Although a new generic name is needed for Pronotogrammus eos, such action is being deferred to Phillip C. Heemstra, who is in the process of revising the subfamily for the entire eastern Pacific. Larvae of three of the four eastern North Pacific anthiins have been described and illustrated (Kendall, 1979), but those of He- manthias signifer have yet to be recognized. KEY TO THE ANTHIINS OF THE EASTERN NORTH PACIFIC la. Pored scales in lateral line 36 to 40; eye large, 12 to 16 percent of SL Pronotogrammus eos 1 b. Pored scales in lateral line 45 or more; eye smaller, less than 1 1 percent of SL 2 2a. Maxillary, mandible, and anterior part of face and head completely scaled; third dorsal spine only slightly longer than second, about equal to fourth; elongate oval patch of teeth on tongue (often difficult to find); vertebrae 10 -F 16. . . Pronotogrammus multifasciatus 2b. Maxillary, mandible, and anterior part of face and head without scales; third dorsal spine about three times as long as second, twice as long as fourth; no teeth on tongue; ver- tebrae 11 + 15 3 3a. Margin of urohyal smooth at all sizes; pored scales in lat- eral line 52 to 59; pectoral rays 16 to 18 (usually 17); total gill rakers 31 to 34 (rarely 34); in fish over 100 mm SL, 1 . Review committee for this Contribution: John E. Graves, Robert J Lavenberg, and C. Richard Robins. 2 Research Associate in Ichthyology and Paleontology, Natural His- tory Museum of Los Angeles County, Los Angeles, California 90007. ISSN 0459-01 13 Table 1. Comparative measurements, in thousandths of standard length, for the four eastern North Pacific anthiins Species and Number of Specimens Measured Measurement P. eos P. multifasciatus H. peruanus H. signifer Number of specimens 30 27 55 35 Standard lengths (mm) 29-153 66.5-205 71.5-267 60-279 Head length 361-422 322-354 326-385 328-405 Horizontal eye diameter 120-159 081 -105 073-102 080-1 10 Snout length 054-079 064-085 065-088 078-104 Maxillary length 152-183 135-151 139-160 137-161 Bony interorbital width 060-087 065-080 063-078 061-075 Length of first gill raker below angle 060-080 052-071 045-062 044-061 Snout to first D insertion 303-345 283-323 289-329 293-341 Snout to A insertion 610-656 604-695 587-656 587-646 Snout to pectoral insertion 341-382 331-378 316-361 306-380 Snout to pelvic insertion 345-397 350-439 353-417 355-412 Length of anal, fin base 167-199 164-199 179-219 175-208 Length of dorsal fin base 476-561 548-622 509-579 488-557 Dorsal peduncle length 168-199 154-189 184-215 177-201 Ventral peduncle length 231-258 221-256 214-252 217-258 Dorsal fin insertion to pelvic insertion 306-369 303-354 300-341 272-344 Caudal peduncle depth 1 11-128 091-119 124-148 101-134 Height of first D spine 041-062 046-073 039-056 031-054 Height of second D spine 073-092 063- 1 14 067-090 061-088 Height of third D spine 094-124 105-147 199-352 176-553 Height of fourth D spine 115-155 107-150 086-136 097-138 Height of fifth D spine 118-158 104-150 086-130 096-128 Longest ray in second D fin 190-220 129-173 176-219 127-197 Longest ray in anal fin 196-230 137-190 176-250 137-187 Length of second A spine 115-150 096-147 069-122 066-113 Length of third A spine 115-149 102-140 092-127 092-120 Pectoral fin length 295-339 262-315 226-274 208-254 Pelvic fin length 310-374 301-466 266-536 229-319 middle rays of both upper and lower lobes of caudal tin longest Hemanthias peruanus 3b. A sharp, anteriorly projecting spine (visible without dissec- tion) on ventral edge of urohyal in fish longer than about 70 mm SL; pored scales in lateral line 59 to 69; pectoral rays 18 to 20 (mostly 19); total gill rakers 33 to 38 (rarely 33); outermost rays of both lobes of caudal fin longest Hemanthias signifer SPECIES ACCOUNTS Pronotogrammus multifasciatus Gill, 1863 Figure 1 Pronotogrammus multifasciatus Gill, 1863:84 (original descrip- tion, type locality: Cape San Lucas, Mexico). Anthias multifasciatus (Gill). Jordan and Gilbert, 1 882:360 (ge- neric reassignment, redescription). Anthias gordensis Wade, 1946:225 (original description, type locality: Inner Gorda Bank, Baja California, Mexico, 70 to 78 fathoms). Holanthias sechurae Barton, 1947:2 (original description, type locality: Talara, Peru, exact location uncertain). Holanthias gordensis (Wade). Hubbs, Follett, and Dempster, 1979:21 (checklist of Californian fishes). MERISTIC DATA. D. X-XI, 14-15; A. Ill, 7-8; P. 18-20; GR 10-12 + 26-30 = 37-42; pored lateral line scales 45-51; ver- tebrae 10+ 16. Twenty-six of the 27 P. multifasciatus from which 1 obtained meristic data had dorsal finray counts of X, 15 and anal finray counts of III, 7. DI AGNOSTIC CHARACTERS. Pronotogrammus multifasciatus is the only eastern North Pacific anthiin with completely scaled maxillary, mandible, and anterior part of the face and head. It also differs from the other three species in having 45 to 5 1 pored lateral-line scales and an elongate, oval patch of granular teeth on the tongue. It is similar to P. eos in lacking an elongate. Contributions in Science, Number 339 Fitch: Eastern North Pacific Anthiins filamentous third dorsal spine but differs in having a shorter maxillary and shorter rays in the second dorsal and anal fins (Table I ). P. multifasciatus also has a greater number of pectoral rays than P eos, although one of the 27 specimens (LAC M 8836) had 18 rays in the left pectoral and 19 in the right. MAXIMUM SIZE. The largest individual 1 could find was 205 mm SL (262 mm total length [TL] ). RANGE. Portuguese Bend, Los Angeles Co., California (34° N), to the region off Talara in northern Peru (4° S); 40 to 205 meters. DISCUSSION. Gill’s (1863) description of P. multifasciatus, based on a 50-mm (2-inch) specimen collected by John Xantus at Cape San Lucas, contained very little information that could be used to distinguish this species from other eastern North Pacific anthiins. He did note that none of the dorsal spines was especially elongate or filamentous, but this would also be true for a 50-mm long Hemanthias and was not of diagnostic value. His pored lateral-line scale count (45) was the only character that could be used to separate this fish from other eastern North Pacific anthiins. Jordan and Evermann (1896) apparently based their generic description on P eos because they reported that Pronotogram- mus lacks teeth on the tongue and has a “naked top of head and maxillary.” Their account of the species duplicates Gill’s (1863) description of the holotype, which- until then (1896) — was the only known specimen. Three years later, Garman (1899) listed three Albatross stations where P. multifasciatus had been taken and noted (for the first time) that the entire head and maxillary were covered with scales and that the tongue had a group of teeth on it. Walford (1974) reported that P multifasciatus was known from “Cape San Lucas to the Galapagos Islands,” but I have been unable to locate a single specimen in a west coast collection that had been identified as P multifasciatus. An inquiry to the National Museum of Natural History re- garding the holotype of P. multifasciatus (USN M 2762) brought the reply that “the brown body, without skin, scales, head or tail is about 30 mm long” (Bruce B. Collette, pers. commun). A radiograph of this fragment revealed 16 caudal vertebrae, which eliminated both species of Hemanthias but not P eos. Gill’s ( 1 863) pored lateral-line scale count eliminated P. eos, however. Anthias gordensis Wade, 1 946, and Holanthias sechurae Bar- ton, 1947, can not be distinguished and are here considered junior synonyms of Pronotogrammus multifasciatus. Pro- notogrammus Gill, 1863, has priority over Holanthias Gunther, 1868. Pronotogrammus eos Gilbert, 1 890 Figure 2 Pronotogrammus eos Gilbert, 1890:100 (original description, type locality: Panama). Anthias eos (Gilbert). Boulcnger, 1895 (generic reassignment, list). MERISTIC DATA. D. X, 14-15; A. Ill, 7-8; P 17-18; GR 1 1- 13 + 27-30 = 38-43; pored lateral-line scales 36-40; vertebrae 10+ 16. Three of the 30 individuals I examined to obtain meristic data had dorsal finray counts of X, 14, and one had an anal finray count of III, 7. DIAGNOSTIC CHARACTERS. Pronotogrammus eos differs from the other three eastern North Pacific anthiins in having only 36 to 40 pored lateral-line scales and by its large eye, comprising 1 2.0 to 1 5.9 percent of SL. MAXIMUM SIZE: The largest individual I examined was 153 mm SL (207 mm TL). RANGE. Mid-Gulf of California (28° N) to Panama (7° N); I 1 5 to 325 meters. DISCUSSION. This species lacks several characters that are of generic magnitude and diagnostic for Pronotogrammus (e.g., fully scaled maxillary, mandible, and anterior face and head; patch of granular teeth on tongue; and salient features on its otoliths). Its proper generic assignment must await a revision of eastern Pacific Anthiinae that currently is in progress ( Phillip C. Heemstra, pers. commun ). Hemanthias peruanus (Steindachner, 1874) Figure 3 Anthias ( Hemanthias ) peruanus Steindachner, 1874:378 (origi- nal description, type locality: Paita, Peru). Pronotogrammus peruanus (Steindachner). Jordan and Eigen- mann, 1890:413 (generic reassignment, redescription). Hemianthias peruanus Steindachner. Jordan and Evermann, I 896: 1 222 (subgenus to generic rank, spelling variant). MERISTIC DATA. D. I X-X, 13-15; A. Ill, 7-9; P 16-18; GR 9 10 + 22-24 = 3 1 -34; pored lateral-line scales 52-59; vertebrae II + 15. One of the 55 specimens from which I obtained meristic data had nine dorsal spines, the rest had ten. Forty-eight of these specimens had 1 4 dorsal soft rays, four had 1 3, and three had 1 5. Fifty of the 55 had anal finray counts of III, 8, two had counts of III, 7, and three had III, 9. DIAGNOSTIC CHARACTERS. Hemanthias peruanus is the only eastern North Pacific anthiin in which the middle rays of each lobe of the caudal fin are the longest (not readily visible in speci- mens smaller than about 90 mm SL). It shares with H. signifer an elongate, filamentous third dorsal spine but lacks the antrorse spine that is found on the urohyal of H. signifer longer than about 70 mm SL. It also can be distinguished from H signifer by gill raker, pectoral ray, and pored lateral-line scale counts, except at overlap numbers (Table 2). MAXIMUM SIZE. The largest individual I examined was 305 mm SL (442 mm TL). RANGE. Hipolito Bank, Baja California Sur (27° N), to Tru- jillo, Peru (8° S); 10 to 1 I 7 meters. Hildebrand ( 1946) reported the southern range as “Chile,” but I could not verify any records south of Trujillo, Peru. Hemanthias signifer ( Garman, 1 899) Figure 4 Centristhmus signifer Garman, 1899:48 (original description, type locality: off Panama at 7 ° 33'N, 78 ° 34'20" W). Hemianthias peruanus ( non Steindachner). Wales, 1932:106 (first California record, misidentified, spelling variant). Contributions in Science, Number 339 Fitch: Eastern North Pacific Anthiins 3 Figure 1. Pronotograrnmus multifasciatus, 176 mm SL, Uncle Sam Bank, Baja California (LACM 38671). Photograph by Paul Gregory. * „ M ifppplpppppii mi lll!|!!ll llll 7 8 0 1 1 1 lllllllll lilllllll III 2 1 3 1 4 1 5 1 6 1 Figure 2. Pronotograrnmus eos, 1 29 mm SL, middle of mouth of Gulf of Nicoya, Costa Rica (LACM 33827). Photograph by Richard Meier 4 Contributions in Science, Number 339 Fitch: Eastern North Pacific Anthiins Figure 3. Hemanthias peruanus, 267 mm SL, Hipoiito Bank, Baja California (LACM W61 124). Photograph by Jack W. Schott. Figure 4. Hemanthias signifer, 201 mm SL, off San Onofre, California (LACM 36401 ). Photograph by Jack W. Schott. Contributions in Science, Number 339 Fitch: Eastern North Pacific Anthiins 5 Table 2. Comparison of pectoral rays, gill rakers and pored lateral-line scales for Hemanthias peruanus and Hemanthias signifer. H. peruanus H. signifer Pectoral rays 16 1 — 17 27 — 18 23 1 1 19 — 22 20 — 2 Total gill rakers 3! 13 — 32 19 — 33 17 1 34 2 14 35 — 9 36 — 10 37 — 0 38 — 1 Pored lateral-line scales 52 1 1 — 53 15 — 54 1 1 — 55 8 — 56 2 — 57 2 — 58 1 — 59 1 2 60 2 61 — 3 62 0 63 — 7 64 — 2 65 — 5 66 — 7 67 — 3 68 — 3 69 — 1 MERISTIC DATA. D. X, 13-14; A. Ill, 7-9; P. 18-20; GR 9- 11 +24-27 = 33-38; pored lateral-line scales 59-69; vertebrae 11 + 15. One of the 35 specimens from which I obtained meristics had a dorsal finray count of X, 1 3; all others had X, 1 4. Except for one individual with seven anal soft rays, and one with nine, this count invariably was eight. DIAGNOSTIC CHARACTERS. Hemanthias signifer longer than about 70mm SL have an anteriorly projecting spine on the urohyal (easily observed by lifting either gill cover) that is unique among anthiins. In the eastern North Pacific, it shares with H. peruanus an elongate, filamentous third dorsal spine, but most individuals can be distinguished by counting pored lateral-line scales, pectoral rays, and gill rakers (Table 2). In addition, the outermost rays of both lobes of the caudal fin are the longest — at all sizes for H. signifer, but only for specimens smaller than about 90 mm SL for H. peruanus. MAXIMUM SIZE. The largest individual I examined was 296 mm SL (385 mm TL). RANGE. Playa del Rey, California (34° N), to off Paita, in northern Peru (5° S); 23 to 306 meters. DISCUSSION. The unique antrorse spine on the urohyal of this species appears to have been Garman’s (1899) primary reason for establishing the genus Centristhmus. His C. signifer, however, is so closely similar to Hemanthias peruanus that, until the spine first appears on the urohyal (at about 70 mm SL), one must count pectoral fin rays, gill rakers, and pored lateral-line scales to deter- mine which of the two species is at hand. In fact, juveniles of both species often have been taken in the same trawl haul, but until now, H signifer seldom was recognized (e.g., SIO 68-66 from La Paz Bay, Gulf of California, in 32 to 42 fathoms [58-77 meters] contained 1 58 fish labeled H. peruanus, but critical examination revealed that only 89 were H. peruanus whereas the other 69 were H. signifer, none of which had an antrorse spine on the urohyal). Obviously, Centristhmus must be considered a junior synonym of Hemanthias. Norma Chirichigno (pers. commun.) described Hemanthias delsolari n. sp. in her unpublished thesis (1970), and subse- quently (Chirichigno, 1974:289) used the name in her “Clave para identificar los peces marinos del Peru.” While her paper was in press, she learned that her H. delsolari was a synonym of Centristhmus signifer (see last page of “addenda” in Chi- richigno, 1974), so the name does not appear elsewhere in reports on the marine fish fauna of Peru. Thus, because H. delsolari fails to meet requirements of establishing a new species as spelled out in the International Code of Zoological Nomenclature, it has no status in taxonomic literature. Wales (1932) reported Hemianthias [sic] peruanus from Cal- ifornian waters (based upon a specimen caught off Redondo Beach), but the reported pored lateral-line scale count was diag- nostic for H. signifer, and examination of the fish in the CAS collection (SU-CAS 24812) revealed the antrorse spine on the urohyal. In February 1977, two additional specimens of H. sig- nifer were caught off California: one off San Onofre (LACM 36401 ) and the other off Playa del Rey (LACM 36944). OTOLITHS In the eastern North Pacific, anthiins apparently are a choice prey for many predators. At least two specimens of Prono- togrammus multifasciatus in west coast collections came from stomachs of larger fish. One (SIO 59-326) came from the stom- ach of a yellowtail, Seriola lalandi lalandi, and the other (LACM W53-313) from a spotted cabrilla, Epinephelus analogus. Otoliths of all four species commonly are found in scats of sea lions, Zalophus calif ornianus, that haul out on Islotes Island (north of La Paz), Gulf of California (unpublished data in my files). When conventional identification characters have disappeared from prey species because of digestive action of the predator, otoliths (sagittae) usually can be found and offer an excellent opportunity to identify the prey to genus or species. Because of this, and the apparent importance of anthiins in the food web. 6 Contributions in Science, Number 399 Fitch: Eastern North Pacific Anthiins sagittae of the four eastern North Pacific anthiins are illustrated (Fig. 5). In identifying otoliths, characters on the inner faces (grooved side) are extremely important for determining family and genus. Overall otolith shape, ratio of height into length, marginal orna- mentation, and similar characters are of importance primarily at species level. The illustrated otoliths (Fig. 5) are all inner faces. Figure 5. inner faces of left otoliths (sagittae) of the four eastern North Pacific anthiins: Part a, Pronotogrammus multifasciatus, 9. 1 by 4.6 mm (length and height), from fish 205 mm SL; Part b, Pronotogrammus eos, 8.7 by 6.3 mm, from fish I 30 mm SL; Part c, Hemanthias peruanus, 10.9 by 6.0 mm, from fish 2 1 3 mm. SL; Part d, Hemanthias signifer, 1 1 .0 by 5.5 mm, from fish 256 mm SL. Photographs by Richard Meier. and all visible features are quite typical of sagittae for other individuals of the same species. Sagittae of all four anthiins have a “channel” just above the ventral margin and roughly parallel to it; the channel is deepest in P multifasciatus and least developed in H. signifer. In all but P. eos , the cauda (posterior portion of sulcus or central groove) flexes slightly downward at its posterior end, and the ostium (anterior portion of sulcus) comprises about 37 to 38 percent of the otolith length. In P. eos, the cauda is expanded and indistinct posteriorly, and the ostium comprises slightly more than 40 percent of otolith length. Also, in P eos, otolith height comprises more than 70 percent of otolith length, compared with 50 to 55 percent in the other three species. In all four species, the crista superior (ridge bordering the sulcus dorsally) is extremely strong and distinct, and the area of the otolith dorsal to it is shallowly to deeply concave. The ros- trum (anteroventral projecting portion of the otolith) is least distinct on sagittae of H. signifer. When placed outer face down on a flat surface, otoliths of Hemanthias are more bowed than those of Pronotogrammus; P. eos sagittae are the least convex when viewed in this aspect. Otoliths from juveniles usually have margins that are much more lobular, incised, or frilly than is indicated on the illustrated sagittae, which are all from adults. Sagittae of P. eos (five pairs measured) comprise about 6.5 to 7 percent of fish SL, whereas those of the other three species (five pairs of each measured) comprise between about 4 and 5.3 percent. ACKNOWLEDGMENTS As with any such project, my investigations of eastern North Pacific anthiins could not have reached fruition without my bor- rowing specimens, library material, work space, ideas, and the special talents of others that are lacking in me. Phillip C. Heemstra, now with the J.L.B. Smith Institute of Ichthyology, Rhodes University, South Africa, encouraged me and provided information that helped me start in the right direction. My sin- cere thanks are extended to him and to Lillian Dempster and W 1 Follett (CAS); Steve Crooke, Paul Gregory, William Max- well, and Jack W. Schott (California Department of Fish and Game); Lawrence L. Jones (CMM); John DeLeon, Teri Kato, Robert J. Lavenberg, Richard Meier, and Camm C. Swift (LACM); Joe Copp, Carl L. Hubbs (deceased), and Richard H. Rosenblatt (SIO); Boyd W. Walker (UCLA); William A. Buss- ing (UCR); and Bruce B. Collette (USNM). If I have forgotten anyone who helped me, it has not been intentional. LITERATURE C ITED Barton, O. 1947. Two new fishes, an Eques and a Holanthias, from Peru. American Mus. Novitates 1350:1-3. Berdegue, Julio. 1956. Peces de importancia comercial en la costa nor-occidental de Mexico. Secretaria de Marina, Dir. Gen. Pesca Industrias Conexas, 345 pp. Boulenger, G. A. 1 895. Catalogue of the fishes in the British Mu- seum (second edition). Catalogue of the perciform fishes. Vol- ume 1, London, xix + 394 pp. Chirichigno, N. 1974. Clave para identificar los peces marinos del Peru. Inst, del Mar del Peru, Informe 44:1-387. Contributions in Science, Number 339 Fitch: Eastern North Pacific Anthiins 7 Garman, S. I 899. Reports on an exploration off the west coasts of Mexico, Central and South America, and off the Galapagos Islands, in charge of Alexander Agassiz, by the U.S. Fish Commission steamer Albatross during 1891, Lieut-Com- mander Z.L. Tanner, U.S.N., commanding. XXVI. The fishes. Mem., Mus., Comp. Zook, Harvard, 431 pp. Gilbert. C.H 1890. A preliminary report on the fishes collected by the steamer Albatross on the Pacific coast of North Amer- ica during the year 1 889, with descriptions of twelve new gen- era and ninety-two new species. U.S. Natl. Mus., Proc. 1 3:49- 126. Gill, T 1 863. Catalogue of the fishes of Lower California, in the Smithsonian Institution, collected by Mr. J. Xantus. Acad. Nat. Sci. Philadelphia, Proc. 15:80-88. Gunther, A.C. 1868. Catalogue of the fishes in the British Mu- seum. Volume 7, Physostomi. London, xx + 512 pp. Hildebrand, S.F. 1946. A descriptive catalog of the shore fishes of Peru. U.S. Natl. Mus., Bull. 189:1-530. Hildebrand, S.F., and O. Barton. 1949. A collection of fishes from Talara, Peru. Smithsonian Miscell. Coll. 111(1 1 0): 1 —36. Hiyama, Yosio. 1937. Marine fishes of the Pacific Coast of Mex- ico. Nissan Fisheries Inst, and Co., Ltd., Odawara, Japan, 75 pp. Hobson, E.S. 1975. First California record of the serranid fish Anlhias gordensis Wade. Calif. Fish and Game 6 1 (2): 111 112. Hubbs, C.L., W.l. Follett, and L.J. Dempster. 1979. List of the fishes of California. Calif. Acad. Sci., Occas. Paper 133:1-51. Jordan, D.S., and C.H. Eigenmann. 1 890. A review of the genera and species of Serranidae found in the waters of America and Europe. U.S. Fish Comm., Bull (for 1888) 8:324-44 1 . Jordan. D.S., and B. W. Evermann. 1 896. The fishes of North and Middle America: a descriptive catalogue of the species of fish- like vertebrates found in the waters of North America, north of the Isthmus of Panama. Part I. U.S. Natl. Mus., Bull. 47:1- 1240. Jordan, D.S., and C.H. Gilbert. 1882. Synopsis of the fishes of North America. U.S. Natl. Mus., Bull. 16:1-1018. Kendall, A. 1979. Morphological comparisons of North Ameri- can sea bass larvae (Pisces: Serranidae). NOAA Tech. Rept. NMFS Circular 428:1-50. Miller, D.J., and R.N. Lea. 1976. Guide to the coastal marine fishes of California (revised edition). Calif. Dept. Fish Game, Fish Bull. 157:1-249. Ramirez, E., and A. Gonzales. 1976. Catalogo de peces marinos Mexicanos. Inst. Nacional Pesca, Sect. Industria y Comercio, 462 pp. Steindachner, Franz. 1874. Ichthyologische Beitrage. Sit- zungsberichte Kaiserlichen Akademie Wissenschaften, Wien. 70:375-390. Thomson, D.S., L.T. Findley, and A.N. Kerstitch. 1979. Reef fishes of the Sea of Cortez. John Wiley & Sons, New York, 302 PP Wade, C.B. 1946. New fishes in the collections of the Allan Hancock Foundation. Allan Hancock Pacific Exped. 9:21 5— 237. Wales, J.H. 1932. An addition to the fish fauna of the United States. Copeia 1 932(2): 1 06. Walford, L.A. 1974. Marine game fishes of the Pacific Coast from Alaska to the Equator (revised edition). Smithsonian Institution, Washington, DC., 19 + 205 pp. Submitted 1 July 1981. Accepted for publication 16 November 1981. 8 Contributions in Science, Number 339 Fitch: Eastern North Pacific Anthiins at irregular intervals in three major series; the articles in each series are numbered individually, and ® Contributions in Science, a miscellaneous series of technical papers describing original research in the life and earth sciences. ® Science Bulletin, a miscellaneous series of monographs describing original research in the life and earth sciences. This series was discontinued in 1978 with the issue of Numbers 29 and 30; monographs are now published by the Museum in the Contributions in Science series. Science Series, long articles on natural history topics, generally written for the layman. Copies of the publications in these series are sold through the Museum Bookshop. EDITORIAL BOARD Robert Gustafson John M. Harris Charles L. Hogue Robert J. Lavenberg EDITOR Camm C. Swift David P. Whistler Edward C. Wilson John W. Wright Robin A. Simpson Printed at The Castle Press, Pasadena, California. . o jfl M 71902 -^/eRARlES^ DESCRIPTION OF A NEW SPECIES OF NET-WINGED MIDGE (DIPTERA: BLEPHARICERIDAE) FROM 1 HE SIERRA NEVADA DE SANTA MARTA, COLOMBIA1 Charles L. Hogue2 ABSTRACT. The status of the genus Limonicola is reviewed, and a preliminary generic diagnosis provided. The genus presently includes the type species, L. plurivectis Lutz, 1928, from Venezuela. L leucoptera Edwards, 1929, from Peru, and a new species named herein L. davila, from the Sierra Nevada de Santa Marta, Colombia. RESliMEN. Se considera el estatus del genero Limonicola y se da una diagnosis generica preiiminar. Actualmente este genero incluye la es- pccie tipo L. plurivectis Lutz, 1 928, procedente de Venezuela, L. leucop- tera Edwards, 1 929, de Peru, y una nueva especie de la Sierra Nevada de Santa Marta, Colombia nobrada aqui L davila. INTRODUCTION The dipteran fauna of torrential waters in South America com- prises several taxa whose immatures are highly adapted to life in swiftly flowing streams. Although widespread and common, two of these taxa, the Blephariceridae (“net-winged midges”) and the genus Maruina Muller, 1895, of the Psychodidae (“lance- winged midges”), are poorly studied. This is unfortunate because their close association with head-water streams, poor dispersal abilities, and great geological age make them particularly useful paleogeographic indicators. Their present distributional patterns can provide data on the positions of tectonic plates and orogenic zones in the past. For this reason, it has long seemed evident to me that these midges might offer significant clues to former relationships between elevated land masses along the Caribbean borderlands of South America, and consequently I have made efforts to collect and analyze material from these areas. One such tectonic unit is the Sierra Nevada de Santa Marta (hereinafter called “Sierra”) in northern Colombia that has been the subject of geologic studies (Tschanz et al., 1 974) and faunistic investiga- tions on butterflies (Adams, 1973; Adams and Bernard, 1977), reptiles and amphibians (Duellman, 1979:382-4), and plants (Sugden and Robins, 1979). The present paper is a partial report on my collections of Blephariceridae and Maruina made in March of 1980. In luture papers, I plan to review my other collections and describe addi- tional species of Blephariceridae ( Paltostoma Schiner, 1 866) and Psychodidae {Maruina). Collection sites pertinent to the present paper are listed by field numbers below with brief habitat descriptions. All distances Contributions in Science, Number 340, pp. 1-10 Natural History Museum of Los Angeles County, 1982 are kilometers by road. All localities are situated in the Depart- ment of Magdalena except CLH 257, which is in Guajira Department. CLH 257. Rio Ancho and Highway No. 2, 1 1 March. A wide, shallow river with a bottom entirely of medium-size, rounded boulders. Collections made in morning on both sides of highway bridge crossing the river about 1 -2 km from its exit to the sea, on coastal plain at an elevation of only 90 m. Water temperature 25°C CLH 259. Quebrada Donama, 5 km northeast of Mamatoco, 1 2 March. A rushing stream (lowing over a hard rock substratum with scattered mixed-sized boulders. Collections made behind a house between road (to Minca) and stream at an elevation of approximately 100 m. Water temperature 25°C. CLH 260. Quebrada Minca, 1 2 March. A wide, shallow, cold water (23. 