' $ : i ' i ‘ ‘ ( i ~ ‘ ‘ a hed at A ee +n Gen. ora.’ SAS dn OF ‘ oe . ‘ . % ite Sie ‘ ere) : Soy ie peat e a arn ec . Soya’ dat pan Ree TNR Ee Og MAS RO es a) ’ Vat i ‘ Staten ay : oa) ee a a. " 1 é denver a: H ao ne nets aod , Monn ofan Tries. : . : we ‘ i : A ee nate betes ticnee Se ANE T. ty EN TS ral yi Lae aa cee ae ie "« Be ee ‘ i PP Pa et ee -_ Yip ae Vee RR OE SS Yad MA Oe Bins ap ees RTA : as : i ‘ ee Lhe ee ik et aes at fet) rr Me Opnt's bebe A ws ‘ 7 . . . See eet Rk: ‘ i oe . aa i . . ey . re ares Lae pera ste eck Laas as a eay pay ‘ wh Pity" ber : cs ts? : een rr . are : oe beh hyt gly tg et eet UNE wes Raven ten 7 ; R . Ba ans & We spate rye ss : Oe 4 _ ¥ x enes ‘. i . Cte mote eA LR tie MEN TRES r rae YS cate. Ses en ar Rie ana ae Bey ae a hea . : Sucusirar de ders : of Foe he 4 pre edt 7 ; Pan ee es oe i oye wnage # : ae, vb ya Sage dts ted ge weomtie 2 , : wpe a pone i . : : sage oi ata ia 2 : . int : Ha me j ie, R nye = Ls x eA ao ver ore tite. : ‘ : pony gs ead pale , : an ‘ : ara Wale RUB. te yay La NE DARE RES Meeps aL foueal ayy ete de mye ey ae he ae he Nea Me OPEN nig ert see Mes qa ese ay pote tT i a NANA XC AT aS OR UN RE oe ERAS NTR era te Ate Oy eM Nag astE ARS Sa Me aN ps Resehcebe Reet Pets BOM PALAIS OARS PoE BILAL yea an ey Serete Feet Ut pataes ar he De DVE SII ree tearays seks Uae REMNCrLa meee AMEN R Winder veer spawinrvphirs rogathe athe ter bie Re PERE Pagans descr ae Aare! PAE Wonk ia aanus aN AAR RIFT INN ed MAME N SSE Me rer saose ha onan VR REA VEN RRS Peni Ng Ae MES Speke rae eee tase MEANS ONBMEMY NNT RAGIN AVE eres Arde RANT EE meat oD cm anne ong eygay tee Vee NT ERLE BE ort Lia aha of CT Fiat wz nian an de te St BESS cnn ssratans Straten f Semel NN shy bye Det en Ne NS foe sana oS FUELS, Oe dea era hc oe RRS RTS Sole tty tear weer SND sete tere Tenec aula ne TAS Re © Wy cae ie an yes Pay Nyt Spire Se Ea Donets Pera Cave de woe MEA ta aae MN Ai Asad nes as a ae, Tae Se carte he NERS toned ye Paes: eer, ae wee ee papers pei na hen we eh gw pes 8 eA or ee ier wk Fate ore wee ey wie eveye ease po eR ee ere ee ERIS SP oe eto ae ae ae aad sly ABE TR ae RED Oe at ateetih are Ne td ty Or tot ib AOAUTAET AO AS PRP PIO NTT Tt ih al dae Reh ied eats Sipe raed: iene Bis eeNeNe Ta 8 Mt ee OS ee OM VEC Ne Tg bp PD AEE IOP Tig ain pone Bt ee eee rs Se aoe ie wuntae ets re ah ew eveyone Wyre ELEN eo ee ee Lae ea vay re ne et Ln eee ” ape SEW ONE E eer ce Me toe feet pede a eR ae sede ye ed eure ae ie BHT mee gee ee earsore eiveaneur tise uy OPEV ABE BOE, HES 4 TET EENS Pm Toward the Phylogeny of a Fossil Species Flock: Semionotid Fishes from a Lake Deposit in the Early Jurassic Towaco Formation, Newark Basin Amy Reed McCune Bulletin 43 Peabody Museum of Natural History Yale University Toward the Phylogeny of a Fossil Species Flock: Semionotid Fishes from a Lake Deposit in the Early Jurassic ‘Towaco Formation ) Newark Basin AMY REED McCUNE Section of Ecology and Systematics Cornell University Ithaca, New York 14853 BULLETIN 43 e 18 DECEMBER 1987 PEABODY MUSEUM OF NATURAL HISTORY YALE UNIVERSITY NEW HAVEN, CONNECTICUT 06511 Bulletins published by the Peabody Museum of Natural History, Yale University, are numbered consecutively as independent monographs and appear at irregular intervals. Shorter papers are published at frequent intervals in the Peabody Museum Postil/a series. The Peabody Museum Bulletin incorporates the Bulletin of the Bingham Oceano- graphic Collection, which ceased independent publication after Vol. 19, Article 2 (1967). Communications concerning purchase or exchange of publications should be ad- dressed to the Publications Office, Peabody Museum of Natural History, Yale University, 170 Whitney Avenue, P.O. Box 6666, New Haven, CT 06511, U.S.A. © Copyright 1987 by the Peabody Museum of Natural History, Yale University. All rights reserved. No part of this publication, except brief quotations for scholarly purposes, may be reproduced without the written permission of the Director, Peabody Museum of Natural History. Printed in the United States of America ISBN No. 0-912532-03-3 CONTENTS us SHUR COA) ET 615 AE, RI do ee We eeetrtepeenn be: CEN AV NOUS tases cot ay ee EEE, geen ee 00, 2: eee tet eee ay eS oe ee, IF as SINT IRONS CIO 9 capes ae aa se i aig eae Ru tA Teh? Beagle ns, os Ue Oe WA RAT SAND) MEE RIOM Sm. te eee eee GROUPS COG ho oo: 2 Ys Se ee es Ss SS ARE, AU rere O25 ol EMATIC DESCRIPTIONS OF SEMIONOTLUS = Miter Scronorus verse). Grolpla ay eet a ett a Species with Wioditied Simple; scales nese oe ee SPEclcsawitmomall Scales\: > .., \ce ee eee ee a ee eee Species with -bhin-spimed) Scales sa,aeae nea ea. 5-- A ek. Mewenuonorustenuiceps Group sae ee ee ee ee Nila RIN Sf@ ESV UAIRTAGLIOIN . tase cole eee. oS eecca ts AG ee 111 LISMOEREIGURES lt. Newark Supergroup-of eastern’ NorthAmerica 5.22 82) a) aoe 6 2 Whe INewaTrki Basin: >, ~ ovs5.4.. 8 oes Rah ae ee eee 7 3:, Generalized: Newark:sedimentany. eycle. = ose sue eee 8 A WWeastirements': 22.00: Suse ices 4 aero, et an ee 14 5, Morphotype-variability plots.> .% 2.22: 22-462 > ede oe 20 6: (Rlotof canonical variables. 2.2... 250A 02 ae ers ee ee Dh 7. Pairwise comparisons of Goodkin viewer tracings ................. 28 8. Comparison of the skulls of Semionotus and Lepidotes .............. 31 9. Morphological: diversity ofdorsal ridge*scalést ty, 29 a2. 42 478 See 32 10; Distribution of ganoime on dorsal ridge’scales, 444-2045: »-- ee aoe 33 ile. SemmtonOlUs RUN SGli we Ae cede toi sti ncest ice WA tol eee Re ey Ce ee 25 IDE Semionotus Olsen eens. 2< 25.0 a eee tno Set eo eee 39 Ie Wemionotus Vir omtaey, in. oa 2 es os + 2S a Bo at eee 41 VAs SenitonmolvwUs THOMSON eich. « s.4. <1 -0e2s whee So ee Cees 4 eee 45 [Dros SCMBLOMOLUSs CUMCTINUS Se Novy 8. oa. gba b ee «Wee, ot ies eee 49 1G. ASemionolUs COMOAIISmEDs Aa iWin) . 222 ae 93 37 System tomrecode dorsaltridge:seale morphology (Fl9 292.0 2. ae 94 a6 hentative cladistic relationships. #25 4155: 9: e2cuen.. 0 40 eee 95 29. ‘Phe semionotid:assemblage from Vowaco cycleiP4 <2... 4p ee 96 40. Holotypes of four new species: S. kirschi, S. thomsoni, S. virginiae, and SOUS ENE en Ls ore rise have GSD a OS Se a Ak RR Se ee 98 41. Holotypes of four new species: S. ewthenius, S. convalis, S. schaeffert, and SEO CLAN ee es oe PEN sel Sty Saat © ese HRN Ae eee SS Ae iar 2) 99 42. Holotypes of five new species: S. johberryi, S. anosteus, S. latheticus, S. melanimus, and Ss Gm BUGEDRAlUS Pe eain ss areas ore ere eee eee 100 AJ Semionotus tenwiceps, MOlotype and NeOtype= 4... ae eee 102 44. Holotypes of four new species: S. greenwoodi, S. profundus, S. decoratus, AU ISS eT AUS ON Serene She xe RTD, oy 25 tert i ees 103 ADM OIOLY PE. Olgse SABINOLUS co ahed.o.d ok eus de eae ee, See eee 104 CNA RON LIST OF TABLES =» Classification functronsHor 8 species of Semionotuss. 4...) ee. 25 Coeticientstorcanonicall variables? (2s 12.4.0 marl M.et Ny 8 as). 26 @anonicalivariables evaluated ‘at-croup means -..2..5.8.5 0.0.4) 2 29 Summary of morphometric data for Semionotus kirscht ............. 36 INESrEssionveqUWations fOrESwisC7eus Wee. rae ane ALi. Wen oy Summarysee. bi Feit noah ies a osmam, Parke Shel MP N42 See Re eae 71 /Summarnry 10.0 13)<.0 VVvSC 3 LO}. 7 ere 10.0 12.0 PLVSC 10 8.2 0.8 Ha) 9.0 ANFSC 10 Gt le) 16.0 20.0 DFSC 10 20.3 aS} 18.0 23.0 CDSC abil 3475: IS 3} 33.0 517/40) AXSC 5 10.6 ALG. 7 8.0 1270 PCTR S 0: 0.0 67/510) 17 7..0 PLVR 4 4.0 0.0 4.0 4.0 ANFR 5 9.4 ies! 8.0 11.0 DFR 7. iba bess) Ws (9) TORO E20 CDDR 5 8.6 ORS 8.0 9.0 CDVR 9 heal Ons 8.0 9.0 DFF1 9 S$} sal ORS 3710 4.0 DFF2 7 5.0 0.6 4.0 6.0 DFF3 10 Yor 0.6 6.0 8.0 AFF1 4 30 0.0 £30) 3°20 AFF2 4 5.0 0.0 50 50 AFF3 7 Whe? 0.9 Uo 9.0 PCTF 4 6.5 0.6 6.0 Uo) PLVF 2 6.0 0.0 6.0 6.0 CDF al 9.0 = 9.0 9.0 CVF 2 als Vis %/ 11.0 12.0 developed spines are fully supported by a bony base. Only the spine and the center of the scale are covered by ganoine; the lateral aspect of the bony base of the scale is not covered by ganoine. Posteriorly, the spines become longer and narrower, but they are still more robust than in Semionotus melanimus or Semiono- tus latheticus. The last scale in the series is a large oval plate that abuts the first (unpaired) fulcrum on the leading edge of the dorsal fin. The total number of scales in the dorsal series, including the predorsal scale, is about 19. The overall shape of this species is fusiform (Fig. 19A), but relative to other species in the S. bergeri group, it has an unusually large head. It is most similar in shape to Semionotus johberryi, but in addition to a longer head, its dorsal fin is positioned more anteriorly. The forehead slopes at an angle of about 30° relative to horizontal. In the holotype, there are 33 lateral line scales. The pelvic fin originates at the 7th scale row, the anal fin at row 19, and the dorsal fin at row 21. In the holotype, there are no intercalated scale rows dorsal to the lateral line and anterior to the dorsal fin. The largest flank scales are in the anterior region near the lateral line. They decrease in size both caudally and anterodorsally. On the caudal peduncle, both the dorsal and ventral median scales are larger than lateral scales. From the lateral line, there are 9 scales to the origin of the anal fin. The dorsal and anal fins are fringed with 8-10 fulcra; the first three of these are basal fulcra, and the bases of the next two lie against the unsegmented portion of the first lepidotrichium. The number of dorsal lepidotrichia is unclear, but there are 9 anal lepidotrichia. The pelvic fins consist of 4 lepidotrichia fringed by 9 fulcra. The pectoral fins are not well preserved in the holotype. In all fins, the tips of the lepidotrichia extend beyond the tip of the last fin fulcrum. The posterior margin of the dorsal fin extends about halfway between the dorsal and caudal fins (YPM 8866). The anal fin and the pelvic fins almost reach the origin of the caudal fin and anal fin, respectively (YPM 8844, 8866). The tips of the pectoral fins are not visible in any specimen. 60 PEABODY MUSEUM BULLETIN 43 Fic. 19. S. anosteus. A, outline of body form traced from YPM 8844; B, camera lucida drawing of the dorsal ridge scales of YPM 8851 (top) and YPM 8844 (bottom). Scale = 5 mm. SPECIES WITH SMALL SCALES Semionotus amplicephalus Figs. 20A, 20B, 42; Table 32 Diagnosis. ‘The body form of this species differs from all species of the S. bergeri group in that the length of the head is much larger relative to standard length, about 33% (Fig. 20A). In addition, this species is distinguished from the S. bergeri group by the morphology of its dorsal ridge scales (Fig. 20B). As in many other species of Semzonotus, the scales of S. amplicephalus are dorsally convex and have posteriorly directed spines. However, the spines of the first eleven scales are very short relative to the size of the scale base, and the dorsal ridge scales are no larger than the flank scales. This distinctive dorsal ridge scale morphology has been termed the “‘small scale” type (Olsen and others 1982). Holotype. YPM 8849 (P4-2442) almost complete fish; partial counterpart; both negatively prepared (Fig. 42) Type locality. Yale excavation in cycle P4, Pompton, New Jersey Formation. ‘Towaco Formation, Newark Basin Age. Hettangian, Early Jurassic Etymology. From amplus, meaning large, and cephal, meaning head Description. The dorsal ridge scales of this species are like those of the small- scale group (Olsen and others 1982). These scales have short, thin spines (Fig. EARLY JURASSIC SEMIONOTID FISHES 61 ——— -7\ Fic. 20. S$. amplicephalus. A, outline of body form; B, camera lucida drawing of dorsal ridge scales. Scale = 5 mm. Both figures based on the holotype YPM 8849. 20B), and the scale bases are the same size or smaller than the flank scales. The holotype of this species is preserved in rock taken from the edge of a slump, and it may be somewhat distorted. However, what is unusual about its shape is that the head is very large in proportion to its body, and the probable direction of distortion would only make the fish appear more slender than it really is, not change the proportion of head to body length. The holotype is very large, about 29 cm standard length, and the head takes up an unusually large proportion of its length, about one-third. The body itself and the caudal peduncle are both stocky (Fig. 20A). Neither the scales nor the fin rays are well enough preserved to count with confidence, but there appear to be about 13 vertical scale rows anterior to the pelvic fin and about 30 lateral line scales. These counts, though only from one individual, are out of the range of variation for most Semzonotus and related genera. The fact that this individual is unusually proportioned, has low scale counts, and has unusual dorsal ridge scales justifies specific distinction even of this single individual. The skull is quite well preserved, with a good view of the skull roof, a single suborbital, open cheek region, the circumorbital series, and the jaw joint, all as described for other Newark semionotids (Olsen and McCune ms). Although I have designated no paratypes, I tentatively refer another specimen, collected by Redfield (YPM 6484), to this species. It is a large individual with a large head like the holotype, but its shape is too distorted for reliable measurement. I refer this specimen to S. amplicephalus on the basis of a single dorsal ridge scale, which is like that described above for S. amplicephalus, and the fact that the pelvic fins originate at about the 13th or 14th scale row. 62 PEABODY MUSEUM BULLETIN 43 SPECIES WITH THIN-SPINED SCALES Semionotus latheticus, new species Pigs 2A iB, 42: Pablew2 Diagnosis. S. latheticus is a moderately deep-bodied species, very similar in shape to S. olseni (Fig. 21A). It is distinguished from the latter by the morphology of its dorsal ridge scale series (Fig. 21B), in which the spines are very long and slender relative to the scale base like those of the Semionotus micropterus group (Olsen and others 1982). Holotype. YPM 8899 (P4-1994) complete fish, part and counterpart; one side negatively prepared; the other side, which shows the dorsal ridge scales in positive view, is not acid prepared (Fig. 42) Paratype. YPM 8841 Type locality. Yale excavation in cycle P4, Pompton, New Jersey Formation. ‘Towaco Formation, Newark Basin Age. Hettangian, Early Jurassic Etymology. From /athetic, meaning likely to escape notice Description. The dorsal ridge scale series of this species is very distinctive. An- teriorly, the bases of the scales are very round, and there is a long, very slender, posteriorly directed spine. Posteriorly, the spines are longer than they are on the anterior scales. The spine and only the central portion of the scale base are covered with ganoine (Fig. 21B). The last scale in the series, a large oval plate, abuts the first (unpaired) fulcrum on the leading edge of the dorsal fin. The total number of scales in the dorsal series, including the predorsal scale, is about 20. The body form is similar to that of S. olseni (Fig. 21A). The body is somewhat deep, with a moderately thick caudal peduncle. The slope of the forehead with respect to horizontal is about 45°. The only paratype, YPM 8841, is not complete, so although the existing portion is consistent with the holotype, it is not definitively the same shape. However, this individual differs in shape from the other two known species having the same type of dorsal ridge scale series. In the holotype, there are 34 lateral line scales. The pelvic, anal, and dorsal fins originate at the 8th, 17th, and 19th scale rows, respectively. The scales in the anterodorsal flank region are not clear enough to determine whether there are intercalated scale rows. The scales are largest in the anterior flank region and around the lateral line. They decrease in size both caudally and anterodorsally. The scales along the dorsal and ventral midline of the caudal peduncle are larger than the lateral caudal scales. In the holotype, there are 13 scales between the lateral line and the dorsal fin, and 10 from the lateral line to the origin of the anal fin. The total number of fulcra on the dorsal and anal fins is uncertain. The dorsal fin fulcra follow the usual Newark semionotid pattern in having three basal fulcra and two fringing fulcra lying against the unsegmented portion of the first lepi- dotrichium. Except in the pelvic fin, which has three lepidotrichia, the fins of the holotype are not well enough preserved to count fin rays. In all fins, the tips of the lepidotrichia extend beyond the tip of the last fin fulcrum. EARLY JURASSIC SEMIONOTID FISHES 63 Fic. 21. S$. latheticus. A, outline of body form traced from YPM 8899; B, camera lucida drawings of YPM 8899 (top) and YPM 8841 (bottom). Scale = 5 mm. Semionotus melanimus, new species Pigss2ZAN 22 B42 lable 32 Diagnosis. 