»am‘ .w- u ‘ J 3,1 3‘35 " ‘6. V ‘.;’g‘ I ' q; I f: ' :1- z.‘ , M12? ‘ 3,! 1 . ‘ f'éfv ”7",! nit-‘3 ,M n.»- n. '"It .1“ "5:5 a” v '15.. :11? d . i... ‘4». ”Is: 1‘ l 38;, I‘LL? " 519691;: E “ mu? . m g ""1332. ...'o«,_ a». - -L‘ .. 900 54505§?3 This is to certify that the dissertation entitled Morphology and phylogenetic implications of Recent and fossil carcharhiniform shark vertebral centra presented by John H. Burris has been accepted towards fulfillment of the requirements for the Ph.D. degree in Geologigal Sciences ’ Michael‘D. Gottfried 28 M 7430" I Date MSU is an Affirmative Action/Equal Opportunity Institution LIBRARY MiChiQan State University PLACE lN RETURN BOX to remove this checkout from your record. TO AVOID FINES return on or before date due. MAY BE RECALLED with earlier due date if requested. DATE DUE DATE DUE DATE DUE 6/01 cJClRC/Dateouepes-p. 15 MORPHOLOGY AND PHYLOGENETIC IMPLICATIONS OF RECENT AND FOSSIL CARCHARHINIFORM SHARK VERTEBRAL CENTRA By John H. Burris A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Geological Sciences 2004 ABSTRACT MORPHOLOGY AND PHYLOGENETIC IMPLICATIONS OF RECENT AND FOSSIL CARCHARHINIFORM SHARK VERTEBRAL CENTRA By John H. Burris The cross—sectional anatomy of secondary calcifications of shark vertebral centra has featured in phylogenetic hypotheses, although never in a rigorous cladistic fi'amework. In this study, the internal calcification patterns, along with the external morphology, of fossil and Recent shark centra of the Order Carcharhiniformes have been coded and subjected to a cladistic analysis to address the utility of centrum features for revealing relationships. Carcharhiniform sharks were selected as a study group because they are a monophyletic clade with reasonably well understood intraordinal relationships, a rich fossil record, and readily available Recent comparative skeletal material. External clmracters include centrum proportions, the presence and distribution of cartilage canals, and the size, shape, and spacing of the foramina for the basidorsal and basiventral arch components. The internal calcification features evaluated include the morphology and spacing of the four intermedialia, the four noncalcified areas, and the four diagonal calcifications. Centrum characters were analyzed both separately and combined with other morphological characters from previous analyses. Results of the cladistic analysis show that shark centrum characters are useful for elucidating phylogeny. Tree topology was very similar for both analyses, and similar to recent molecular databased phylogenies. The addition of centrum data to shark phylogenetic analyses will allow for a more objective means of determining the interrelationships of fossil and extant carcharhiniform sharks than studies based on teeth alone, with their well—documented difficulties. The data gathered will also be important for future studies to interpret the relationship between centrum morphology and swimming characteristics in extant and, ultimately, extinct taxa. ACKNOWLEDGMENTS This dissertation was completed as a result of the assistance and support of many people and institutions, and I am grateful for all of their help. I would like to thank Dr. Michael D. Gottfried for his assistance on this project. I appreciate his guidance and suggestions, his review of this dissertation, and for giving me the opportunity to pursue this degree as his student. I wish to thank Drs. Robert L. Anstey, Gary S. Weissmann, and Thomas G. Coon for serving on my committee and for their suggestions and criticisms of this project. I am indebted to Mr. John Grove at the MSU Department of Radiology for the donation of his time, equipment, and materials for creating the x—radiographs of my centra. I wish to express my gratitude to the many institutions that generously loaned specimens to the MSU Museum for my study, including the American Museum of Natural History, Academy of Natural Sciences, Philadelphia, The Natural History Museum (London), California Academy of Sciences, Calvert Marine Museum, San Diego Natural History Museum, University of California Museum of Paleontology, Florida Museum of Natural History, Smithsonian Institution, National Museum of Natural History, and Dr. Gordon Hubbell. Funding was provided from a variety of sources, and I am gratefiil for their support: Stephen J. Gould Student Grant (The Paleontological Society), UCMP Welles Fund, MSU College of Natural Sciences Dissertation Completion Fellowship, MSU iv Graduate Office Fellowship, MSU Graduate Travel Fellowship, and the Lucile Drake Pringle and Gordon H. Pringle Endowed Fellowship. Special thanks to Dr. Danita Brandt for always having an open door to talk and her help in so many different ways, to Lisa Whitenack and Joann Labs for their help in obtaining specimens, to Yasemin Tulu for her help in measuring specimens and putting up with dead sharks in the office, and the rest of the staff and students at the MSU Department of Geological Sciences for their fi'equent and varied help. Finally, I am thankful for the support and love my family has given me through the years. I am especially grateful to my wife, Carol, for her continued good humor and patience as I continued to spend evenings and weekends working on my project instead of with her, and for helping me relax when it all got to be a bit too much. She has given me more than she realizes. I would especially like to thank her for waiting until the day after my defense for going into labor with our son, William. TABLE OF CONTENTS LIST OF FIGURES ........................................................................................................ ix INTRODUCTION ........................................................................................................... l SHARK CENTRA .......................................................................................................... 