5°C) stream flowing through village. Collections made immediately west of village, on medium to small, rounded, loose boulders. Elevation 800 m. CLH 263. Rio Gallina (tributary of Rio Sevilla) near Finca California, 11.2 km northeast of Palmar, 13 March. Large stream flowing with considerable velocity through a steep canyon of large, fixed boulders and intermittent rapids. Collections made near bridge just above ranch house at approximately 1 ,500-m elevation. Water temperature 2 1 °C. CLH 264. Quebrada Charua (tributary of Rio Gallina), near Finca California, 1 1 .2 km northeast of Palmar, 1 3 March. Small, steep, stepped stream with intermittent dams of large, fixed boul- ders. Collections made near confluence with Rio Gallina. CFH 267. Rio Sevilla, 8.7 km southeast of Palmar, 1 4 March. Farge rushing stream flowing through steep gorge, inclinous, intermittently forming slips over bare rock exposures, falls through gigantic boulder dams, and large pools. Water tempera- ture 24°C. Elevation approximately 800 m. The water at all localities was free of visible pollution and generally in good ecological condition. All streams appeared to 1. Review committee for this Contribution: Julian P. Donahue, F. Christian Thompson, and D. M. Wood. 2. Curator of Entomology, Natural History Museum of Los Angeles County, 900 Exposition Boulevard, Los Angeles, California 90007. ISSN 0459-81 13 be near their lowest normal flows in response to the dry season, which reaches its height on the Sierra in March. Rock types forming their beds are varied, but light-colored granites, quartz diorites, and hard metamorphics predominate. TERMINOLOGY Special morphological and descriptive terms applied herein are explained in the text and figures. For the genitalia, I follow the interpretations of homologies as given in my most recent paper (Hogue, 1981). Reference to my papers of 1973 (general) and 1978 (larval chaetotaxy) will provide the reader with further explanations of abbreviations and the particular nomenclature applied to the family. 1 refer to the variously shaped, short sensilla of the larval dorsum as “tuberculiform” (sensilla tuberculiforma) to dis- tinguish them collectively from the hairlike sensilla (sensilla chaetica). The tuberculiform types may be further characterized by specific shapes such as conoid (acorn-shaped), clavate, or ovoid; these may be straight or sharply curved at the base (falcate). SYSTEMATIC'S Blepharicerinae and Paltostomatini sensu Zwick, 1977, and pers. comm. Genus Limonicola Lutz HISTORY Lutz proposed the genus Limonicola in 1928 to include an un- usual blepharicerid that he collected in the Limon River near Maracay, Venezuela In his publication (Lutz, 1928), he noted that the unique characteristic separating this new entity from other genera was the male's very long hind legs, with swollen basal tarsal segments. He described the type species, plurivectis, from complete material of all stages, but his specimens have become lost. A year later, Edwards recognized a new Andean species as belonging to Limonicola and described it in his monu- mental treatment on the Diptera of Patagonia and South Chile (Edwards, 1929:73-74). The species was named leucoptera and was based on a single male, the type of which I have not examined. Because of the absence of good material from l utz’s and Ed- wards’s work for comparative and diagnostic reference, no fur- ther species have been added to date (Hogue, 1971), and Limonicola has become an enigmatic taxon with undefined boundaries and cryptic affinities. I have material now on hand of these and several additional new species, excluding that from the Sierra described below. From this and good series of all stages from the Sierra, it is now realized that the genus is truly distinct in many characters and contains numerous species widely dis- tributed over the mountainous terrain of northwestern South America. A full generic definition will not be possible without further study. However, in order to provide a basis for particularization of the new species, 1 am providing here a preliminary definition of the genus based on its salient features. DESCRIPTION Adult Small, short-bodied, fragile Blephariceridae. Coloration. Body integument well sclerotized, dull, velvety- black with opalescent reflections, especially on abdomen; some sclerites and membranes dark brown. Wing membrane infuscate in both sexes. Size. Small, wing length ranges, male 3.5 to 9.5 mm; female 2.8 to 9.0 mm. Male slightly larger than female. Structure. Head: Colocephalous [amandibulate, mispropor- tioned head, as defined by Hogue, 1970:7] in both sexes. Ocellar lobes slight, ocelli sessile. Eyes disjunct, entire, and without callis oculi [unfacetted strip]; ommatidia subequal, large in male, somewhat smaller in female. Proboscis very short, mouthpart elements greatly reduced and modified: labrum narrowly lan- ceolate; mandibles totally absent in both sexes; hypopharynx vestigial, slightly larger in female but always less than half the length of labrum; maxillary palpus composed of a single diminu- tive segment, lacinia similar to palpus but weaker and with obliq- uely rounded apex; labellum composed of deeply furcate, elongate lobes. Antenna short, 1 5-segmented; flagellomeres subglobose. Thorax: Sclerites generally typical for Blepharicerinae; scutum somewhat larger than in most genera and projecting far- ther forward, giving entire thorax an oblique set in lateral view; katepisternum truncate ventraliy; pleural suture sinuous, forked (posterior branch separating dorsal portion of meron, itself bro- ken into two portions, a small triangular lower plate and a larger, upper lobate swelling; the convex anterior part of the latter ex- tends over the pleural suture, forming a pitlike depression beneath). Wing: Broad as in other Paltostomatini but with a hyperex- panded anal lobe. Venation typical for tribe, with reduced longi- tudinal veins: subcosta evanescent; fork of Rs near wing apex; M3 absent; M, + 4 (false crossvein) lacking; 1A sometimes not reaching wing margin. Posterior veins decidedly more weakly sclerotized than radius and costa. Legs: Strongly sexually dimorphic: in males very long and slender, especially the hind leg, which is twice the length of oth- ers and has inflated tibial apex and tarsal segments 1 and 2; fore femur sinuously curved; all tarsal segments 5 with ventrobasal lobes bearing large setae (calcipalae); tarsal claws elongate and slender, with subbasal tooth. Legs in female short and normally proportioned except fore femur, which is strongly upcurved basally and abruptly swollen apically; hind femur moderately upcurved, hind tibia slightly curved sinuously; tarsal segments 5 and claws simple. Tibial spurs usually lacking from all legs in both sexes, hind tibia of female sometimes with pair of equal spurs. Abdomen: Short, strongly tapered posteriorly. Sclerites mod- erately sclerotized, especially sternites; pleural membranes wide. Tergite 1 very narrow, well sclerotized laterally only. 2 Contributions in Science, Number 340 Hogue: New Species of Limonicola Male genitalia: Segment VIII ligulate. Tergite IX broad, pos- terior margin unmodified. Lobes of Xth tergite well developed, ventrolateral portions lobate, converging. Tegmen complex, dor- somesal portion convex, raised into a longitudinal slitlike fold. Vesica large, spherical, with transverse, dorsal, lobate apodcme. Aedeagal rods three, short, equal, attenuate filaments. Lateral tine undeveloped. Genital capsule (fused hypandrium and gonocoxite) broadly membranous mesally. Outer lobe of gonostylus variously shaped, generally with dorsal and ventral lobes; inner lobe of gonostylus transverse, with ventral extension, which is fused with the posterior margin of the genital capsule. Female genitalia: Tergite VIII narrow, distinct from tergite IX. Sternite VIII well sclerotized medially, developed into a deeply emarginate (bilobed) plate. Oviscapt (hypogynial plate) complex, posterior lobes directed mesad, apices spiculate (and with a few minute setae), midportion constricted. Genital fork well defined, a transverse stirrup-shaped plate. Lateral lobes of sternite IX (normally present in the Blepharicerinae) absent. Proctiger sclerotized basally to form a transverse plate similar to genital fork (but with shorter arms). Lobe of tergite X (cercus, auct.) deeply cleft, with a distinct ventral secondary lobe. Three ovoid spermathecae. Pupa Depressed. Outlined shape oval, lateral margins of abdominal segments moderately convex, that of segment 1 1 strongly convex (more so in female), thoracic region abruptly narrowed. Coloration. Upper surface evenly black (in life); lower surface white. Size. Small to medium-sized, body lengths 2.8 to 7.0 mm. Male pupa slightly larger than female and more elongate. Structure. General dorsal cuticle and under folded lateral mar- gins of abdomen coarsely and densely verrucose. Scutellar scle- rite broad, reaching nearly to lateral extreme of metascutai sclerite; abdominal segments I and II curving strongly anterad laterally. Branchiae with heavy, expanded bases; that of anterior- most lamella bullate laterally; branchial posture porrect, all plates elongate, subtriangular, rigid, with acute apices, and of about equal length but inner pair much narrower than outer. Antennal and mouthpart cases ill-defined. Ventrolateral adhesive organs present on abdominal segments 1 1— V (four pairs). Fourth-Instar Larva Size. Small to medium-sized, body lengths 3.7 to 9.0 mm. Structure. Intercalary convexities (Fig. 1 I ) undeveloped. Head capsule weekly sclerotized (hind margin evanescent), with deep lateral insertions and lateralia. Antenna short, 2-segmented, the proximal segment about one-third length of distal. Dorsal pseudopod absent from segments of anterior and median divi- sions; present and similar to pseudopods, but smaller, on terminal segment. Dorsal sclerotized plates, tubercles, or other processes entirely absent. Ventral gill filaments digitiform, arranged in two slightly disjunct, stellate clusters, the anterior usually with six filaments, the posterior with four filaments. Pseudopods elon- gate, directed laterad, only slightly downcurved; subequal in length (about one-quarter width of segment) except that of seg- ment I, which is one-half the size of others. Ventral integument of each abdominal segment basal to pseudopod, dorsal pseudopod of terminal segment, and sclerotization of terminal lobe with numerous denticles grading to papillae mesad. Cliaetotaxy. Primary trunk scnsilla as follows: ic slightly ventral on anterolateral corner of segment, minute and difficult to see because of sclerotization in this area; it in normal position, min- ute; is double, associated with anteromesa! invagination of each segment; iM-T forming a quadrate constellation in center of dorsum, chaetiform; tl-Vlll normal in position, tuberculiform; stPl-VIII normally positioned, chaetiform; stM-T just lateral to t, tuberculiform; pdpod not evident; dpod on dorsal pseudopod of segment V 1 1 only, both chaetiform; p not evident; sp obscured by proliferation; ssP-T in line, large chaetiform; pdl-VI normal. Secondary sensilla of small tuberculiform and chaetiform types generally and evenly distributed over dorsum, latter most numer- ous laterally, former absent from pseudopods but forming a no- ticeably denser field middorsally (circular concentration); ante- rior and posterior lobular portions of each segment often free of sensilla; no conspicuously large fiat or specially modified sensilla in linear series or otherwise. Terminal setae very small and few in number. Limonicola davila new species (Figs. 1 12) DIAGNOSIS Limonicola davila is most similar to plurivectis. From that spe- cies and the only other described species in the genus, leucoptera, it may be identified by the following primary structural dif- ferences in all known stages. Adult Male The transverse, strongly asymmetrical, V-shaped dististyle of davila is unique; in the two other species, this structure is subhex- agona! and roughly symmetrical. Adult Female The tarsal claws of both davila and plurivectis are similar, short and broad, whereas those of leucoptera are long and slender. Paralleling this condition is the length of tarsal segment 5 and the presence or absence of calcipalae: long (over twice length of segment 4) and without calcipalae in davila and plurivectis, short (about equal to length of segment 4) and with calcipalae in leuco- ptera. The shape of the oviscapt lobes of the genitalia also dis- tinguish davila from plurivectis: with an evenly narrowed apex in the former, broad and basally constricted in the latter. Pupa The pupae of davila and plurivectis are similar but that of the former may be distinguished from the latter most easily by the more extensive verrucae on the scutum and branchial sclerite: in davila, the verrucose area of the scutum extends medially to the metascutai sclerite and covers the entire branchial sclerite pos- Contributions in Science, Number 340 Hogue: New Species of Limonicola Figures 1 and 2. Limonicola davila. Heads, frontal view. Figure 1, male; Figure 2, female. terior to the branchioalar suture; in plurivectis, the scutal ver- rucae are restricted to the central area and cover only the anterior half of the branchial sclerite. Also, the bullate posterior basal extension of the anterior branchial lamella is much larger in davila than in plurivectis. L. davila differs from leucoptera by its smaller size (3.0 to 3.4 mm body length versus 3.4 to 4.0 mm), its relatively larger and more widely spaced integumentary ver- rucae, and the virtual absence of a bullate swelling on the pos- terior extension of the anterior branchial sclerite base. Larva Gross differences separate the larvae of the three species: davila and leucoptera are alike in the sparcity of conspicuous, large, dorsal, integumentary, tuberculiform sensilla, which are present in plurivectis; the larva of leucoptera is easily distinguished from that of davila by its significantly larger dorsal pseudopod of the terminal division; the organ also is well separated and extrorse in davila. nearly reaching the level of the posterior margin, but is almost adnate to the terminal margin in leucoptera. DESCRIPTION Adult (Figs. 1-9) Size. Small, measurements as follows: wing length male (range, N = 10) 3.5 to 3.9 mm, mean 3.7 mm; female 2.8 to 3.8 mm, Figures 3 and 4. Limonicola davila. Hind tarsal segments 4 and 5 and tarsal claws. Figure 3, male; Figure 4, female. 4 Contributions in Science, Number 340 Hogue: New Species of Limonicola I- 2 mm Figure 5. Limonicola davilci. Wing. Figures 6 through 8. Limonicola davila. Male genitalia. Figure 6, entire structure, dorsal view; F’igure 7, inner and outer gonostyles, ventral view; Figure 8, phallic complex, dorsal view. Contributions in Science, Number 340 F!ogue; New Species of Limonicola 5 Figure 9. Limonicola davila. Female genitalia, ventral view. mean 3.2 mm. Leg segment lengths given in Table I . Structure. Head (Figs. 1-2): Ocular bristles short (about 1.5 width of ommatidium) and single over nearly all of eye. Mouth- parts: Labellar lobe short, barely exceeding apex of labrum in male, slightly short of this point in female; lobe with only a few apical setae in both sexes. Legs: Female tarsal segment 5 over twice as long as segment 4 and without setae except for single, small apicodorsal bristle (Fig. 4). Tarsal claws and calcipalae normal for genus (Figs. 3- 4). Tibial spurs entirely lacking. Wing (Fig. 5): Venation typical. Abdomen: Setae of sternite VII lacking mesally in both sexes. Male genitalia (Figs. 6-8): Setae of IXth tergite in two dis- junct groups, a posterior group of several medium-sized hairs and an anterolateral group of fewer and generally heavier hairs. Lobe of Xth tergite about as long as broad, apex slightly incurved. inner margin angled; internal, ventrolateral lobe well developed and with rounded apex. Outer lobe of gonostylus (Fig. 7) trans- verse, strongly asymmetrical, open V-shaped in mesal view. Female genitalia (Fig. 9): Lobe of plate formed from sternite VIII narrowly rounded at apex; anterior border of plate emargi- nate; 29-30 setae on each side of plate. Base of oviscapt (hypo- gynial plate) only slightly broader than lobular portion; lobe with evenly narrowed, spiculate apex. Necks of spermathecae short, 1.5-2 times width. Transverse inner plate of proctiger (Fig. 9, “transverse plate”) asymmetrically bifurcate laterally, the pos- terior fork much longer than the anterior. Pupa (Fig. 10) General shape and morphology typical for genus. Size. Small, male body length range 2.9 to 4. 1 , width 1 .7 to 2.5 mm (N = 20; means 3.5 and 2.0 mm); female body length range 6 Contributions in Science, Number 340 Hogue: New Species of Limonicola Table 1. Leg segment lengths of male and female Limonicola ilavilu. LInit of measurement is millimeters; values are ranges (N = 14) followed by means. Male Female Fore leg Femur 1.2-1. 5, 1.3 0.66-0.96,0.79 Tibia 1.4-1. 7, 1.6 0.74-1.06,0.86 Tarsus 1 1.04-1.30, 1.21 0.48-0.65, 0.5 1 2 0.48-0.59, 0.55 0.20-0.30,0.25 3 0.29-0.34, 0.31 0.17-0.23,0.20 4 0.20-0.23, 0.21 0.15-0.20,0.17 5 0.21-0.26, 0.24 0.30-0.39,0.34 Mid leg Femur 1.3-1. 5, 1.4 0.70-0.90,0.82 Tibia 1.3-1 6, 1.4 0.68-0.92,0.79 Tarsus 1 1.06-1.16, i ll 0.41-0.58,0.48 2 0.44-0.57, 0.51 0.20-0.27,0.23 3 0.26-0.32, 0.30 0.17-0.23,0.19 4 0.19-0.24, 0.21 0.14-0.20,0.16 5 0.21-0.25, 0.23 0.31-0.41,0.34 Hind leg Femur 3.1 3.9, 3.6 1. 4-1.9, 1.6 Tibia 3. 7-4.5, 4.2 1. 4-2.0, 1.7 Tarsus 1 1.44 1.68, 1.56 0.44-0.60,0.50 2 0.68-0.76, 0.74 0.21-0.30,0.25 3 0.51-0.58, 0.55 0.17-0.25,0.20 4 0.32-0.37,0.34 0.15-0.20,0.17 5 0.23-0.27,0.25 0.30-0.40, 0.34 2.8 to 3.8 mm, width 1 .6 to 2.2 mm (N = 20; means 3.2 and 1 .9 mm). Proportions length/width, male 1.75, female 1.68. Structure. Verrucae of dorsal integument relatively large and widely spaced (6 in line per 0.1 mm); extensive on thorax; those of median portion of scutal field extending posteriorly to meta- scutal sclerite, and those of lower portion of branchial sclcrite covering entire sclerite posterior to the branchioalar suture. Bui 1- ate posterobasal portion of anterior branchial lamella large. Fourth-Instar Larva (Figs. 11-12) General shape and morphology typical for genus. Coloration. Generally yellowish gray-brown or medium-brown, intermediate forms with varying amounts of brown mottling on light background: pigmented areas usually confined to escharae (sclerotizations externally marking points of muscle insertions) but may expand transversely over dorsum, often selectively col- oring segments III and V, giving larva a double-banded appearance. Size. Small, body length: range of random specimens (N = 54) 3.7 to 5.5 mm, mean 4.6 mm; of mature specimens only (i.e. pupal branchiae visible) (N = 24) 3.7 to 5.5 mm, mean 4.6 mm; head capsule width, range (N = 54) 0.90 to 1 .30 mm, mean 1.11 mm. Structure and C'haetotaxy (Figs. 11-12). Secondary tuberculiform sensilla of dorsum sparse and mostly very small, conoid to ovoid in shape, including those of median circular group. Dorsal Figure 10. Limonicola davila. Pupa, female (dorsal left, ventral right), satellite details of adjacent male features. pseudopod of segment VII relatively large and well separated and directed away from the terminal lobe. MATERIAL Type Material HOL.OTYPE male (completely dissected and mounted on slides Nos. CLH 80-112 a-d): COLOMBIA, Department of Mag- dalena, Sierra Nevada de Santa Marta, Rio Gallina at Finca California, 1,500 m, 13 March 1980, C. L. Hogue, CLH 263. ALLOTYPE female (completely dissected and mounted on slides Nos. CLH 80-85 a-d): same data as holotype. 21 male, 23 female PARATYPES (pinned, in alcohol, and mounted on slides): same data as holotype. Holotype and allotype deposited in the collection of the In- Contributions in Science, Number 340 Hogue: New Species of Limonicola 8 Contributions in Science, Number 340 Hogue: New Species of Limonicola stituto Nacional de los Recursos Naturales Rcnovables y del Ambiente, Bogota (INDERENA). One paratype male and female sent to National Museum of Natural History, Wash- ington, D.C. (USNM) and British Museum (Natural History) (BMNH), London. All others in Natural History Museum of Los Angeles County (LACM). Additional Specimens COLOMBIA, Department of Guajira, Sierra Nevada de Santa Marta, Rio Ancho and Highway No. 2, 1 1 March 1980, C. L. Hogue, CLH 257 (325 larvae, 75 pupae: INDERENA, USNM, BMNH, LACM). Department of Magdalena. Quebrada Do- nama, 5 km NE Mamatoco, 1 2 March 1 980, C. L. Hogue, CLH 259 (3 larvae, 4 pupae, 15 males, 1 female: LACM); Quebrada Minca, Minca, 12 March 1980, C. L. Hogue, CLH 260 (142 larvae, 47 pupae, 21 males: LACM); Rio Gallina, Finca Califor- nia, 13 March 1980, C. L. Hogue, CLH 263 (2 larvae: LACM). Quebrada Charua, tributary to Rio Gallina, Finca California, 1 3 March 1980, C. L. Hogue, CLH 264 (8 larvae, 4 pupae, 1 male: LACM); Rio Sevilla, 8.7 km SW Palmar, 14 March 1980, C. L. Hogue, CLH 267 ( 1 4 larvae, 2 pupae: LACM ). ETYMOLOGY The species is named for Sr. Francisco E. Davila Riascos, long time resident and prominent rancher of the Sierra whose gener- ous assistance during my collecting made the discovery of this species possible. DISTRIBUTION Limonicola davila is known only from the Sierra Nevada de Santa Marta and appears to be endemic within that mountain system. ECOLOGY In all localities where I found L. davila, the larvae and pupae were attached to relatively small (10- to 45-cm diameter), smooth stones that were detachable from the substratum; none appeared on vertical faces of large, implanted boulders or por- tions of exposed basement rock as is the habitat of Paltostoma and other blepharicerid genera elsewhere. Larvae exhibit a dis- tinct preference for lodging in small crevasses or notches on the stone’s surface. Most collections were of single specimens or groups of only a few individuals. Water temperatures of streams at collecting sites varied from 22° to 25°C. Although I was not able to verify it by direct observation, the (light posture of the males and their morphology leads me to the hypothesis that they capture females in Hight for mating. Lim- onicola males are unique in possessing swollen hind tarsi on ex- tremely long hind legs, twice the length of the others; all the legs of the female are of normal proportions and without special mod- Figure 1 2. Limonicola davila. Larva, fourth instar, dorsal view of left half of abdominal segment 1 1 Figure 1 1. Limonicola davila. Larva, fourth instar (dorsal left, ventral right), abbreviated sensillar designations explained in Hogue ( 1 978:3-4). Contributions in Science, Number 340 Hogue: New Species of Limonicola ilications. These features apparently contribute to the male’s tendency to fly with the hind legs extended far below the plane of the body in which position they may be used to snare females in (light in a manner similar to the prey capture techniques of cer- tain Mccoptera. This would also explain the sinuously curved fore femora of the females, a configuration that permits them to hold their fore legs upward directly over the body where the male’s tarsi could most easily grapple them. That the tarsi of the male are capable of grasping in such a fashion was well demon- strated during field collecting by the way in which adult midges of both sexes become tenaciously entangled by the tarsi of males in a killing tube. The elongate tarsal claws fold back against the calcipalae at the base of the last tarsal segment, creating a very effective clasping device. The female lacks this arrangement, logically, being the passive sex in the precopulatory capture scenario. Because of their very reduced mouthparts, neither sex could possibly prey on other insects as is normal, at least with the female, in most other Blephariceridae. This, plus their overall small size, indicates an ephemeral, possibly nonfeeding, adult life, paralleling that of other diminutive, short-lived torren- ticolous midges, such as the Deuterophlebiidae, Nympho- myiidae, and Thaumaliidae. ACKNOWLEDGMENTS 1 wish to express my very great thanks to the following indi- viduals and institutions for their support and assistance in making possible my held trip and the completion of this report: The Instituto Nacional de los Recursos Naturales Renovables y del Ambiente (INDERENA) for extending official permission to conduct scientific research in Colombia (by instrument of communication 02475 dated 7 March 1980); Drs. Sergio Duran (president) and Jorge Hernandez C. (chief of the Division de la Fauna Silvestre) of this agency for their numerous courtesies, including the authorization of said permission and logistical ad- vice; Dr. Gustavo Maldonado of the Santa Marta office of INDE- RENA for assistance with transportation; Sr. and Sra. Francisco Davila of Santa Marta and Dr Gerardo Reichel-Dolmatoff of Bogota for their personal interest in my work, reflected in many generous favors during the course of my stay in Colombia; and Sr. Wilfredo Silva for his field assistance and services as chauffeur. In addition I would like to acknowledge the financial aid ex- tended by the Natural History Museum of Los Angeles County Foundation as well as thank the review committee and my col- leagues Peter Zwick, Julian P. Donahue, and Roy R. Snelling for criticism of the manuscript. LITERATURE CITED Adams, M. J. 1973. Ecological zonation and the butterflies of the Sierra Nevada de Santa Marta, Colombia. Journal of Natu- ral History 7:699-718. Adams, M. J., and G. E Bernard. 