8. melanimus has dorsal ridge scales similar to those of the S. micropterus group (Olsen and others 1982). The form of the caudal peduncle of this species is unusual among semionotids, as there is almost no constriction of the body posterior to the dorsal fin (Fig. 22A). Instead, the sides taper almost straight back to the minimum point at the posterior end of the caudal peduncle, the overall shape being more like a cone than an hourglass. Holotype. YPM 8840 (P4-2368) complete fish, part and counterpart; one side negatively prepared (Fig. 42) Paratypes. YPM specimens 8758, 8877, 8979 Type locality. Yale excavation in cycle P4, Pompton, New Jersey Formation. ‘Towaco Formation, Newark Basin Age. Hettangian, Early Jurassic Etymology. From melanimum, meaning black-clad, in reference to the black- colored bone characteristic of this and most fishes from the P4 excavation Description. The dorsal ridge scales of this species are like those described for S. latheticus. Anteriorly, the bases of the scale are very round with long, very slender, posteriorly directed spines that are longer on posterior scales than on anterior 64 PEABODY MUSEUM BULLETIN 43 Fic. 22. S$. melanimus. A, outline of body form traced from YPM 8840; B, camera lucida drawing of the dorsal ridge scales of YPM 8840. Scale = 5 mm. scales. Only the spine and the central portion of the scale base are covered with ganoine (Fig. 22B). A large oval plate at the end of the series abuts the first (unpaired) fulcrum on the leading edge of the dorsal fin. The total number of scales in the dorsal series, including the predorsal scale, is about 19. S. melanimus is a moderately deep-bodied species, with a large head and thick caudal peduncle tapering in an almost straight line from the dorsal and anal fins back to the minimum point of the caudal peduncle (Fig. 22A). The slope of the forehead with respect to horizontal is about 30—40°. In the holotype, there are 38 lateral line scales. The origin of the pelvic fin is not visible in the holotype, but the anal fin originates at the 21st scale row, and the dorsal fin at row 24. At least in the holotype, there are intercalated scale rows dorsal to the lateral line and anterior to the dorsal fin. Near the lateral line, anterior flank scales are largest, decreasing in size both caudally and anterodor- sally. In the caudal region, dorsal and ventral median scales are larger than lateral scales. The dorsal fin is fringed with 8 fulcra; the first three are basal fulcra, and the bases of the next two lie against the unsegmented portion of the first lepidotrich- ium. ‘There are 11 dorsal and 10 anal lepidotrichia in the holotype. The pelvic fins are not sufficiently well preserved to count fulcra or fin rays. The pectoral fin has 14 lepidotrichia, but the number of fulcra is uncertain. The caudal fin has 19 lepidotrichia, 10 of which insert dorsal to the lateral line. In all fins, the tips of the lepidotrichia extend beyond the tip of the last fin fulcrum. The caudal fin appears to be weakly emarginate (YPM 8979), as shown in Figure 22A. The tips of the dorsal and paired fins are not visible in any specimen, but as in other Newark semionotids, the anal fin reaches almost to the beginning of the caudal fin. EARLY JURASSIC SEMIONOTID FISHES 65 Semionotus johberryi, new species Figs’ 239A. 23'B, 42; Leable 32 Diagnosis. This species is distinguished from most other species of Semionotus by the morphology of its dorsal ridge scales (Fig. 23B), which are similar to those of the S. micropterus group (Olsen and others 1982). ‘The dorsal ridge scales of this species differ from other S. micropterus group scales in that the spines are wider and the supporting flange of bone not covered by ganoine is more robust. Compared with S. melanimus and S\ latheticus, which also have S$. micropterus group scales, S. johberry: is more slender in body form and its dorsal fin is more posterior. Compared with S. anosteus, which is most similar in body shape, the spines of 8. johberry: are much broader and the flange of bone underneath the spine is much larger. In addition, the dorsal fin of §. johberry: is more posterior than that of S. anosteus and the caudal peduncle is thus shorter. Holotype. YPM 8800 (P4-2612) complete fish, negatively prepared (Fig. 42); counterpart shows no detail Type locality. Yale excavation in cycle P4, Pompton, New Jersey Formation. ‘“Towaco Formation, Newark Basin Age. Hettangian, Early Jurassic Etymology. Named for John Strong Newberry, author of the classic monograph on semionotid fishes from Connecticut and New Jersey. Contraction of his first and last names was necessary because the name Semionotus newberryi has already been published (Loper 1893). Description. ‘The dorsal ridge scales are round at the base, with long, narrow spines on all but the first three or four scales (Fig. 23B). Certain features of the dorsal ridge scales, such as the roundness of the scale bases and the long, slender spines extending posteriorly well beyond the scale base resemble features of S. micropterus group scales (Olsen and others 1982). However, the spines on the dorsal ridge scales of YPM 8800 are supported by a bony extension of the scale base, whereas in fishes of the S. mzcropterus group the spines extend unsupported beyond the base. The spines on posterior scales are longer than they are on anterior scales. ‘The last scale in the dorsal scale series is a large oval plate that abuts the first (unpaired) fulcrum on the leading edge of the dorsal fin. The total number of scales in the dorsal series, including the predorsal scale, is about 21. The shape of this specimen is unique among the Semionotus complex (see Fig. 23A). It is probably most similar to that of S. anosteus, but in S. johberryi the body is deeper through the belly, the tail is shorter or the dorsal fin more posterior, and the head is shorter. The slope of the forehead relative to horizontal is about 305: In the holotype, there are 34 lateral line scales. The origin of the pelvic fin is not visible, but the anal fin originates at row 20 and the dorsal fin at row 21. In the holotype, a few intercalated scale rows are present dorsal to the lateral line and anterior to the dorsal fin. The scales are largest in the anterior flank region and near the lateral line. They decrease in size both caudally and anterodorsally. Dorsal and ventral median scales on the caudal peduncle are larger than lateral scales. From the lateral line, there are 8 scales to the dorsal fin and 7 to the origin of the anal fin. The fin fulcra can be counted only on the anal fin, which has 7. Pelvic and 66 PEABODY MUSEUM BULLETIN 43 a Rr Fic. 23. S. johberry:. A, outline of body form traced from YPM 8800; B, camera lucida drawing of the dorsal ridge scales of YPM 8800. Scale = 5 mm. pectoral lepidotrichia are not well preserved in the holotype, but in the dorsal and anal fins there are 15 and 10, respectively. The posterior portions of all fins are truncated well before their tips. THE SEMIONOTUS TENUICEPS GROUP One of the first semionotids from North America (Hitchcock 1819; Agassiz 1835), S. tenuiceps (Agassiz) is one of the most distinctive semionotids, distinguishable by its very elaborate dorsal ridge scales, previously termed robust or globular scales (Fig. 9E, F). Semionotids having such scales have been found in the Jurassic of both eastern North America and China (Olsen and others 1982). In North America, they have been found in the Turner’s Falls Sandstone in the Deerfield Basin of Massachusetts; the Feltville and ‘Towaco Formations in the Newark Basin of New Jersey (Olsen and others 1982); and the Waterfall Formation in the Culpeper Basin of Virginia (Hentz 1981). From China, isolated robust dorsal ridge scales have been found in the lower Lufeng (Olsen and others 1982). The morphology of these elaborate dorsal ridge scales is sufficiently distinctive to suggest that semionotids having these scales form a monophyletic group. I refrain from giving this complex a formal subgeneric or generic designation, however, because to do so would result in Semionotus becoming paraphyletic. Instead, I refer informally to species in this group as the S. tenuwiceps group. The derived characters that unite members of this group are 1) anterior (be- ginning with 2nd to 6th), dorsal ridge scales greatly enlarged, 2) base of dorsal ridge scale accounting for bulk of scale, 3) spines on anterior dorsal ridge scales point dorsally before angling posteriorly, and 4) spines barely extending beyond the scale base. Within the S. fenuiceps group there are three variants on the generally elaborate morphology of the dorsal ridge scale series. The first pattern is best exemplified by S. tenuiceps (Fig. 24A). The first two scales are small and spineless, but beginning with about the third scale, there are EARLY JURASSIC SEMIONOTID FISHES 67 Fic. 24. Variation in dorsal ridge scales of the S. tenuiceps group. A, classic “tenuiceps” dorsal ridge scales; camera lucida drawing of the neotype, S. tenuiceps, YPM 8162; Sunderland, Massachusetts. B, hypermorphic “‘tenuiceps” dorsal ridge scales, YPM 6960; Turner’s Falls, Massachusetts. C, moderate or globular “‘tenuiceps” dorsal ridge scales (YPM 8932); Yale excavation in Wayne, New Jersey. Note the size of the dorsal ridge scales relative to the flank scales in each type. Scale = 1 cm. about 6 to 9 robust, angular scales. These scales are broad at the base in an anterior-posterior direction and 2 to 3 times the width of a normal flank scale. Dorsally and along an anterior-posterior axis, the spines are convex. In lateral view, they are undercut slightly beneath the spine and the bases of the scales expand ventrally. The posteriormost scales are simple and convex with well- developed spines, each one overlapping the scale behind it. Very much like this robust series are hypermorphic scales, which differ from robust scales by their much larger size relative to flank scales and the fish itself (Fig. 24B). Semionotus with these exaggerated dorsal scales are known from the Feltville Formation in the Newark Basin (Olsen and others 1982) and the Tur- ner’s Falls Sandstone in the Deerfield Basin, but not from the Towaco Formation of the Newark Basin. Thus, no descriptions of species with hypermorphic scales are included here. In a third version of the S. tenwiceps-type dorsal ridge scale series, the globular form, enlarged scales do not begin until about the 6th scale posterior to the extrascapulars (Fig. 24C). Fish with this moderate scale pattern seem to be missing the several most exaggerated scales (3rd to 6th) in the robust series. From about the 6th to 12th scales, the scale bases are large and bulbous and the spines point 68 PEABODY MUSEUM BULLETIN 43 Fic. 25. Skull of S. tenuiceps. Camera lucida drawing of neotype, YPM 8162; Sunderland, Mas- sachusetts. Scale = 1 cm. Abbreviations: br, branchiostegal rays; d, dentary; ce, ceratohyal; cl, cleithrum; dpt, dermopterotic; dsp, dermosphenotic; ect, ectopterygoid; ex, extrascapular; fr, frontal; io, infraor- bitals; iop, interopercular; mpt, metapterygoid; op, opercular; p, parasphenoid; par, parietal; pmx, premaxilla; pop, preopercular; pt, posttemporal; q, quadrate; sub, suborbital. first dorsally, then posteriorly. Again, the posteriormost scales have flatter bases and longer spines, each one overlapping the scale behind. Skull The pattern of dermal bones in the skull of the S. tenuiceps group shows little deviation from the general pattern described for the S. elegans group (Olsen and McCune ms), and for Semzonotus (Schaeffer and Dunkle 1950); it is described only briefly here. Skulls from individuals of two different species of the group, S. tenuiceps (Fig. 25), with robust dorsal ridge scales, and Semionotus saginatus (Fig. 26), with globular dorsal ridge scales, are figured here. The skull roof shows the only deviation from the general Newark semionotid pattern. The parietals are rectangular, almost square, and join medially. ‘The frontals are narrow anteriorly in the S. tenwiceps group (Olsen and McCune ms), and the S. bergert group including S. (Lepidotes) minor (Woodward 1916-19; McCune 1986), and Semionotus normanniae (Larsonneur 1964). This contrasts to the broad frontals found in Semionotus brauni (Olsen and others 1982), Lep- idotes elvensis (Deschaseaux 1943; Wenz 1967), L. mantelli (Woodward 1916- 19) and L. laev. (MNHP 1917-05). In the S. tenwiceps group, the suture between the frontals is digitate as in the S. elegans group, but between the frontals and parietals the suture is almost straight rather than digitate as in the S. elegans EARLY JURASSIC SEMIONOTID FISHES 69 Fic. 26. Skull of S. saginatus. Camera lucida drawing of holotype, YPM 8932 from Yale P4 excavation in Wayne, New Jersey. Scale = 1 cm. For abbreviations, see legend for Fig. 25. group (Olsen and McCune ms). The pattern of the snout is like that described for the S. elegans group (Olsen and McCune ms). The circumorbital series is complete and composed of 2 or 3, usually 3, su- praorbitals (which may or may not be tuberculated), the dermosphenotic, 5 or 6 infraorbitals, the lachrymal, and 2 infraorbitals between the lachrymal and the antorbital. The infraorbitals within the circumorbital ring seem to be slightly thinner than in the S. e/egans group, but the lachrymal and anterior