5 DEVELOPMENT ..................................................................................................... 6 CENTRUM MORPHOLOGY ................................................................................... 8 AGE DETERMINATION USING CENTRUM GROWTH RINGS ........................ 16 REGIONS OF THE VERTEBRAL COLUMN ........................................................ 18 COMPOSITION AND FUNCTION ........................................................................ 20 NOMENCLATURE ................................................................................................ 21 MATERIALS AND METHODS RADIOGRAPHS ..................................................................................................... 25 PHOTOGRAPHS .................................................................................................... 25 CENTRUM MEASUREMENTS ............................................................................. 25 SPECIMENS AVAILABLE FOR STUDY .............................................................. 28 SPECIMEN CATALOGING SYSTEM ................................................................... 28 ABBREVIATIONS OF INSTITUTIONS ................................................................ 29 SYSTEMATIC DESCRIPTION FAMILY TRIAKIDAE TRIAKIS Description of Modern T riakis centra ........................................................... 30 Description of Fossil Triakis centra .............................................................. 34 MUSTEL US Description of Modern Mustelus centra ........................................................ 35 Description of Fossil Mustelus centra ........................................................... 38 GALEORIHNUS Description of Modern Galeorhinus centra ................................................... 39 Description of Fossil Galeorhinus centra. ..................................................... 42 SUMMARY OF TRIAKIDAE ........................................................................... 44 FAMILY HEMGALEIDAE HEMIPRIST IS ................................................................................................... 46 Description of Fossil Hemipristis centra ....................................................... 46 FAMILY CARCHARHINTDAE GALEOCERDO Description of Modern Galeocerdo centra .................................................... 49 Description of Fossil Galeocerdo centra ....................................................... 52 RHIZOPRIONODON Description of Modern Rhizoprionodon centra ............................................. 53 Description of Fossil Rhizoprionodon centra ................................................ 56 vi PRIONA CE Description of Modern Prionace centra ........................................................ 57 Description of Fossil Prionace centra ........................................................... 6O NEGAPRION Description of Modern Negaprion centra ..................................................... 62 Description of Fossil Negaprion centra ........................................................ 65 CARCHARHINUS Description of Modern Carcharhinus centra ................................................ 67 Description of Fossil Carcharhinus centra ................................................... 70 SUMMARY OF CARCHARHINIDAE ............................................................. 72 FAMILY SPHYRNTDAE SPHYRNA Description of Modern Sphyma centra ......................................................... 74 Description of Fossil Sphyma centra ............................................................ 77 INDETERMINATE CARCHARHINIFORM 1 ....................................................... 79 INDETERMINATE CARCHARHINIFORM 2 ....................................................... 82 DISCRIMINANT ANALYSIS TESTING CLASSIFICATION OF FOSSIL CENTRA INTRODUCTION TO DISCRIMINANT ANALYSIS ............................................ 86 METHODS ............................................................................................................. 89 DISCRIMINANT ANALYSIS TESTING IDENTIFICATION OF MODERN CENTRA ........................................................................................................... 92 DISCRIMINANT ANALYSIS INCLUDING FOSSIL CENTRA ............................ 97 CALVERT MARINE MUSEUM FOSSIL CENTRA .............................................. 98 FLORIDA MUSEUM OF NATURAL HISTORY FOSSIL CENTRA ..................... 99 THE NATURAL HISTORY MUSEUM (LONDON) FOSSIL CENTRA .............. 101 SAN DIEGO NATRUAL HISTORY MUSEUM FOSSIL CENTRA .................... 103 ACADEMY OF NATURAL SCIENCES OF PHILADELPHIA FOSSIL CENTRA .......................................................................................................... 106 NATIONAL MUSEUM OF NATURAL HISTORY FOSSIL CENTRA ............... 108 UNIVERSITY OF CALIFORNIA MUSEUM OF PALEONTOLOGY FOSSIL CENTRA ......................................................................................................... 