1977. Pronophiline butterflies (Satyridae) of the Sierra Nevada de Santa Marta, Colom- bia. Systematic Entomology 2:263-281. Duellman, W. E. 1979. The herpetofauna of the Andes: Patterns of distribution, origin, differentiation, and present commu- nities. Chap. 15 in W. E. Duellman, ed., The South Ameri- can herpetofauna: Its origins, evolution, and dispersal. Museum of Natural History, University of Kansas, Mono- graphs 7:1-484. Edwards, R. W. 1929. Blepharoceridae. Vol. 2, pp. 1-75, figs. 1- 3, pis. V-VIII in British Museum (Natural History), Dip- tera of Patagonia and South Chile. British Museum (Natu- ral History), London. Hogue, C. L. 1970. Description of a new species of net-winged midge from the Great Basin, with a key to the North Ameri- can species of the genus Dioptopsis (Diptera: Blephariceridae). Natural History Museum of Los Angeles County, Contributions in Science. 178:1-10. . .1971. Family Blephariceridae. Number 8, pages 8.1- 8.12 in N. Papavero, ed., A catalogue of the Diptera of the Americas south of the United States. Museu de Zoologia, Uni- versidade de Sao Paulo, Sao Paulo. .1973. The net-winged midges or Blephariceridae of California. California Insect Survey, Bulletin 15:1 -83. .1978. The net-winged midges of eastern North Amer- ica; with notes on new taxonomic characters in the family Blephariceridae (Diptera). Natural History Museum of Los Angeles County, Contributions in Science 29 1 : 1 -4 1 . .1981. Blephariceridae, Pages 191-197 in J. F. McAlpine, B. V. Peterson, G. E. Shewell, H. L. Teskey, J. R. Vockeroth, and D. M. Woods, eds., Manual of Nearctic Dip- tera, vol. 1. Agriculture Canada Monographs 27:1-674. Lutz, A. 1928. Blefaroceridae. Estudios de Zoologia y Para- sitologia Venezolanas, pp. 65-69, 2 pi. and table. Caracas. Sugden, A. M., and R. J. Robins. 1 979. Aspects of the ecology of vascular epiphytes in Colombian cloud forests, 1 . The distribu- tion of the epiphytic flora. Biotropica 1 1 : 1 73-1 88. Tschanz, C. H., R. F. Marvin, J. Cruz, H. H. Mehnert, and G.T. Gebula. 1974. Geologic evolution of the Sierra Nevada de Santa Marta, northeastern Colombia. Geological Society of America, Bulletin 85:273-284. Zwick, P. 1977. Australian Blephariceridae (Diptera). Aus- tralian Journal of Zoology, Supplementary Series 46: 1-1 21 . Submitted 7 April 1981; accepted for publication 16 February 1982. 10 Contributions in Science, Number 340 Hogue: New Species of Limonicola Number 341 2 8 May 1982 FOSSIL BIRDS FROM TERTIARY MARINE BEDS AT OCEANSIDE, SAN DIEGO COUNTY, CALIFORNIA, WITH DESCRIPTIONS OF TWO NEW SPECIES OF THE Hildegarde Howard Natural ''History Museum of Uk Aiigdtt Court}' » 900 Exposition Boulevard • Los Angeles, California 90007 The scientific publications of the Natural History Museum of Los Angeles County have been issued at irregular intervals in three major series; the articles in each series are numbered individually, and numbers run consecutively, regardless of subject matter. ® Contributions in Science, a miscellaneous series of technical papers describing original research in the life and earth sciences. ® Science Bulletin, a miscellaneous series of monographs describing original research in the life and earth sciences. This series was discontinued in 1978 with the issue of Numbers 29 and 30: monographs are now published by the Museum in the Contributions in Science series. ® Science Series, long articles on natural history topics, generally written for the layman. Copies of the publications in these series are sold through the Museum Bookshop. Leon G. Arnold Lawrence G. Barnes Robert L. Bezy Kenneth E. Campbell EDITORIAL BOARD Robert Gustafson John M. Harris Charles L. Hogue Robert J. Lavenberg EDITOR Robin A. Simpson Camm C. Swift David P. Whistler Edward C. Wilson John W. Wright FOSSIL BIRDS FROM TERTIARY MARINE BEDS AT OCEANSIDE, SAN DIEGO COUNTY, CALIFORNIA, WITH DESCRIPTIONS OF TWO NEW SPECIES OF THE GENERA URIA AND CEPPHUS (AVES: ALCIDAE)' Hildegarde Howard1 2 ABSTRACT. The San Luis Rey River Local Fauna, in the lowermost rocks in Lawrence Canyon that are referred to the San Mateo Formation, yielded bones of six species of fossil birds: a murre, Uria paleohesperis n. sp.; a guillemot, Cepphus olsoni n. sp.; a flightless auk, Praemancalla cf. P. wetmorei; a loon, Gavia sp.; an albatross, Diomedea sp.; and an auklet, ? Aethia sp. The holotype of U. paleohesperis has shoulder and wing bones of stockier proportions than in Recent species of Uria and has distinctive characters of the scapula, coracoid, and ulna. The holotype humerus of C. olsoni is shorter than in Recent C. columba and has distinctive characters of the deltoid crest and ectepicondylar process. The partial skeleton of Praemancalla cf. P. wetmorei includes the first leg bones known for the genus. In the Lawrence Canyon Local Fauna, higher in the formation, abundant fragmentary bones of Mancalla occur. Mancalla milleri, M . diegensis, and M. cf. M . cedrosensis are recog- nized. The presence of Praemancalla in the lower beds and Mancalla in the upper indicates a major time interval within the San Mateo Forma- tion at this Oceanside site, with a range from latest Miocene to earliest Pliocene. INTRODUCTION During the past several years the University of California Mu- seum of Paleontology at Berkeley (UCMP) and the Natural History Museum of Los Angeles County (LACM) have collected vertebrate fossils in Lawrence Canyon at Oceanside, San Diego County, California. Over one hundred bird bones in these collec- tions can be assigned to genus, and 13 taxa are recognized. Per- sonnel of the San Diego Society of Natural History (SDSNH) also continue to collect in the area, and further reports are anticipated. The Lawrence Canyon localities are at the extreme southern extent of Woodford’s (1925:217) mapped outcrop of the San Mateo Formation, a rock unit that he grouped among “Post Cap- istrano formations.” Later, Vedder (1972:167) suggested that Woodford’s type section of the San Mateo Formation “may be a channel deposit within the lower part of the Capistrano Formation.” Tentatively retaining the San Mateo Formation designation, Barnes et al. (1981) recognized two separate local faunas based on vertebrate fossils (see Fig. 1 ). The faunal assemblage in the Contributions in Science, Number 341, pp. 1-15 Natural History Museum of Los Angeles County, 1982 ISSN 0459-8113 lowermost beds (fine gray to white sands that immediately over- lie the San Onofre Breccia) was named the San Luis Rey River Local Fauna. The fossils referred to this fauna were collected at locality LACM 4297 ( = UCMP V68147) and nearby localities LACM 4298 ( = UCMP V68144) and LACM 4299 ( = UCMP V68145). The fossil assemblage from a coarser sand and gravel matrix at locality LACM 4301 ( = UCMP V68106) and UCMP V6880, some 25 feet higher and near the top of the exposed stratigraphic section, was named the Lawrence Canyon Local Fauna. Both local faunas were assigned to the Hemphillian North American Land Mammal Age, approximately 3.5 to 8.5 million years B.P, which includes the earliest Pliocene and latest Miocene epochs (Repenning and Tedford, 1977, table 1). Bones of sharks, fish, birds, and terrestrial and marine mammals were found in both local faunas. The 26 bird bones from the lower horizon represent approximately half of the vertebrate fossil specimens recognized in the San Luis Rey River Local Fauna and are assigned as follows (number of bones in parentheses): Gavia sp. — loon (1), Diomedea sp. — albatross (1), Uria pal- eohesperis new species — extinct murre (9), Cepphus olsoni new species — extinct guillemot (1), ? Aethia sp. — auklet (1), and Praemancalla cf. P. wetmorei Howard — extinct flightless auk (13). Most of the vertebrate bones in the Lawrence Canyon Local Fauna from the upper horizon represent flightless auks of the genus Mancalla Lucas, 1901, though only 24 of the 81 so as- signed are specifically determinable. The following avian taxa are recognized from this local fauna: Family Sulidae — booby or gannet (1), Family Accipitridae — eagle (1 ), IFalco sp. — falcon (1), Cepphus sp. — guillemot (1), Mancalla milleri Howard — extinct flightless auk (9), Mancalla diegensis (Miller) — extinct 1. Review committee for this Contribution: Kenneth E. Campbell, Storrs L. Olson, and David W. Steadman. 2. Chief Curator Emeritus, Natural History Museum. Section of Ver- tebrate Paleontology, Natural History Museum of Los Angeles County, 900 Exposition Boulevard, Los Angeles, California 90007. flightless auk (3), and Mancalla cf. M. cedrosensis Howard- extinct flightless auk ( 1 2). METHODS AND MATERIALS Abbreviations The specimens cited in text are deposited in the following institu- tions: Academy of Natural Sciences of Philadelphia (ANSP), Natural History Museum of Los Angeles County (LACM), Mu- seum of Vertebrate Zoology, University of California, Berkeley (MVZ), San Diego Society of Natural History, Natural History Museum (SDSNH), University of California Museum of Paleon- NORTHWEST tology, Berkeley (UCMP), and United States National Museum (USNM). Methods The specimens illustrated were coated with a sublimate of am- monium chloride before photographing to present an evenly colored surface. Materials FOSSIL. The following holotypes were available for this study: Mancalla milleri Howard, 1970 (LACM 2185), M. cedrosensis Howard, 1971 (LACM 1 5373), Praemancalla lagunen*is SOUTHEAST < oo < o o LU oc CD UJ OC W I I Figure 1. Diagrammatic geologic cross section of part of the San Mateo Formation and the San Onofre Breccia exposed in Lawrence Canyon at Oceanside, showing the stratigraphic occurrence of some fossils making up the two local faunas. Relative thicknesses of the beds are estimated, and the amount of displacement on the fault has not been measured. There is an unconformity at the contact between the San Mateo Formation and the San Onofre Breccia, and another within the San Mateo Formation at the base of the sand and gravel bed that produced the Lawrence Canyon Local Fauna. The San Luis Rey River Local Fauna is derived from the entire thickness of the lowest sand unit resting uncomformably upon the San Onofre Breccia. Modified from Barnes et al, 1 98 1 , fig. 1 . Based on notes and sketches made in the field by Barnes. 2 Contributions in Science, Number 341 Howard: Tertiary Marine Birds Howard, 1966a (LACM 15288), P. wetmorei Howard, 1976 (LACM 42653), Uria brodkorbi Howard, 1981 (PB7960). Casts of the following holotypes were also available: Mancalla califor- niensis Lucas, 1901 (USNM 4976), M. diegensis (Miller, 1937) (UCMP 33409), Uria antiqua (Marsh, 1 870) (ANSP 1 3357), U. affinis (Marsh, 1872) (ANSP 13358). The following referred material was examined: Australca sp., 1 1 ulnae from the Lower Pliocene Yorktown Formation, Lee Creek, North Carolina (USNM 177833, 179270, 179311, 181106, 183497, 192706, 206457, 210456, 215723, 215905, 242223); Miocepphus mc- clungi Wetmore, 1940, ulna (USNM 237219) from the Miocene Calvert Formation of Westmoreland County, Virginia, and humerus (USNM 25668) from the same formation in Maryland. In addition, large collections of avian fossils from the San Diego Formation (LACM and UCMP), Almejas Formation (LACM), Monterey Formation (LACM), and Repetto Formation (LACM), containing specimens discussed herein, were at hand. Included with the LACM and UCMP specimens from the San Mateo Formation in the discussions herein are two humeri (SDSNH 119165 and 23568) and a femur (SDSNH 21101) from the Oceanside locality collected by the San Diego Society of Natural History. RECENT. All skeletal material of Recent species referred to herein is in the LACM collections, with the exception of speci- mens of the genus Aethia Merreni, 1788, which were on loan from MVZ, and Cepphus carbo Pallas, 1811, and C. grylle Lin- naeus, 1758, from USNM. SYSTEM ATICS San Luis Rey River Local Fauna Order Gaviiformes Family Gaviidae Forster, 1788 Loons Genus Gavia Forster, 1788 Gavia sp. REFERRED SPECIMEN. Left tarsometatarsus eroded distally, UCMP 88656. From locality V68145, collected by UCMP field party, August 1 968. DESCRIPTION. In length, the fossil tarsometatarsus is within the size range of tarsometatarsi of the living Pacific Loon, Gavia pacifica (Lawrence, 1858) (66.2-75.3 mm, mean 70.1 mm) and close to that of the Red-throated Loon, Gavia stellata (Pontop- pidan, 1 763) (68.9-7 1 .4 mm, mean 70.2 mm), though below the mean for either (based on four specimens of each of the Recent species). Relative breadths of shaft and proximal end are within the range of G. pacifica, which slightly exceeds G. stellata in these measurements. The only notable character of the fossil specimen is found in the very high posterior position of the distal foramen, which is near the level of the proximal edge of the internal trochlea and 6.9 mm above the external intertrochlear notch. In G. pacifica and G. stellata, the foramen is well below the proximal level of the internal trochlea and only 3.6-5. 1 mm above the intertrochlear notch. Of five available specimens of the much larger G. immer ( Briinnich, 1 764), the maximum distance of the foramen above the intertrochlear notch is only 6.4 mm. Measurements (in mm): length to center of eroded middle trochlea 68.6, breadth of proximal end 12.0, least breadth of shaft 3.8, greatest depth of shaft 6.9. DISCUSSION. Four extinct species of the genus Gavia have been recorded from the Tertiary of North America: Gavia brodkorbi Howard, 1978, from the Upper Miocene Monterey Formation of California; G. palaeodytes Wetmore, 1943, from the early Pliocene (Hemphillian age), Bone Valley Formation of Florida; G. concinna Wetmore, 1940, from the early-middle, Pliocene Etchegoin Formation of California with referred mate- rial from the Bone Valley Formation and the late Pliocene, San Diego Formation of California, and G. howardae Brodkorb, 1953, from the San Diego Formation. A fifth species, G. portisi (Regalia, 1902), is recorded from the Pliocene of Italy, based only on a cervical vertebra. Three modern species of loons are recorded from the Pleistocene of California (Brodkorb, 1963:225). The tarsometatarsus has not been recorded for any of the ex- tinct species. However, Robert Chandler of the San Diego Natu- ral History Museum has called to my attention a small tarsometatarsus (SDSNH 22916) from the San Diego Forma- tion that he is referring to G. howardae (Chandler MS). It bears the same size relationship to the tarsometatarsus of Recent G. stellata (14% shorter) as was previously observed (Howard, 1 978:5) for one of the referred humeri of G. howardae. Not only is it 8 mm shorter than the Oceanside specimen, it does not have the high position of the distal foramen. The very small size of the holotype ulna of G. brodkorbi, which is 23.6 percent shorter than the minimum for G. stellata (Howard, 1978), suggests that this species also cannot be represented by the tarsometatarsus from the San Mateo Formation. Those elements recorded for G. concinna and G. palaeodytes appear to fall within the size range of G. stellata and G. pacifica. Since this is also true for the Oceanside tarsometatarsus, it would be ill-advised to attempt a specific assignment of the specimen at this time. Order Procellariiformes Fiirbringer, 1888 Family Diomedeidae (Gray, 1840) Albatrosses Genus Diomedea Linnaeus, 1758 Diomedea sp. REFERRED SPECIMEN. Distal end of left tibiotarsus, LACM 119353, from locality LACM 4297. Collected by Robert M. McKenzie 9 March 1979. DESCRIPTION. The fossil tibiotarsus resembles the compara- ble element in the Black-footed Albatross, Diomedea nigripes Audubon, 1839, in having a slight notch on the distal surface of the internal condyle and a rounded contour of the external con- dyle (viewed laterally). The fossil is narrower in the area of the ligmental bridge and differs in relative depth to breadth of the distal end (depth of distal end 13.5mm, anterior breadth of distal end 1 3.6 mm; the same dimensions in D. nigripes (LACM 86350) are 15.1 mm and 14.7 mm). Contributions in Science, Number 341 Howard: Tertiary Marine Birds 3 DISCUSSION. A number of albatross bones have been recorded previously from the Miocene of California, although only two species have been described: Diomedea californica Miller, 1962 (a large form intermediate in size between the Recent D. exulans Linnaeus, 1758 and D. albatrus Pallas, 1769) and D. milleri Howard, 1966b (a form notably smaller than D. nigripes). Both were described from the middle Miocene of the Sharktooth Hill Bonebed in Kern County. Diomedea californica is also tentatively recorded from the late Miocene of Orange County (Howard, 1978). Diomedea milleri, known only from the holotype ulna, appears to have been even smaller than the species represented by the tibiotarsus at hand, based on the comparable element of D. nigripes. Bones assignable to Diomedea, and apparently inter- mediate in size between the two described species, have been found in three California late Miocene sites: in the Valmonte diatomite at Lomita, Los Angeles County (Miller, 1935); and in the Monterey Formation in Laguna Hills, Orange County, at locality LACM 1945 (Howard, 1968) and Laguna Niguel, Or- ange County, at localities LACM 6902 and 6906 (Howard, 1978). A single species may be represented by these several oc- currences, possibly including the Oceanside specimen. Two other fossils species of Diomedea have been described: D. thryidata Wilkinson, 1 969 from the Miocene of Australia, known only from the holotype rostrum; and D. anglica Lydekker, 1891 from the early Pleistocene of England (with referred specimens from the Pliocene of England and Florida (Brodkorb, 1 963:242)). The Florida specimen of D. anglica is a tibiotarsus of much larger size than LACM 1 19353 from Oceanside. Extant species of the genus have been recorded from the Pleistocene of southern California (Brodkorb, 1963:242). Order Charadriiformes (Huxley, 1867) Family Alcidae Vigors, 1825 Auk-like Birds Subfamily Alcinae (Vigors, 1825) Genus Uria Brisson, 1760 — Murres DISCUSSION. The coracoid, ulna, and radius of an incomplete skeleton (UCMP 88704) from locality UCMP V68147 more closely resemble the corresponding bones of Recent species of Uria than of any other Recent alcid genus in the following char- acters: coracoid combining a prominent procoracoid process and well-formed foramen with an anteroposteriorly narrow sternal facet; ulna with a stout olecranon, separated from the cotylar rim by a deep notch or channel (as viewed externally); and radius with the ulnar surface of the shaft above the distal end broad, with a distinct, round ligamental attachment on the ligamental prominence. The associated scapula, however, has a narrower, less prominently projected glenoid facet than that of Uria or the closely related Razorbill ( Alca torda Linnaeus, 1758), and, in these characters, more closely resembles the murrelets of the genus Endomychura Oberholser, 1899. Although the scapula suggests generic distinction from Uria, 1 have chosen to place this incomplete skeleton in the genus Uria, in view of the similarities observed in the coracoid, ulna, and radius. Uria paleohesperis n. sp. Fig. 2, a-f HOLOTYPE. UCMP 88704, associated elements including ar- ticular end of right scapula, left and right coracoid (both lacking head), complete left ulna and distal end of right ulna, distal ends of left and right radius, distal end of left carpometacarpus. and wing phalanx digit II, phalanx 2. Collected by University of California Museum of Paleontology field party, August 1968. LOCALITY. UCMP V68147, Loretta Street, Lawrence Can- yon, Oceanside, San Diego County, California. FORMATION AND ACE. San Luis Rey River Local Fauna, San Mateo Formation. Late Miocene. ETYMOLOGY. The specific name is derived from the Greek pa- laeo — ancient, and hesperos — of the west, thus indicating an ancient western rnurre. DIAGNOSIS. Distinguished from Recent species, Uria aalge (Pontoppidan, 1 763) and U. lomvia (Linnaeus, 1758) as follows: coracoid relatively stouter, with well-formed procoracoid more distinctly pointed at the tip, foramen small, and scapular facet deeply cup-shaped and sharply rimmed; scapula more concave dorsally between acromion and glenoid facet, acromion longer, and glenoid facet narrower and not markedly projected from shaft; ulna relatively stouter, with rounded shaft (less blade-like than in Recent Uria), attachment of anterior articular ligament more broadly oval and flatter, and brachial impression broader, extending more than half the depth of the shaft; radius more distally attenuated on internal side of distal articulation; car- pometacarpus poorly preserved, but contour of metacarpal 3 in symphysial region seemingly more rounded than angular; digit II, phalanx 2 less deeply depressed posteriorly and with less sharply developed ridge anteriorly than that of U. aalge, closer to that of U. lomvia, but with ridge slightly more marked anteriorly; distinguished from extinct species Uria antiqua (Marsh, 1870), U. affinis (Marsh, 1872), and U. brodkorbi Howard, 1981, by smaller size, and also from U. brodkorbi by scapula being more concave dorsally and having narrower glenoid facet. MEASUREMENTS. See Table 1. DISCUSSION. The proximal end of a humerus (LACM 52018) from the late Miocene, Monterey Formation of Orange County, California, was assigned to lUria sp. (Howard, 1978). Although only a small fragment, it suggests the stockier proportions of the shaft, in comparison with humeri of Recent species of Uria, that are noted in the ulna of Uria paleohesperis. Possibly the same species is represented. Four other extinct species of the genus Uria have been de- scribed: Uria antiqua (Marsh, 1 870), from the Lower Pliocene of North Carolina, U. affinis (Marsh, 1 872), from the Pleistocene of Maine, and U. ausonia Portis, 1888, from the Pliocene of Italy, each based on a humerus; and U. brodkorbi Howard, 1981, from the Miocene, Sisquoc Formation in California, based on an in- complete skeleton. The small size of the elements of U. pa- leohesperis precludes their assignment to either U. antiqua or U. 4 Contributions in Science, Number 341 Howard: Tertiary Marine Birds affinis, the holotypes of which are larger than humeri of living species of Uria. Furthermore, according to Olson and Gillette (1978), U. antiqua should be referred to the genus Australca Brodkorb, 1 955, rather than to Uria. This may be true, as well, of U. ausonia. Although the holotype humerus of U. ausonia is only a fragmentary distal end, the illustration (Portis, 1891, table I ) that appeared subsequent to the original description suggests that the tricipital grooves are equal in size as in Alca torda Lin- naeus, 1758, and Australca grandis Brodkorb, 1955. The exter- nal tricipital groove is notably narrower than the internal groove in the genus Uria. The holotype of U. brodkorbi consists of the impressions (in diatomite) of an incomplete skeleton, the wing and girdle elements of which are generally larger than those of U. paleohesperis. The details of the scapula are best compared in the two species. The scapula in U. brodkorbi resembles that of Recent species of Uria and differs from that of U. paleohesperis in the greater breadth of the glenoid facet and less dorsal concavity. Ulnae (USNM) from Lee Creek, North Carolina, which Olson (USNM, pers. comm.) assigns to Australca. were loaned for this study. All have a more raised and distally pointed attach- ment for the anterior articular ligament than that of U. pa- leohesperis. This is also true of an ulna (USNM 237219) from the Miocene Calvert Formation of Virginia, which Olson (pers. comm.) assigns to Miocepphus mcclungi Wetmore, 1940. This ulna is also notably smaller and more curved than