infraorbitals are similarly deep. The dermopterotic and the pattern of the lateral line canals are like those of the S. elegans group. Like other non-Lepidotes semionotids, including the S. elegans group, Semionotus capensis (Olsen and McCune ms), 8. normanniae, and S. bergeri (McCune 1982), the cheek region is completely open except for an oval suborbital positioned between the circumorbitals and preoperculum and below the dermopterotic. The Jaw joint is like that described by Patterson (1973) for L. toombsi, and all other Newark semionotids (Olsen and McCune ms), S$. normanniae, S. bergeri (McCune 1982). The palate, the hyoid arch, the opercular series, and the pectoral girdle show no deviation from the pattern seen in the S. elegans group. I have seen very little of the gill arches, braincase, and other endoskeleton and will not attempt to describe them. 70 PEABODY MUSEUM BULLETIN 43 o : ee. ‘ Fic. 27. Variation in body shape of species in the S. tenwiceps group. Outline drawings based on complete specimens, in most cases the holotype. A, S. greenwoodi YPM 8970; B, S. decoratus YPM 8957; C, S. fundus YPM 8927; D, S. tenuiceps YPM 8162; E, S. saginatus YPM 8932. F, S. profundus YPM 8944. Body Shape The several species in the S. ¢enuiceps group described here vary substantially in body shape (Fig. 27), from slender to very deep bodied, ranging in body depth from about 26-52% of standard length. The slope of the forehead relative to horizontal is 30—50° but usually less than 40°. Some species have a pronounced dorsal hump directly behind the head, whereas in others the outline of the back curves gently from the skull to the dorsal fin. Head length is somewhat variable over all species (27-35% of standard length), but most are in the range of about 30-35% of standard length. Caudal peduncle depth tends to be variable (25-54% of DFCD), and correlated with body depth. Of the fish measured, primarily from the ‘Towaco Formation of the Newark Basin in New Jersey, the maximum size is about 17 cm standard length. These data are summarized in Table 25. Squamation Species in the S. tenuiceps complex are completely covered by a fabric of inter- locking rhomboid scales with smooth posterior margins. The scales are generally largest anteriorly near the lateral line. They decrease in size ventrally, caudally, and dorsally. In some species, for example Semionotus fundus and Semionotus decoratus, there may be intercalated scale rows in the epaxial region anterior to the dorsal fin. Unlike the intercalary scale rows in macrosemiids described by Bartram (1977), these extra scale rows are not intraspecifically consistent nor is the pattern the same on both sides of the same individual. They are, however, more common in some species than others. The lateral line is complete and relatively straight and passes through midflank. Lateral line scales number 32- 36. The median scales on the caudal peduncle are enlarged relative to the lateral caudal scales. These scales are in linear series with the paired fulcra fringing the caudal fin. Both the dorsal and anal fins are preceded by a large oval scale, which EARLY JURASSIC SEMIONOTID FISHES Wi Table 25. Summary of morphometric data for the Semionotus tenuiceps group. MAXCD** = MAXCD x 100/DFCD. MAXCD* = MAXCD x 100/AFCD. All other variables expressed as % SL, for SL = 7.4 to 16.8 cm. Variable N Mean Standard Minimum Maximum deviation value value DPTH 22 40.06 7.49 26.72 52.38 HDL 20 315.136 2.26 26.60 34.65 PDL 20 64.96 3.46 5922 70.65 DFPV 20 Si2Zicigie 6.23) PRK val 42).55 DFAN 20 Sassou! 4.78 21.82 40.43 DFCD 22 42.29 2.64 36.90 AF Dy, AFCD 20 28.50 1.89 24.14 32.10 HDD 22 ZT tee sya 7/S3 19.66 34/052 MAXCD 21 i. Sif SiGiaal 10.91 23.81 MINCD 22 15310 2.68 10.91 20.24 PTAL 20 47.03 SOW 44.14 50.42 DFPT 22 44.17 5.04 36.03 52.78 DFB 5 G6 Di) 2.30 lp ya tsy72 20.59 AFB 18 10.43 1.48 Votl® 375i! MAXCD** 29 42.30 7.64 25.00 54.05 MAXCD* 28 61.74 9.84 35.29 80.95 is about twice the length of an ordinary flank scale. The predorsal scale abuts the first fulcrum of the dorsal fin, and the fin itself originates between the 19th and 21st vertical scale row, and 9-12 scales above the lateral line. The enlarged preanal scale lies a bit in front of the anal fin and vent. The anal fin originates behind the 17th to 19th vertical scale row, but usually two rows in front of the dorsal fin and from 9-12 scales below the lateral line. The pelvic fins originate behind rows 7-9. Fins All fins are fringed by fulcra and are composed of paired lepidotrichia that are unsegmented at their bases and segmented distally. The ends of the segmented rays are difficult to see in most specimens because they branch at least 2 times and feather out toward their tips, but they always extend beyond the last fringing fulcrum. The length of the dorsal and anal fin bases is about 14-19% and 9-12% of standard length, respectively. Although the dorsal fin originates about 2 scale rows behind the anal fin, it is placed slightly in front of the anal in lateral view. In both the dorsal and anal fins, there is a single unpaired basal fulcrum, two paired basal fulcra, and up to nine paired fringing fulcra, of which two or three lie against the basal unsegmented portion of the first lepidotrichium. Lepidotrichia number 7 to 9 in both fins. The pectoral fins have up to 17 lepidotrichia and are fringed by about 7 fulcra. The pelvic fins, originating between scale rows 7-9, are composed of 3—5 lepi- dotrichia and fringed by up to 6 fulcra. There are 15-18 caudal lepidotrichia in the caudal fin. Lepidotrichia dorsal to the lateral line are usually equal in number or greater than lepidotrichia ventral to the lateral line. The most dorsal ray is probably a continuation of the most posterior scale in the axial lobe of the tail (Olsen and McCune ms). Semionotus tenuiceps Agassiz Figs. 24, 27, 28, 29, 43; Tables 26, 32 Eurynotes tenuiceps Agassiz (1835) Paleoniscus latus J. H. Redfield (1837) Wo: PEABODY MUSEUM BULLETIN 43 Paleoniscus agassiz W. C. Redfield (1841) Ischypterus latus Egerton (1850) Ischypterus tenuiceps ceratocephalus Emmons (1857) Eurinotus ceratocephalus Emmons (1860) Ischypterus latus Traquair (1877) Ischypterus tenuiceps Newberry (1888) Ischypterus latus Newberry (1888) Semionotus tenuiceps Woodward (1895) Type locality. Whitmore’s Ferry, Sunderland, Massachusetts Collector. Edward Hitchcock Formation. ‘Turner’s Falls Sandstone, Deerfield Basin, Massachusetts Age. Early Jurassic Holotype. Agassiz did not designate a holotype for “Eurynotes tenuiceps” ex- plicitly, but the specimen he figured (Fig. 43) from Murchison’s private collection (collected by Edward Hitchcock) has served as the holotype (Woodward 1895). According to Agassiz (1835) and later Woodward (1895), the specimen, which by then had become S. tenuiceps, had been deposited in the Geological Society of London (now BGS.GSM) by Murchison. Today, however, there is no record of its ever having been there (the fate of a fair portion of Murchison’s collection is unknown (Andrews 1982)). Without success, I have searched for Agassiz’s figured specimen of Semionotus (= Eurynotes) tenuiceps at the Geological Society of Lon- don, the British Museum (Natural History), the Institut de Geologie de 1’Uni- versite de Neuchatel, the Museum of Comparative Zoology, and the Pratt Mu- seum, Amherst College (which holds Hitchcock’s collection). In addition, I have examined Agassiz’s notes and the manuscript for Recherches sur les Poissons Fossiles at l’Archiv de l’Etat in Neuchatel, as well as selected Agassiz correspondence there and at the Museum of Comparative Zoology and the Houghton libraries of Harvard University for mention of the whereabouts of the specimen. The holotype of S. tenuiceps must be considered lost. No syntypes are available, as Agassiz used only the holotype for his description. Therefore, I designate a neotype specimen to serve as the standard of comparison for S. tenuiceps (see below). Neotype. YPM 8162 (Fig. 43; see also fig. 11b in Olsen and others 1982) complete specimen, negatively prepared, counterpart unknown Collector. Unknown, possibly J. H. Redfield Neotype locality. Whitmore’s Ferry, Sunderland, Massachusetts Formation. Lower Turner’s Falls Sandstone Age. Hettangian, Early Jurassic Neoparatypes. YPM (all from the P4 excavation) 8921, 8945, 8946, 8947, 8948, 8949, 8950, 8961, 8972, 8981 Diagnosis. A moderately deep-bodied Semionotus with a pronounced dorsal hump and very large helmet-shaped scales along the dorsal midline. Relative to others in the S. tenuiceps group, this species is deeper bodied than Semionotus greenwood, S. fundus, and §. decoratus but more slender than S. saginatus and Semionotus profundus (all described below; Fig. 27). S. tenuiceps is distinguished from the species most similar in body form, S$. fundus, by its pronounced dorsal hump and robust anterior dorsal ridge scales. The largest dorsal ridge scales begin two to EARLY JURASSIC SEMIONOTID FISHES UE Fic. 28. S. tenuiceps dorsal ridge scales from YPM 8162. three scales posterior to the extrascapulars rather than at about the sixth scale behind the extrascapulars, as is the case in S. greenwoodi, S. fundus, and S. saginatus. Distribution. Sunderland, Massachusetts, and Turner’s Falls, Massachusetts, in the Deerfield Basin, Hettangian; Pompton (and probably Martinsville and East Round Top), New Jersey, in the Newark Basin, Hettangian; Haymarket, Vir- ginia, in the Culpeper Basin (Hentz 1981), Sinemurian. Description. The neotype of S. tenuiceps has well-preserved robust dorsal ridge scales (Fig. 28; Olsen and others 1982). The first two scales behind the extrascapu- lars are unmodified. ‘The next three are very large helmet-shaped scales (Fig. 28) that expand basally along an anterior-posterior axis. The ganoine-covered spines point first dorsally and then curve posteriorly. Except for its very tip, the spine is not a free-standing extension of the scale base; rather, it has the appearance of being elevated dorsally by a large mass of bone not covered by ganoine. From about the 6th to 12th dorsal ridge scale, the scale base is reduced in size. The spines still point in a dorsoposterior direction, but more posteriorly and less dorsally than the large helmet-shaped scales in front of them. The most posterior scales are simple, convex, and have spines extending beyond the scale base and overlapping the scale behind it in the series. The total number of dorsal ridge scales is about 19, including the predorsal scale. S. tenuiceps is relatively deep bodied, with a pronounced dorsal hump and a deep belly (Fig. 29D). The forehead slopes at about 30—40° relative to horizontal. The caudal peduncle is narrow relative to S. profundus and S. saginatus and narrows more rapidly from the dorsal and anal fins toward the caudal fin than does S. profundus. The species most similar in body shape to S. tenwiceps are easily distinguished by their globular dorsal ridge scales, and they also differ slightly in body shape. Figure 29 (C, D) illustrates the difference in body shape between S. tenuiceps and S. fundus as well as S. saginatus, while taking into account intraspecific variability by showing a scatter plot of the distribution of fin positions. Morphometric data are summarized in Table 26. A caveat is necessary at this point: In compliance with Art. 75c-5 (ICZN 1964), the neotype of S. tenuiceps is a specimen from Sunderland in the Deerfield Basin. However, the neoparatypes are all from the Newark Basin. The body depth relative to standard length for the neotype is on the high side of the range for body depth among the neoparatypes from the Newark Basin. With more specimens, the depth-to-length ratio might 74 PEABODY MUSEUM BULLETIN 43 Fic. 29. Comparisons of body shape, S. tenwiceps group. For each comparison, the species that corresponds to the solid line and solid symbol is given first. A, S. greenwood: versus S. decoratus; B, S. fundus versus S. decoratus; C, S. fundus versus S. tenuiceps; D, S. saginatus versus S. tenuiceps; E, S. saginatus versus S. profundus. actually be bimodal or trimodal, and S. tenuiceps as described here could include two species. Given the geographic and stratigraphic limitations of this work, uncertainty is unavoidable. Flank scales covering the anterior portion of the body and near the lateral line are the largest. Caudally, dorsally and ventrally, the scales decrease in size. On the neotype, some of the scales in the predorsal region dorsal to the lateral line were lost in preservation, but on the area remaining there are no intercalary scale rows. Extra rows do occur rarely in S. tenuiceps from the Yale P4 excavation in Pompton, New Jersey (for example YPM 8949). However, as some semionotids have different numbers of intercalary scale rows on the left and right side of the same fish (Olsen and McCune ms), the presence of intercalary rows is not tax- onomically significant at the species level (which does not preclude ecological significance between localities or the significance of a developmental tendency to produce extra scale rows that is statistically demonstrable at some higher taxo- nomic level). Lateral line scales number 33-36, usually about 34. There are nine vertical scale rows anterior to the pelvic fin, 18-19 to the anal fin, and 21 to the dorsal fin. Between the 21st lateral line scale and the dorsal fin there are 10 scales. Below the lateral line, there are 9 scales to the origin of the anal fin. The dorsal fin has 10-12 rays and is fringed by about 9 fulcra. The first three are basal fulcra, the next three le against the unsegmented portion of the first ray, and the remaining three or more fringe the segmented portion of the first lepidotrichium. Anal rays number 7-10. On the anal fin there are three basal fulcra and about 6 fringing fulcra, 3 of which fringe the unsegmented lepidotrichi- um base. The best view of a pectoral fin (YPM 8948) shows 17 lepidotrichia and about 7 fulcra. The pelvic fins on this same specimen appear to have 5 rays fringed with 6 fulcra. When appressed to the body, the tips of the pectoral fins reach the origin of the pelvic fins, and the pelvic fins extend as far as the preanal scale. EARLY JURASSIC SEMIONOTID FISHES WS Table 26. Summary of morphometric data for Semionotus tenuiceps . MAXCD* = MAXCD x 100/DFCD. MAXCD** = MAXCD x 100/AFCD. All other variables expressed as % SL, for SL = 10.1 to 16.8 cm. Variable N Mean Standard Minimum Maximum deviation value value DPTH 4 39.18 3162 3662/1 44.05 HDL 4 S746 Sial. 1.84 30.36 34.65 PDL 3 66.61 oe 64.66 67.86 DFPV 3 325310 2.59 29.31 33.93 DFAN 4 30.97 0.94 29.63 31.68 DFCD 4 41.08 2.79 36.90 42.59 AFCD 4 CM Th} 0.82 26.72 28.70 HDD 4 PAU GAS) 1.79 25.00 28.71 MAXCD 4 16.88 ba Yy/ TS 52 18.45 MINCD 4 14.60 0.96 LS a9. 