that of U. paleohesperis. One species of murre, Uria aalge (Pontoppidan, 1 763), occurs today as far south as Newport Beach, Orange County, California, and has been recorded from the Pleistocene of Los Angeles County (Howard, 1936). Genus Cepphus Pallas, 1769 Guillemots DISCUSSION. A humerus (LACM 107032) resembles that of the guillemots (genus Cepphus) as distinguished from all other genera of the family Alcidae in having a rounded, rather than compressed or bladelike, shaft that describes a broad S curva- ture from proximal to distal end as viewed both laterally and anconally. Cepphus olsoni n„ sp. Fig. 2, g, h HOLOTYPE. Right humerus, LACM 107032. Collected by Robert M. McKenzie, 19 March 1975. LOCALITY. LACM 4297, Loretta Street. Lawrence Canyon, Oceanside, San Diego County, California. LORMATION AND AGE. San Mateo Formation; San Luis Rey River Local Fauna. Late Miocene. ETYMOLOGY. The species is named in honor of Dr. Storrs L. Olson, of the National Museum of Natural History, in recogni- tion of his extensive studies of fossil Alcidae. Figure 2. Uria paleohesperis n. sp. and Cepphus olsoni n. sp. natural size Parts a through f, U. paleohesperis n. sp., holotype, UCMP 88704. Parts a and b, left ulna, external and internal views. Part c, right coracoid, dorsal (posterior) view. Part d, distal end of left carpometacarpus, external view. Part e, distal end of right radius, palmar view. Part f, right scapula, dorsal view. Parts g and h, C. olsoni, n. sp., holotype, LACM 107032, humerus, anconal and palmar views. Contributions in Science, Number 341 Howard: Tertiary Marine Birds 5 Tabic I. Measurements I in mm) of Uria paleohesperis n. sp. compared to Recent Uria aalge californica (7 specimens). Uria paleohesperis U. a. Max. californica Mean Min. Scapula Breadth proximal end 10.6 I 1.5 1 1.1 1 1.0 Glenoid facet Length 5.9 5.7 5.5 5.0 Breadth 3.7 4.3 3.8 3.6 Shaft near proximal end Breadth 4.3 5.2 4.6 4.1 Depth 2.5 2.25 2.15 2.1 Coracoid Length from sternal facet to lip of scapular facet 23.4 26.4 25.0 24.3 Breadth of shaft at level of foramen 4.8 4.7 4.5 4.4 Breadth sternal end 14.6 16.2 15.1 14.6 Depth of sternal facet 4.5 5.1 4.5 4.2 Ulna Greatest length 62.6 69.5 65.3 61.3 Proximal end Breadth 8.5 8.6 8.3 8.1 Depth internally 9.7 10.15 9.9 9.8 Middle shaft Breadth 3.85 4.3 4.0 3.8 Depth 5.35 6.6 6.1 5.75 Greatest breadth of brachial impression 3.5 3.4 2.9 2.6 Radius, breadth distal end 5.7 5.9 5.7 5.6 DIAGNOSIS. Compared with the three living species of the genus (in increasing order of size of the humerus), Cepphus grylle Linnaeus, 1758, C. columba Pallas, 1811, and C. carbo Pallas, 1811: between C. grylle and C. columba in length; rela- tively stouter of shaft; deltoid crest longer and less abruptly ter- minated distally; pectoral attachment shorter, with surface concave and more clearly defined; shaft mediad of pectoral at- tachment sharply ridged, suggestive of condition in C. grylle, but space between ridge and median crest narrower and more de- pressed; bicipital surface distinctly bordered medially and raised from level of shaft; ectepicondylar prominence more evenly rounded in contour, lacking prominent proximal tip; tricipital grooves less distinct (bordering ridges less acute) than in C. grylle and C. carbo, closer to C. columba. MEASUREMENTS. See Table 2. DISCUSSION. The genus Cepphus was tentatively recorded from locality LACM 6906 in the late Miocene Monterey Forma- tion of Orange County (Howard, 1 978:21 ). Although that speci- men, an ulna (LACM 47045), suggests a species comparable in size to Cepphus olsoni, there is no clear indication that the diffe- rent elements represent the same species. Apart from these two specimens, there is no other previous fossil record of the genus Cepphus. In the material from the upper level of the San Mateo Formation, however, a fragment of a coracoid is herein identified as Cepphus sp. According to Storrs Olson (pers. comm.) Cepphus has not been found in the abundant alcid material known from the Miocene and Pliocene deposits of the western Atlantic. That Cepphus olsoni is in no way related to Miocepphus mc- clungi Wetmore, 1940, from the Miocene Calvert Formation of Maryland is evident from the illustration of the holotype humerus of the latter species (Wetmore, 1940:36, figs. 11 and 12), and the shaft measurements provided, as well as by compari- son with another humerus (USNM 25668), from a nearby lo- cality in the type formation, which Storrs Olson (USNM, pers. comm.) refers to this species. The shaft in Miocepphus is more laterally compressed, the ectepicondylar process more promi- nent proximally, and the pectoral attachment more elliptical with space between attachment and median crest less sharply de- pressed than in the holotype of Cepphus olsoni. These observa- tions are in keeping with an opinion previously provided by Olson (Howard, 1978:21) that Miocepphus is more closely related to Uria and Alca than to Cepphus. Genus Aethia Merrem, 1788 Auklet ? Aethia sp. REFERRED SPECIMEN. Left humerus lacking proximal end, LACM 107031, from locality LACM 4297. Collected by Law- rence G. Barnes, 19 March 1975. DISCUSSION. This small auklet humerus resembles the re- ferred humerus (LACM 18949) collected and described with the holotype ulna of Aethia rossmoori Howard, 1968, from its type locality (LACM 1945) in the late Miocene, Monterey Formation of Laguna Hills, Orange County, California. It also resembles a humerus (LACM 37686) from a slightly later horizon of the Monterey Formation in Orange County (locality LACM 6906) recorded (Howard, 1 978:2 1 ) as ? Aethia sp. All three specimens differ from Recent specimens of Aethia pusilla (Pallas, 1 8 1 1 ) and A. pygmaea (Gmelin, 1789) by having a more rounded shaft and a greater depression of the brachial area and of the attachment for the anterior articular ligament, with the attachment facing more palmad than laterad. Of the three fossils, only LACM 107031 is complete in the region of the tricipital ridges. It differs from specimens of Recent Aethia in that the internal tricipital groove is narrower than the external and lacks the deep depression above the distal edge. In view of this notable distinction, plus the fact that several very small coracoids previously recorded from locality LACM 1945 (Howard, 1968:17) have a well-developed procoracoid with well- formed foramen (unlike the short procoracoid and lack of a fora- men characteristic of Recent Aethia ), it is suggested that an extinct genus of murrelet or auklet is represented in the late Miocene. It is possible that even A. rossmoori should be gener- 6 Contributions in Science, Number 341 Howard: Tertiary Marine Birds ically reassigned. It is hoped that further finds of this small alcine will be forthcoming. Measurements (in mm) of LACM 107031, with those of fossil humeri from Orange County ( LACM 1 8949 and LACM 37686, respectively) in parentheses: breadth of distal end 5.0 (4. 1:5.3); least breadth of shaft 2.0 ( 1 .8; 2.3); depth of shaft at same place 2.7 (2.6; 3.3); distance from distal condyle to proximal tip of ectepicondylar process 4.9 (4.5; 5.2). Subfamily Mancallinae (Miller, 1946) Extinct Flightless Auks DISCUSSION. The subfamily Mancallinae is best known from the type genus, Mancalla Lucas, 1901, in which the wing bones were modified as swimming paddles. Fossil bones of Mancalla are abundant in Pliocene deposits in southern California and are recorded herein from the Lawrence Canyon Local Fauna (de- scribed later in this paper). Prior records of the genus Praeman- calla are restricted to the Upper Miocene Monterey Formation in Orange County. The generic name reflects its possible phy- logenetic status relative to Mancalla. Praemancalla has wing bones less highly modified for swimming. Genus Praemancalla Howard, 1966a DISCUSSION. Two species of Praemancalla have been re- corded from the Monterey Formation, Orange County, Califor- nia: P. lagunensis Howard, 1966a, and P. wetmorei Howard, 1976. The type locality (LACM 1945) of P. lagunensis repre- sents a slightly earlier and stratigraphically lower horizon of the Monterey Formation than that of P. wetmorei (locality LACM 6906). Only bones of the wing and shoulder girdle and one frag- ment of mandible were known previously for Praemancalla. Praemancalla cf. P. wetmorei Howard, 1976 Fig. 3 REFERRED MATERIAL. Partial skeleton (LACM 107028) con- sisting of left ulna, left carpometacarpus, right tibiotarsus, left tarsometatarsus, left pedal digit 11, phalanx 1, and thoracic ver- tebra 3. From locality LACM 4297, collected by Robert M. McKenzie, 19 March 1975. DESCRIPTION OF REFERRED MATERIAL. The ulna resembles the paratype of Praemancalla wetmorei (LACM 32429) and is distinguished from the ulna in the several species of Mancalla in having (1) olecranon prominent, set off from the cotylae by a deep groove externally and a depression internally (although in the paratype the depression is deeper and more pitlike than in LACM 107028); (2) brachial impression partially palmad and bordered by a long, heavy ridge; and (3) proximal radial depres- sion broad. The distal end, which is missing in the paratype, is distinguished from specimens of the several species of Mancalla by having a greatly enlarged projection overhanging the tendinal pit externally and a deep groove separating the carpal tuberosity from the trochlea internally. Although the paratype of P. wet- morei lacks the distal end, the length of the Oceanside specimen (36.7 mm) conforms well with the length of the previously re- Table 2. Measurements (in mm) and ratios (in percent) of humeri of Cepphus olsoni n. sp., C. grylle (2 specimens), and C. columba (3 specimens). C. grylle C. columba C. olsoni Female Male Min. Max. Greatest length 61.6 59.8 60.6 66.2 69.2 Breadth of proximal end across bicipital crest 15.4 14.4 14.0 14.7 15.1 Breadth of distal end through condyles 7.1 7.0 7.1 7.4 8.0 Breadth of shaft (middle) 4.8 4.3 4.0 4.3 4.8 Depth of shaft from anconal to palmar surface (middle) 3.8 3.4 3.4 3.4 3.7 Ratio, breadth of shaft to length of humerus 7.8 7.2 6.6 6.5 6.9 ferred radius of P. wetmorei (LACM 53907, 35.8 mm). Neither ulna nor radius is known for P. lagunensis. The carpometacarpus resembles that of both species of Prae- mancalla, and is distinguished from all species of Mancalla, in having a distinct, blunt pisiform process and a more flared inter- nal trochlear crest. It resembles the referred carpometacarpus of P. wetmorei (LACM 52216), and is distinguished from the para- type carpometacarpus of P. lagunensis (LACM 15287), by hav- ing a greater proximal extension of the trochlear crest above metacarpal I, a longer process of metacarpal I (proximodistally), and a less deeply depressed surface above the pisiform process. Distally, it is distinguished from carpometacarpi of Mancalla by having a more posteriorly extended distal surface of metacarpal 1 1 toward Mill, resulting in limited extent of the groove between metacarpal II and III distally. The available carpometacarpi of both P. wetmorei and P. lagunensis lack the distal end. The con- tours of the referred specimen of P. wetmorei ( LACM 52216), as compared with the complete bone from Oceanside, suggest that only 1 mm is lacking from that of P. wetmorei and that it was approximately equal in length to the one now at hand. The tibiotarsus is similar to that of the several known species of Mancalla in its general shape and curvature, being relatively broader and flatter of shaft than in species of the subfamily Alcinae, such as the murres (genus Uria) or the Great Auk, Pinguinus impennis (Linnaeus, 1758). It is distinguished from tibiotarsi of Mancalla spp. by having a broader shaft that is less convex anteriorly below the cnemial crests, relatively longer cnemial crests, and a more markedly flared distal end. The tarsometatarsus is similar to that of Mancalla spp. in general shape, with the shaft’s anterior surface depressed and Contributions in Science, Number 341 I loward: Tertiary Marine Birds 7 bordered externally by a ridge. It is distinguished from tar- sometatarsi of Mancalla spp. by the more abrupt narrowing of the shaft distal to the center; more broadly and evenly rounded anterior face of the shaft above the trochleae (lacking the acutely raised area proximal to the middle trochlea); more proximal posi- tion of the distal foramen, which is set in a shallower groove; and more distal position of the internal trochlea. No leg bones have been previously recorded for either species of Praemancalla. The pedal phalanx conforms in general characters to this ele- ment in Recent alcids, such as the murres (genus Uria), but is markedly larger: length 19.9 mm, proximal breadth 5.7 mm, distal breadth, 3.4 mm. The specimen articulates suitably with Figure 3. Praemancalla cf. P. wetmorei, natural size, a-e and g-i, LACM 107028, f, j, k, SDSNH 21101. Part a, carpometacarpus, internal view. Parts b and c, ulna, external and internal views. Parts d and e, tarsometatarsus, posterior and anterior views. Parts f, j, and k, femur, external, anterior, and posterior views. Parts g, h, and i, tibiotarsus, internal, anterior, and external views. 8 Contributions in Science, Number 341 Howard: Tertiary Marine Birds the internal trochlea of the tarsometatarsus of LACM 107028. This element has not been previously recorded for either Man- calla or Praemancalla. The thoracic vertebra was compared with the 3rd thoracic vertebra of Mancalla cedrosensis Howard, 1971 (LACM 15425 from locality LACM 65144), the best preserved of the available mancalline vertebrae. The Oceanside specimen resembles LACM 15425 and is distinguished from thoracic vertebrae of Recent species of Uria in having a longer costal attachment and a straighter ventral border of the posterior articular surface of the centrum. Anteriorly, the articular surface of the centrum is broadly concave as in Mancalla and Uria, but it is distinct from both in having a greater dorsoventral dimension at the center. Measurements (in mm): length of centrum 11.9; anterior breadth of centrum 8.4; posterior breadth of centrum 6.5; pos- terior height of centrum 5.1. Same measurements in M. cedrosensis (LACM 15425): 10.0, 7.3, 4.9, and 4.2 respectively. No vertebrae of either species of Praemancalla have been pre- viously recorded. Measurements of wing and leg bones of LACM 107028 are compared with those available for the two species of Praeman- calla and the largest specimens of Mancalla diegensis in Table 3. TENTATIVELY REFERRED MATERIAL. From locality LACM 4297 and its equivalent, UCMP V68147: scapular end of cor- acoid (LACM (107030); synsacrum with incomplete pelvic bones of right side including acetabulum and antitrochanter (LACM 1 19406); axis vertebra (UCMP 102428). From locality LACM 4298 and its equivalent, UCMP V68 1 44: right humerus, abraded distally and in the area of the internal tuberosity (LACM 107029); pedal digit III, phalanx 1 (UCMP 88640). From locality UCMP V68145: poorly preserved proximal end of left scapula (UCMP 95119); left femur poorly preserved both proximally and distally (UCMP 95 1 1 8). A complete right femur (SDSNH 21101), with characters similar to those of UCMP 95118, was loaned for this study by the San Diego Natural His- tory Museum (locality SDSNH 3003). Judging from the preser- vation of the specimen and its adhering matrix, it is from the lower level deposits of the San Luis Rey Local Fauna. DESCRIPTION OF TENTATIVELY REFERRED MATERIAL. Coracoid LACM 107030 resembles this element of Praeman- calla and is distinguished from that of Mancalla by the more medially, less posteriorly oriented triosseal canal and the broad, flat coracohumeral attachment, which is not twisted anteriorly. The specimen is poorly preserved, and the characters that dis- tinguish P. wetmorei from P. lagunensis are not clearly observ- able. The only measurement possible (the span from the furcular facet to the top of the glenoid facet) agrees with this dimension in the referred coracoid of P. wetmorei (LACM 37637), 10.4 mm. The same dimension in the coracoid referred to P. lagunensis (LACM 15289) is 8.4 mm. The scapula (UCMP 95119) is distinguished from that of the several species of Mancalla by the flatter, less concave ventral surface of the proximal end and the longer, more distal extension of the acromion. These characters are suggested in the single, poorly preserved referred scapula of P. lagunensis (LACM 15294). The scapula has not been recorded previously for P. wetmorei. The proximal breadth of UCMP 951 19 (13.8 mm) is greater than that of P. lagunensis (12.5 mm). The following characters of the humerus (LACM 107029) agree with those attributed to the genus Praemancalla (Howard, 1976:142): deltoid crest weakly developed; absence of papilla proximal to internal condyle; and presence of a groove separating the base of the ectepicondylar process from the external condyle. It is difficult to determine the profile of the capital groove be- cause the internal tuberosity is broken away. The protrusion of the head over the groove, however, appears to be less than in Mancalla spp. and similar to the condition in Praemancalla wet- morei. The proximal end of the humerus is not known for P. lagunensis. The relative breadth to depth of the shaft above the distal end is 53 percent, as in P. wetmorei, contrasted with 66 percent in P. lagunensis. For measurements, see Table 3. The two femora, SDSNH 21 101 and UCMP951 18, undoubt- edly represent a single species. SDSNH 21101 is complete; UCMP 95118 lacks the proximal surface and has eroded distal contours. Both show the mediad thrust of the distal end, sug- gestive of the holotype femur of Mancalla diegensis (Miller, 1937). They are distinguished from this and all other species of Mancalla by the marked depression of the external side of the shaft adjacent to the fibular condyle, which results in emphasiz- ing the prominence of the fibular and external condyles. The popliteal area is deeper than in M. diegensis but resembles speci- mens of M. milleri Howard, 1970 as well as those of the Great Auk, Pinguinus impennis (Linnaeus, 1758), in this respect. The size of the femur is also similar to that of P. impennis. Proximally, the head is large and tilted proximally, and the shaft recedes more abruptly from the head than in Mancalla diegensis or Pinguinus impennis. The shaft is depressed posteriorly adjacent to the obturator ridge, giving the ridge added prominence. Ante- riorly, the trochanter is also prominent and deeply depressed along the internal edge, resembling in this respect the femur of Uria aalge, although the trochanter is longer in that species. Both obturator ridge and trochanter are more prominent than in Man- calla diegensis. (For measurements, see Table 3.) The mancalline features of the femora and their size, which is proportionate to that of the tibiotarsus and tarsometatarsus associated with the wing elements herein assigned to Praemancalla cf. P wetmorei. justify tentative assignment of the specimens to P. wetmorei. The ratio of the length of femur SDSNH 21101 to the length of tibiotarsus LACM 107028 assigned to P. cf. P. wetmorei. is 65.9 percent. The same ratio in the holotype skeleton of Mancalla cedrosensis Howard, 1971 (LACM 15373) is 65.1 percent. The synsacral section of the pelvis (LACM 1 19406) resembles a similarly preserved specimen assigned to Mancalla diegensis (LACM 2340) from the San Diego Formation in being generally heavier than that of members of the subfamily Alcinae, such as Uria aalge and Alca torda. It is, however, even heavier than that of M. diegensis (LACM 2340), and the median dorsal ridge is broader (greatest breadth of dorsal ridge 3.5 mm; 2.0 mm in LACM 2340). The axis vertebra (UCMP 102428) was compared with an unrecorded axis of Mancalla diegensis (UCMP 45892) from the San Diego Formation. The Oceanside specimen has the facets of the postzygapophyses rounded in contour as in M. diegensis, rather than oval as in the Recent murre, Uria aalge. These facets, are however, more concave than in M. diegensis, and the sides of the vertebra are more deeply depressed toward the anterior end. Contributions in Science, Number 341 Howard: Tertiary Marine Birds 9 In this latter respect, the condition is more suggestive of that found in Uria than in Mancalla. The specimen is markedly larger than M. diegensis UCMP 45892 and is in keeping with the large leg bones (LACM 107028) herein assigned to Praemancalla cf. P. wetmorei. Measurements (in mm): greatest length exclusive of the dorsal spine, 18.5 (A/, diegensis UCMP 45892, 13.4); breadth and height of posterior surface of centrum, 3.1 and 5.5 respectively (UCMP 45892, 2.7 and 4.0). The pedal digit III, phalanx 1 (UCMP 88640) suggests in general size digit II, phalanx 1 of the partial skeleton (LACM 107028) assigned herein to Praemancalla cf. P. wetmorei. Lawrence Canyon Local Fauna Order Pelecanifornies Sharpe, 1891 Family Sulidae (Reichenbach, 1849) Boobies and Gannets Sulidae gen. and sp. indet. REFERRED SPECIMEN. Digit II, phalanx 1, LACM 119312, from locality LACM 4301. Fable 3. Measurements (in mm) of Oceanside Praemancalla bones* and those of P. wetmorei, P. lagunensis, and Mancalla diegensis (maximum) Oceanside Praemancalla Praemancalla Mancalla Praemancalla wetmorei lagunensis diegensis Humerus Length to intercondylar sulcus 80.8 81.1 — 83.4 Proximal breadth 20.5 22.2 — 20.3 Greatest depth of shaft 9.8 9.6 7.7 10.3 Breadth of shaft at point of greatest depth 5.2 5.1 5.1 5.2 Ulna Length to intercotylar ridge 36.7 — — 32.0 Proximal breadth 7.5 7.5 — 6.6 Proximal depth 1 1.0 11. 3 — 9.3 Breadth shaft (middle) 4.4 4.2 — 3.9 Depth shaft (middle) 6.6 6.2 — 6.8 Carpometacarpus Length 36.3 36.3(est) — 37.2 Breadth proximal trochlea 5.2 5.3 5.2 4.7 Proximal depth through M 1 1 1.8 12.1 1 1.7 1 1.0 Length process of M 1 15.2 15.7 14.0 15.5 Breadth of shaft 4.0 4.0 4.5 3.7 Femur Greatest length (external) 69.3 — — 57.0 Distal breadth 14.4 — — 11.1 Proximal breadth 14.6 — — 1 1.7 Tibiotarsus Length to proximal articular surface 105.1 — — 98.5 Distal breadth 13.5 — — 10.7 Distal depth 12.1 — — 9.7 Tarsometatarsus Greatest length 47.5 — — 43.6 Proximal breadth 12.1 — — 10.7 Distal breadth 11.6 — — 9.4 Shaft breadth (middle) 6.1 — — 5.4 *Humerus LACM 107029; Femur SDSNH 21 101; all others LACM 107028. 10 Contributions in Science, Number 341 Howard: Tertiary Marine Birds DISCUSSION. This incomplete phalanx resembles the com- parable element in the gannet, Morns bassanus (Linnaeus, 1758), in general characters but is 2 mm (29 percent) broader. Although the family Sulidae is no longer represented on the west coast north of Mexico, it is well recorded in the Tertiary and Quaternary of California (Brodkorb, 1963:258-261 and Howard, 1978: 16-19). It is impossible to provide definite identi- fication for this fragment, although its large size is suggestive of Morns magnus Howard, 1978, from the Monterey Formation. Order Falconiformes Seebohm, 1890 Family Accipitridae (Viellot, 1816) Eagles and Hawks Accipitridae gen. and sp. indet. REFERREDSPECIMEN. Proximal end of pedal digit I, phalanx 1 , LACM 119310, from locality LACM 4301 . DISCUSSION. In size, this specimen suggests assignment to a small eagle or large hawk. Greatest breadth of proximal end is 13.2 mm; the same measurement in the Golden Eagle, Aquila chrysaetos (Linnaeus, 1758), is 1 6.7 mm, and in the Ferruginous Hawk, Buteo regalis (Gray, 1844), 1 1.9 mm. The shape of the proximal end more closely resembles that of the eagle. There are only two previous records of the family Accipitridae for the Tertiary of California, both of early Miocene age: Miohierax stocki Howard, 1944, and Buteo indet. (Merriam, 1919). Neither is from a marine deposit. The family is well repre- sented in the Pleistocene ( Brodkorb, 1 964:269-27 1,281 -284). Family Falconidae (Vigors, 1824) — Falcons Genus Falco Linnaeus, 1758 1 Falco sp. REFERRED SPECIMEN. Fragment of right clavicle, UCMP 88597 from locality UCMP V6880. DISCUSSION. This incomplete specimen is suggestive of a falconid clavicle in its prominent, round coracoidal facet. It is similar in size to that of a male Peregrine Falcon, Falco per- egrinus Tunstall 1771, a species known in California today. CORRECTION. Owing to a misunderstanding regarding the ex- act location of UCMP V6880, this specimen was incorrectly cited as being from the San Luis Rey River Local Fauna in Barnes et al. (1981:61 ). Order Charadriiformes (Huxley, 1867) Family Alcidae Vigors, 1825 — Auk-like Birds Subfamily Alcinae (Vigors, 1825) Genus Cepphus Pallas, 1769- Guillemots Cepphus sp. REFERREDSPECIMEN. Fragment of right coracoid including glenoid facet and portion of procoracoid, LACM 1 19260, from locality LACM 4301. DISCUSSION. The procoracoid is notched 4.6 mm below the upper surface, resembling in this character specimens of living Pigeon Guillemot, Cepphus columba Pallas, 1811, a species known today from the coast of California. The few measure- ments possible on this fragment, compared to those of C. co- lumba, are (in mm): length through glenoid facet and scapular facet 9.8 (8.7-9. 1 in C. columba): depth from tip of procoracoid to shaft 7.8 (7.7-8. 