15.84 PTAL 4 47.48 4 45.54 48.81 DFPT 4 43.05 Bh5 ALS 40.59 47.62 DFB 4 15.94 Zs aleve alo) 18.10 RAFB 4 10.05 0.63 aaa Oval MAXCD* 9 42.24 5 1'6 SOri4i2 50.00 MAXCD** 9 60.89 3755 54.84 65.96 Good skull material is limited. YPM 8162 affords a particularly good view of a reasonably complete articulated skull (Fig. 25). See the description (above) or the description of the S. elegans group (Olsen and McCune ms). Semionotus greenwoodi, new species Figs. 27, 29, 30, 44; Tables 27, 32 Diagnosis. S. greenwood: is distinguished from all other species of the S. tenuiceps group by its very slender body form, short head, and smoothly sloping profile from the nape onto the skull roof. The slender body form of S. greenwoodi relative to all others is reflected quantitatively by the values of eight linear measurements, taken as percentages of standard length. The values of all these ratios are smaller than corresponding ratios for all other species of the S. tenuiceps group. Except for head length (HDL), all measurements (DFPT, DFPV, HDD, MINCD, MAXCD, DPTH, DFAN) are highly correlated with body depth and also related to fin position along the anterior-posterior axis. The largest dorsal ridge scales of this species are similar to those of §. fundus and S. saginatus and not as large as S. tenuiceps, S. decoratus, or S. profundus. Holotype. YPM 8970 (P4-2180) complete fish, mechanically prepared; coun- terpart is a negative impression (Fig. 44) Paratypes. YPM 6586, 8959, 8960, 8967, 8968, 8969, 8971, 8972, 8973, 8974, 8976, 8977 Type locality. Yale excavation in cycle P4, Pompton, New Jersey Formation. ‘Towaco Formation, Newark Basin Age. Hettangian, Early Jurassic Etymology. Named for P. H. Greenwood, in honor of his outstanding contri- butions to the study of the species flocks of cichlid fishes living in the African rift lakes Description. The dorsal ridge scales of $. greenwoodi are large relative to the flank scales, as in S. tenuiceps, but the largest scales in the series are more posterior 76 PEABODY MUSEUM BULLETIN 43 Fic. 30. S. greenwood. A, dorsal ridge scales of S$. greenwoodi, holotype YPM 8970; B, frontals and parietals of YPM 8970. Scale = 1 cm. than they are in S. tenuiceps (Fig. 30). The first 3 scales lack spines completely, and the 4th and 5th scales have very broad, short spines. ‘The 6th through 12th scales have large spines and bulbous bases, with only the spines covered by ganoine. Posterior to the 12th dorsal ridge scale, the scale bases become more rounded and flattened with very prominent long spines, very much like the posterior scales of the §. micropterus group (Olsen and others 1982). The last scale in the series is a large oval plate that abuts the first unpaired fulcrum on the leading edge of the dorsal fin. The total number of scales in the dorsal series, including the predorsal scale, is about 22. This species is the most slender and streamlined of the S. tenuiceps group. ‘The head is small relative to others in the S$: tenuiceps group, and the slope of its forehead with respect to horizontal is about 30—40°. Overall body form is most clearly seen by comparison of outline drawings (Fig. 27) of each of the six species in this genus. By visual inspection or by superimposing a tracing of S. greenwoodi on outlines of the other species (all at the same standard length), one can easily see that S$. greenwoodi is more slender than any other species (Fig. 29). These morphometric data are summarized in Table 27. There are 33 or 34 (usually 34) lateral line scales. The pelvic fin originates at the 8th or 9th (usually 9th) scale row, the dorsal fin at row 20 or 21, and the anal fin at row 18 or 19. The scales along the dorsal and ventral midline of the caudal peduncle are enlarged. Rarely, there are intercalated scale rows dorsal to the lateral line and anterior to the dorsal fin. The scales are largest in the anterior flank region and around the lateral line. ‘They decrease in size both caudally and anterodorsally. From the lateral line, there are 8 scales to the dorsal fin and about 9 to the origin of the anal fin. The dorsal and anal fins are fringed with 7—9 fulcra; the first 3 are basal fulcra; the bases of the next 2 lie against the unsegmented portion of the first lepidotrichi- um. There are about 7-11 dorsal and 7-8 anal lepidotrichia. The pectoral and pelvic fins are fringed by at least 6 fulcra. Pelvic lepidotrichia number 3-4, and pectoral lepidotrichia number about 11. In all fins, the tips of the lepidotrichia extend beyond the tip of the last fin fulcrum. The holotype shows all of the fins well except the pectorals. The pectoral fins, when appressed to the body, extend almost to the origin of the pelvics (YPM 6568), and the pelvics extend just to the origin of the anal fin. The anal fin in turn extends just beyond the ventral origin of the caudal fin. The posterior margin of the dorsal fin reaches at least to the region where the scales along the dorsal midline of the caudal peduncle begin to enlarge, or about halfway between the dorsal and caudal fins. EARLY JURASSIC SEMIONOTID FISHES Gi Table 27. Summary of morphometric data for Semionotus greenwoodi. MAXCD* = MAXCD x 100/DFCD. MAXCD** = MAXCD x 100/AFCD. All other variables expressed as % SL, for SL = 11.0 to 11.7 cm. Variable N Mean Standard Minimum Maximum deviation value value DPTH 3 28.27 1.60 26%.,72 29.91 HDL 3 28.55 1.48 Ua 30.17 PDL 3} 61.48 2.28 60.00 64.10 DFPV 3 23%5311 1.29 22 4! 24.79 DFAN 3 24.44 24d PRL e374 26.72 DFCD 3 473 1.76 AOA 7. 43.64 AFCD 3 27.18 3.44 24.14 30.91 HDD 3 20.42 0.67 19.66 20.91 MAXCD 3 kato} 0.96 10.91 D282 MINCD 3 IEA SKS) 0.54 10.91 alae e)7/ PTAL 3 47.18 Bats) 44.55 50.00 RDFPT 3 39.06 1.98 237/ Go}7/ 41.03 RDFB 2 14.60 2.52 P2382 16.38 RAFB 3 36 7/83 Pe si4 TaV/® 10.26 MAXCD* 3 28.69 3.48 25.00 Syabs evil MAXCD** 3 44.56 8.07 35129 50.00 The skull does not differ significantly from the S. elegans group (Olsen and McCune ms) or from Semionotus (McCune 1986). The frontals are narrow anteriorly, with the preorbital portion longer than the posterior portion. The parietals are simple and rectangular (Fig. 30B). The teeth are simple and conical. Semionotus decoratus, new species igs2 6.29.15. 44- WablesiZ8, 32 Diagnosis. Body deeper (about 35% of standard length) than S. greenwood: but more slender than all other species. Head longer than S. greenwoodi. Robust scales appear to occur almost immediately behind the extrascapulars, but as the anterior dorsal scales are very tiny, the first robust scale is probably the fourth dorsal scale. In lateral view, the slight hump of the anterior, dorsal portion of the body also distinguishes this species from the two species most similar in body shape, S. greenwoodi and S. fundus Holotype. YPM 8957 (P4-222a); complete fish, mechanically prepared (Fig. 44). I have chosen YPM 8957 as the holotype over YPM 8953, though in many ways the latter is a better specimen. YPM 8953 has been negatively prepared in hydrochloric acid and therefore shows the details of the skull, squamation, and fins especially well. However, the view of the dorsal hump and of the dorsal ridge scales, which are important species-level characters, is superior in YPM 8957 Paratypes. YPM 8953, 8955, 8956 Type locality. Yale excavation in cycle P4, Pompton, New Jersey Formation. ‘Towaco Formation, Newark Basin Age. Hettangian, Early Jurassic Etymology. From decoratus, meaning adorned, for the elaborate dorsal ridge scales of this species Description. Superficially, robust scales appear to begin almost immediately pos- terior to the extrascapulars, but on close inspection the nape region is populated by many tiny scales and the S. tenuiceps-type scales do not begin until about the 78 PEABODY MUSEUM BULLETIN 43 Fic. 31. S. decoratus. Dorsal ridge scales of holotype specimen YPM 8957. Scale = 5 mm. Ath scale in the dorsal series (Fig. 31). This position corresponds to the first or second complete vertical scale row behind the cleithrum. There are some (probably 3) incomplete scale rows dorsal to the lateral line and anterior to the first large dorsal ridge scale. Following that there are about 5 robust scales, and then from about scales 11-18 the bases are flat and the elongated spines of each scale overlap the scale behind. S. decoratus is an ordinary fusiform-shaped fish, with a caudal peduncle that is narrow even close to the anal fin (Fig. 27B). It is slightly deeper bodied than S. greenwood and more slender than all other species of in the S. tenuiceps group. Figure 29 (A, B) illustrates the difference in form of this species relative to S. greenwood and S. fundus. The maximum depth of the caudal peduncle relative to both standard length and tail length is more slender than in all but S. greenwood. Ratio of head length to standard length is large compared with that of S. green- woodi, and the slope of the forehead is about 30—35° relative to horizontal. Mea- surements of overall body form are summarized in Table 28. There are 33-34 lateral line scales. The pelvic fin originates at the 8th—9th scale row, the anal fin originates at row 18-19, and the dorsal fin at row 20-21. The scales along the dorsal and ventral midline of the caudal peduncle are enlarged and form a series with the fulcra fringing the caudal fin. There are frequent intercalated scale rows (9 in YPM 8953) in the area dorsal to the lateral line and anterior to the dorsal fin, especially in the nape region, where the scales are very tiny. The flank scales are largest in the anterior flank region and decrease In size caudally, dorsally and ventrally. Both the dorsal and anal fins are fringed with 7—9 fulcra, 3 of which are basal fulcra, and two more lie against the unsegmented portion of the first lepidotrichi- um. At least 5 fulcra fringe the pectoral fin and 4 fulcra fringe the pelvic fin, but these are minimum counts as no specimen of this species has sufficiently well- preserved paired fins to merit such confidence in fulcra counts. There are 9-12 lepidotrichia in the dorsal fin, 7—9 in the anal fin, 3 in the pelvic fins, and about 13 in the pectoral fins. The pectoral fin count is based on only one, albeit very well-preserved, specimen. If this count is correct, the pectoral fin ray count is very different from the usual 17. Dorsal and pectoral fin lengths for the outline drawing (Fig. 27B) are from YPM 8955; caudal, pectoral, and pelvic fin lengths are from YPM 8953. The skull of the holotype is not informative. However, the detail of the skull of YPM 8953 is excellent. It clearly shows the single suborbital characteristic of Semionotus (McCune 1986) and S. elegans group (Olsen and McCune ms). EARLY JURASSIC SEMIONOTID FISHES 79 Table 28. Summary of morphometric data for Semionotus decoratus . MAXCD* = MAXCD x 100/DFCD. MAXCD** = MAXCD x 100/AFCD. All other variables expressed as % SL, for SL = 11.1 to 13.6 cm. Variable N Mean Standard Minimum Maximum deviation value value DPTH 3 34.89 1.40 33.35 36.04 HDL 3 32152 1.42 30.88 3313/3 PDL 3 62.93 QSL 7 61.26 6S ii DFPV 2 26.83 155 25.74 29'S DFAN 3 29.08 1.40 27.93 30.63 DFCD 3 44.19 0.83 43.38 45.05 AFCD 3 27.69 0.99 27.103 28.83 HDD 3 26.60 1595 25'.23 28.83 MAXCD 3 14.92 eS 13.24 LG 22 MINCD 3 14.38 1.86 12. 50 UG6.i22 PTAL 3 44.68 0.47 44.14 45.05 DFPT 3 38.74 2.39 36.03 40.54 DFB 3 18.87 baie! Lefeealia: 20.59 AFB 3 10.58 0.59 9.91 11.03 MAXCD* 3 SISO FAS) Srenl2 30.51 BeI5 f/S) MAXCD** 3 53.92 le 7h?) 48.65 60.00 Semionotus fundus, new species Figs, 27, 29A, 29B, 32, 44 aables, 29, 52 Diagnosis. S. fundus is deeper bodied than S. greenwood and S. decoratus and more slender than either S. profundus or S. saginatus (Figs. 27, 29B, C). This species differs from the two species most similar in shape, S. decoratus and S. tenuiceps, by having the globular variety of S. tenwiceps-type dorsal ridge scales and by lacking the dorsal postcranial hump shared by S. decoratus, S. tenuiceps, and S. profundus. It differs from all other species in the S. tenwiceps group by having only 32 lateral line scales. Holotype. YPM 8926 (P4-1707); complete fish, no counterpart, mechanically prepared (Fig. 44D) Paratypes, YPM 8925, 8927 Type locality. Yale excavation in cycle P4, Pompton, New Jersey Formation. ‘Towaco Formation, Newark Basin Age. Hettangian, Early Jurassic Etymology. From fundus, meaning foundation, in acknowledgment of funds for this research provided by the National Science Foundation Description. The dorsal ridge series is the moderate, globular form (Fig. 32). The 5 anterior scales are slightly enlarged and rounded but spineless. Enlarged scales begin about the 5th or 6th scale, and beyond the 10th—12th the scale bases become smaller and flatter. As the scale bases decrease in size posteriorly, the spines elongate and extend far beyond the scale margin, overlapping the scale posterior to it. The total number of dorsal ridge scales, including the predorsal scale, is 19. The deeper body of this species relative to S. decoratus and S. greenwoodi is reflected in larger values of DF PV and DFAN relative to standard length (Table 29; Figs. 27, 29B, C). The back slopes gently to the skull roof, with no postcranial dorsal hump, and the slope of the forehead is about 35—40° relative to horizontal. 