1 in C. columba). It is impossible to assess the relationship of this fragment to the new species, Cepphus olsoni. described herein from a humerus in San Luis Rey River Local Fauna. Subfamily Mancallinae ( Miller, 1946) Extinct Flightless Auks Genus Mcmcalla Lucas, 1901 DISCUSSION. Five species of Mancalla are known: M. califor- niensis Lucas, 1901, from the early Pliocene (Repetto Forma- tion) in Los Angeles, California, with 1 1 referred specimens (Howard, 1949:196 and 1970:3) from the Repetto Formation in Corona del Mar, Orange County, California; M. cedrosensis Howard, 1971, the holotype a nearly complete skeleton, and many referred specimens, all from the Almejas Formation (late Hemphilhan) of Cedros Island, Baja California, Mexico; M. di- egensis (Miller, 1937) and M. milleri Howard, 1970, both de- scribed and well represented by all principal skeletal elements from the late Pliocene, San Diego Formation in San Diego, Cal- ifornia; and M. emlongi Olson, 1981, based on an ulna from the San Diego Formation at Pacific Beach, San Diego County. Eighty bones from locality LACM 4301 and one from locality UCMP V68106 are assignable to this genus. Also available for this study were two complete humeri from locality SDSNH 2643 (equivalent of locality LACM 4301 ). Preservation of the LACM and UCMP material is poor, but the characters that distinguish Mancalla from Praemancalla are observable in the coracoids, scapulae, humeri, ulnae, radius, carpometacarpi, tibiotarsi, and tarsometatarsi, as follows: coracoids with scapular facet placed laterally and triosseal canal posterointernally; scapulae concave ventrally, with short acromion, and glenoid facet projecting me- diad; proximal ends of humeri with notch-like capital groove, and distal ends with papilla above distal condyles; ulna with olecranon not projecting proximad; radius bladelike; car- pometacarpi without distinct pisiform process; tibiotarsi without flaring distal end; tarsometatarsi with anterior face of shaft de- pressed along external side but raised above middle trochlea. Several bones fall within the size range of the small M. milleri. The others are commensurate in size with those of M. diegensis, M. cedrosensis, and M. calif orniensis. Mancalla emlongi, known only from the holotype ulna, was a larger species. Twenty-six bones, including the two from SDSNH, are as- signed, at least tentatively, to three species. The other 57 remain as Mancalla species indeterminate. Mancalla milleri Howard, 1970 REFERRED MATERIAL. From locality SDSNH. 2643, com- plete humerus (SDSNH 23568); from locality LACM 4301, Contributions in Science, Number 341 Howard: Tertiary Marine Birds 11 distal ends of 2 humeri (LACM 1 19166 and 1 19272), 2 ulnae (LACM 119283 and 119286), incomplete radius (LACM 119292), proximal end of carpometacarpus (LACM 119290) and proximal ends of 3 scapulae (LACM 119262, 119264, 119265). DESCRIPTION. The complete humerus agrees in size with that of M. milleri. Measurements (in mm): length to internal condyle 62.7 (M. milleri 56.4-66.6), proximal breadth 15.7 ( M . milleri 14.2-16.4), distal breadth 6.0 (M. milleri 5. 1-6.0) The humerus also has the additional character of the angular distal contour of the bicipital crest, which distinguishes it from the humerus of M. cedrosensis. The other specimens are assigned entirely on the basis of small size. Measurements (in mm): distal breadths of humeri 5.4 and 5.7, proximal breadths of ulnae 5.3 and 5.5 (M. milleri 4. 5-5. 7) greatest shaft depth of radius 5.3 (M. milleri 4. 6-5. 6), proximal depth of carpometacarpus through metacar- pal I 8.8 (M. milleri 8. 8-9. 3), proximal breadth of complete scapular end 1 1.6 (M. milleri 10.2-1 1.8). Mancalla diegensis (Miller, 1937) REFERRED MATERIAL; From locality SDSNH 2643, com- plete left humerus (SDSNH 23567); from locality LACM 4301 , 2 proximal ends of humeri (LACM 119301 and 119276). An ulna (LACM 1 19279) is tentatively referred. DESCRIPTION: The complete humerus has the very angular distal contour of the bicipital crest characteristic of Mancalla diegensis, M. milleri, and M. californiensis, and is further dis- tinguished from M. cedrosensis by the broader, more shallow area below the head. It is distinguished from M. milleri by larger size and from M. californiensis by the muscle scar at the median border of the pneumatic fossa, which does not protrude into the fossa as a distinct groove (Miller and Howard, 1949:209). Al- though incomplete, the two proximal fragments have the small scar at the distal edge of the bicipital crest that provides the angular contour of the crest as noted above. The olecranon of the ulna is incomplete but appears to be straighter in anconal contour than in M. cedrosensis, and the adjacent depression is small. Measurements (in mm) compared with those of M. diegensis (Howard 1970, table 3): humerus (SDSNH 23567), length to internal condyle, 75.9, proximal breadth, 18.9 (M. diegensis, 71.0-85.2 and 17.3-20.3, respectively); ulna (LACM 119279), length 31.7 (M. diegensis 28.0-32.1), proximal breadth 6.1 (M. diegensis 5. 9-6. 6). Mancalla cf. M. cedrosensis Howard, 1971 REFERRED MATERIAL: From locality 4301: a nearly com- plete, but poorly preserved humerus (LACM 1 19165); proximal ends of 5 humeri (LACM 119222, 119224, 119269, 119273, 1 19372); complete tarsometatarsus (LACM 1 19298) and shaft of tarsometatarsus (LACM 119174). In addition, 4 complete ulnae (LACM 119280, 119281, 119287, 119288) are tentatively referred. DESCRIPTION. The humeri resemble this element of M. cedrosensis and are distinguished from those of M. diegensis, M. milleri and M. californiensis by the characters described (Howard, 1971:12) for the holotype of M. cedrosensis (LACM 1 5273): “ . . . internal contour from shaft through bicipital crest broadly and gradually curved; area below head between pectoral attachment and pneumatic fossa oval and deeply depressed.” All are within the size range for M. cedrosensis except LACM 1 19372, which falls between M. cedrosensis and M. milleri in proximal breadth (16.6 mm in LACM 1 19372; M. milleri max- imum 16.4; M. cedrosensis minimum 17.0 mm). The complete tarsometatarsus has the proximal and distal ends Haring as in both M. cedrosensis and M. diegensis, as con- trasted with the more columnar shape in M. milleri. In both LACM 1 19298 and the incomplete LACM 1 19174, the shaft is more depressed anteriorly than in either M. diegensis or M. cal- iforniensis, and both borders of the anterior face of the shaft are sharply defined as in M. cedrosensis. Length (in mm) of LACM 1 1 9298, 41.1, proximal breadth 1 0. 1 (same measurements in M. cedrosensis, 37.3-42.0 and 9.9-10.9, respectively). In the tentatively referred ulnae, the olecranon extends be- yond the shaft in anconal contour, and there is a deep depression adjacent to the olecranon on the internal side as in M. cedrosen- sis. The depression is also present in the ulnae of M. milleri, but the olecranon is less protruded in the smaller species. The four Oceanside specimens fall within the size range of M. cedrosensis in length (28.3-30.7 mm; M. cedrosensis 28.3-21.7 mm) but are somewhat more slender. Mancalla sp. indet. REFERRED MATERIAL. From locality UCMP V68106, in- complete left coracoid (UCMP 88614). From locality LACM 4301, 20 fragmentary humeri (LACM 1 19167-1 19171, 1 19268, 119270, 119271, 119274-119278, 119302-119306, 119407, 121530); 6 incomplete ulnae (LACM 1 19282, 1 19284, 1 19285, 119289, 119311, 119408); 2 carpometacarpi (LACM 119291, 121530); 9 fragments of coracoids (LACM 1 19223, 119253- 119259, 119307); 6 proximal fragments of scapulae (LACM 1 19172, 1 19261, 119263, 119266, 1 19267, 119373); 6 fragments of tibiotarsi (LACM 119173, 1 19178, 1 19294-1 19296, 1 19371, 121531); 3 fragments of tarsometatarsi (LACM 1 19297, 1 1 9299, 1 1 9300); 1 wing phalanx (LACM 1 1 9293); 2 pedal pha- langes (LACM 1 19308, 1 19309); 1 thoracic vertebra (LACM 1 19252). DISCUSSION. These fragments probably represent one or more of the species herein specifically assigned with the genus, but it would be ill-advised to attempt to assign them to species. CONCLUSIONS This study of the avifaunas of the San Mateo Formation at Oceanside has resulted in the recognition of two new species of the family Alcidae, Uria paleohesperis and Cepphus olsoni, from the San Luis Rey River Local Fauna in the lower level of the formation. Both species add support to the tentative earlier rec- ord (Howard, 1978) of the genera Cepphus and Uria from the slightly older Upper Miocene Monterey Formation located far- ther north, in Orange County, California. The only other con- 1 2 Contributions in Science, Number 341 Howard: Tertiary Marine Birds firmed Tertiary record of the murres, genus Uria, is from the late Miocene Sisquoc Formation, near Lompoc, California (Howard, 1981). The guillemots, genus Cepphas, have no prior fossil re- cord. A loon (genus Gavia), an albatross (genus Diomedea) and a murrelet (genus lAethia) are also known from both the Monterey Formation of Orange County and the lower part of the San Mateo Formation, although there is no proof of specific identities. Also significant is the occurrence of the extinct flightless auk genus Praemancalla in both the lower level of the San Mateo Formation at Oceanside and in the Monterey Formation in Or- ange County. The partial skeleton from Oceanside is assigned to P. cf. P. wetmorei, the species described from the uppermost horizon of the Monterey Formation at locality LACM 6906. The Oceanside specimen includes wing elements, previously re- corded for the species, associated with the first tibiotarsus and tarsometatarsus known for the genus. In addition, two isolated femora are the first of this element to be assigned to Praemancalla. In the Lawrence Canyon Local Fauna, in the upper part of the San Mateo Formation at Oceanside, the predominance of the genus Mancalla is typical of the marine Pliocene of California. Although it was impossible to specifically identify all of the more than 80 specimens recorded here, three species are noted: M. milleri and M. diegensis (the predominant species of the late Pliocene, San Diego Formation) and M. cedrosensis of the latest Miocene to early Pliocene (Repenning and Tedford, 1977) of Cedros Island, Baja California, Mexico. This is the first record of the association of these three species. The additional fragmen- tary records from the upper beds add little to the picture, except that the presence of eagle and falcon suggests near-shore deposition. The occurrence of the highly specialized flightless auk genus Mancalla in the upper part of the San Mateo Formation and the related, but more primitive Praemancalla in the lower part is important in providing information as to the relative ages of the beds at the two levels. Mancalla is the most abundantly repre- sented genus of fossil birds in the Pliocene marine formations (including part of the Capistrano Formation) of the southwest coast. One specimen has also been recorded (Howard, 1970) from Humboldt County in northern California, in beds pur- ported to be as young as Pleistocene in age (Kohl, 1974) with an amino-acid age estimate of 470,000 years B.P. (Wehmiller et al., 1978). Praemancalla, on the other hand, has been recorded pre- viously only from the late Miocene, Monterey Formation. If Ved- der’s ( 1 972: 1 67) postulation that the San Mateo Formation may be a channel deposit within the Capistrano Formation is correct, it is not surprising to find Praemancalla in the lowermost beds. For, as he also notes (Vedder, 1972: 165-1 66), in its type-area, the Capistrano Formation is in gradational contact with the underly- ing Monterey Formation, indicating that there may be no hiatus between these formations. It would appear that the interval be- tween the lower and upper stratigraphic levels of the San Mateo Formation could be of considerably greater temporal extent, al- lowing time for the demise of the Praemancalla and the evolution of Mancalla. As the upper part of the Monterey Formation is of Clarendon- ian age, ranging roughly from 8.5 to 1 2 million years B.P, and the lower parts of both the San Mateo and Capistrano Formations are early Hemphillian in age, the geochronologic age of Prae- mancalla is now extended from the Clarendonian into the Hemphillian age. ACKNOWLEDGMENTS The donation of extensive fossil collections by Brian Brockmeier and Larry Danielson to SDSNH and LACM respec- tively greatly aided this study. My thanks are extended to the following persons and institu- tions for their cooperation during the preparation of this paper: Storrs L. Olson (USNM) for discussion regarding fossil Alcidae and for the selection and loan of appropriate comparative mate- rial from the U.S. National Museum collections; Robert M. Chandler (SDSNH) for the opportunity to examine material at the San Diego Natural History Museum and for the loan of fossil material; the Museum of Paleontology and the Museum of Ver- tebrate Zoology, University of California, Berkeley, for the loan of fossil and recent skeletal material; Lawrence G. Barnes (LACM) for critical comments regarding the manuscript and for his readiness to confer on matters pertaining to the geologic as- pects of the California fossil deposits; Kenneth E. Campbell, Jr. (LACM) for helpful suggestions and for assistance in pho- tographing the specimens; Henry Anson Wylde, my husband, for preparing the layout for the illustrations. I am continually grateful for the opportunity to study the avian fossil collections at the Natural History Museum of Los Angeles County and for the cooperation of the staff. The photo- graphs were taken by Richard Meier, Staflf Photographer, the figures were prepared by Mary Butler, Illustrator. LITERATURE CITED Audubon, J.L. 1 839. Ornithological biography, or an account of the habits of the birds of the United States of America. 5 vols. Edinburgh (vol. 5, p. 327). Barnes, L.G., H. Howard, J.H. Hutchison, and B.J. Welton. 1981. The vertebrate fossils of the marine Cenozoic San Mateo Formation at Oceanside, California. Pages 53-70 in PL. Abbott and S. Dunn, eds., Geologic investigations of the San Diego Coastal Plain. San Diego Association of Geologists, San Diego. Brisson, M.J. 1 760. Ornithologia sive synopsis methodica sistens Avium divisionem in ordines. 6 vols. Paris (vol. I , p. 52 and vol. 6, p. 70). Brodkorb, P. 1953. A review of the Pliocene loons. The Condor 55:21 1-214. 1955. The avifauna of the Bone Valley Formation. Florida Geological Survey, Report of Investigations No. 1 4, 57 pp. 1963. Catalogue of fossil birds, part 1 (Archaeop- terigiformes through Ardeiformes). Bulletin of the Florida State Museum, Biological Sciences 7(4): 1 79-293. 1964. Catalogue of fossil birds, part 2 (Anseriformes through Galliformes). Bulletin of the Florida State Museum, Biological Sciences 8(3): 195—335. _ _ . 1967. Catalogue of fossil birds, part 3 (Ralliformes, Contributions in Science, Number 341 Howard: Tertiary Marine Birds 1 3 Ichthyornithiformes, Charadriiformes). Bulletin of the Flor- ida State Museum, Biological Sciences 1 1 ( 3 ):99— 220. Briinnich, M.T. 1764. Ornithologia Borealis. 80 pp. Flafnioe. Forster, J.R. 1788. Enchiridion Historiae Naturali. 224 pp. Haloe. Gmelin, J.F. 1789. Systema Naturae, ed. 1 3, vol. 1 , pt. 2, pp 233— 1032. Lipsiae. Gray, G.R. 1844. The genera of birds; comprising their generic characters and an extensive list of species, 3 vols. London (vol. 1, pt. 1). Howard, H. 1936. A new fossil bird locality near Playa del Rey, California, with a description of a new species of sulid. The Condor 38:21 1-214. 1944. A Miocene hawk from California. The Condor 46:236-237. 1949. New avian records for the Pliocene of Califor- nia. Contribution to Paleontology, Carnegie Institution of Washington Publication 584:177-199. . 1966a. A possible ancestor of the Lucas Auk (Family Mancallidae) from the Tertiary of Orange County, California. Los Angeles County Museum Contributions in Science 101:1 — 8. . 1966b. Additional avian records from the Miocene of Sharktooth Hill, California. Los Angeles County Museum Contributions in Science 1 14:1-1 1 . 1968. Tertiary birds from Laguna Hills, Orange County, California. Los Angeles County Museum Contribu- tions in Science 142: 1 -2 1 . 1970. A review of the extinct avian genus, Mancalla. Los Angeles County Museum Contributions in Science 203:1- 12. 1971. Pliocene avian remains from Baja California. Los Angeles County Museum Contributions in Science 217:1 — 17. 1976. A new species of flightless auk from the Miocene of California (Alcidae: Mancallinae). Pages 141-146 in S.L. Olson, ed.. Collected papers in avian paleontology honoring the 90th birthday of Alexander Wetmore. Smithsonian Contri- butions to Paleobiology 27. 1978. Late Miocene marine birds from Orange County, California. Natural History Museum of Los Angeles County Contributions in Science 290:1-26. 1981. A new species of murre, genus Uria, from the late Miocene of California (Aves: Alcidae). Bulletin of the Southern California Academy of Sciences 80( 1 ): 1 -1 2. Kohl, R.F. 1974. A new late Pleistocene fauna from Humboldt County, California. The Veliger 1 7(2):2 1 1-219. Lawrence, G.N. 1858. In Baird, Cassin, and Lawrence, Cata- logue of North American birds. Reports of explorations and surveys to ascertain the most practicable and economical route for a railroad from the Mississippi River to the Pacific Ocean, vol. 9, pt. II, pp. 889-890. Linnaeus, C. 1758. Systema Naturae, ed. 10, vol. 1, v + 823 pp. Lucas, F.S. 1901 . A flightless auk, Mancalla californiensis, from the Miocene of California. Proceedings of the United States National Museum 24:133-134. Lydekker, R. 1891. Catalogue of the fossil birds in the British Museum (Natural History). British Museum of Natural His- tory, London, xxvii + 368 pp. Marsh, O.C. 1870. Notice of some fossil birds from the Cre- taceous and Tertiary formations of the United States. Ameri- can Journal of Science series 2, vol. 49( 1 46):205— 2 1 7. _. 1872. Notice of some new tertiary and posttertiary birds. American Journal of Science series 3, vol. 4(22):256- 262. Merrem, B. 1788. Versuch eines Grundrisses zur allegemeinen Geschichte und naturlichen Eintheilung der Vogel, Bd. 1:7, 1 3, 20, Leipzig. Merriam J.C. 1919. Tertiary mammalian faunas of the Mojave Desert. University of California Publications, Department of Geological Sciences Bulletin 1 1:437-585. Miller, L. 1 935. New bird horizons in California. Publications of the University of California at Los Angeles in Biological Sci- ences 1 ( 5 ): 7 3—80. . 1937. An extinct puffin from the Pliocene of San Di- ego, California. Transactions of the San Diego Society of Nat- ural History 8(29):375-378. 1946. The Lucas Auk appears again. The Condor 48:32-36. . 1 962. A new albatross from the Miocene of California. The Condor 64(6) 47 1-472. Miller, L , and H. Howard. 1949. The flightless bird Mancalla. Contributions to Paleontology, Carnegie Institution of Wash- ington Publication 584:201-228. Oberholser, H.C. 1899. Some untenable names in ornithology. Proceedings of the Academy of Natural Sciences of Phila- delphia 51 : 20 1 2 1 6. Olson, S.L. 1981. A third species of Mancalla from the late Pliocene San Diego Lormation of California (Aves: Alcidae). Journal of Vertebrate Paleontology 1(1 ):97— 99. Olson, S.L., and D.G. Gillette, 1978. Catalogue of type speci- mens of fossil vertebrates. Academy of Natural Sciences, Phil- adelphia. Part III: Birds. Proceedings of the Academy of Natural Sciences of Philadelphia 1 29( 7):99— 1 00. Pallas, PS. 1769. Spicilegia Zoologica, 2 vols. Berlin (vol. 1 ( 5 ):28 and 33). 1811. Zoographia Rosso-Asiatica, vol. 2, p. 348. Pontoppidan, E. 1763. Danske Atlas e Kongeriget Dannemark med dets naturliche Egenskater. 3 vols. Kiob (vol. 1 :621). Portis, A. 1888. Contribuzioni alia Ornitolitologia Italiana, Teil 2. Memorie della R. Accademia della Scienze di Torino ser. 2. 38:181-203. 1891. Gli ornitoliti del Valdarno superiore e di alcune altra localita plioceniche di Toscana. Memorie del Istituto su- periore di perfeziamento. Florence. 20 pp. Regalia, E. 1902. Sette uccelli pliocenici del Pisano e del Val- darno superiore. Palaeontographia italica VI 1 1:2 19-238. Repenning, C.A., and R.H. Tedford. 1 977. Otarioid seals of the Neogene. United States Geological Survey Professional Paper 992:1-93. Tunstall, M. 1771. Ornithologia Britannica, pp. 6 fol. London. Vedder, S.D. 1 972. Review of the stratigraphic names and mega- 14 Contributions in Science, Number 341 Howard: Tertiary Marine Birds faunal correlation of Pliocene rocks along the southeast mar- gin of Los Angeles Basin, California. Pages 158- 172 in Proceedings of the Pacific Coast Miocene Biostratigraphic Symposium: Pacific Section Society of Economic Paleontolo- gists and Mineralogists, Bakersfield, California. Wehmillcr, J.F., k.R. Lajoie, A.M. Sarno-Wojcicki, R.F. Yerkcs, C.L. Kennedy, T.A. Stephens, and R.F. Kohl. 1978. Amino- acid racemization dating of Quaternary mollusks. Pacific Coast United States. Pages 445-448 in R.E. Zartman, ed.. Short papers of the Fourth International Conference, Geo- chronology, Cosmochronology, lsotypc Geology, United States Department of Interior Geological Survey Open-File Report 78-701. Wetmore, A. 1940. Fossil bird remains from tertiary deposits in the United States. Journal of Morphology 66( 1 ):25— 37. _ _. 1943. Fossil birds from the Tertiary deposits of Florida. Proceedings of the New England Zoological Club 22:59-68. Wilksinon, FEE. 1969. Description of an Upper Miocene al- batross from Beaumaris, Victoria, Australia, and review of fossil Diomedeidae. Memoirs of the National Museum of Vic- toria 29:41-51. Woodford, A.O. 1925. The San Onofre Breccia, its nature and origin. University of California Publications, Bulletin of the Department of Geological Sciences 15:159-280. Submitted 24 September 1981; accepted for publication 10 March 1982. Contributions in Science, Number 341 Howard: Tertiary Marine Birds 15 LARGE ARCMIBENTHAL GASTROPODS OF CENTRAL CHILE: COLLECTION'S FROM AN EXPEDITION OF THE James H. McLean and Hector Andrade V. still ( \ :v . ; iipl It m i|rai : , ’’V \ L: ■ .!?r\: 1$:: ■ li'ip-i 7 T;jd? ••• / : ■ mm ni • * , 1 ' ji '! ”• , 1 f H . ' ;:.:i Natural History Museum of Los Angeles County « ]&$ Angeles, California 90007 i MMMiiwiiir'’ « WSffiAl SERIAL PUBLICATIONS OF THE NATURAL HISTORY MUSEUM OF LOS ANGELES COUNTY The scientific publications of the Natural History Museum of Los Angeles County have been issued at irregular intervals in three major series; the articles in each series are numbered individually, and numbers run consecutively, regardless of subject matter. * Contributions in Science, a miscellaneous series of technical papers describing original research in the life and earth sciences. * Science Bulletin, a miscellaneous series of monographs describing original research in the life and earth sciences. This series was discontinued in 1978 with the issue of Numbers 29 and 30; monographs are now published by the Museum in the Contributions in Science series. * Science Series, long articles on natural history topics, generally written for the layman. Copies of the publications in these series are sold through the Museum Bookshop. Leon G. Arnold Lawrence G. Barnes Robert L. Bezy Kenneth E. Campbell EDITORIAL BOARD Robert Gustafson John M. Harris Charles L. Hogue Robert J. Lavenberg Camm C. Swift David P. Whistler Edward C. Wilson John W. Wright ■ EDITOR Robin A. Simpson li [wWfcj §1 Printed at The Castle Press, Pasadena, California 11 LARGE ARCHIBENTHAL GASTROPODS OF CENTRAL CHILE: COLLECTIONS FROM AN EXPEDITION OF THE R/V ANTON BRUUN AND THE CHILEAN SHRIMP FISHERY1 James H. McLean2 and Hector Andrade V.3 ABSTRACT. Fifteen species of large gastropods from off central Chile collected both by an expedition of the R/V ANTON BRUUN in 1966 and by commercial trawling for the shrimp Heterocarpus reedi are treated. Nine previously described species are reviewed: Bathybembix macdonaldi, B humboldti, Calliostoma chilena, Capulus ungaricoides, Fusitriton magellanicus, Aeneator fontainei, A. loisae, Miomelon alar- coni, and Ptychosyrinx chilensis. Six species are described as new: Di- odora codoceoae, Calliostoma delli, Trophon bahamondei, Colum- barium tomicici, Aeneator castiilai , and Cancellaria stuardoi. RESUMEN. Se estudiaron quince especies de macrogastropodos ob- tenidos en faenas de pesca camaronera por arrastre y por la expedicion del B/l ANTON BRUUN en la zona central de Chile. De estas, nueve especies habian sido ya descritas y se hace la revision de cada una de ellas: Bathybembix macdonaldi, B. humboldti, Calliostoma chilena, Capulus ungaricoides, Fusitriton magellanicus, Aeneator fontainei, A. loisae, Miomelon alarconni y Ptychosyrinx chilensis. Las seis restantes son consideradas como nuevas y se entregan sus descripciones: Diodora codoceoae, Calliostoma delli, Trophon bahamondei, Columbarium tomicici, Aeneator castiilai y Cancellaria stuardoi. INTRODUCTION The marine invertebrate fauna of the lower continental shelf and slope off central Chile is poorly known. This region is one of the few areas of the world not covered by the great expeditions of the last century. In recent years, however, collections from two sources have become available. An expedition of the R/V AN- TON BRUUN sampled the region in 1966, and, in subsequent years, the shrimp fishery for Heterocarpus reedi Bahamonde, 1955, has yielded abundant material of the larger species. Over the last 14 years, six mollusks commonly taken by the shrimp fishery have been described from central Chile: Berry (1968) described Ptychosyrinx chilensis-, Rehder (1971) de- scribed Limopsis ruizana, Bathybembix humboldti, Calliostoma chilena, and Aeneator loisae-, Stuardo and Villarroel (1974) de- scribed Miomelon alarconi. Other new species have been recog- nized by marine biologists in Chile but have remained undescribed until now. From 1976 through 1980, Andrade obtained extensive mate- rial of mollusks and other invertebrates incidental to the shrimp Contributions in Science, Number 342, pp. 1-20 Natural History Museum of Los Angeles County, 1982 fishery. Specimens were saved by crew members of the trawling vessel GODEN WIND and other vessels based in Quintero. These vessels worked the Chilean coastline to the north and south of Valparaiso between Los Vilos (31°56'S) and Constitucion (35°20'S). The following contributions treating the echinoderms and crustaceans from the Chilean shrimp fishery have been pub- lished: Andrade (1980), Andrade and Baez (1977, 1980), Baez and Andrade (1977, 1979), Codoceo and Andrade (1978, 1980, in press), Codoceo et al. (1978), Revuelta and Andrade (1978). The offshore fauna of northern Peru has been sampled by expe- ditions of the ANTON BRUUN and by expeditions conducted by the Instituto del Mar, Callao, Peru. These expeditions have produced specimens of some of the species known from central Chile. The Peruvian records of these species are given in this paper. The present paper is limited to the large gastropods of central Chile that have been taken by the shrimp fishery. Nine previously described species are reviewed, and six additional species are described. For most of the species, we have included a photo- graph of the radular ribbon as an aid in identification. Full discus- sion of radular features is beyond the scope of this paper. A more detailed study of radular structure would require the use of scan- ning electron microscopy. OCEANOGRAPHIC CONDITIONS OFF CENTRAL CHILE Depths of samples reported on here from off central Chile are archibenthal, corresponding to the upper part of the continental slope. The benthic fauna is mainly under the influence of Antarc- 1. Review committee for this contribution: William K. Emerson, Robert .1 Lavenberg, and Joseph Rosewater. 