80 PEABODY MUSEUM BULLETIN 43 Fic. 32. S. fundus. Dorsal ridge scale series of S. fundus, holotype YPM 8926. Scale = 1 cm. The caudal peduncle is deeper anteriorly, and its maximum depth relative to standard length is greater than that of either S. greenwood: or S. decoratus. In all individuals, lateral line scales number 32, fewer than any other species in the §. tenuiceps group. There are 7-8 scale rows to the pelvic fins, 17 rows anterior to the anal fin, and 19 rows to the dorsal fin. Between the lateral line and the dorsal fin there are 9 scales; between the lateral line and the anal fin there are 8. As in some other semionotids (Olsen and McCune ms), the pattern of intercalated scale rows is not the same on both sides of the fish. For example, in YPM 8927, there are 9 on one side of the fish and 6 on the other. In addition, the holotype shows the rare instance of intercalated rows ventral to the lateral line. The scales along the ventral and dorsal midline of the caudal peduncle are enlarged. Flank scales are largest in the central anterior flank region and become smaller dorsally, ventrally, and caudally. The dorsal and anal fins are fringed with about 7-8 fulcra, the first three being basal fulcra and two more lying against the unsegmented portion of the first lepidotrichium. There are about 13 lepidotrichia in the dorsal fin and 10 in the anal fin. The caudal fin consists of 16 lepidotrichia, split equally in number between dorsal and ventral. The pectoral fin consists of 3—4 lepidotrichia, fringed with about 5 fulcra. The best skull is found on YPM 8927, and it does not differ from other species in the group in any significant respect. Table 29. Summary of morphometric data for Semionotus fundus. MAXCD* = MAXCD x 100/DFCD. MAXCD** = MAXCD x 100/AFCD. All other variables expressed as % SL, for SL = 9.5 to 13.5 cm. Variable N Mean Standard Minimum Maximum deviation value value DPTH 3 38.64 0.87 37.84 39.56 HDL 2 32325 1.02 shal s} 32.97 PDL 2 64.10 3.63 61.54 66.67 DFPV 3 SVG! Pha S YH: 29.63 SOW DFAN 2 31.90 SIR (O)7/ 29.73 34.07 DFCD 3 Abels aoe 40.00 42.86 AFCD 2 29.15 0.82 28.57 PASSE TS! HDD 3 Po Alt 355 23.70 SHO 7/7/ MAXCD 2 Ores 5 0.61 18.92 19.78 MINCD 3 14.44 onda SSS 16.48 PTAL Z 45.60 0.78 45.05 46.15 DFPT 3 laset/ at Ike Az 41.76 44.14 DFB 2 15.45 1.46 14.41 16.48 AFB 2 ata bes (aXo) 1.54 9.91 12.09 MAXCD* 2 46.41 0.36 46.15 46.67 MAXCD** 2 66.43 3.96 63.64 69.23 EARLY JURASSIC SEMIONOTID FISHES 81 Semionotus profundus, new species Figs:2 7/429; 33,0445 lables 30,32 Diagnosis. $. profundus is the most deep-bodied species of the S. tenuiceps group. In lateral view, the large dorsal hump posterior to the head and the prominent robust dorsal ridge scales distinguish it from S. saginatus (compare Figs. 33 and 34). Holotype. YPM 8944 (P4-2071); complete fish, part and counterpart, negatively prepared (Fig. 44) Raratypes. ~YEM-38913,8920), 8933, 8954, 8935, 6937, 8938, 8939; 89408941, 8943, 8978 Type locality. Yale excavation in cycle P4, Pompton, New Jersey Formation. ‘Towaco Formation, Newark Basin Age. Hettangian, Early Jurassic Etymology. From profundus, meaning deep, for the gibbose form of this species Description. S$. profundus is very deep bodied, with a pronounced dorsal hump (Figs. 27, 44B). The forehead slopes more steeply than in all others in the S. tenuiceps group, at about 50° from horizontal, so that the curvature is almost continuous with the anterior descending slope of the dorsal hump. Head depth relative to standard length is greater than in all other species in this group except S. saginatus. Maximum depth of the caudal peduncle is also deeper than in all other species, and its outline is almost straight from the dorsal and anal fins to the minimum point of the caudal peduncle, rather than a gently concave curve as in other species. All of these features are most clearly seen in Figures 27 and 29E. Morphometric data are summarized in Table 30. Commensurate with its deeper body, this species has more horizontal scale rows than other species in the S. tenuiceps group except for S. saginatus. ‘There are 11 between the lateral line and the dorsal fin and 10-11 between the lateral line and the anal fin, whereas others in the S. tenwiceps group have 8-9 scales between the lateral line and either fin. Flank scales are largest anteriorly, near Fic. 33. Dorsal ridge scales of S. profundus. Camera lucida drawing. Scale = 1 cm. The scales are pushed to the left side of center, giving a top view on the left side of the fish. This view is somewhat difficult to interpret as the dorsal-ventral height of robust scales is collapsed. However, the base of the scale is clearly expanded along the anterior-posterior axis, and the ganoine-covered spine does not extend much beyond the scale base. Below one of the scales, I give a reconstruction of what I think the scale would look like in side view. 82 PEABODY MUSEUM BULLETIN 43 Table 30. Summary of morphometric data for Semionotus profundus . MAXCD* = MAXCD x 100/DFCD. MAXCD** = MAXCD x 100/AFCD. All other variables expressed as % SL, for SL = 7.4 to 10.8 cm. Variable N Mean Standard Minimum Maximum deviation value value DPTH 2 48.19 Iba AAS 47.30 49.07 HDL 2 30.82 Vis Si7/ 30.56 31.08 PDL 2 66.69 2.58 64.86 68.52 DFPV 72 38.58 0.87 37.96 39.19 DFAN 1 39.19 = 39.19 39.19 DFCD 2 43.81 3.03 41.67 45.95 AFCD al 28.38 = 28.38 28.38 HDD 2 30.36 0S 29.63 31.08 MAXCD 2 PAPAS ANS} ibe ee) 21.30 PRAT MINCD 2 18.26 0.94 ily Sisx) 18.92 PTAL iL 50.00 = 50.00 50.00 DFPT 2 B65 7k 2.92 48.65 52.78 DFB i ALT) 6 {2y7/ — ALF Gi Sy¢/ IT eiay7/ AFB al Ihe} 4 'sal - als}oiul iksyo yal MAXCD#* 3 48.60 3.44 44.68 yal oalal MAXCD** 2 Ua 2.24 Uitlodhs) 80.95 the lateral line. They decrease in size caudally, dorsally, and ventrally. On the dorsal flank, anterior to the dorsal fin, there are occasional intercalary scale rows. Lateral line scales number 33-34. The pelvic fin originates after the 7th—9th vertical scale row, and the anal and dorsal fins originate after rows 17-19 and 19-21, respectively, there being at least one row between them. As in other Semionotus, the scales along the dorsal and ventral midline of the caudal peduncle are enlarged. The dorsal fin of this species is placed a bit forward of that on either S. fundus or S. saginatus. This fin has 3 basal fulcra and at least 5 fringing fulcra, of which the first two lie against the unsegmented portion of the first ray. Dorsal fin lepidotrichia number about 12. The anal fin also has 3 basal fulcra and at least 5 fringing fulcra, of which two lie against the unsegmented portion of the first ray. The pectoral fins have 17 lepidotrichia and about 7 fringing fulcra. The pelvic fins are composed of 3-4 rays fringed by about 7 fulcra. The holotype has the best skull of the type series. The frontals are as in other species in the S. tenuiceps group, and the teeth are simple and conical. Semionotus saginatus, new species Bigs26;.27;20; 34; 45; Tables: 31,32 Diagnosis. S. saginatus is distinguishable from all other species except S. profundus by its deep body. It differs from S. profundus in that a greater portion of its depth is distributed below rather than above its lateral line. In addition, it differs from S. profundus, S. tenuiceps, and S. decoratus by having moderate globular scales. Holotype. YPM 8932 (P4-735); complete fish, part mechanically prepared and counterpart negatively prepared (Fig. 45) Paratypes. YPM 8928, 8929, 8930, 8931 Type locality. Yale excavation in cycle P4, Pompton, New Jersey Formation. ‘Towaco Formation, Newark Basin Age. Hettangian, Early Jurassic Etymology. From saginatus, meaning fattened EARLY JURASSIC SEMIONOTID FISHES 83 Fic. 34. S. saginatus. A, camera lucida drawing of dorsal ridge scales of YPM 8932, holotype; B, camera lucida drawing of frontals, YPM 8929. Scale = 1 cm. Description. The dorsal ridge scales of S. saginatus are globular (Fig. 34). In the holotype, the first 3 dorsal ridge scales are only slightly enlarged, slightly convex, and almost radially symmetrical with no spine. The 4th scale is like the first three but much larger and more convex, radially symmetrical, and lacking a spine. The 5th through 12th scales are large, with bulbous bases and prominent spines covered with ganoine. The spines do not overlap the next scale in the series. Scales 13- 20 have flatter, smaller bases and more prominent spines. The 21st scale is the spineless predorsal scale. In most individuals, the posterior 5 scales in the series have spines that look like the anterior scales in the S. micropterus group (Olsen and others 1982). At least one spine in the posterior region of YPM 8932 is elongated, and several are definitely short (not broken). Whether the remaining posterior scales are short spined or just broken is ambiguous. If real, the short spines may be an anomalous variation as the spines are prominent on posterior scales in other specimens of S$. saginatus. The general form of the dorsal ridge scales is similar to those of §. greenwoodi and S. fundus but less prominent than those of S. profundus, S. tenuiceps, and S. decoratus. S. saginatus is deeper bodied than all but S. profundus, but in the former, a greater proportion of this body depth is ventral to the lateral line (Figs. 27, 29D, E). The slope of the skull roof is about 35—40° relative to horizontal. The dorsal fin is more posteriorly placed than in S. profundus. Morphometric data are sum- marized in Table 31. There are 33-35 lateral line scales. ‘The pelvic, anal, and dorsal fin positions are at rows 7-9, 17-19, and about 21, respectively. There are about 11 scales between the lateral line and the dorsal fin and 10-12 below the lateral line at the anal fin. Scale size decreases dorsally, ventrally, and caudally. There do not seem to be any intercalary scale rows on the flank, although this tendency is so variable in other species that their apparent absence may not be significant. The dorsal fin is preceded by 3 basal fulcra and fringed with about 5 more. Three of the fringing fulcra lie against the unsegmented portion of the first fin ray. The rays or lepidotrichia number 10-11 in the dorsal fin and about 9 in the anal fin. Like the dorsal fin, the anal fin is preceded by 3 basal fulcra and fringed with 5 more, two of which lie against the unsegmented portion of the first lepidotrichium. The pectoral fin comprises about 16 lepidotrichia and is fringed 84 PEABODY MUSEUM BULLETIN 43 Table 31. Summary of morphometric data for Semionotus saginatus . MAXCD* = MAXCD x 100/DFCD. MAXCD** = MAXCD x 100/AFCD. All other variables expressed as % SL, for SL = 10.8 to 11.9 cm. Variable N Mean Standard Minimum Maximum deviation value value DPTH 2 46.55 4.29 43.52 49.58 HDL 2 33°52 05 S2tedad 34.26 PDL 2 69.55 1.46 68.52 70.59 DFPV 2 38.73 1.08 37.96 39.50 DFAN 2 31.66 IGS i/ 30.56 Seni DFCD 2 38.35 0. 76 37.82 38.89 AFCD 2 4, ta 1022 26.85 28s Di HDD 2 29.94 0.44 29.63 30525 MAXCD 2 18.46 23 17.59 HOR 33 MINCD 2 16.23 2.00 14.81 Was PTAL 2 49.75 0.95 49.07 50.42 DFPT 2 49.79 0.30 49.58 50.00 DFB iL 13.45 - T3545 S45 AFB 2, 9.63 1.83 Sass 10.92 MAXCD* 2 48.37 S42 43.48 54.05 MAXCD** 2 TOD 5 8.18 7/5 a let 80.95 by an undetermined number of fulcra. The pelvics, as in other species of the S. tenuiceps group, have about 4 rays, and again the number of fulcra is not visible. The caudal fin has 8 rays ventrally and 8-9 dorsally. The skull is very well preserved in the holotype (Fig. 26) and shows nothing unusual for this group. One of the paratypes (YPM 8929) shows the frontals particularly well (Fig. 34). As in other Newark semionotids, the teeth are simple and conical. EARLY JURASSIC SEMIONOTID FISHES 85 eID RIN SS ORV AA iON Most of the observed variation in semionotids is confined to two complexes of characters—dorsal ridge scales and body shape. Although a few other characters, such as the shape of the caudal fin or the number of lateral line scales, are variable, only a few species deviate from the common condition for these characters. These rarely varying characters, which have little more taxonomic value than in the definition of particular species, are reviewed below, followed by a discussion of the more variable characters, body shape, and morphology of the dorsal ridge scales. FINS AND FLANK SCALES Data for Semionotus from cycle P4 and Europe are summarized in Table 32. The ranges of variability for meristic characters overlap in most species. Only a few species have ranges for a given variable that are outside the range of all or most other species. For example, S$. amplicephalus has 13-14 vertical scale rows anterior to the insertion of the pelvic fin and only 30 lateral line scales, whereas the total range of variation in these variables for all other species is 7-11 and 31-40, respectively. This case is notable because the two scale counts are skewed in opposite directions: the pelvic count is high but the lateral line count is low. The number of horizontal scale rows (from dorsal fin to lateral line and from anal fin to lateral line) is usually high in deep-bodied species such as S. schaefferi, S. latheticus, S. convalis, S. minor, S. profundus, and S. saginatus and low in slender species such as S. kirschi and S. johberryi. Similarly, the slope of the forehead in deep-bodied species (S. schaefferi, S. olseni, S. minor, S. convalis, S. profundus) is usually higher than it is in slender-bodied species (S$. kirschi, S. anosteus, S. johberryi). To some degree, both the slope of the forehead and number of horizontal scale rows are subsumed by the character body shape (see below). ‘The fact that deep- and slender-bodied fish have differing scale counts independently supports the arguments made earlier that distortion does not account for the variation in body shape. The number of lateral line scales characteristic of S. kapffi overlaps the ranges of only two other species, S$. ewthenius and S. thomsoni (aside from the anomalous S. amplicephalus). In the latter two species, the mean number of lateral line scales is higher than in S. kapffi (34.6 and 34.5 versus 32). Although the shape of the caudal fin is not easily determined in many cases, it is forked in S. virginiae and S. thomson but only weakly emarginate in all other species when it is clearly visible. The range of variation in number of lepidotrichia and fulcra is the same in a given fin for all species. The only exceptions are the fin fulcra of two European species, S. berger: and S. normanniae. In these two species, there are more basal fin fulcra, more fulcra lying against the unsegmented portion of the first lepi- dotrichium, and more fulcra altogether on the dorsal fin (and probably the anal fin) than on any semionotid from P4. Unfortunately, the number of specimens of European species on which this statement is based is small, particularly in the case of S. normanniae. S. minor presents an intermediate case. It has an unusually high number of basal fulcra on the dorsal fin (five rather than three to four, usually three), but it also has the five additional long, slender, fringing fulcra seen in all Newark semionotids discussed here. Intercalated scale rows in the epaxial region anterior to the dorsal fin are 86 PEABODY MUSEUM BULLETIN 43 Table 32. Summary of Morphological Data for Semionotus from Europe and from Towaco cycle P4 Scales Species drs plvsc anfsc dfsc cdsc vdsc vvsc S. berger - 11-11 17-21 20-24 34-37 - ~ S. ka = - 18-22 21-23 32-33 - - S. normanniae - - - 21-21 - - - S. minor 22-22 9-10 19-20 21-24 35-38 10-11 8-12 S. kirschi 19-19 8-11 18-22 20-24 33-39 7-8 8-8 S. olseni 20-20 8-10 17-19 18-22 33-37 9-11 9-9 S. virginiae 18-19 7-10 17-21 18-26 33-40 7-10 7-11 S. thomsoni 20-20 7-10 16-20 19-22 31-36 8-10 7-10 S. convalis 18-18 9-9 17-17 20-20 36-36 11-11 11-11 S. redfieldiit 19=19) 507-10 17=—20) 19=24°035=37 ~“S9—1ae Se s—10 S. euthenius 20-20 7-10 16-21 19-24 32-38 8-12 7-11 S. schaefferi 18-18 7-9 16-20 18-23 33-37 10-13 10-12 S. latheticus 20-20 8-8 17-17 19-19 34-34 13-13 10-10 S. melanimus 21-21 - 21-21 24-24 38-38 - - S. anosteus 19-19 7-7 19-19 21-21 33-33 9-10 9-9 S. johberryi 21-21 - 20-20 21-21 34-34 8-8 7-7 S. amplicephalus = 13-14 = - 30-30 - = S. tenuiceps 19-19 9-9 18-19 21-21 33-36 10-10 9-9 S. greenwoodi 22-22 8-9 18-19 20-21 34-34 8-8 9-9 S. decoratus 19-19 9-9 18-19 20-21 33-34 9-9 8-8 . fundus 19-19 7-8 17-17 19-19 32-32 8-8 8-8 S. profundus 21-21 7-9 17-19 19-21 33-34 11-11 10-11 S. saginatus 21-21 7-9 17-19 21-21 33-35 12-12 12-12 Continued on next page difficult to interpret. Bartram (1977) believed that the presence of these extra scale rows was a synapomorphy of the Macrosemidae, the sister group to the Semionotidae (Olsen and McCune ms). However, in many species of semionotids from P4 and also in the S. elegans group, the occurrence of these extra scale rows is not only variable within species but may differ on the right and left sides of the same individual. Most semionotids with intercalary scale rows have only a few extra rows, but one individual has as many as nine extra rows (S. decoratus, YPM 8957), and in two individuals, the predorsal scales above the lateral line are completely doubled [Semzonotus sp. (YPM 8846) from P4, and Semzonotus sp. (AMNH 3210) from Sunderland, Massachusetts]. Therefore, neither the presence nor frequency of intercalary scale rows is useful for relating species within the family. Paradoxically, though the presence of intercalary scale rows is often not consistent on both sides of the same individual, the tendency to produce these scale rows could be considered a synapomorphy uniting macrosemiids and semionotids (as variants indicating a higher level homology sensu Roth (1984) of the developmental pathway that produces scales). DORSAL RIDGE SCALES One of the more prominent fields of variability in Newark semionotids is the morphology of the dorsal ridge scale series. Seven morphologies of dorsal ridge scales are exhibited by semionotids from cycle P4 (Fig. 9). ‘These have been described in detail on pages 30-34. In all types, the morphology of individual scales changes with position along the anterior-posterior axis of the series. The anterior scales in each series are always most distinctive, allowing definite iden- tification of dorsal ridge scale type on the basis of these anterior scales alone. Modified simple, small, thin-spined, and concave scales (Fig. 9B, C, D, G) can EARLY JURASSIC SEMIONOTID FISHES 87 Table 32--continued Fin fulcra dffl1 Gftf2 5 IGEE3S affil ati2) waters Species fh cf ismadrst S. berger 30-40 - - s 4-6 10-10 15-15 - - 14-14 S. kapffi = SES S|) = 4-4) aa=7 = = = - S. normanniae - - - s 5-5 >8 = - = = S. minor 50-50 = - s-ms 5-5 6-7 >10 3-3 5-5 >7 S. kirschi 15-25 we rare s 3-4 4-5 6-8 3-3 5-6 6-7 S. olseni 50-60 we none s 3-4 5-5 7-8 3-3 6-6 8-8 S. virginiae 35-40 f rare s 3-4 4-6 6-8 2-3 4-5 6-8 S. thomsoni 30-40 £) occ Ss) -3=4 4-6 5-8 2=3 4—5)5) 56-9 S. convalis 50-50 - none s 3-3 - 6-6 3-3 5-5 8-8 S. redfieldii 35-40 we none s 3-4 5—6 @=8 2-4 5-6) 216-8 S. euthenius 40-50 we occ s 3-4 4-6 5-8 2-3 4-5 5-9 S. schaefferi 50=65" we oce® msi) —3—4)-4=6 6-8 3-3) + 5=5)) 7-9 S. latheticus 45-45 - - m 3-3 5-5 - = = = S. melanimus 35-35 - occ m 3-3 5-5 8-8 3-3 - 7-7 S. anosteus 30-30 - none ms_ 3-3 5-6 7-8 3-3 5-5 8-10 S. johberryi 30-30 - occ ms - - - - - 7-7 S. amplicephalus = = = ms = = = = = = S. tenuiceps 30-40 we none rr 3-4 5-6 8-11 3-3 5-6 9-12 S. greenwoodi 30-40 we rare 3=3 5-6 DT 3=3 5-5 ded S. decoratus 30-35 we freq 75 3-3 5-5 7-9 3-3 5-5 7-9 S. fundus 35-40 we occ g = 8-3 5-5 7-7 3=3) 5=5 8-8 S. profundus 50-50 we occ re. 3=3 UT 8-8 = = 9-9 S. saginatus 35-40 we rare Gees 6-6 8-8 - 5-5 = Abbreviations: cf=caudal fin shape (we=weakly emarginate, f=forked), drst=dorsal ridge scale type (c=concave, g=globular, r=robust, ms=modified simple, s=simple, sm=small, t=thin-spined), fh=angle of forehead in degrees from horizontal, is=intercalated scales (none, rare, occasional, frequent). Continued on next page be recognized with reasonable confidence on the basis of posterior scales. Simple, globular, and robust scales (Fig. 9A, E, F) cannot be identified from only the posterior scales in the series. Different dorsal ridge scale morphologies cannot be accounted for by ontogenetic change, because the size range of fishes having different dorsal ridge scale types is comparable (most between 7 and 12 cm). There are no very small individuals (less than 5 cm) in which very early stages of dorsal ridge scale development can be seen, but in very large individuals the morphology of individual scales is no different from that of smaller individuals. In addition to the types of dorsal ridge scales previously described, there are occasional anomalous variants on these themes. Rarely, a dorsal ridge scale series of one type will include one or a few “sports” of another type (Fig. 35). Another kind of variant involves the doubling or tripling of scales in the dorsal ridge series (Fig. 35). Several variants involve doublings of only one or two scales in the series (Fig. 35C, G, M) although in a few a large portion of the series is doubled (Fig. 35A, B, D, F) or tripled (Fig. 35E). Most specimens showing dorsal ridge scale anomalies are incomplete and therefore indeterminate at the species level, but supernumerary dorsal ridge scales do not seem to be characteristic of a particular species. There are supernumerary scales in concave (Fig. 35M), modified simple (Fig. 35B), simple (Fig. 35A), and globular (Fig. 35D) dorsal ridge scale series. There are also at least two cases of double dorsal ridge scales in S. capensis from South Africa (Fig. 35E). The generation of scale doublings may occur in several ways. One specimen has a hoof-shaped dorsal scale (Fig. 351), suggesting that the rudiment of the median dorsal ridge scale is producing two scales. In some individuals with concave scales, the scales lateral to the dorsal ridge scale series develop spines (Fig. 35G, 88 PEABODY MUSEUM BULLETIN 43 Table 32--continued Fin fulcra Fin rays Species peti plvi idors anal pet Ply cddr cdvr S. bergen - - 13-15 7-10 - 4-4 8-8 8-9 S. kapffi - - - - - - - - S. normanniae >17 = - - Sals' - 10-10 7-7 S. minor 9-9 >4 12-12 10-10 17-17 4-7 #+410-10 8-8 S. kirschi 5-7 6-7 9-14 8-10 11-16 3-5 7-9 8-8 S. olseni 10-10 5-8 12-12 10 15-15 5-5 - 8-8 S. virginiae 5-7 4-6 9-12 8-11 15-17 3-6 9-9 8-8 S. thomsoni 5-7 4-6 10-13 8-10 13-16 3-6 9-9 8-8 S. convalis 6-6 4-4 13-13 9-9 13-13 5-5 9-9 8-8 S. redfieldii - 5-5 10-12 9-10 - 3-7 9-9 8-9 S. euthenius 5-7 5-8 10-13 8-12 14-17 3-8 8-9 8-8 S. schaefferi C=], nG—6 = 10-1 2am) Stee 11717 ac Oe S. latheticus - 6-6 - - - 3-3 - 8-8 S. melanimus - - 11-11 10-10 14-14 - 10-10 9-9 S. anosteus 9-9 6-6 13-13 13-13 - ~ 9-9 8-8 S. johberryi = = 10-10 15-15 - ~ 9-9 9-9 S. amplicephalus = = = = - = = = S. tenuiceps >7 6-6 10-12 7-10 17-17 5-5 8-9 8-9 S. greenwoodi 6-6 6-6 11-11 8-8 - - 8-8 T=7 S. decoratus 5-5 4-4 10-12 7-9 - - - = S. fundus 7-7 - 11-13 9-10 17-17 3-4 8-8 8-8 S. profundus >7 Ey yal! - iy/ 37) = = = S. saginatus >5 7, 9-9 10-11 11-16 4-4 8-8 9-9 H). In some individuals, in which all of the series is doubled or tripled (Fig. 35D, E), extra scales appear to be crammed along the dorsal midline rather than spreading laterally and ventrally away from the median series of scales, suggesting that scales lateral to the dorsal ridge series are not appropriated into the median series. The occasional or complete doubling of the scales in the dorsal ridge scale series is particularly interesting in light of the tendency of macrosemiids (Bartram 1977) and semionotids to develop intercalary scale rows. Flank scale doubling appears to be concentrated in the anterior epaxial region of the fish, and exclusively in the region anterior to the dorsal fin, which is in the same region of the fish as the dorsal ridge scales. It may be significant that the anterior flank is the last region to develop scales in Lepisosteus (personal observation), as well as in a variety of paleoniscids (D. Bardack, personal communication). BODY SHAPE Newark semionotids exhibit a wide diversity of body form. Such diversity of body form is relatively rare among semionotids from other parts of the world, the one exception being the American Western ‘Triassic (Schaeffer 1967). I recognized 22 different shapes, described either by classification functions from the discrim- inant analysis, Goodkin viewer tracings, or both. ‘Twenty-two different shapes (including European species) may seem an excessive number, but, in addition to the justifications I have given earlier for defining these groups, the relationship between shape and other characters increases my confidence that the groups are a reasonable first solution to the problem of sorting these fishes by shape. Particular shapes are not unique within scale groups, but at the same time neither are all shapes duplicated in every scale group. For example, there are deep-bodied fish with both robust scales and modified simple scales, but there are no similar fish with simple or concave scales. Overall, 3 of 22 shapes are duplicated EARLY JURASSIC SEMIONOTID FISHES 89 \ Mga t BSR MW, Fic. 35. Dorsal ridge scale anomalies. A, YPM 8865; B, YPM 8758; C, YPM 8861; D, YPM 8847; E, YPM 8848; F, S. capensis BMNH P.6861; G, YPM 8719 (note flank scales with “spines”); H, YPM 8812 (note flank scales with “spines’’); I, YPM 8845 (split dorsal ridge scales); J, YPM 8750 (note two concave spines anteriorly); K, YPM 8869 (dorsal ridge series is a combination of thin- spined, simple and concave scales); L, YPM 8894 (dorsal ridge series is a combination of thin-spined, simple, and concave scales); M, YPM 8836 (dorsal ridge series is a combination of concave and an uncategorizable convex scale); N, YPM 8932 (dorsal ridge series is a combination of globular, thin- spined and small scales). Scale = 5 mm. in one or two other scale groups. Species that share the same shape but are distinguished by their dorsal ridge scales are as follows: S. kirschi and an unde- scribed species in the S. elegans group (McCune 1982); S. olseni and S. latheticus; S. greenwoodi, S. virginiae, and another undescribed species in the S. elegans group (McCune 1982). In addition, some groups of species are very similar in shape: S. decoratus, S. thomson, and S. anosteus; S. fundus and S. euthenius. The shapes of all other species are unique. The relationship between shape and dorsal ridge scale morphology was par- ticularly interesting within the S. fenwiceps group. For each of three shape groups in this complex, there was another shape group that was similar though not the same, in my estimation. ‘hat each member of these three pairs of similarly shaped 90 PEABODY MUSEUM BULLETIN 43 groups should be recognized separately was reinforced by the fact that the two members of each similar pair had different dorsal ridge scales. Three of these SIX shapes are very similar to, but not quite the same as, the shapes of fishes in other scale groups. EARLY JURASSIC SEMIONOTID FISHES 91 8. RELATIONSHIPS Beyond initial character selection, schemes of relationship are most heavily de- termined by assignments of character polarities. Therefore, before discussing the relationships of semionotids, I will explicate my rationale for assigning particular character polarities. For most meristic characters, the polarity of character states is reasonably straightforward. Serrated scales are found in several actinopterygians, such as Pteronisculus (Lehman 1952) and a number of macrosemiids (Bartram 1977), but flank scales with smooth posterior margins are much more widespread among actinopterygians, including Acentrophorus (Gill 1923), parasemionotids (Lehman 1952), and a wide variety of paleoniscids. Therefore, I consider serrated flank scales in S. normanniae and S. minor to be derived. It is more ambiguous whether the forked caudal fin in S. virginiae and S. thomson is primitive or derived. On the basis of commonality in semionotids, one would conclude that a forked caudal fin is derived. Both forked and weakly emarginate caudal fins are found among macrosemiids (Bartram 1977), but para- semionotids (Lehman 1952), L. elvensis (personal observation), and Acentrophorus (Gill 1923) have weakly emarginate caudal fins. Because of these last outgroup comparisons, I suggest that a weakly emarginate caudal fin is primitive for Se- muonotus. The low lateral line scale counts (30-32) in S. fundus, S. kapffi, and S. ampli- cephalus are derived. Macrosemiids have 34—50 lateral line scales (Bartram 1977). Lepisosteids have 53-62 (Suttkus 1963), and L. elvensis has 37-38. In general, the deeper-bodied and slender-bodied forms are probably derived, though spec- ifying a particular form as primitive hardly seems possible. The probable primitive condition for dorsal ridge scale morphology in semiono- tids is convex with posteriorly directed spines. Similar dorsal scales are known in paleoniscids and elonichthyids (Orlov 1967), but not generally throughout the actinopterygians. In the paleoniscids and elonichthyids that have dorsal ridge scales, the form of the scales is simple and convex with posteriorly directed spines. Dorsal ridge scales of most Lepidotes, all European Semionotus, and some American semionotids also have this morphology. L. mantelli, L. toombsi, L. laevis, and nearly all species of Semzonotus from sediments outside the Newark Supergroup (the exception being some S. tenuiceps scales from the lower Lufeng of China (Olsen and others 1982) have simple or modified simple, convex dorsal ridge scales. However, the dorsal ridge scales of many American Semionotus are modified in different ways (Olsen and others 1982). With semionotids it is particularly difficult to determine the polarities for the two most important character complexes, dorsal ridge scale morphology and body shape. The problem stems from the fact that both characters have many more than two states—22 in the case of shape and 7 in the case of dorsal ridge scales— and also from operational difficulties relating to coding the characters (or character complexes). In the case of body shape, it is possible to code the single character ‘“‘body shape” as the ranges of the ratios of two measurements, such as depth/standard length or head length/standard length. Aside from the obvious difficulty of defining discrete character states from continuous variables, the high correlation of these body measurements would severely bias the outcome toward a solution based entirely on shape. Another possibility would be to recognize formally the gen- eralization that both extremes of shape (generally slender or deep) are derived relative to the fishes of middle body depths, but it is misleading to label two deep- 92 PEABODY MUSEUM BULLETIN 43 bodied but differently shaped fish as sharing uniquely derived character states. The most desirable solution would be to understand the transformation from one shape to another developmentally or to describe it mathematically. The former is impossible with fossils, and the latter analysis is beyond the scope of the present work, but a solution along the lines discussed by Bookstein (1978), Humphries and others (1981), or Raup (1966) is certainly possible (cf. Bookstein and others 1985, published after this study was completed). However, a sufficient mathe- matical description would not necessarily describe the true developmental (Raff and Kauffman 1982) or evolutionary transformation. In the absence of a mathematical or developmental rationale for the polarity of shape transformation in these fishes, I have chosen to consider the many states of the character “shape” as equally derived. At present, this is the least misleading null hypothesis. A future analysis may resolve this difficulty. The criterion for recognizing a shared shape was the same as for incorporating an individual fish into a Goodkin viewer group (see Chapter 4). This convention leaves only the loose end of determining which of the 22 shapes is primitive. However, if no transformational scenario about shape is advanced, the taxonomic value of shape is limited to describing species; for that function, both primitive and derived shapes are equally useful. Devising a transformation series of the different dorsal ridge scale series appears more tractable than it was for body shape. There are fewer patterns of dorsal ridge scales, more easily