2. Malacology Section, Natural History Museum of Los Angeles County, Los Angeles, California 90007. 3. Instituto de Oceanologia, Universidad de Valparaiso, Casilla 1 3-D, Vina del Mar, Chile (present address: Station Marine d’Endoume, Rue de la Batterie des Lions, 13007 Marseille, France). ISSN 0459-8113 tic Intermediate Water and to a lesser degree of the overlaying Equatorial Subsurface Water, as discussed by Andrade and Baez (1980). Detailed accounts of the physical and chemical proper- ties of the water masses in central Chile are given by Sievers and Silva (1975) and Silva and Sievers (1981). MATERIALS AND STATION DATA Station data for the collections reported in this paper are given in Table 1 . Localities and depths for material from the shrimp fish- ery are necessarily less accurate than data for material collected by scientific expeditions. Depths and coordinates are not re- peated in the text, except for type localities of the new species. For lots that define the northern and southern records, the corre- sponding latitudes are given in the distribution heading. Abbreviations of museums mentioned in the text are as fol- lows: BM(NH), British Museum (Natural History); LACM, Los Angeles County Museum of Natural History, Los Angeles, Cal- ifornia, U.S.A.; MNHN, Museo Nacional de Historia Natural, Santiago, Chile; MZICB, Instituto Central de Biologia, Univer- sidad de Concepcion, Concepcion, Chile; USNM, United States National Museum of Natural History, Washington, D.C., U.S.A. Type material of the six new species described herein is dis- tributed among these institutions. Collections upon which this report is based have been obtained from four sources: Table 1. ANTON BRUUN, SNP-1, and Chilean shrimp fishery stations arranged north to south. Location Depth (m) PERU W of Lobos de Tierra, SNP-1, sta. 26 06 26 S 81 05 W 1,025 S of Lobos de Tierra, SNP-1, sta. 25 06 42 80 59 785 S of Lobos de Afuera, SNP-1 , sta. 13 07 07 80 46 1,200 N of lsla Macabi, SN P- 1 , sta. 28 07 44 80 30 750-760 W of lsla Macabi, ANTON BRUUN, sta. 754 07 49 80 38 605-735 CHILE Junquiilar, ANTON BRUUN, sta. 714 25 00 70 40 950 Punta Mar Brava, ANTON BRUUN, sta. 710 29 21 71 25 65-95 Punta Hornos, ANTON BRUUN, sta. 709 29 38 71 21 110 Coquimbo, trawler not identified* 29 58 — 200-400 Los Vilos, GODEN WIND 31 56 71 54 240-400 Pichidanqui, GODEN WIND 32 08 71 54 300-400 Punta Salinas, ANTON BRUUN, sta. 703 32 09 71 43 960 Caleta Molles, ANTON BRUUN, sta. 702 32 17 71 40 580 Papudo, GODEN WIND 32 31 71 54 200-350 Zapallar, GODEN WIND 32 33 71 43 350-450 Quintero, GODEN WIND 32 42 71 48 200-280 Algarrobo, trawler not identified 33 22 71 55 360 Punta Penablanca, ANTON BRUUN, no sta. 33 22 71 54 260-280 Punta Panulcillo, ANTON BRUUN, sta. 701 32 32 71 35 180-175 Puerto San Antonio, ANTON BRUUN, sta. 699 33 39 72 10 1,170-1,480 Punta Toro, trawler not identified 33 06 72 03 270 Bahia Navidad, ANTON BRUUN, sta. 686 33 58 72 05 140 Topocalma, trawler not identified 34 06 72 14 180-360 Punta Topocalma, ANTON BRUUN, sta. 687 34 07 72 19 750-730 Piehilemu, trawler not identified 34 27 72 24 240-350 Mataquito, trawler not identified 35 01 72 10 300 Constitucion, trawler not identified 35 20 72 55 260 Cabo Carranza, ANTON BRUUN, sta. 697 35 27 73 01 290-450 * For Coquimbo only, the locality refers to the home port, not the actual station, which is unknown. 2 Contributions in Science, Number 342 McLean and Andrade: Gastropods of central Chile Figures 1 and 2, Diodora codoceoae new species. Figure 1, three views of holotype, LACM 1 979, 260 m off Constitucion, Chile, length 40.7 mm. Figure 2, LACM 72485, 360 m off Topocalma, Chile, Length 40.6 mm. 1. Cruises of the ANTON BRUUN in Peru and Chile during 1966. This material is deposited in the LACM and USNM. 2. A cruise of the Peruvian naval vessel SNP-1 to the vicinity of the Lobos Islands in northern Peru in January 1974 on which McLean was present. Material from that cruise is deposited in the LACM and the Instituto del Mar, Callao, Peru. 3. Shrimp trawling localities (GODEN WIND and other ves- sels) in central Chile, 1977-1980. McLean visited the Montemar Marine Laboratory of the Universidad de Valparaiso in August 1978 and examined all of the collections then received by An- drade. This material is now deposited in the LACM, the MNHN, and the Museo Comparativo, Instituto de Oceanologia, Univer- sidad de Valparaiso. 4. Shrimp trawling vessels based in Coquimbo, Chile, 1964- 1975; collections deposited at LACM by the late Harvey McMillin, Luis Ferreira Osses, and Jorge Tomicic K. In some cases, Coquimbo (29°58'S) is cited as the northern range limit of a species, although it should be understood that the actual occur- rence for such a record could have been either to the north or south of Coquimbo. SYSTEMATIC ACCOUNT Superfamily Fissurellacea Family Fissurellidae Subfamily Diodorinae Genus Diodora Gray, 1821 Type species (monotypy): Patella apertura Montagu, 1803 ( = P. graeca Linnaeus, 1758). Recent, Europe. Diodora species occur in most temperate and tropical regions of the world in intertidal, sublittoral, and, less frequently, archi- benthal depths. Although there is considerable diversity of form among the species, an adequate subgeneric classification has never been offered. Diodora codoceoae new species Figures 1-3 DESCRIPTION. Shell large for the genus, thin, conical, basal outline elongate-oval, anterior end slightly narrower than pos- terior, basal margin lying flat, with ends slightly elevated. Lateral slopes straight, anterior and posterior slopes slightly concave. Foramen nearly central, oval, length of foramen about 10% of shell length, highest point of shell at the posterior rim of the foramen. Periostracum thin, brown, worn away over most of the surface; underlying surface chalky, worn near summit. Sculpture of numerous fine radial ribs; under magnification, the ribs are rounded, broader than the interspaces, crossed by raised growth lamellae to give a minute beaded appearance; primary ribs re- main stronger so that every fourth rib is slightly more prominent than the others. Interior chalky white, the margin finely crenu- lated and grooved by the ribs; position of the stronger primary ribs marked by corresponding internal grooves. Under magnifi- cation, the crossed-lamellar aragonitic structure near the margin is visible. Muscle scar weakly discernible. Callus surrounding the foramen has an oval outline and is slightly truncated and depressed at its posterior edge. Thickness of shell at summit (depth of foramen) approximately equal to the width of the inter- nal callus. Dimensions; length 40.7 mm, width 27.0 mm, height 1 1.4 mm (holotype. Fig. 1 ); length 40.6 mm, width 26.9 mm, height 13.7 mm (Fig. 2). Animal of preserved specimen (Fig. 2): colorless except for darkly pigmented eyes; contracted to fit completely within the shell; uppermost fold of mantle with a very finely scalloped edge, surface of the mantle folds and the foot sides with negligible development of papillae; cephalic tentacles and epipodial tenta- cles well developed. Contributions in Science, Number 342 McLean and Andrade: Gastropods of central Chile 3 Radula (Fig. 3): typical of the genus, laterals four, rachidian three times wider than the adjacent laterals; first marginal bi- cuspid as in other species of Diodora. MATERIAL. Chile: Pichidanqui, Zapallar, Topocalma (LACM, Fig. 2), Constitucion (LACM, holotype. Fig. 1 ). Speci- mens examined: 1 5. TYPE MATERIAL. Three specimens from the type locality, col- lected 25 March 1976, by Andrade, shrimp trawler GODEN WIND. Holotype, LACM 1979; one paratype MNHN 200488, one paratype MZICB 15.527. TYPE LOCALITY. 260 m off Constitucion, Chile (35°20'S, 72°55'W). DISTRIBUTION. Piehidangui (34°06'S) to Constitucion, Chile (35°20'S). Depth range 180-360 m. DIAGNOSIS. A species of Diodora characterized by its large size, thin shell, oval aperture, and numerous radial ribs. It most resembles D. tanneri (Verrill, 1883), known from archibenthal depths in the western North Atlantic, from Delaware to the Ca- ribbean (see Perez-Farfante, 1943:19). D. codoceoae differs in having a less conical profile, a proportionately larger foramen, and a more chalky shell surface. We have no information about the animal of D. tanneri. Diodora codoceoae also resembles Stromboli beebei (Hertlein and Strong, 1951), an offshore spe- cies in the tropical Panamic Faunal Province ranging from the Gulf of California to Ecuador (see McLean in Keen, 1971:318, fig. 29). Unlike D. codoceoae. S. beebei has a fleshy mantle that fully envelops the edge of the shell and radial ribs that are uni- formly fine, with no distinction between primary and secondary ribs; in addition, S. beebei has a sturdier shell than that of D. codoceoae. REMARKS. The internal callus surrounding the foramen is curved on the posterior side in D. codoceoae. D. tanneri , and 5. beebei , not straight-edged as in most species of Diodora. How- ever, this is a variable feature among species of Diodora , particu- larly those with oval foramena. ETYMOLOGY. We are pleased to name this species in honor of Prof. Maria Codoceo of the Museo Nacional de Historia Natu- ral, Santiago. Superfamily Trochacea Family Trochidae Subfamily Eucyclinae Genus Bathybembix Crosse, 1893 Type species (original designation): Bembix aeola Watson, 1 879. Recent, Japan. Bathybembix species are large mud-ingesting trochids occurring on the outer continental shelf and slope (Merriman, 1967; Hick- man, 1981). They are members of the subfamily Eucyclinae, which appeared in the Mesozoic. Such recent genera as Bathybembix Crosse, 1893, Calliotropis Seguenza, 1903, and Figure 3, Diodora codoceoae new species, radular dentition of holo- type, rachidian and lateral teeth, maximum dimension 0.8 mm. Cidarina Dali, 1909, have been recognized as members of the same group (McLean, 198L335).4 The genus Bathybembix occurs in the northern Pacific, with several species in Japan, and one, B. bairdii (Dali, 1889), in the northeastern Pacific ranging from the Bering Sea, Alaska, to the Gulf of Tehuantepec, Mexico (McLean in Keen, 1971:331, fig. 62). The two species treated here are common off the coasts of both Peru and Chile. These species are members of the same typical subgenus in which the shells are large, the umbilicus closed, and the periostracum thick and colored greenish brown. Bathybembix macdonaldi (Dali, 1890) Figures 4-6 Turcicula macdonaldi Dali, 1890:348, pi. 7, fig. 7; 1908:349, pi. 19, fig. 7. Bathybembix macdonaldi , McLean in Keen, 1 97 1:331, fig. 63. DESCRIPTION. Shell large, thin, high-spired; final whorl rounded, early whorls strongly carinate. Periostracum thin, yel- lowish or greenish brown. Protoconch missing, teleoconch whorls eight on intact specimens; early whorls often eroded. Spiral sculpture of two cords per whorl, a prominent projecting pe- ripheral cord, with short projecting spines, and another weaker, noded cord just below the suture, becoming weaker and disap- pearing on the final whorl. Base with prominent spiral cords. Axial sculpture lacking except for fine growth increments on the periostracum. Outer lip thin, interior nacreous; operculum large, multispiral. Dimensions: height 69.3 mm, diameter 49.0 mm (Fig. 4); height 75 mm, diameter 60 mm (holotype. Fig. 5). Radula (Fig. 6): rachidian tooth broadly flanged, overhanging tip bearing numerous fine serrations along the sides; lateral teeth three, marginal teeth numerous. MATERIAL. Peru: W of Lobos de Tierra (LACM), S. of Lobos 4. McLean (1981) used the subfamily Amberleyinae for this group. That name was proposed at the family level by Wenz, 1938. The older Eucyclinae must be used, a name proposed at the family level by Koken, 1896. 4 Contributions in Science, Number 342 McLean and Andrade: Gastropods of central Chile de Tierra (LACM), S. of Lobos de Afuera (LACM). Chile: Jun- quillar (LACM), Los Vilos (LACM), Punta Salinas (LACM, fig. 4), Caleta Molles (LACM), Papudo, Zapallar, Quintero, Algar- robo, Punta Penablanca (LACM), Punta Topocalma (LACM). Specimens examined: more than 100. TYPE MATERIAL AND TYPE LOCALITY. Holotype, USNM 96559 (Fig. 5). Type locality: 401 fm (733 m) off Manta, Ecuador. DISTRIBUTION. Off Punta Mala, Panama (LACM 10369) (7°28'N) to Punta Topocalma, Chile (34°7'S). As is the case with B. humboldti , northern specimens occurring from Panama to Peru are known only from depths of 780-1,200 m, whereas Figures 4 through 9, Bathybembix macdonaldi and B. humboldti. Figures 4 through 6, B. macdonaldi. Figure 4, LACM 66-1 57, 960 m off Punta Salinas, Chile, height 69.3 mm. Figure 5, holotype, USNM 96559, 733 m off Manta, Ecuador, height 75 mm. Figure 6, radula ribbon, LACM 66-1 52, 750-730 m off Punta Topocalma, Chile, width of field 0.6 mm. Figures 7 through 9, B. humboldti. Figure 7, radula ribbon, LACM 66-171, 605-735 m, W of Isla Macabi, Peru, width of field 0.7 mm. Figure 8, LACM 66-154, 1 ,1 70-1 ,480 m off Puerto San Antonio, Chile, height 61.7 mm. Figure 9, two views, LACM 66-1 71, 605-735 m W of Isla Macabi, Chile, height 39.7 mm. Contributions in Science, Number 342 McLean and Andrade: Gastropods of central Chile 5 those from Chile occur in similar depths and also as shallow as 200 m. REMARKS. Bathybembix macdonaldi differs from B. hum- boldti in having strong peripheral nodes and a much lighter col- ored periostracum. Northern specimens tend to have a fainter subsutural cord and more numerous peripheral nodes, as in the holotype (Fig. 5). Bathybembix humboldti Rehder, 1971 Figures 7-9 Bathybembix humboldti Rehder, 1971:578, fig. 4. DESCRIPTION. Shell large, thin, suture deeply impressed, whorls rounded. Periostracum thin, brown, varying from yellow- ish brown to dark brown. Protoconch missing, teleoconch whorls seven on perfect specimens, but usually the early whorls are missing and sealed over with a chalky shell layer. Axial sculpture of low, narrow ribs on the upper half of the whorl, base nearly smooth except for broadly spaced, faintly marked spiral cords. Sculpture of three spiral cords on early whorls, forming elongate nodes where intersecting the axial ribs; one cord just below the suture and two others closer together at the periphery. Aperture nearly circular, outer lip thin, interior nacreous; operculum multispiral. Dimensions: height 61.7 mm, diameter 46.9 mm (Fig. 8); height 39.7 mm, diameter 40.8 mm (Fig. 9); height 53.1 mm, diameter 44.3 mm (holotype). Radula (Fig. 7): similar to that of B. macdonaldi. MATERIAL. Peru: N of Isla Macabi (LACM), W of Isla Ma- cabi (LACM, Fig. 9). Chile: Coquimbo (LACM), Los Vilos, Cal- eta Molles (LACM), Papudo, Zapallar, Algarrobo, Puerto San Antonio (LACM, Fig. 8), Topocalma, Pichilemu. Specimens ex- amined: more than 100. TYPE MATERIAL AND TYPE LOCALITY. Holotype, USNM 701665, 200 m, 17-18 km NW of Valparaiso, Chile. DISTRIBUTION. Isla Macabi, Peru (7°44'S), to Pichilemu, Chile (34°27'S). Depth range: 200-1,480 m. In northern Peru, the depth range is 605-760 m; in Chile, the depth range is 200- 1 ,480 m. REMARKS. This species is characterized by its dark brown periostracum and the predominance of axial sculpture. Most specimens have the apical whorls badly eroded, although such specimens may seal the apical area with internally deposited shell layers (Fig. 9). Populations of B. humboldti have uniformly colored periostraca, some darker than others. Subfamily Calliostomatinae Genus Calliostoma Swainson, 1840 Type species (subsequent designation Hermannsen, 1846): Tro- chus conulus Linnaeus, 1758. Recent, Europe. Subgenus Otukaia Ikebe, 1943 Type species (original designation): Calliostoma kiheiziebisu Otuka, 1939. Japan. The calliostomatine trochids have a characteristic radula, modi- fications in the reproductive system, and a channel in the ventral lip — the pseudoproboscis — directed to the right (Fretter and Graham, 1962, fig. 92). Clench and Turner (1960) reviewed the Atlantic species, but a worldwide generic review has not been offered, and there are varying concepts of genera and subgenera in the family. The two species of Calliostoma in the present material are members of the subgenus Otukaia Ikebe, 1943, a group of rela- tively large-shelled species broadly distributed in deep water throughout the world. They are characterized by a silky white surface layer, rather than a variegated color pattern, and sculp- ture of three spiral cords on the early whorls, which may or may not persist on later whorls. Alertalex Dell, 1956 (type species: A. blacki Dell, 1956, from New Zealand), is regarded as a synonym. Calliostoma ( Otukaia ) chilena Rehder, 1971 Figures 10-12 Calliostoma chilena Rehder, 1971:590, figs. 2, 5. DESCRIPTION. Shell large, thin, nonumbilicate; whorls flat to slightly convex, early spire profile slightly concave; color silky white with a gray-green nacreous sheen. Protoconch smooth, rounded, teleoconch whorls nine, early whorls with three spiral cords and axial ribs that produce square cancellations; these cords fade altogether by the fifth whorl. The suture is laid just below the basal keel; base with fine and even spiral cords, more prominent near the columella. Columella thick, slanted, outer lip sharp, thin. Operculum corneous, multispiral. Dimensions: height 31.7 mm, diameter 26.5 mm (Fig. 10); height 32.7 mm, diameter 29.9 mm (Fig. 1 1); height 36.1 mm, diameter 3 1 .9 mm (holotype). Animal: cephalic lappets lacking, left and right neck lobes broad, rolled to form incurrent and excurrent siphons; epipodial tentacles four pairs, two in line with the neck lobes and two adjacent to the operculum; cephalic tentacles broad at the base, eye peduncles short, eyes large. Radula (Fig. 12): the rachidian tooth is more than twice as broad as any of the lateral teeth; there are at least six pairs of lateral teeth, and the serrate, overhanging cusps of the lateral teeth are exceptionally long. MATERIAL. Peru: W of Isla Macabi (LACM, Fig. 11). Chile: Coquimbo (LACM), Los Vilos, Punta Salinas (LACM, Fig. 10), Papudo, Zapallar, Punta Topocalma (LACM). Specimens exam- ined: 12. TYPE MATERIAL AND TYPE LOCALITY. Holotype, USNM 701669, 17 km NW of Valparaiso, Chile, 200 m. DISTRIBUTION. Off Cebaco Island, Gulf of Panama (LACM 10370) (7°32'N), to Punta Topocalma, Chile (34°7'S). Depth range: 200-750 m. The minimum depth in Panama and northern Peru is 560 m; in Chile, the depths range from 200 to 750 m. REMARKS. Calliostoma chilena was described originally from a single specimen. It is still known from rather few specimens. As Rehder noted, C. chilena most closely resembles C. platinum Dali, 1890, which ranges from British Columbia to southern California at depths of 150-700 m. Calliostoma platinum is broader and has convex rather than flat-sided whorls. 6 Contributions in Science, Number 342 McLean and Andrade: Gastropods of central Chile 12 Figures 10 through 15, Calliostoma chilena and C. delli new species. Figures 10 through 12, C. chilena. Figure 10, LACM 66-157, 960 m off Punta Salinas, Chile, height 31.7 mm. Figure 1 1, LACM 66-171 , 605-735 m, W of Isla Macabi, Peru, height 32.7 mm. Figure 12, radular ribbon of specimen in Fig. 10, width of field 0.6 mm. Figures 13 through 15, Calliostoma delli new species. Figure 13, holotype, LACM 1980, 400 m off Los Vilos, Chile, height 29.6 mm. Figure 14, paratype, LACM 1 981a, same locality, height 24.3 mm. Figure 15, paratype, LACM 1981b, same locality, height 29.0 mm. Calliostoma (Otukaia) delli new species Figures 13-15 DESCRIPTION. Shell large, thin, silky white, nonumbilicate, spire whorls convex. Protoconch of one and one-fourth whorls, teleoconch whorls seven, first teleoconch whorl with three spiral cords, noded to produce square cancellations; sculpture chang- ing by the second whorl to three prominently projecting spiral cords, the uppermost cord beaded, the other two smooth. Mature sculpture of three prominent cords, subsutural cord the least prominent, remaining close to the suture and losing its beading by about the third whorl; second cord sharply defined and sepa- rated from the subsutural cord by a broad, smooth area; third cord equally strong and projecting to form the peripheral extent of the whorl. Basal keep sharp; suture laid directly on its lower surface, not forming a channel. Basal cording of about three fine cords on the outer edge and two to three bordering the columellar wall; intermediate area of base smooth except for fine spiral striae. Columellar wall thickened, slanted, forming a spur at the base; outer lip thin. Operculum corneous, multispiral. Animal and radula as in C. chilena. Dimensions: height 29.6 mm, diameter 30.9 mm (holotype. Fig. 13): height 24.3 mm, diameter 23.2 mm (paratype. Fig. 14); height 29.0 mm, diameter 26.0 (paratype, Fig. 15). MATERIAL. Chile: Los Vilos (LACM, type lot. Figs. 13-15), Papudo, Zapallar, Algarrobo, Punta Penablanca (LACM), Pichilemu, Constitucion. Specimens examined: 1 14. TYPE MATERIAL. Thirty-three specimens from the type lo- cality, collected 29 May 1977, by Andrade, shrimp trawler GODEN WIND. Holotype, LACM 1980; paratypes, LACM 1981; paratypes, MNHN 200489; paratypes, MZICB 15.528; paratypes, USNM 784738. TYPE LOCALITY. 400 m off Los Vilos, Chile (31°56'S; 71 °54'W). DISTRIBUTION. Los Vilos (31°56'S) to Constitucion, Chile (35°20'S). Depth range 200-450 m. DIAGNOSIS. A species of the subgenus Otukaia characterized by having three spiral cords prominent at all growth stages. It Contributions in Science, Number 342 McLean and Andrade: Gastropods of central Chile 7 differs from the similarly sculptured C. blacki (Dell, 1956) from New Zealand (see Dell, 1956:46, pi. 7, fig. 6) in being lower spired, and in having a weaker subsutural (first) cord and a stronger second cord. REMARKS Calliostoma delli tends to be broader than high; one of the figured paratypes (Fig. 14) is unusually narrow, compared to most specimens in the type lot. ETYMOLOGY. We are pleased to name this species in honor of Dr. Richard K. Dell of the National Museum of New Zealand, Wellington. Superfamily Hipponicacea Family Capulidae Genus Capulus Montfort, 1810 Type species (original designation): Patella ungaricus Linnaeus, 1767. Recent, Europe. Capulus species are sedentary, protandric, brooding limpets that are usually attached to shells of living bivalves. Most commonly, they are attached to pectinid bivalves, although the European type species may attach to stones or to Turritella (Thorson, 1 965). Some of their nutrition is derived from filter feeding, as in the calyptraeids, but they also use the pseudoproboscis, an openly grooved proboscis, to take food from the host (Fretter and Graham, 1962). Recently, some species have been found to bore holes in the host shell, inserting the pseudoproboscis directly (Orr, 1962; Matsukuma, 1978). The capulid in the present material from central Chile was attached to a large bivalve of the family Limidae, identified as Acesta patagonica (Dali, 1902). This is the second capulid spe- cies associated with a limid rather than a pectinid bivalve. Re- cently, Dell (1978) described Capulus novaezelandiae , attached to an unidentified species of Acesta from New Zealand waters. [For a review of Acesta, see Vokes, 1963]. Capulus ungaricoides (Orbigny, 1841) Figures 16-19 Pileopsis ungaricoides Orbigny, 1 841 :457, pi. 78, fig. 4. Capulus ungaricoides , Dali, 1909:234; Keen, 1966:3, pi. 1, figs. 14a, 14b; Keen, 1971:467, fig. 833. "Capulus chilensis Dali,” Carcelles, 1944:2, fig. 1. DESCRIPTION. Shell large for the genus, thin but sturdy; outline nearly circular, irregular, conforming to the attachment surface. Apex narrow, strongly incurved, overhanging the pos- terior margin. Sculpture of fine radial ribs, rib interspaces con- sisting of narrow grooves. Periostracum and all traces of the radial ribs worn away on apical region but persisting near the margin; periostracum slightly overhanging the shell edge. Inte- rior glossy white; muscle scar horseshoe-shaped. Dimensions: diameter 39.6 mm, height 15.7 mm (Fig. 18); diameter 24.4 mm, height 9.0 mm (Fig. 16). Animal: preserved specimens (Fig. 19) have the brood sac with a folded edge unlike that shown in other published illustrations of Capulus species. Animal not boring into the shell of its host but producing a notch in the shell edge (Fig. 18, left). The specimen in Figure 18 (right) is shown attached to the left valve of A. patagonica, and in Figure 18 (left), the attachment scar area is shown. The area corresponding to the posterior edge of the foot is eroded in the host shell; this is not a result of boring by the Capulus, because it is too far from the position of the head. Radula (Fig. 17): rachidian and lateral strongly cusped and serrate on the edges, marginals uncusped. MATERIAL. Chile: Coquimbo (LACM, Figs. 1 8, 1 9), Los Vilos (LACM, Fig. 16), Zapallar, Quintero. Specimens examined: six. TYPE MATERIAL AND TYPE LOCALITY. Flolotype, BM(NH) 54.12.4.554. Type locality: Paita, Peru. DISTRIBUTION. Paita, Peru (5°5'S), to Isla de Los Estados, Argentina (54°47'S). Depth range off central Chile: 200-450 m. REMARKS. Orbigny’s species Capulus ungaricoides, de- scribed from Paita, Peru, has not been reported (other than in faunal lists) subsequent to its original description. The present material agrees with the original description and with illustra- tions of the holotype given by Keen (1966), except for lacking reddish rays, which may not be significant. Although the mate- rial treated here may not be positively identified with Orbigny’s species, it is premature to distinguish separate species until mate- rial conforming to Orbigny’s holotype can be shown to represent a separate species. Orbigny’s holotype was probably not associ- ated with an Acesta species, but because many species of Ca- pulus are not host-specific, that need not preclude it being con- specific with the present material. Capulus ungaricoides has the fine spiral sculpture present in C. chilensis Dali, 1908, and C. novaezelandiae Dell, 1978, but differs from both in having a narrower, more rapidly expanding apex. Capulus chilensis has regular early coiling of three distinct whorls, very unlike the narrow recurved apex of C. ungaricoides. Type material of Capulus chilensis has not previously been fi- gured. The holotype measures 25 mm in diameter; a smaller paratype specimen is figured here (Fig. 20), to show the coiling of the early whorls. Although Carcelles (1944) reported upon a specimen from Isla de los Estados (east of Tierra del Fuego), Argentina, identi- fied as C. chilensis, his description agrees with the present mate- rial and forms the basis for the distributional record of C. ungaricoides cited above. Superfamily Tonnacea Family Cymatiidae Genus Fusitriton Cossmann, 1903 Type species (monotypy): Triton cancellatus Lamarck, 1816 [ = Neptunea magellanica Roeding, 1798], Recent, Magellanic. Fusitriton is noted for its bipolar distribution; species are com- mon in shallow water in the North Pacific, South Pacific, and South Atlantic Oceans. At lower latitudes, all of the species become deeply submergent, which helps to explain the presence of the genus in cold waters of both the Northern and Southern Hemispheres. The genus has been thoroughly reviewed by Smith (1970), who discussed the broadly distributed species in the present ma- 8 Contributions in Science, Number 342 McLean and Andrade: Gastropods of central Chile Figures 16 through 20, Capulus ungaricoides and C. chilensis. Figures 16 through 19, C. ungaricoides. Figure 16, preserved but retracted animal attached to shell, LACM 72487, 300-400 m off Los Vilos, Chile, diameter 24.4 mm. Figure 17, radular ribbon, LACM 72487, 300-400 m off Los Vilos, Chile, both pairs of marginal teeth on right side of ribbon folded to the right, width of field 0.9 mm. Figure 18, two views, LACM 72488, attached to left valve of Acesta patagonica, right view with Capulus in place, left view showing attachment scar and indentation in host shell corresponding to position of pseudoproboscis, 450 m off Coquimbo, Chile, length of Acesta 107.3 mm, diameter of Capulus 39.6 mm. Figure 19, body of specimen in Fig. 1 8, showing the folded edge of the brood sac. Figure 20, C. chilensis , paratype, USNM 96926, "Albatross” station 2781, 636 m “off the Chilean coast,” diameter 8.4 mm. ferial at length; her synonymy is not repeated here. Although two names for the Chilean species, F. magellanicus and F. cancellatus, have been used inconsistently by subsequent authors, Cer- nohorsky (1977) has recently shown that F. magellanicus , cred- ited to Roeding, 1 798, has priority over F. cancellatus, credited to Lamarck, 1816. Fusitriton magellanicus ( Roeding, 1 798) Figures 21-23 Neptunea magellanica Roeding, 1798:1 16. Fusitriton magellanicus, Cernohorsky, 1977:107, fig. 3. Triton cancellatus Lamarck, 1816:4. Fusitriton cancellatus, Smith, 1970:475, pi. 42, figs. 4-10. DESCRIPTION. Shell large, whorls six, rounded. Periostracum thick and bearing projecting hairs. Varices irregular, more fre- quent in juvenile stages, the mature lip marked by a final varix. Axial sculpture strongest on early whorls, nearly lacking on final whorl. Spiral cords low, with broad, shallow interspaces. Dimensions: height 94.0 mm, diameter 47.3 mm (Fig. 21); height 85.1 mm, diameter 43.6 mm (Fig. 22). Animal: Smith ( 1 970, text fig. 2c) gave a sketch of the mantle cavity. Radula (Fig. 23): typically taenioglossate, rachidian and lat- Contributions in Science, Number 342 McLean and Andrade: Gastropods of central Chile 9 eral finely denticulate. See also Smith (1970, text fig. 3d). MATERIAL. Chile: Los Vilos (LACM), Punta Salinas (LACM), Caleta Molles (LACM, Fig. 21), Pupudo, Zapallar, Algarrobo, Punta Penablanca (LACM), Punta Topocalma (LACM). Specimens examined: 32. TYPE MATERIAL AND TYPE LOCALITY. Type material un- known, type locality presumed to be the Strait of Magellan. DISTRIBUTION. Los Vilos, Chile (31°56'S), to Tierra del Fuego; north in the Atlantic to Sarita, Rio Grande do Sul, Brazil (Smith, 1970). Depth range in central Chile: 180-960 m. REMARKS. Fusitriton magellanicus is the only member of the present assemblage of large gastropods also to occur in relatively shallow water in southern Chile. McLean has collected speci- mens at several localities in the Gulf of Corcovado east of Chiloe Island by diving in depths as shallow as 5 m (Fig. 22). Superfamily Muricacea Family Muricidae Subfamily Trophoninae Genus Trophon Montfort, 1810 Type species (original designation): Murex magellanicus Gmelin, 1791 (= Buccinum geversianum Pallas, 1774). Re- cent, Magellanic. The muricid subfamily Trophoninae, of which Trophon is the type genus, comprises white-shelled forms with axial lamellae rather than varices, open canals, and simple apertures. It is the least known group in the Muricidae; only the type species of recognized genera were treated by Radwin and D’Attilio ( 1 976). Trophon is primarily an austral genus with numerous species in the Antarctic and subantarctic region. The type species, which is relatively large, has both axial and spiral sculpture and is com- mon in the intertidal zone in the Fuegian and Patagonian regions. An illustration given by Radwin and D’Attilio (1976, fig. 130), though identified as Stramonitrophon laciniatus, is actually T. geversianus. Trophon bahamondei new species Figures 24-25 DESCRIPTION. Shell moderately large, thin but sturdy, white, spire high, canal long, recurved, open; length of aperture and canal slightly greater than height of spire, aperture shape quad- rate, canal constricted. Protoconch eroded, teleoconch whorls six, early whorls bulging at midwhorl, having about 12 thick axial ribs. Mature sculpture of 10 to 13 axial lamellae, sup- pressed on the shoulder, at the periphery producing open, raised spines that rise above the level of the suture; lamellae sharply raised on the body whorl but suppressed on the siphonal canal; siphonal fasciole with overlapping lamellae; axial sculpture lack- ing. Aperture simple, lacking denticles or columellar callus. Dimensions: height 49.4 mm, diameter 26.1 mm (holotype. Fig. 25). Radula (Fig. 24): typical for the genus, rachidian plate broad and shallow, having five cusps, a strong central cusp and two on either side, the outermost the larger; lateral teeth sickle-shaped. MATERIAL. Chile: Coquimbo (LACM), Los Vilos (LACM), Papudo, Zapallar, Quintero, Algarrobo, Pichilemu (LACM, ho- lotype, Fig. 25). Specimens examined: 73. TYPE MATERIAL. Twenty-four specimens from the type lo- cality, collected 25 May 1976, by Andrade, unidentified shrimp trawler. Holotype, LACM 1982; paratypes, LACM 1983, para- types, MNHN 200490; paratypes, MZICB 15.529; paratypes, USNM 784739. TYPE LOCALITY. 340 m off Pichilemu, Chile (34°27'S). DISTRIBUTION. Coquimbo (29°58'S) to Pichilemu (34°27'S), Chile. Depth range: 200-450 m. DIAGNOSIS. A species of Trophon characterized by its quad- rate aperture, and sculpture of axial lamellae, which are spinose at the periphery. It most resembles Trophon ( Stramonitrophon ) plicatus (Solander in Lightfoot, 1 786),5 a common shallow-water species from southern Chile in which spiral sculpture is lacking, but is smaller with the aperture more quadrate and the canal more constricted. The two species are not closely related, for Trophon plicatus has a unique radula in which there are ac- cessory cusps (the feature upon which the subgenus Stramonitrophon is based). REMARKS. There is virtually no variation in Trophon baha- mondei; all specimens examined are very similar. ETYMOLOGY. We are pleased to dedicate this species to Dr. Nibaldo Bahamonde N., of the Museo Nacional de Historia Nat- ural, Santiago. Family Columbariidae Genus Columbarium Martens, 1881 Type species (original designation): Pleurotoma ( Columbarium ) spinicincta Martens, 1881. Recent, Queensland, Australia. Columbarium is one of several genera in the Columbariidae, a family restricted to moderately deep water. Shell form and struc- ture of the protoconch resembles that of Fusinus Rafinesque, 1 8 1 5, in the Fasciolariidae, but the radula relates the genus to the Muricacea. Living and fossil Columbariidae of the world were reviewed by Darragh (1969). Japanese species were recently re- viewed by Habe (1979). Although the family is represented in the western Atlantic (Clench, 1944; Bayer, 1971), no species until now has been re- ported from the eastern Pacific. Authors are not agreed as to whether the available taxa should be genera or subgenera, but the new species that follows is clearly a member of Columbarium, sensu stricto, a genus well repre- sented in the Tertiary and Recent of New Zealand, Australia, and Japan. Columbarium tomicici new species Figures 26-30 DESCRIPTION. Shell fusiform, length of aperture and canal about two-thirds the total length, white under a thin brown per- iostracum. Protoconch eroded, teleoconch whorls six, early 5. Trophon plicatus is better known as T. laciniatus (Gmelin, 1791) as used by Dell (1971). However, Cernohorsky (1977) showed that T. plicatus , credited to Solander in Lightfoot, 1786, is the prior name. 10 Contributions in Science, Number 342 McLean and Andrade: Gastropods of central Chile Figures 21 through 25, Fusitriton magellanicus and Trophon bahamondei new species. Figures 21 through 23, Fusitriton magellanicus Figure 21, LACM 66-1 56, 580 m off Caleta Molles, Chile, height 94.0 m. Figure 22, LACM 75-46, 5 m, Isla Laitec, Chiloe Province, Chile, height 85. 1 mm. Figure 23, radula ribbon, LACM 66-156, 580 m off Caleta Molles, Chile, width of field 1 .2 mm. Figures 24 and 25, Trophon bahamondei new species. Figure 24, radular ribbon, LACM 72491 , 240-400 m off Los Vilos, Chile, width of field 0.35 mm Figure 25, holotype, LACM 1982, 340 m off Pichilemu, Chile, height 49.4 mm. whorls with a sharp median carina, at first with weak projections, but changing to thin, triangular, posteriorly directed spines, 9 to 1 6 on the final whorl. Shoulder and spire whorls smooth or finely striate; suture laid upon, or just anterior to, a stout cord (the anterior carina of Darragh, 1969). Base and canal with promi- nent, rounded, nonscabrous spiral cords, with interspaces of nearly equal or lesser width, about five across the base and ten more on the canal. Columellar callus thin, not raised to form a columellar lip; aperture subquadrate, lip thin. Dimensions; height 46.8 mm, diameter 22.8 mm (holotype. Fig. 26); height 78.4 mm, diameter 33.2 mm (Fig. 29). Radula (Fig. 30): rachidian plate with curved base and three cusps that project over the basal plate of the next row, the middle cusp the longest; lateral tooth with large base and curved tip. The radula is similar to that of other species in the family (see Bayer, 1971:172). MATERIAL. Peru: S of Lobos de Afuera (LACM, Fig. 29), N of Isla Macabi (LACM). Chile: Junquillar (LACM, holotype, Fig. 26), Coquimbo (LACM, Fig. 27), Los Vilos (LACM), Punta Salinas (LACM), Papudo, Algarrobo. Four other specimens are in the LACM collection from depths of 520-1 ,200 in in northern Peru (Banco de Mancora, Fig. 28; Chilca, Fig. 30; Mollendo), received from Dr. Enrique del Solar of Lima, Peru. Specimens examined: 30. TYPE MATERIAL. Thirteen specimens from the type locality, collected by R/V ANTON BRUUN, station 714, 16 August 1966. Holotype, LACM 1984; paratypes, LACM 1985; para- type, MNHN 200491; paratypes, MZICB 15.530; paratypes, USNM 784740. Although there are single larger specimens from other localities in the material at hand, this lot was selected as the type lot because it contains 13 specimens (only 2 live- collected) of about the same size as the holotype, enabling dis- tribution of paratypes from the type locality. TYPE LOCALITY. 950 m, W of Junquillar, Chile (25°0'S, 70°40'W). DISTRIBUTION. Banco de Mancora, Peru ( 3 ° 25'S), to Algar- robo, Chile (33°22'S). Depth range: 240-1,200 m. DIAGNOSIS. A species of Columbarium characterized by its sculpture of triangular spines at the periphery and regular, non- scabrous cords on the base and canal. C. tomicici most resembles C. veridicum Dell, 1963, from New Zealand (see Powell, 1979:169, pi. 37, fig. 1), which entirely lacks spiral sculpture on the base and canal. General proportions are similar to those of the Japanese C. pagoda (Lesson, 1840), which has a projecting Contributions in Science, Number 342 McLean and Andrade: Gastropods of central Chile 1 1 columellar lip and may have spinose sculpture on the canai, as figured by Habe (1979). REMARKS. Columbarium tomicici has the essential features of Columbarium, sensu stricto, as diagnosed by Darragh (1969), except that the parietal or inner lip callus is not raised, the spiral cords of the base and canal are nonscabrous, and there is no tooth on the outer lip at the position of the anterior (basal) carination. ETYMOLOGY. This species is dedicated to Prof. Jorge Tomicic K., of the Universidad de Antofagasta, Antofagasta, Chile. Superfamily Buccinacea Family Buccinidae Subfamily Buccinulinae Powell ( 1 929, 1951, 1 979) has discussed the higher classification of buccinid whelks with particular reference to southern genera. In early publications, he advocated the use of several families based on radular characters, but, more recently (Powell, 1979), he placed genera with a tricuspid rachidian and a tricuspid lat- eral in the subfamily Buccinulinae, of which Aeneator is a member. Genus Aeneator Finlay, 1927 Type species (original designation): Verconella marshalli Mur- doch, 1924. Pleistocene, New Zealand. Rehder (1971) introduced the New Zealand genus Aeneator to the Chilean fauna with»his description of Aeneator ( Ellicea ) loisae, although he did not mention the most common Chilean species of the group, A. fontainei (Orbigny, 1841). The latter species was assigned by Dali (1909) and Keen (1966) to Aus- trofusus Kobelt, 1 879. Austrofuscus , however, pertains to a New Zealand group with broad apertures and short, twisted canals. New Zealand species of Aeneator have been discussed by Dell, (1956, 1963), Beu (1979), and Powell (1979). The occurrence of the genus in Chile has not been mentioned by these authors. The Chilean species seem to differ in having a thickened or expanded final lip. However, the living representative of the Pleistocene type species, Aeneator marshalli separabilis Dell, 1956, from Figures 26 through 30, Columbarium tomicici new species. Figure 26, 2 views of holotype, LACM 1984, 950 m off Junquillar, Chile, height 46.8 mm. Figure 27, LACM 75-88, depth unknown off Coquimbo, Chile, height 73.7 mm. Figure 28, LACM 71-234, 520 m. Banco de Mancora, Peru, height 65.6 mm. Figure 29, LACM 74-9, 1 ,200 m S of Isla Lobos de Afuera, Peru, height 78.4 mm. Figure 30, radular ribbon, LACM 72-187, 800 m S of Chilca, Peru, width of field 0.2 mm. 1 2 Contributions in Science, Number 342 McLean and Andrade: Gastropods of central Chile Figures 31 through 39, Aeneator fontainei. Figure 31,LACM 75-21, 20 m off Mejillones, Chile, height 70.0 mm. Figure 32,LACM 75-21, same locality, height 66.3 mm. Figure 33,LACM 54737, 40 m off Montemar, Chile, height 39.5 mm. Figure 34,LACM 66-161 , 56-95 m off Punta Mar Brava, Chile, height 46.6 mm. Figure 35, radular ribbon, LACM 75-21, 20 m off Mejillones, Chile, width of field 0.5 mm. Figure 36, radular ribbon, LACM 66-1 50, 260-280 m off Punta Penablanca, Chile, width of field 0.35 mm. Figure 37, LACM 66-159, 1 10 m off Punta Hornos, Chile, height 67.0 mm. Figure 38, LACM 66-159, same locality, height 64.9 mm. Figure 39, LACM 72493, 270 m off Punta Toro, Chile, height 58.9 mm. New Zealand, has an expanded outer lip. The allocation of the Chilean species to Aeneator is therefore followed here. Ellicea Finlay, 1927 (type species: Siphonalia orbita Hutton, 1855), used by Dell (1956) and Rehder (1971) as a subgenus to distinguish species with strong spiral sculpture, was reduced to synonymy by Powell (1979), who found no clear separation of species on that character. The following diagnosis of Aeneator is offered: fusiform buc- cinids of moderate size, canal plus aperture more than half the length of the shell, whorls rounded but for a subsutural con- cavity; lip with a broad shallow sinus below the suture; sculpture of strong axial ribs overridden by spiral cords, rachidian and laterals tricuspid. Aeneator fontainei (Orbigny, 1841 ) Figures 31-39 Fusus fontainei Orbigny, 1841:447, pi. 63, fig. 2. Austrofusus fontainei, Dali, 1909:213; Keen, 1966:4. Fusus alternatus Philippi, 1847, pi. 4, fig. 6; Reeve, 1847, pi. 2, fig. 6. Siphonalia alternata, Tryon, 1881:137. DESCRIPTION. Shell large, covered by periostracum, length of aperture and canal more than half the length of the shell. Whorls six, convex except for a concave subsutural area, suture not deeply impressed. Axial sculpture on penultimate whorl of 15-17 rounded ribs, interspaces slightly narrower; axial sculp- Contributions in Science, Number 342 McLean and Andrade: Gastropods of central Chile 1 3 ture usually lacking altogether on the final half whorl. Spiral sculpture of major and minor cords; major cords four on early whorls, raised, dark brown, the interspaces twice as broad as these cords, cords overriding the axial ribs. Minor cords between each brown cord about five, separated by incised grooves. Lip lirate within, edge sharp, thick behind the edge, edge marked with brown at the termination of the cords. Dimensions: height 70.0 mm, diameter 38.0 mm (Fig. 31); height 67.0 mm, diameter 32.0 mm (Fig. 37). Radula (Figs. 35, 36): typical for the genus, rachidian and laterals tricuspid. MATERIAL. Chile: Punta Mar Brava (LACM, Fig. 34), Punta Hornos (LACM, Figs. 37, 38), Coquimbo (LACM), Los Vilos (LACM), Papudo (LACM), Quintero (LACM), Punta Pen- ablanca (LACM), Punta Panulcillo (LACM), Punta Toro (LACM, Fig. 39). Other records: Bahia Independencia and Bahia San Juan, Peru, 20-50 m, collected by the Hancock Expe- ditions in 1938; Mejillones, Chile (LACM, Figs. 31, 32), col- lected by McLean in 1975 in 20 m by scuba diving. Specimens examined: more than 100. TYPE MATERIAL AND TYPE LOCALITY. Five specimens, BM(NH) 54.2.4.517 (Keen, 1966). Type locality: Callao, Peru. Type locality for F. alternatus: Mejillones, Chile. DISTRIBUTION. Independencia Bay, Peru (LACM) (14°13'S), to Punta Toro, Chile (33°1'S). Depth range: 20-350 m. REMARKS. Aeneator fontainei is the only offshore species treated here to occur also in relatively shallow water (at depths of 20 m or more) in central Chile. Three extremes in shell form are noted. Most specimens from shallow water (Figs. 31, 32) have relatively thick shells with pronounced development of siphonal fasciole; the darkly marked cords are prominent. Specimens from archibenthal depths (Figs. 37-39) have a thinner shell and a straighter canal, and the dark cords are less prominent than those of the shallow-water form. A fine, dark intercalary cord may appear in the interspace between major cords, as in Figure 38. At first glance, especially with periostracum intact as in Fig- ure 37, these specimens appear very different from the shallow- water form. However, specimens such as that in Figure 39 seem to be intermediate between the shallow- and deep-water forms. Finally, there is a dwarf form occurring at intermediate depths (Fig. 33), in which there is little subsutural concavity, a more crowded condition of the axials, and more numerous, more closely spaced dark cords. This form shows complete intergrada- tion (Fig. 34) with the usual shallow-water form. Aeneator loisae Rehder, 1971 Figures 40-44 Aeneator (Ellicae) loisae Rehder, 1971:593, figs. 7, 8. DESCRIPTION. Shell large for the genus, fusiform, surface chalky white, covered by periostracum, length of aperture and canal more than half the length of the shell. Whorls seven, con- vex, except for a concave subsutural area; suture not deeply impressed. Axial sculpture on early whorls of about 14 low ribs running from suture to suture, interspaces narrower than the ribs; ribs lacking altogether on body whorl, somewhat indistinct on penultimate whorl but prominent on earlier whorls. Spiral sculpture of major cords alternating with secondary cords, finer tertiary cords may appear in the interspaces and may be super- imposed on the major cords. Major cords about nine to ten on the penultimate whorl, continuing with regular spacing across the body whorl and canal. Lip sinuate on upper part, weakly lirate within, scalloped to correspond with the spiral cords; final lip flared. Canal long, parietal and columellar area well defined, glazed. Dimensions: height 91.8 mm, diameter 43.7 mm (Fig. 40); height 88.5 mm, diameter 37.9 mm (Fig. 43); height 74.7 mm, diameter 37.3 mm (holotype. Fig. 42). Radula (Fig. 44): central and lateral teeth tricuspid; see also Rehder, 1971, fig. 8. MATERIAL. Chile: Coquimbo (LACM), Los Vilos (LACM), Caleta Molles (LACM, Fig. 43), Papudo, Zapallar, Quintero (LACM), Algarrobo, Cabo Carranza (LACM, Figs. 40-41). Specimens examined: more than 100. TYPE MATERIAL AND TYPE LOCALITY. Holotype, USNM 701667. Type locality: 200 m, 17-18 km NW of Valparaiso, Chile. DISTRIBUTION. Coquimbo (29°58'S), to Cabo Carranza, Chile (35°27'S). Depth range: 200-450 m. REMARKS. Aeneator loisae differs from both A. fontainei and A. castillai new species in lacking any brown coloration to the shell. It also has more numerous primary cords than does A. fontainei. Its closest relative, as noted by Rehder (1971), is the New Zealand species A. benthicola Dell, 1973 (see Powell, 1979:202, pi. 41, fig. 3). Aeneator loisae is highly variable in proportions. The holotype (Fig. 42) is relatively broad; specimens from Cabo Carranza (Figs. 40, 41) are more slender, and a single specimen from Cal- eta Molles (Fig. 43) is still more slender. Aeneator castillai new species Figures 45-50 DESCRIPTION. Shell large, fusiform, length of aperture and canal more than half the length of the shell; canal relatively short, twisted. Shell light brown under a thin periostracum, often with a lighter colored band coinciding with the periphery. Whorls six, protoconch and early whorls eroded; whorls convex but for a concave subsutural area; suture not deeply impressed. Axial sculpture on penultimate whorl of about 16 low ribs, strong across the periphery but faint in the subsutural area; final whorl with weak sculpture; axial sculpture lacking altogether on the final half whorl. Spiral sculpture of fine cords alternating in strength, all cords in the concave subsutural area fine; those of the periphery and base coarser, frequently darker in color. Lip faintly lirate within, sinuate on upper part, final lip not expanded. Parietal and columellar area well defined, glazed. Dimensions: height 78.7 mm, diameter 39.8 mm (holotype, Fig. 45); height 75.7 mm, diameter 37.8 mm (Fig. 47). Radula (Fig. 50): typical for the genus, central and lateral teeth tricuspid. MATERIAL. Chile: Coquimbo (LACM, Fig. 49), Los Vilos (LACM, Figs. 46-48), Papudo (LACM, type lot. Fig. 45), Zapal- lar (LACM), Quintero (LACM), Punta Penablanca (LACM). Specimens examined: 23. 14 Contributions in Science, Number 342 McLean and Andrade: Gastropods of central Chile Figures 40 through 44, Aeneator loisae. Figure 40,LACM 66-153, 290-450 m off Cabo Carranza, Chile, height 91.8 mm. Figure 41, LACM 66-153, same locality, height 70.0 mm. Figure 42, holotype, USNM 701667, 200 m off Valparaiso, Chile, height 74.7 mm. Figure 43, two views, LACM 66-156, 580 m off Caleta Molles, Chile, height 88.5 mm. Figure 44, radular ribbon, LACM 66-1 53, 290-450 m off Cabo Carranza, Chile, width of field 0.5 mm. TYPE MATERIAL. Nine specimens from the type locality, col- lected 29 March 1977, by Andrade, shrimp trawler GODEN WIND. Holotype, LACM 1986; paratypes, LACM 1987; para- types, MNHN 200492; paratypes, MZICB 15.532; paratypes, USNM 784741. TYPE LOCALITY. 