ordered, than there are shapes. Even so, the correct way to code the dorsal ridge scale series is not obvious. Does one type of dorsal ridge scale series constitute a single character state, or should the dorsal ridge scale series be treated as a complex of characters and coded as several characters? If the dorsal ridge scale series were treated as a single character, then there would be two relatively simple hypotheses of character transformation. If different scale types were independently derived, then the resulting single character cladogram would be a hexachotomy (Fig. 36A). Alternatively, the dorsal ridge scales can be ordered in a relatively straightforward morphological series (Fig. 36B), resulting in a better-resolved (but not necessarily more correct) one-character cladogram (Fig. 36C). A third alternative is to atomize the morphology of the dorsal ridge scale series into five two-state characters (Fig. 37). The many-character cladogram (Fig. 38) was constructed using this system of coding for the dorsal ridge scale complex. Aspects of dorsal ridge scale morphology dominate the structure of the cladogram because body shape, and a large portion of other characters, serve only to define species, not to relate them to each other. Species having identical shapes but different dorsal ridge scales are grouped by the latter, and shared shapes are represented as homoplastic (shapes 15, 16, and 17 in Fig. 38). All other shapes are shown as autapomorphies defining species. It is possible, given the patterns of variation discussed earlier, that considerable homoplasy has remained undetected in dorsal ridge scale characters. ‘Vhe occur- rences of supernumerary spines and dorsal ridge scale “sports” misplaced in series of differing morphologies suggest that all or many semionotid species have the genetic/developmental potential necessary to generate any of the dorsal ridge scale patterns actually observed (in the sense of atavisms; see Hall 1984), though the expression of this potential may usually be hidden. One of the most striking features of the cladogram is the number of unresolved polychotomies. Of the eight nodes defined by one or more synapomorphies, five of these are polychotomies—nodes that include between three and nine branches. Although there is nothing intrinsically wrong with polychotomies, they are often EARLY JURASSIC SEMIONOTID FISHES 93 CSSM GR MS T ganoine convex to concave SQ” anterior \ scales scales enlarge enlarge — ———_ A & begin earlier spines narrow CSSMG RMS T CS SMGRMS T Fic. 36. Hypotheses of character transformations. S = simple scales, MS = modified simple, G = globular, R = robust, C = concave, SM = small scale, T = thin-spined. A, A hexachotomy, the simplest hypothesis of transformation for the seven types of dorsal ridge scales in semionotids from P4. B, Hypothesized morphological transformation for dorsal ridge scales in semionotids from P4. The dotted line indicates ambiguity as to whether ““R” scales were derived from “G” scales or directly from “MS” scales. The transformation of simple to modified simple scales involves a decrease in the area of the dorsal surface covered by ganoine and slight undercutting lateral to the spine. Extreme reduction of the bone under the scale spine yields thin-spined from modified simple. Increase in the mass of the anterior scales (except the four most anterior ones) of modified simple and continued recession of ganoine would result in globular scales. Robust scales might derive from either an anterior shift in the position of enlarged scales and a general increase in size from globular scales or from an independent size increase of anterior scales (starting with more anterior scales than in the globular type) from modified simple scales. Small scales could be derived by shortening of the spine from simple scales. C, Two alternative character branching sequences based on hypothesized morphological transfor- mation in B. The difference between these alternatives is the derivation of ““R” scales from either “G” or “MS” scales. 94 PEABODY MUSEUM BULLETIN 43 SCALE TYPE CHARACTER CHARACTER STATES S C MS G RSM T shape convex/concave Oo Pe Dect i Oheloeed lateral undercutting no/yes 0 0 1 scl wtOhel size anterior scales small/large 0..0. 0.1, .2., 07a spine length long/short 0.0... 0.0.0 ao scale size normal/small 0 0.0 0 0. ita ganoine cover complete/partial 0: 0 led deve Opa spine width normal/narrow 0°-O 0 TO tra Fic. 37. System to recode the seven-state dorsal ridge scale character into seven two-state (or three- state) characters. Abbreviations are the same as those given for Fig. 36. Under character state, the primitive state (0) is listed first, the derived states (1, 2) second and third, respectively. Simple scales are taken to be primitive for all seven new characters. thought to arise from “partial ignorance” (Nelson 1980), owing to shortages of specimens, taxa, or characters, particularly in paleontological studies (Patterson and Rosen 1977). However, in Newark semionotids, the kinds of characters as well as the general completeness and high quality of specimens require us to accept the present character distributions until contrary evidence accrues. A large proportion of living species of fishes (within a particular family) can be recognized on the basis of the sorts of characters I have used for semionotids—scale counts, scale morphology, fin ray counts, teeth, fin characters, and body proportions. Color, behavior, and ecology are notably absent from fossils, but they are also absent in the museum specimens of extant species that have accumulated over many decades and on which much of fish taxonomy has been based. Furthermore, as discussed by McCune and others (1984), in cladograms of species, especially diverse groups, polychotomies can represent true character distributions and not simply incomplete analyses. EARLY JURASSIC SEMIONOTID FISHES 95 o & ets 2 oO < = = = ri Sl (Se), fol ae) g8ae : eae ihr ek Se reu -s* “tmnt 2 PGC, & ook J q Pp =F 22 27 928 30 32 515 514 c314 29 13 C7 =)\] 15 317 =i 20 12 10 7 8 9 6 5 4 3 2 1 Fic. 38. ‘Tentative cladistic relationships for semionotids from P4 (descriptions for some species unpublished; see McCune 1982) and European semionotids (McCune 1986). Species a through u (also pictured in Fig. 39) are as follows: a = S. kirschi, b = S. redfieldiu, c = S. euthenius, d = S. olseni, e = S. virginiae, f = S. thomsoni, g = S. convalis, h = S. anosteus, i = S. schaefferi, } = S. johberryi, k = S. latheticus, | = S. melanimus, m = undescribed species in S. elegans group, n = undescribed species in S. elegans group, o = S. amplicephalus, p = S. greenwoodi, q = S. fundus, r = S. saginatus, s = S. decoratus, t = S. tenuiceps, and u = S. profundus. Large black bars are synapomorphies. Small black bars are autapomorphies. White represents homoplasies. Primitive character states are not shown. Key to derived character states is as follows: 1, Dorsal ridge scales present, premaxillae with long rostral process, epiotic with posteriorly directed process, lachrymal serially subdivided. 2, Fringing fulcra reduced in number. 3, Suborbitals number greater than one. 4, Basal fin fulcra reduced in number. 5, Dorsal ridge scales laterally undercut. 6, Dorsal ridge scales concave. 7, Dorsal ridge scales with short spines, dorsal ridge scales small relative to flank scales. 8, Dorsal ridge scale spines very narrow and separated distally from base. 9, Anterior dorsal ridge scales enlarged. 10, Anterior dorsal ridge scales even larger. 11, Flank scales serrated. 12, Caudal fin forked. 13, Horizontal flank scale rows reduced in number. 14, Vertical flank scale rows reduced in number. 15, Shape “1.” 16, Shape “2.” 17, Shape “3.” 18-33, Autapomorphic shapes. Reproduced from McCune and others, 1984, Semionotid fishes from the Mesozoic great lakes of North America, p. 27-44. In A. A. Echelle and I. Kornfeld [eds.] Evolution of fish species flocks. University of Maine Press, Orono. 96 PEABODY MUSEUM BULLETIN 43 9. CONCLUDING REMARKS Lake cycle P4 exhibits a diversity of species within a single family heretofore unknown in fossil fishes. There are 21 species of semionotid fishes that once dominated this single lake (Fig. 39), and the diversity of P4 semionotids is only one small portion of the total semionotid diversity in eastern North America (most of it undescribed). Most aspects of the morphology of Semzonotus from P4 are not different among species. In the fauna as a whole, only two fields of mor- phology—dorsal ridge scales and body shape—are significantly variable. Diversity of body shape is quite common in fishes (McCune 1981, and references therein), but even so the range of body shape within the North American representatives of Semionotus is far greater than the variation in most genera of living fishes. Variation in dorsal ridge scale morphology is known in semionotids and to a limited extent in a few other fossil groups (see Orlov 1967). Although the func- tional significance of dorsal ridge scales is not yet understood, the pattern of anomalous variation in relation to consistent patterns of variation of dorsal ridge scales may have general significance for systematics and evolution. That many semionotids and their sister-group macrosemiids do funny things with their scales in the anterior epaxial region suggests that the developmental system controlling these scales is particularly labile or uncanalized. This is a restatement of the fact that the morphology of dorsal ridge scales is variable in semionotids; something like it could be said about the bills of honeycreepers and finches or the teeth of cichlids. From the fact that thin-spined or small scales may Fic. 39. The semionotid assemblage from P4. Species are grouped by their dorsal ridge scales. Species with simple scales are a = S. kirschi, b = S. redfieldii, c = S. euthenius, d = S. olsent, e = S. virginiae, f = S. thomsoni, g = S. convalis. Species with modified-simple scales are h = S. anosteus, i = S. schaeffert. Species with thin-spined scales are j = S. johberryi, k = S. latheticus, 1 = S. melanimus. Species with concave scales are m = undescribed species in S. elegans group, n = undescribed species in S. elegans group. Species with globular scales are p = S. greenwoodi, q = S. fundus, r = S. saginatus. Species with robust scales are s = S. decoratus, t = S. tenuiceps, u = S. profundus. EARLY JURASSIC SEMIONOTID FISHES OF appear (rarely) in a globular series, I begin to suspect that many semionotids may have more or less the same hidden potential (in the sense of atavisms; see Hall 1984) to produce most scale types. For systematics, this hidden potential may imply a high probability of widespread parallelism and a low confidence in the most parsimonious cladogram. ‘Vhe implication from the pattern of variability of dorsal ridge scales for evolution as a process is that evolution is not simply a linear accumulation of phenotypic/genotypic acquisitions. If genetic-develop- mental potentials are shared (primitively) by large groups (see Roth 1984), the generation of diversity, especially in speciose groups, may come less from the accumulation of mutations in reproductively isolated populations than from the selective unmasking and mixing of already present and cryptically accumulating (see Rachootin and Thomson 1981) genetic potentials. Although the manipulative experiments that one might want to conduct to understand the generation of variability in dorsal ridge scales or body shape are clearly impossible, a long series of historical experiments has proceeded over and over again through time, and simultaneously in a number of geographically distinct lake basins. We need only to collect and analyze data and interpret the results. In this work I have begun to develop a species-level taxonomy that I hope will be a useful vocabulary for asking questions about the paleoecology and evolution of semionotid fishes in Newark lakes. MUSEUM BULLETIN 43 PEABODY 98 "UD J = JBIS “OPL8 WdA Guida ‘§ ‘ f8€S9 WdA 1U85]0 “S ‘OD SOFL8 Wd A Mosmoy? “§ “Gf S81L8 WdA 2Y95424 “SOV ‘saioads Mau sno} Jo sadAjojopY “Op “DIY 99 os) INOTID FISHEI SEMIC ARLY JURASSIC 4 / ol i ‘Cs-9S88 WdA afJanyos “s ‘ D) ‘OLL8 WdA 51/DaU02 “UID 'S “@ Sh088 Wd Sniuayjna “5 “yw ‘satoads c = 23[k9 C= MOU amoj jo sadAjo I OH S “CLL8 WdA "UPI9YP '§ “Ty “S19 MUSEUM BULLETIN 43 PEABODY 100 101 ARLY JURASSIC SEMIONOTID FISHES E "WW J = IRIS “Gb88 INA SrVYdardun ‘§ “FT ‘OP88 WdA Sruiunjaw '§ “J ‘6688 Nd A SrINAYID) ‘SO ‘bh88 Wd Snaisoun “S “gf (0088 Wd Muaquol “5 ‘ y V ‘satoads Mau aay jo sadAjojoPY ‘Cy OIA 102 PEABODY MUSEUM BULLETIN 43 Fic. 43. S. tenuiceps. A, original of S. tenuiceps (Agassiz) = “Eurynotes tenuiceps” Agassiz. From Agassiz (1836) pl. 4, fig. 4. B, neotype of S. tenuiceps YPM 8162, from Sunderland, Massachusetts. Specimen was negatively prepared. Scale = 2 cm. 103 28 SEMIONOTID FISHE ARLY JURASSIC ‘b+68 WdA snpunfosd *§ ‘ ‘wo 7 = d -0L68 WdA PIS “9768 Wd A Smurf 'S “Cl ‘8568 Wd A 87704 Mau INO} Jo sadAiojoFy poomuaals '§ ‘y ‘satoads )2ap aA 4 © 0) ‘O17 104 PEABODY MUSEUM BULLETIN 43 Fic. 45. Holotype of S. saginatus YPM 8932. Scale = 2 cm. EARLY JURASSIC SEMIONOTID FISHES 105 LITERATURE CITED Agassiz, L. 1832. Untersuchungen uber die fossilen Fische der Lias-Formation. Jahrb. Mineral. Geogn. Geol. Petrefakt. 1832, part 3: 139-149. 1833. Recherches sur les poissons fossiles, Livraison I. Imprimerie de Petitpierre, Neuchatel. — 1834. Recherches sur les poissons fossiles, Livraison II. Imprimerie de Petitpierre, Neuchatel. — 1835. Recherches sur les poissons fossiles, Livraison V. Imprimerie de Petitpierre, Neuchatel. 1836. Recherches sur les poissons fossiles, Livraison VI. Imprimerie de Petitpierre, Neu- Andrews, S. M. 1982. The discovery of fossil fishes in Scotland up to 1845, with checklists of Agassiz’s figured specimens. R. Scott. Mus., Edinburgh. Bartram, A. W. H. 1977. The Macrosemiidae, a Mesozoic family of holostean fish. Bull. Brit. Mus. (Geol.) 29(2): 137-234. Bock, W. 1959. New Eastern Triassic fishes and Triassic correlations. Acad. Nat. Sci., Vol. 1 (Geol. Cent. Res. Ser.). Bookstein, F. L. 1978. The measurement of biological shape and shape change. Lecture notes in biomathematics 24. Springer-Verlag, New York. 191 p. Bookstein, F. L., B. Chernoff, R. Elder, J. Humphries, G. Smith, and R. Strauss. 1985. Morpho- metrics in evolutionary biology. Acad. Nat. Sci. (Philadelphia), Spec. Publ. No. 15. Cornet, B. 1977. The palynostratigraphy and age of the Newark Supergroup. Ph.D. thesis, Penn- sylvania State Univ. Cornet, B., A. Traverse, and N.G. McDonald. 1973. Fossil spores, pollen and fishes from Connecticut indicate Early Jurassic age for part of the Newark Group. Science 182: 1243-1247. Davies, G. R.,andS. D. Ludlam. 1973. Origin of laminated and graded sediments, Middle Devonian of western Canada. Geol. Soc. Am. Bull. 84: 3527-3546. Deschaseaux, C. 1943. Contribution a l'étude du genre Lepidotes. Ann. Paleontol. 1943: 3-13. Dixon, W. J.,and M. B. Brown. 1979. BMDP-79: Biomedical Computer Programs P Series. Univ. California Press, Berkeley. 