300 m off Papudo, Chile (32°31'S; 71°54'N). DISTRIBUTION. Coquimbo (29°55'S), to Punta Penablanca, Chile (33°22'S). Depth range: 200-450 m. DIAGNOSIS. A species of Aeneator characterized by its rather short, twisted canal and the absence of axial ribs in the concave subsutural area. It differs from both A. fontainei and A. loisae in these features. Additionally it differs from the offshore form of A. fontainei in having more numerous spiral cords and a flesh colored surface. From A. loisae, it also differs in having a brown rather than white shell coloration. REMARKS. The surface layers of the shell of A. castillai are particularly prone to erosion, leaving an unsculptured, chalky shell surface upon loss of the sculptured layer. A number of the specimens are partially or completely eroded (Fig. 49). ETYMOLOGY. We are pleased to name this species after Dr. Juan Carlos Castilla, of the Universidad Catolica, Santiago. Superfamily Volutacea Family Volutidae Subfamily Odontocymbiolinae Genus Miomelon Dali, 1907 Type species (original designation): Volutilithes philippiana Dali, 1890. Recent, central Chile. Contributions in Science, Number 342 McLean and Andrade: Gastropods of central Chile 15 Miomelon is known from three species occurring in the region from central Chile to the vicinity of the Falkland Islands. The species in the present collection was identified by Weaver and Dupont (1970) and Rehder (1971) as M. philippiana (Dali, I 890). Stuardo and Villarroel (1974) showed that Dali’s species, which is known only from the single holotype from abyssal depths off central Chile ( 1,238 m, 38°8'S), differs from the more common archibenthal species, which they described as M. alarconi. Stuardo and Villarroel treated the radula and anatomy of Miomelon alarconi, confirming that the genus should be assigned to the volutid subfamily Odontocymbiolinae. Miomelon alarconi Stuardo and Villarroel, 1974 Figures 51, 52 Miomelon philippiana (Dali, 1890), of Weaver and DuPont, 1970:132, pi. 56, figs. C, D; Rehder, 1971:594. Not Dali, 1890. Miomelon alarconi Stuardo and Villarroel, 1974:140, figs. 4a, 4b, 5a, 5b. DESCRIPTION. Shell moderately large, aperture length equal to spire height; shoulder concave, siphonal canal broad, outer lip thin. Sculpture of fine axial ribs and less prominent spiral cords, interspaces broad. Columella with three or four plaits, the ante- riormost the strongest. Color light brown under a thin brown periostracum; parietal glaze light brown; surface often chalky in specimens that have lost the periostracum. Dimensions: height 74.9 mm, diameter 32.0 mm (Fig. 52); height 89.7 mm, diameter 39.4 mm (holotype). Radula (Fig. 5 1 ): base of rachidian elongate, strongly tricuspi- date. See also Stuardo and Villarroel ( 1 974, fig. 2). MATERIAL. Chile: Coquimbo (LACM), Los Vilos (LACM), Papudo, Zapallar, Algarrobo, Pichilemu (LACM), Cabo Car- ranza (LACM, Fig. 52). Specimens examined: more than 100. TYPE MATERIAL AND TYPE LOC ALITY. Holotype, MZICB 5553; paratypes, MZICB 5554, 5555, 5556. Type locality: 125 m, Chanco Bay, Chile (35°45'S). DISTRIBUTION. Coquimbo (29°58'S), to 37°51'S (Stuardo and Villarroel, 1974), Chile. Depth range: 125-450 m. REMARKS. Miomelon alarconi is larger, heavier, and more broadly inflated and has coarser spiral sculpture than M. philippiana. Superfamily Cancellariacea Family Cancellariidae The large new species of Cancellaria in the present material was unexpected, for the previously known austral representatives of the family include such genera as Admete Kroyer, 1842, and related small-shelled genera (see Carcelles and Williamson, 1951; Powell, 1960, 1979). The cancellariid radula is unlike that of stenoglossate neo- gastropods (Olsson, 1 970; Keen, 1971), yet other features of can- cellariid anatomy are typical of those of higher neogastropods (Graham, 1966; Harasewych and Petit, 1982). Genus Cancellaria Lamarck, 1799 Type species (monotypy): Voluta reticulata Linnaeus, 1767. Re- cent, Florida. Subgenus Crawfordina Dali, 1919 [ = Crawfordia Dali, 1918, not Pierce, 1908] Type species (original designation): Cancellaria crawfordiana Dali, 1891. Recent, California. The subgenus Cancellaria, sensu stricto, comprises moderately large shelled forms with reticulate sculpture and prominent col- umellar plaits. Cancellaria stuardoi new species is most closely related to the type species of the subgenus Crawfordina. Grant and Gale (1931:614) diagnosed Crawfordina as follows: “This section differs from Cancellaria, s. s., in the more elongate shape, lighter weight, smaller, more oblique plaits, and shorter col- umella.” Another feature to be mentioned is the minimal devel- opment of the parietal callus. There are three members of the subgenus Crawfordina: the offshore Californian type species, the new Chilean species de- scribed here, and the connecting link, the deep-water Panamic species, Cancellaria io Dali, 1896 (see Keen, 1971:654, fig. 1477), from 589 m, Panama Bay. Cancellaria ( Crawfordina ) stuardoi new species Figure 53 DESCRIPTION. Shell large, thin but sturdy, light brown under a thin but persistent light brown periostracum; whorls six, convex and slightly shouldered, suture deeply impressed. Protoconch lost; sculpture eroded on first two whorls. Axial ribs 1 8 to 20 per whorl, except on the final half whorl, where the ribs become more irregular and tend to be more broadly spaced from growth pauses; ribs crossing the whorls completely; ribs narrower than the interspaces; ribs posteriorly flexed near the suture. Spiral sculpture of broad, low ribs with more or less equal interspaces, 9 on the penultimate whorl and about 20 on the body whorl; spiral cords somewhat indistinct on the shoulder and narrower near the columella. Aperture ovate, columella incurved, columellar callus thin, parietal callus not thick enough to obliterate the sculpture, the anterior end of the columella consisting of a curved fold; two oblique folds higher on the columella, the posteriormost the largest, the folds hardly showing on apertural view, but much stronger within (shell viewed obliquely). Outer lip thin, lirate within. Operculum lacking. Dimensions: height 61.3 mm, diameter 31.5 mm (holotype, Fig. 53); height 62.9 mm, diameter 31.0 mm (paratype); height 57.5 mm, diameter 30.8 mm (paratype). MATERIAL. Chile: Coquimbo, Papudo, Pichilemu (LACM, type lot. Fig. 53). 16 Contributions in Science, Number 342 McLean and Andrade: Gastropods of central Chile Figures 45 through 50, Aeneator castillai new species. Figure 45, two views of holotype, LACM 1 986, 300 m off Papudo, Chile, height 78.7 mm. Figure 46, LACM 72499a, 400 m off Los Vilos, Chile, height 69.3 mm. Figure 47, LACM 72499b, same locality, height 75.7 mm. Figure 48, LACM 72499c, same locality, height 56.6 mm. Figure 49, LACM 61-12, 1 1 0 m off Coquimbo, Chile, height 79.8 mm. Figure 50, radula ribbon, LACM 72499, 400 m olf Los Vilos, Chile, width of field 0.5 mm. TYPE MATERIAL. Three specimens from the type locality col- lected by Andrade, 25 May 1976, unidentified shrimp trawler. Holotype, LACM 1988; paratype, MNHN 200493; paratype, MZICB 15.532. Each specimen has the soft parts separately preserved. TYPE LOCALITY. 240-350 m off Pichilemu, Chile (34°27'S, 72°24'W). DISTRIBUTION. Coquimbo (29°58'S), to Pichilemu (34°27'S), Chile. Depth range: 200-350 m. DIAGNOSIS. A species of the subgenus Crawfordina charac- terized by its large size, high spire, convex whorls, and sharp clathrate sculpture. It is remarkably similar to the type species of Crawfordina, C. crawfordiana from California (Fig. 54). The latter species has a more fibrous periostracum, is somewhat more slender and smaller (attaining a length of about 50 mm), has a slightly more constricted tip to the canal, and has fine pustules on the anterior region of the columellar callus. Despite these dif- ferences, the two species are clearly related, having similar pro- portions, sculpture, and the columellar plications visible in oblique view. Cancellaria io, the third member of the subgenus, differs from both C. stuardoi and C. crawfordiana in having sub- dued spiral sculpture. ETYMOLOGY. We are pleased to name this species in honor of Dr. Jose Stuardo of the Universidad de Concepcion, Concepcion, Chile. Superfamily Conacea Family Turridae Subfamily Turrinae Genus Ptychosyrinx Thiele, 1925 Type species (original designation): Pleurotoma (Subulata) bisinuata Martens, 1901. Recent, East Africa. Contributions in Science, Number 342 McLean and Andrade: Gastropods of central Chile 1 7 Figures 51 through 56, Miomelon alarconi, Cancellaria stuardoi, C. crawfordiana, and Ptychosyrinx chilensis. Figures 51 and 52, Miomelon alarconi. Figure 51, radular ribbon, LACM 73502, 200-400 m off Coquimbo, Chile, width of field 0.6 mm. Figure 52, LACM 66-1 53, 290-450 m off Cabo Carranza, Chile, height 74.9 mm. Figure 53, Cancellaria stuardoi new species, two views of holotype, LACM 1988, 240-350 m off Pichilemu, Chile, height 61.3 mm. Figure 54, Cancellaria crawfordiana. two views, LACM 72505, 50-100 m off Redondo Beach, California, height 49.8 mm. Figures 55 and 56, Ptychosyrinx chilensis. Figure 55, two views, LACM 72506, 350 m off Pichilemu, Chile, height 42.4 mm. Figure 56, radula ribbon, LACM 72506, same locality, width of field 0.4 mm. Ptychosyrinx chilensis is the only eastern Pacific member of its genus, a member of the subfamily Turrinae, in which the excur- rent sinus is located at the peripheral keel (see Powell, 1966; McLean, 1971). According to Powell ( 1 966), other species occur in the "deep ocean basins of the Indian Ocean, Natal, East Af- rica and the East Indies, the north Atlantic, off Bermuda.” Hick- man (1976) illustrated P. chilensis and recorded species of Ptychosyrinx from the Oligocene of Oregon. Ptychosyrinx chilensis Berry, 1968 Figures 55, 56 Ptychosyrinx chilensis Berry, 1968:158; Hickman, 1976;89, pi. 7, figs. 1 , 6. DESCRIPTION. Shell moderately large, length of aperture and canal less than half the length of the shell; color light brown under a fine brown periostracum; shoulder concave, growth line deeply sinuate, periphery marked by strong axial projections, about 15 per whorl; base with three strong spiral cords, the pos- teriormost retained just above the suture; fasciole with more subdued spiral sculpture. Dimensions: height 42.4 mm, diameter 15.8 mm (Fig. 55). Radula (Fig. 56): rachidian rectangular, unicuspid; marginals of “wishbone” type. This is typical for the genus as described by Powell (1966). MATERIAL. Chile: Punta Mar Brava (LACM), Punta Hornos (LACM), Coquimbo (LACM), Punta Panulcillo (LACM), Bahia Navidad (LACM), Pichilemu (LACM, Fig. 55), Cabo Carranza 18 Contributions in Science, Number 342 McLean and Andrade: Gastropods of central Chile (LACM). Specimens examined: more than 100. TYPE MATERIAL AND TYPE LOCALITY. Holotype, LACM 1912. Type locality: “about 200 fms.,” (366 m) off Coquimbo, Chile. DISTRIBUTION. Punta Mar Brava (29°21'S) to Cabo Car- ranza, Chile (35°27'S). Depth range: 65-400 m. REMARKS. Ptychosyrinx chilensis resembles the type species of Ptychosyrinx but has stronger spiral sculpture on the base and lacks the sinuate projection of the lower lip of that species. ACKNOWLEDGMENTS We are most grateful to the captains and the crews of the trawl- ing vessels, in particular Sr. Alejandro Gonzalez, of the GODEN WIND, for collecting the majority of the specimens upon which this report is based. Andrade thanks Prof. Maria Codoceo R., of the Museo Nacio- nal de Historia Natural, Santiago, Prof. Pedro Baez, Instituto Profesional de Iquique, and Prof. Fernando Alcazar, Universidad de Valparaiso, for their continued assistance with many phases of the investigation of the offshore fauna of central Chile. McLean thanks the director and staff members of the Instituto del Mar in Lima, Peru. Through arrangements made by Dr. Enrique M. del Solar, of Lima, McLean was a participant in the cruise of the SNP-1 to the Lobos Islands in northern Peru in 1974. McLean is grateful to staff members of the Smithsonian Oceanographic Sorting Center for forwarding the material from cruises of the ANTON BRUUN, and to the committee members who designated the LACM, in addition to the USNM, as the final repository for mollusk collections made by the ANTON BRUUN in the southeast Pacific. McLean is also grateful to many Chilean biologists and mal- acologists who aided him during his visits to Chile in 1975 and 1978, including Dra. Marta Bretos Alarcon, Prof. Nibaldo Bahamonde N., Dr. Juan Carlos Castilla, Prof. Maria Codoceo R., Prof. Miguel Padilla G., Prof. Luis Ramorino M., Sr. Jaime Ramirez Boehme, Dr. Patricio Sanchez, Dr. Jose Stuardo, Prof. Jorge Tomicic K., and Dr. C. A. Viviani. We thank the photography department of the LACM for as- sistance in the preparation of the illustrations. Radula slides were prepared and photographed by museum volunteer Jo-Carol Ramsaran. LITERATURE CITED Andrade, H. 1980. Nueva especie de Paralomis en aguas de Chile: Paralomis chilensis n. sp. (Crustacea, Anomura, Lithodidae). Bol. Mus. Nac. Hist. Nat. Chile 37:269-273. Andrade, H., and P. Baez. 1977. Calastacus rostriserratus n. sp. (Crustacea, Decapoda, Macrura, Axiidae). An. Mus. Hist. Nat. Valparaiso 10:65-67. 1980. Crustaceos decapodos asociados a la pesqueria de Heterocarpus reedi Bahamonde 1955 en la zona central de Chile. Bol. Mus. Nac. Hist. Nat. Chile 37:261-267. Baez, P, and H. Andrade. 1977. Geryon affinis Milne-Edwards y Bouvier, 1894, frente a las costas de Chile (Crustacea, De- capoda, Brachyura, Geryonidae). An. Mus. Hist. Nat. Val- paraiso. 10:215-219. 1979. Crustaceos decapodos arquibentonicos fre- cuentes frente al la costa de Chile central. An. Mus. Hist. Nat. Valparaiso 12:219-231. Bayer, F.M. 1971. New and unusual mollusks collected by R/V JOHN ELLIOTT PILLSBURY and R/V GERDA in the tropical western Atlantic. Bull. Marine Sci. 2 1 ( 1 ): 1 1 1-236. Berry S.S. 1968. Notices of new Eastern Pacific Mollusca. VII. Leaflets in Malacology 1 (25): 1 55—1 58. Beu, A.G. 1979. Bathyal Nukumaruan Mollusca from Oaro, southern Marlborough, New Zealand. New Zealand Jour. Geol. Geophysics 22( 1 ):87— 1 03. Carcelles, A. 1944. Las especies del genero Capulus Montfort, 1810, en aguas argentinas. Comunic. Zool. Mus. Hist. Nat. Montevideo 1(9): 1-5. Carcelles, A.R., and S.I. Williamson. 1951. Catalogo del los moluscos marinos de la provincia Magallanica. Rev. Inst. Nac. Invest. Cienc. Nat. Buenos Aires 2(5): 225-383. Cernohorsky, W.O. 1977. The taxonomy of some southern ocean Mollusca (Gastropoda) mainly Antarctic and subantarctic. Rec. Aukland Inst. Mus. 14:105-1 19. Clench, W. J. 1944. The genus Columbarium in the Western Atlantic. Johnsonia 1 ( 1 5): 1 —4, pi. 1 Clench, W.J., and R.D. Turner. 1960. The genus Calliostoma in the western Atlantic. Johnsonia 4( 40): 1 —80. Codoceo, M., and H. Andrade. 1978. Asterozoos arquiben- tonicos de Chile Central. An. Mus. Hist. Nat. Valparaiso I 1:153-174. 1980. Solanometra antarctica (Carpenter, 1888) en la region arquibentica de Chile Central. Bol. Mus. Nac. Hist. Nat. Chile, 37:229-234. (In press.) Nuevo asteroideo para Chile: Cryp- topeltaster philippii n. sp. (Goniasteridae, Hippasteriinae). Rev. Biol. Mar. Inst. Oceanol. Univ. Valparaiso 17(3). Codoceo, M., P. Baez, and H. Andrade. 1978. Segundo registro Ophiomastus molinae Castillo, 1968 (Equinodermata, Ophiuroidea). Not. Mens. Mus. Nac. Hist. Nat. 261:10. Dali, W.H. 1980. Scientific results of explorations by the U.S. Fish Commission steamer Albatross. VII. Preliminary report on the collection of Mollusca and Brachiopoda obtained in 1887-88. Proc. U.S. Nat. Mus. 1 2(773):2 1 9-362. 1 908. Reports on the dredging operations off the west coast of Central America to the Galapagos, to the west coast of Mexico, and in the Gulf of California. XIV. The Mollusca and Brachiopoda. Bull. Mus. Comp. Zool. Harvard Univ. 43(6):205— 487. 1909. Report on a collection of shells from Peru, with a summary of the littoral marine Mollusca of the Peruvian Zoological Province. Proc. U. S. Nat. Mus. 37 ( 1 074): 1 47— 294, pis. 20-28. Darragh, T.A. 1969. A revision of the family Columbariidae (Mollusca: Gastropoda). Proc. Royal Soc. Viet. 3( 1 ):63— 1 1 9, pis. 1-6. Dell, R.K. 1956. The archibenthal Mollusca of New Zealand. Dominion Mus. Bull. 18:1-235. _______ 1963. Archibenthal Mollusca from northern New Zealand. Trans. Royal Soc. New Zealand, Zool. 3(20):205— 216, pi. 1. 1971. The marine Mollusca of the Royal Society ex- Contributions in Science, Number 342 McLean and Andrade: Gastropods of central Chile 1 9 pedition to southern Chile, 1958-59. Rec. Dominion Mus. 7( 1 7): 1 55—233. 1978. Additions to the New Zealand Recent mol- luscan fauna with notes on Pachymelon (Palomelon). Nat. Mus. New Zealand Rec. 1(1 1):161 — 176. Fretter, V., and A. Graham. 1962. British prosobranch molluscs: their functional anatomy and ecology. Ray Society, London, xiv + 755 pp. Grant, U.S., IV., and H.R. Gale. 1931. Catalogue of the marine Pliocene and Pleistocene Mollusca of California and adjacent regions. Mem. San Diego Soc. Nat. Hist. 1, 1,036 pp., 32 pis. Graham, A. 1966. The fore-gut of some marginellid and can- cel lari id prosobranchs. Studies Tropical Oceanography, Miami 4:134-151. Habe, T. 1979. The Western Pacific columbariid species. Venus, Jap. Jour. Malac. 38( 1 ): 1 —4, pi. 1. Harasewych, M.G., and R.E. Petit. 1 982. Notes on the morphol- ogy of Cancellaria reticulata (Gastropoda: Cancellariidae). Nautilus 96(3): 104-113. Hickman, C. S. 1976. Bathyal gastropods of the family Turridae in the early Oligocene Keasey Formation in Oregon, with a review of some deep-water genera in the Paleogene of the east- ern Pacific. Bulls. Amer. Paleont. 70(292): 1 — 1 19, pis. 1-7. 1981. Selective deposit feeding in the deep-sea ar- chaeogastropod Bathybembix aeola. Mar. Ecol. Prog. Ser. 6:339-432. Keen, A.M. 1966. West American mollusk types at the British Museum (Natural History), 111. Alcide d’Orbigny’s South American collection. Veliger 9(1): 1-7, pi. 1. 1971. Sea shells of tropical west America — Marine mollusks from Baja California to Peru, 2nd ed. Stanford Univ. Press, Calif, xiv + 1,064 pp. Lamarck, J. P. B. 1816. Tableau encyclopedique et methodique des trois regnes de la nature. Pt. 23, pp. 1-16, pis. 391-488. Matsukuma, A. 1978. Fossil boreholes made by shell-boring predators or commensals. 1. Boreholes of capulid gastropods. Venus, Jap. Jour. Malac. 37(1 ):29— 45. Merriman, J. A. 1967. Gut content and radular illustration of Bathybemix bairdii (Dali, 1889). Veliger 9:355-356. McLean, J.H. 1971. A revised classification of the family Tur- ridae, with the proposal of new subfamilies, genera, and sub- genera from the eastern Pacific. Veliger 24(1 ):1 14-130, 4 pis. 1981. The Galapagos Rift limpet Neomphalus : rele- vance to understanding the evolution of a major Paleozoic- Mesozoic radiation. Malacologia 23(1 — 2):29 1—336. Olsson, A. A. 1970. The cancellariid radula and its interpreta- tion. Paleontogr. Americana 7(43): 19-27, pis. 4-6. Orbigny, A. d'., 1834-1847. Voyage dans PAmerique Merid- ionale. Mollusques. Paris, vol. 5, pt. 3, xliii + 758 pp.; atlas, 85 pis. Orr, V. 1 962. The drilling habit of Capulus danieli (Crosse) (Mol- lusca: Gastropoda). Veliger 5(2):63-67. Perez-Farfante, I. 1943. The genera Fissurella, Lucapina, and Lucapinella in the western Atlantic. Johnsonia 1(10): 1-20. Philippi, R.A. 1845-51. Abbildungen und Beschreibungen neuer oder wenig gekannter Conchiylien. Kassel, 3 vols. 24 parts. Powell, A.W.B. 1929. The Recent and Tertiary species of the genus Buccinulum in New Zealand, with a review of related genera and families. Trans. New Zealand Inst. 60:57-98. 1951. Antarctic and subantarctic Mollusca: Pel- ecypoda and Gastropoda. Discovery Repts. 26:47-196. . 1960. Antarctic and subantarctic Mollusca. Rec. Auck. Inst. Mus. 5:1 17-193. . 1966. The molluscan families Speightiidae and Tur- ridae. Bull. Auckland Inst, and Mus. 5, 184 pp., 23 pis. 1 979. New Zealand Mollusca. Collins, Auckland, xiv + 500 pp. Reeve, L. 1847-48. Monograph of the genus Fusus. Con- chologica Iconica, 4, 21 pis. Radwin, G.E. and A. D’Attilio. 1 976. Murex shells of the world, an illustrated guide to the Muricidae. Stanford Univ. Press, 284 pp., 32 pis. Rehder, H.A. 1971. A molluscan faunule from 200 meters off Valparaiso, Chile, with descriptions of four new species. Proc. Biol. Soc. Wash. 83(5 1 ):585-596. Revuelta G., and H. Andrade. 1978. Neuva localidad para Lithodes murrayi Henderson, 1888, en el Pacifico Sur Orien- tal. Nat. Mens. Mus. Nac. Hist. Nat. 261:3-4. Roeding, P.F. 1798. Museum Boltenianum. Pars II. Hamburg, 1 99 pp. Sievers, H., and N. Silva. 1 975. Masas de agua y circulacion en el Pacifico Suroriental. Latitudes 18 S - 33 S (Operacion Oceanografica MARCHILE VIII). Cienc. y Tec. del Mar. Contrib. CON A 1:7-67. Silva, N., and H. Sievers. 1981. Masas de agua y circulacion costera de la corriente de Humboldt latitudes 18 S - 33 S (Operacion Oceanografica MARCHILE X y ERFEN I). Cienc. y. Tec. del Mar, Contrib. CONA 5:5-50. Smith, J.T. 1970. Taxonomy, distribution, and phylogeny of the cymatiid gastropods Argobuccinum, Fusitriton, Mediargo, and Priene. Bulls. Amer. Paleont. 56(254):445— 573. Stuardo, J., and M. Villarroel. 1974. On some living and fossil volutes referred to Miomelon Dali, 1907, and Proscaphella von Ihering, 1907. Veliger 1 7(2): 1 39—1 55. Thorson, G. 1 965. A neotenous dwarf-form of Capulus ungaricus (L.) (Gastropoda, Prosobranchia) commensalistic on Tur- ritella communis Risso. Ophelia 2(1): 175-2 10. Tryon, G. W. 1881. Manual of Conchology. 3, Tritonidae, Fusidae, Buccinidae. Philadelphia, 310 pp., 97 pis. Vokes, H.E. 1963. Studies on Tertiary and Recent giant Limidae. Tulane Stud. Geol. 1 1 (2):73— 92, pis. 1-2. Weaver, C.S., and J.E. DuPont. 1970. Living volutes, a mono- graph of the Recent Volutidae of the world. Delaware Mus. Nat. Hist., Monogr. Ser. 1, 375 pp., 79 pis. Received 15 April; accepted for publication 12 July 1982. 20 Contributions in Science, Number 342 McLean and Andrade: Gastropods of central Chile Wl W' S co '" m xv\OSi^/ co x^Ssv^X m g m ^ riON” NO!iniiISNi"WIMOSHIf^St;5S3!HVHan“UBRAR!ES^SMITHSONIAN““iNSTfTUTjONt/'NOIiniliSNl'‘NVINOSHilWe co z «2 ^ z -s <2 z co z -Cl ac CO >” !'* ' - | I' • • | vpr | Ntas^ | »gr |.. s I a I I^LI B R AR I ES^SMITHSONIAN INSTITUTION ^NOliniUSNI NVINOSHiliAIS^Sl 1 8VH8 H^LI BRAR I ES^SMITHSONIAN “ c o ._ co 5 \ m ^ ^ 5 co TION NOIlfUllSNI NVIN0SH1IWS S3IHVH8n LIBRARIES SMITHSONIAN INSTITUTION NOIilUUSNI NVINOSHilWS z r- z r* 2 «“ Z r ° m 2 m 2 . .-ffi /ffiSwhSX 2 ™ .<^c m 'tr m x^y x; m z' nmuiisSJ'' fm (O \ £ CO — CO ” — to fail LIBRARIES SMITHSONIAN INSTITUTION NOliniUSNI NVIN0SH1MS S3IHVH8n LIBRARIES SMITHSONIAN z .... CO 2 ^ CO Z _ — _ CO Z , . CO ? xJ^rx < JkSs s A /?S$X i . ** B Jp. g |pw|| o ilfc 3 " vX K " x o x h- ^ X£> 2 r CO X to O Z > z V Z £ \ z, ^ to z CO X _ TION NOIlfUllSNI NVIN0SH1IIMS S3IBV0an LIBRARIES SMITHSONIAN INSTITUTION NOUOLSlSNi NVIN0SH1IWS CO — co 5 co 5 CO — av'^'-mX £ ijJ X\v*4'TT”X to W #« IS 0£ m w^mw.s-1 — « o XgN PCX “ o 'Xi£C>' “ N^uas^X Q I a 11 “'Ll BRAR I E S2 SMITHSONIAN”1 INSTITUTION ^ NO!iniilSNI~JNVINOSHlI WS^S 3 IBVHail^LIBRARf ES^SMITHSONIAM z >2 z 1 * z , - r- z l0/ m ' ■" m m >W g S wsmvnv -- CO 2 CO £ CO _ to _ TION NOIIDIUSNI NVIN0SH1IWS SBIBVaail LIBRARIES SMITHSONIAN INSTITUTION NOilfllUSNI NVIN0SH1IIAIS co z co z , w z * co z ^ ^ >• 5 *\sr > 2 - )^v > '■ S J a n~LI B R AR I ES” SMITHSONIAN_ INSTITUTION NOlinillSNI^NVINOSHlIWS^Sa I B VH 8 Il^LI B RAR I ES^SMITHSONIAN ~ co r; co 2 X w .— - 5 co co q xyost^- — yvory u> _ v»»yy q 'sJXpS^JX “ ‘ Z J 2 ' ' j 2 j TION NOliniUSNI NVIN0SH1SWS S3 1 HVH 0 H LIBRARIES SMITHSONIAN INSTITUTION NOUfULLSN! NVIN0SH1IINS Z r- z *" 2 *“ Z J a II I! B RAR I ES^SMITHSON IAN ~ INSTITUTION 00 NOlinillSNrNVINOSHlIWS S3 I HVH8 n“LI B RAR ! ES^SMITHSONIAh m TION NOliniUSNI NVIN0SH1IWS S3IBVH0I1 LIBRARIES SMITHSONIAN INSTSTUTSON NOIlfUllSNI NVIN0SH1IIM CO — to 5 c/s 5 - co 5 Mv - AaSKLA g .JBa W w . A '± W , ., s I# | '<§gP/ S V^P>' 1* 3 x^ I xITH'sONIAN:1NST.TUTlON%OliniaSNr NWNMI1IM S3 « »*** '^UIBRAR. ESgW™™“"* - s 3 | ■%% 1 Ifws I JP"% g jSSlN § 3lfik 1 c c vlw § m »\w > * & s s <*? 2 « l0N%0Un!L,SNV.N<>sJs3IBV8anfu8RAR^^^^^ ,« = •« «« ^«5#^ 2 H! a ?? >f^TBSx. w «» « z OT C0 to® 1 MMffci J|4|f%l®l 0 ,;y S 5 2' /'* | nsjos^' £ Vsasx o ; a n=u B RAR . ESZSM.THSON1AN JNST.TUT.O^NO.inl.lSW^.WSHi.W^a IBVBBIl^LIB BAR • E* SMITHSONIAN *x 2 4% p „,«/ ap S /fipSSft s ifli 5 5 pr ^ « 3 p 50 f?lF *aia s J/sP'JM %. life 2 fel? Us! » - -■• »ei£ ^SB > s 1“ m z y vuMC.y ».» ^55^ TO « Xggmssy w z 5 ,0N“ M0linlUSNl"NVIN0SHllWs"s3 1 aVHBI l^U B RAR I ES^THSONIANJNSTITUT.O^ NOIllUUSN.gNV.NOSHU«S § t P ' li p 1 i wm | 1 w i X i 1 ^ 1 anStiBRtRIEs”SMITHSONIA!«3lNSTlTyTION"MO«iniIlSW“MVfflOSHiWS S318VB3I1 UBRARIES .SMITHSONIAN °* *— m 55 S ^a-s^ a jots > “.» ^sssss^,. is w , ... .* z toJ m m a z § 2 ««'»» « ia»i“8fi « e a «b 9 e ewiif h§OSS!SAN ^SNSTITUTIOI^ NOIXnillSNI NVIN0SH1IWS ■ION NOSiniSASNi^ NVIN0SH1IWS SilHVMflJl LSBRARSES SMITHSONiftN^iwaii^BU'w.H E . 5 ..,,,. C «.. § = ,.Ao*e, 9 A 5 .<£■&&•>, 2 « A., mk s » s» » »“ <£» z mSU8RARI ES^SMITHSONIAN^NSTlTUTION^NOIXnillSNrNVINOSHUWS^SamVMa II ~LI BR AR I ES SMITHSONiAt l0N"N0u.nmsNiSNviN0SHiiws"s3 1 ava a nf librar i E8"smithsoniah jhstitution "noiiriusni nvinoshxiw < z 3 8 2 2 «« U! z m m 2 2 «» nan libraries S“ z SMITHSONIAN INSTITUTION NOUIUUSNI NWN0SH1IWS SliaVMBIT ^UBRARIES ^SWTHSONIA m » s© a > n " - W7 ^ p| '■cyity g W ^®v g __ UT10H“ N0lilUUSNI=N»m0SHilWs"s3 1 BVB 8 n~Ll B RAR I ES>ITHSONIAN;.NSTTOTION^ NOUIUIlSNI^NW^lll I ,1 .1 *!fe 1 1 I /Be S il x*i£&^ & a \ s jf, *.5 w ,V 2 L 1 8 R AR I eS^SMITOSONIAN* INSTITUTION^NOlinmSN^NVINOSHXI NS^S 3 1 8 VM 8 .1 _ i • 6RAR> E S^SMITHSONU 2 < ? iMeriTHTinM umimilSMI WWINOSHill