840 p. Eastman, C. R. 1905. The Triassic fishes of New Jersey. Rep. Geol. Surv. New Jersey 1904: 67- 102. Egerton, Sir P. de M. G. 1850. Paleoichthyologic notes. No. 3. On the Ganoidei Heterocerci. Q. J. Geol. Soc. 6: 1-10. Eldredge, N., and J. Cracraft. 1980. Phylogenetic patterns and the evolutionary process. Columbia Univ. Press, New York. 341 p. Emmons, E. 1857. American geology, Part VI. Sprague & Co., Albany. 142-145. 1860. Manual of geology. Sower, Barnes, & Co., Philadelphia. p. 186 and fig. 164. Ferson, S., J. E. Rohlf, and R. K. Koehn. 1985. Measuring shape variation of two-dimensional outlines. Syst. Zool. 34: 59-68. Gaudant, M. 1978. Recherches sur l’Anatomie et la Systematique de Ctenothrissiformes et des Pattersonichthyiformes (Poissons Teleosteens) du Cenomanien du Liban. Mém. Mus. Nat. Hist. Nat. Ser. C, Tome XLI. 124 p. + 10 pl. Gill, E. L. 1923. The Permian fishes of the genus Acentrophorus. Proc. Zool. Soc. London 1923: 19-40. Grande, L. 1982a. A revision of the fossil genus Diplomystus, with comments on the interrelationships of clupeomorph fishes. Am. Mus. Novit. 2728: 1-34. 1982b. A revision of the fossil genus Anightia, with a description of a new genus from the Green River Formation (Teleostei, Clupeidae). Am. Mus. Novit. 2731: 1-22. Gratacap, L. P 1886. Fish remains and tracks in the Triassic rocks at Weehawken, N.J. Am. Nat. 10: 243-246. Hall, B. 1984. Developmental mechanisms underlying the formation of atavisms. Biol. Rev. 59: 89- 124. Helwig, J. T., and K. A. Council, eds. 1979. SAS user’s guide. SAS Institute Inc., Cary, North Carolina. 494 p. Hennig, W. 1966. Phylogenetic systematics. University of Illinois Press, Urbana. 263 p. Hentz, T. F. 1981. Sedimentology and structure of Culpeper Group lake beds (Lower Jurassic) at Thoroughfare Gap, Virginia. Masters thesis, University of Kansas. 166 p. Hitchcock, E. 1819. Remarks on the geology and mineralogy of a section in Massachusetts on the Connecticut River, with a part of New Hampshire and Vermont. Am. J. Sci. 1: 105-116. 1823. Geology, mineralogy and scenery of the Connecticut Valley, with a geological map and drawings of organic remains. Am. J. Sci. 6: 1-85. Hubbs, C. L., and K. F. Lagler. 1947. Fishes of the Great Lakes Region. Bull. 26, Cranbrook Inst. Science, Bloomfield Hills, Michigan. 213 p. Humphries, J. M., F. L. Bookstein, B. Chernoff, G. R. Smith, R. L. Elder, and S. G. Poss. 1981. Multivariate discrimination by shape in relation to size. Syst. Zool. 30(3): 291-308. International Commission on Zoological Nomenclature. 1964. International Code of Zoological Nomenclature. Int. Trust for Zoological Nomenclature, London. Larsonneur, C. 1964. Semionotus normanniae du Trias superieur de Basse-Normandie (France). Ann. Paleontol. 50(2): 101-117. 106 PEABODY MUSEUM BULLETIN 43 Lehman, J.-P. 1952. Etude complementaire des poissons de |’Eotrias de Madagascar. K. Sven. Vetenskapsakad. Handl. 4 2(6): 1-201. 48 pl. Loper, S. W. 1893. On a new fossil fish (Uschypterus newberry1). Pop. Sci. News 27: 75. 1899. Fossiliferous black shales. Pop. Sci. News 33: 97-98. Manspeizer, W., H. L. Cousminer, and J. H. Puffer. 1978. Separation of Morocco and eastern North America: a Triassic-Liassic stratigraphic record. Geol. Soc. Am. Bull. 89: 901-920. Manspeizer, W., and P. E. Olsen. 1981. Rift basins of the passive margin: tectonics, organic-rich lacustrine sediments and basin analysis, p. 25-103. In G. W. Hobb [ed.]. Field guide to the geology of the Paleozoic, Mesozoic and Tertiary rocks of New Jersey and the Central Hudson Valley. Petroleum Exploration Society of New York, New York. McCune, A. R. 1981. Quantitative description of body form in fishes: implications for species level taxonomy and ecological inference. Copeia 1981(4): 897-901. 1982. Early Jurassic Semionotidae (Pisces) from the Newark Supergroup: systematics and evolution of a fossil species flock. Ph.D. dissertation, Yale University. 371 p. 1985. Semuionotus Agassiz 1832 (Pisces): designation of Semionotus berger: as the type species under the plenary powers. Bull. Int. Comm. Zool. Nomencl. 42(4): 371-373. 1986. A revision of Semionotus (Pisces: Semionotidae) from the Triassic and Jurassic of Europe. Palaeont. 29(2): 213-233. McCune, A. R., K. S. Thomson, and P. E. Olsen. 1984. Semionotid fishes from the Mesozoic great lakes of North America, p. 27-44. Jn A. A. Echelle and I. Kornfield [eds.]. Evolution of fish species flocks. Univ. Maine Press, Orono. McDonald, N. G. 1975. Fossil fishes from the Newark Group of the Connecticut Valley. Masters thesis, Wesleyan University, Middletown, Connecticut. 230 p. McLaughlin, D. B. 1939. A great alluvial fan in the Triassic of Pennsylvania. Mich. Acad. Sci. Pap. 24: 59-74. Neff, N., and L. Marcus. 1980. A survey of multivariate methods for systematics. For a workshop, “Numerical Methods in Systematic Mammalogy.” American Society of Mammalogists Annual Meeting 1980. Am. Mus. Nat. Hist., New York. 243 p. Nelson, G. 1980. Multiple branching in cladograms: two interpretations. Syst. Zool. 29(1): 86-91. Nelson, G., and N. Platnick. 1981. Systematics and biogeography. Columbia University Press, New York. 567 p. Newberry, J. S. 1888. Fossil fishes and fossil plants of the Triassic rocks of New Jersey and the Connecticut Valley. U.S. Geol. Surv. Monogr. XIV. 152 p. Olsen, P. E. 1975. The microstratigraphy of the Roseland Quarry (Early Jurassic, Newark Su- pergroup). Open file report to Essex County Park Commission. 87 p. 1978. On the use of the term Newark for Triassic and Early Jurassic rocks of eastern North America. Newsl. Stratigr. 7: 90-95. 1980a. Triassic and Jurassic Formations of the Newark Basin, p. 2-39. Jn W. Manspeizer [ed.]. Field studies of New Jersey geology and guide to field trips. 52nd Annu. Meet., N.Y. State Geol. Assoc. Newark College of Arts and Sciences, Rutgers University, Newark, N.J. 1980b. Fossil Great Lakes of the Newark Supergroup in New Jersey, p. 352-398. In W. Manspeizer [ed.]. Field studies of New Jersey geology and guide to field trips. 52nd Annu. Meet., N.Y. State Geol. Assoc. Newark College of Arts and Sciences, Rutgers University, Newark, N.J. 1980c. A comparison of the vertebrate assemblages from Newark and Hartford Basins (Early Mesozoic, Newark Supergroup) of eastern North America, p. 35-53. Jn L. L. Jacobs [ed.]. Aspects of vertebrate history, essays in honor of Edwin Harris Colbert. Northern Arizona Press, Flagstaff. 1983. On the non-correlation of the Newark Supergroup by fossil fishes: biogeographic, structural, and sedimentological implications. Geol. Soc. Am. Abst. 15: 122. 1984a. Periodicity of lake-level cycles in the Late Triassic Lockatong Formation of the Newark Basin (Newark Supergroup), New Jersey and Pennsylvania, p. 129-146. Jn A. Berger, J. Imbrie, J. Hays, G. Kukla, and B. Saltzman [eds.]. Milankovitch and climate. Reidel, Boston. 1984b. Comparative paleolimnology of the Newark Supergroup: a study of ecosystem evolution. Ph.D. dissertation, Yale University. 726 p. 1984c. The skull and pectoral girdle of the parasemionotid fish Watsonulus eugnathoides from the Early Triassic Sakema Group of Madagascar with comments on the relationships of the holostean fishes. J. Vert. Paleo. 4: 481-499. Olsen, P. E., C. L. Remington, B. Cornet, and K.S. Thomson. 1978. Cyclic change in Late Triassic lacustrine communities. Science 201: 729-733. Olsen, P. E., A. R. McCune, and K.S. Thomson. 1982. Correlation of the Early Mesozoic Newark Supergroup by vertebrates, principally fishes. Am. J. Sci. 282: 1-44. Olsen, P. E., and A. R. McCune. ms. Morphology of Early Jurassic fishes from the Newark Supergroup of eastern North America with comments on the family Semionotidae and description of a new genus. Orlov, Y. A., ed. 1967. Fundamentals of paleontology, Vol. XI, Agnatha, Pisces. Translated from Russian, Israel Program for Scientific Translation. Smithsonian Institution and the National Science Foundation, Washington, D.C. Patterson, C. 1973. Interrelationships of holosteans, p. 233-305. Jn P. H. Greenwood, R. S. Miles, and C. Patterson [eds.]. Interrelationships of fishes. Academic Press, London. EARLY JURASSIC SEMIONOTID FISHES 107 1975. The braincase of pholidophorid and leptolepid fishes, with a review of the actinop- terygian braincase. Philos. Trans. R. Soc. London 269: 275-579. 1981. Significance of fossils in determining evolutionary relationships. Annu. Rev. Ecol. Syste 12 195-223: Patterson, C., and D. E. Rosen. 1977. Review of ichthyodectiform and other Mesozoic teleost fishes and the theory and practice of classifying fossils. Bull. Am. Mus. Nat. Hist. 158(2): 83-172. Pearson, D. M. 1981. Functional aspects of the integument in polypterid fishes. Zool. J. Linn. Soc. 72: 93-106. Rachootin, S. P., and K. S. Thomson. 1981. Epigenetics, paleontology and evolution, p. 181-193. In G. G. E. Scudder and J. L. Reveal [eds.]. Evolution today. Proc. Int. Cong. Syst. Evol. Biol. Hunt Institute for Botanical Documentation, Carnegie-Mellon University, Pittsburgh. Raff, R. A., and T. C. Kauffman. 1982. Embryos, genes and ancestors. Macmillan Publishing Co., Inc., New York. 395 p. Raup, D. M. 1966. Geometric analysis of shell coiling: general problems. J. Paleontol. 40: 1178-1190. Redfield, J. H. 1837. Fossil fishes of Connecticut and Massachusetts, with a notice of an undescribed genus. Ann. Lyc. Nat. Hist. N.Y. 4: 35-40. Redfield, W. C. 1839. Fossil fishes in the red sandstone of New Jersey. Am. J. Sci. 35(Misc. No. 1D) al 925 1841. Short notices of fossil fishes. Am. J. Sci. 41: 24-28. 1843. Fish beds and a fossil foot mark in New Jersey. Am. J. Sci. 44: 134-136. Reeside, J. B., P. L. Applin, E. H. Colbert, J. T. Gregory, H. D. Hadle, B. Kummel, P. J. Lewis, J. D. Love, M. Maldonado-Koerdell, E. D. McKee, D. B. McLaughlin, S. W. Muller, D. A. Reinemund, J. Rodgers, J. Sanders, N. J. Siberling, and K. Waage. 1957. Correlation of the Triassic formations of North America, exclusive of Canada. Geol. Soc. Am. Bull. 68: 1451-1514. Roth, V. L. 1984. On homology. Biol. J. Linn. Soc. London 22: 13-29. Sanders, J. E. 1974. Guidebook to field trip in Rockland County, N.Y. Petro. Explor. Soc. New York, New York. 87 p. Schaeffer, B. 1948. A study of Diplurus longicaudatus with notes on the body form and locomotion of the Coelacanthini. Am. Mus. Novit. 1378: 1-31. 1952. The palaeoniscoid fish 7urseodus from the Upper Triassic Newark Group. Am. Mus. Novit. 1581: 1-24. 1967. Late Triassic fishes from the western United States. Bull. Am. Mus. Nat. Hist. 135(6): 287-342. Schaeffer, B., and D. Dunkle. 1950. A semionotid fish from the Chinle Formation, with considerations of its relationships. Am. Mus. Novit. 1457: 1-29. Schaeffer, B., D. Dunkle, and N.G. McDonald. 1975. Ptycholepis marshi Newberry, a chondrostean fish from the Newark of eastern North America. Fieldiana, Geol. 33(1): 205-233. Schaeffer, B., and N. G. McDonald. 1978. Redfieldiid fishes from the Triassic-Liassic Newark Supergroup of eastern North America. Bull. Am. Mus. Nat. Hist. 159(4): 131-173. Siegel, A. F., and R. H. Benson. 1982. A robust comparison of biological shapes. Biometrics 38(2): 341-350. Silliman, B. 1821. Miscellaneous observations relating to geology, mineralogy and some connected topics, in extracts of letters from Mr. Alexander Brongniart, member of the Royal Academe des Sciences, Engineer of mines, & c., of Paris with remarks by the editor to whom the letters were addressed. Am. J. Sci. 3: 216-227. Suttkus, R. D. 1963. Order Lepisostei, p. 61-88. Jn Fishes of the Western North Atlantic. Memoir, Sears Foundation for Marine Research, No. 1, Part 3. Sears Foundation for Marine Research, New Haven, Connecticut. Thomson, K. S., and A. R. McCune. 1984. Scale structure as evidence of growth patterns in fossil semionotid fishes. J. Vertebr. Paleontol. 4(3): 422-429. Traquair, R. H. 1877. On the Agassizian genera Amblypterus, Paleoniscus, Gyrolepis, and Pygopterus. Q. J. Geol. Soc. London 33: 548-578. Twenhofel, W. H. 1939. Environments of origin of black shales. Am. Assoc. Pet. Geol. Bull. 23: 1178-1198. Van Houten, F. B. 1962. Cyclic sedimentation and the origin of analcime-rich Upper Triassic Lockatong Formation, west-central New Jersey and adjacent Pennsylvania. Am. J. Sci. 260: 561-576. 1964. Cyclic lacustrine sedimentation, Upper Triassic Lockatong Formation, central New Jersey and adjacent Pennsylvania, p. 497-531. Jn Symposium on Cyclic Sedimentation. Kansas State Geol. Surv. Bull. 169(2). 1965. Composition of Triassic Lockatong and associated formations of the Newark Group, central New Jersey and adjacent Pennsylvania. Am. J. Sci. 263: 825-863. 1969. Late Triassic Newark Group, north central New Jersey and adjacent Pennsylvania and New York, p. 314-347. Jn S. Subitzki [ed.]. Geology of selected areas in New Jersey and Eastern Pennsylvania. Rutgers University Press, New Brunswick, New Jersey. 1977. ‘Triassic-Liassic deposits of Morocco and eastern North America: comparison. Am. Assoc. Petrol. Geol. 61: 79-99. Wenz, S. 1967. Complements a [etude des poissons actinopterygiens du Jurassique frangaise. C.N.R.S., Paris. 276 p., 48 pl. 108 PEABODY MUSEUM BULLETIN 43 Wiley, E. O. 1976. The phylogeny and biogeography of fossil and recent gars (Actinopterygii: Lepisosteidae). Misc. Publ. Univ. Kansas Mus. Nat. Hist. 64: 1-111. 1981. Phylogenetics: the theory and practice of phylogenetic systematics. Academic Press, New York. 439 p. Woodward, A. S. 1895. Catalogue of the fossil fishes in the British Museum. British Museum (Natural History), London. 4 vols. 1916-19. The fossil fishes of the English Wealden and Purbeck Formations. Palaeontogr. Soc. (monogr.). 148 p., 26 pl. ISBN No. 0-912532-03-: i beet Pov tense 1078 Pe Ct one Mot Cited nade a foots o ‘ wa were ore . fal “a> oe . 5 Dent Ott HIS Binet eh te e 1 PPT ay ee eed OMe ah eh 299 DALES PRT ron perarvesy wee ey peer ee ee tiok ERMA OF Tee kites orl Parts ie Deke ONAL ee AGS ee weg eens atanetreee nt s our he et heer se cnehoneher Jor rmsgye eee © Pee a np Fab ne et oANeee of Ree & a aeat rte fo ee ered ee Lhe Cth ot Leeann Pita nee ee OP aids ahr nne Pans Tri mat ae eth seen’ = ane St a ae et FRED OF Sane eyes pe peer td Pov ee oat penerene er a ae re od, ate ree nae Mya a att Oe vert wre regnet pe were ee eth ht et ott! Per ae yes es eet eee 7 s AIA Ae eA Vewy Ore Me ee Wy ao Per pera a Aare ee PO Hee eS eee keer 0 Ly ery erat wey Wee Ue NS oe oe we Aar Oras es avg ey Ae CUTE IES & Teewre en Fey yaNDA Ger ALY EN PPAR Te AES yueeet ube ta enh ree B F AU Nay CATON Ny PU GHB ee HSE EY iv AUER ERT LIS Wiinamarg eva oy aay Pa ey, Se ak ea event a did cL gere beurre AS yy WW SNES ve Wee © es ee CIA T INTE S ernie Gs Co kee Hs UARSENGIE NG Fs a! MLW IV eG, A ae PE ee pores vane ’ - aa Sew ew, GD MEE VENER MT ee Tov aeU i AU NAN Lea a Te ok EA Be Je VAR OE wae >, wee Ny acing 0 ae : Sard egy EE LOD T De ee Ne veg Bewik Tew re hy re paraey STYRIA eae PEGs bre Oh Yast) pre a ee eneyeve WS wurden Grey ¥ Wye UAT Eira INTE ewe Br od wees Oe NPE PALLY eee FUENTES 7a) WE ANY EY RYE VEN a sg) Y, wey ear NP RU MOEN AI PORANY WHEN pests parr pe eae % PE MTT EV! a oy ea ; ae