110 PHYLOGENETIC ANALYSIS .................................................................................. 113 RESULTS ............................................................................................................. 114 RELATIONSHIPS OF CARCHARHINIFORM SHARKS USING CENTRUM DATA .............................................................................................................. 119 RELATIONSHIPS OF CARCHARHINIFORM SHARKS USING ALL AVAILABLE MORPHOLOGICAL DATA .................................................... 124 CONCLUSIONS IDENTIFICATION OF CARCHARHINIFORM VERTEBRAL CENTRA ........... 126 TAXONOMIC VALUE OF SHARK VERTEBRAL CENTRA ............................. 128 FUTURE WORK .................................................................................................. 131 APPENDD( A: SHARK CENTRA MEASUREMENTS ............................................. 134 vii APPENDIX B: PEARSON CORRELATION MATRIX ............................................. 176 APPENDIX C: DISCRIMINANT ANALYSIS RESULTS ......................................... 178 APPENDIX D: PHYLOGENETIC ANALYSIS CHARACTERS ............................... 198 REFERENCES ............................................................................................................ 214 viii LIST OF FIGURES Number Page 1. Ordinal relationships of Carcharhiniformes and cross—sectional anatomy of a typical carcharhiniform centrum ............................................................................................. 3 2. lateral view of a section of a typical carcharhiniform vertebral column ....................... 9 3. Four typical carcharhiniform centrum shapes ............................................................. 11 4. Variations in cartilage canals on vertebral centra ....................................................... 13 5. Trend in cross—sectional calcification pattern in carcharhiniform shark centra ........... 24 6. Carcharhiniform centrum measurements .................................................................... 26 7. Photographs and x—radiographs of fossil and Recent T riakis centra ........................... 31 8. Photographs and x—radiographs of fossil and Recent Mustelus centra ........................ 36 9. Photographs and x—radiographs of fossil and Recent Galeorhinus centra ................... 40 10. Photographs and x—radiographs of fossil Hemipristis centra .................................... 48 11. Photographs and x—radiographs of fossil and Recent Galeocerdo centra .................. 50 12. Photographs and x—radiographs of fossil and Recent Rhizoprionodon centra ........... 54 13. Photographs and x—radiographs of fossil and Recent Prionace centra ...................... 58 14. Photographs and x—radiographs of fossil and Recent Negaprion centra .................... 63 15. Photographs and x—radiographs of fossil and Recent Carcharhinus centra ............... 68 16. Photographs and x—radiographs of fossil and Recent Sphyrna centra ....................... 75 17. Photographs and x—radiographs of fossil Indeterminate Carcharhiniform 1 centra... 80 18. Photographs and x—radiographs of fossil Indeterminate Carcharhiniform 2 centra... 83 l9. Consensus tree and possible tree morphology showing relationships for Carcharhiniformes based on centrum data .............................................................. 1 16 20. Consensus tree and possible tree morphology showing relationships for Carcharhiniformes based on centrum data .............................................................. 117 21. Consensus tree and single most parsimonious tree showing relationships for Carcharhiniformes based on whole specimen characters and combination of whole specimen and centrum characters ........................................................................... 1 18 22. Naylor (1992) consensus trees and Distance Wagner tree showing relationships of Carcharhiniformes ................................................................................................. 121 INTRODUCTION F ossilized sharks have been studied for over a hundred and seventy years, (e. g. Agassiz, 1833—1843), though most of these studies focus only on isolated teeth, particularly for Mesozoic and Cenozoic sharks. The shark skeleton is composed entirely of cartilage, with a low potential for preservation. While articulated fossil shark skeletons are known from the Paleozoic (e.g. Coates and Sequeira, 2001; Lund, 1985; Maisey, 1989), skeletons from the Mesozoic and Cenozoic are rare. Fossilized shark teeth, however, are the most common and abundant vertebrate fossil in the entire fossil record (Maisey, 1984). These teeth are abundant due to their hard enamel covering, and the tremendous numbers produced during the life span of these animals. Fossil shark teeth present many difficulties for study and identification, especially due to the fact that nearly all sharks display heterodonty in their dentition. This heterodonty is ofien monognathic, where the teeth in a single jaw have different morphologies depending on position within that jaw (Compagno, 1970). Most sharks display dignathic heterodonty, where teeth of the upper and lower jaws lave different morphologies (the most commonly observed form of heterodonty in Carcharhiniformes [Cappetta, 1987], the focus group of this study). In many cases, a single individual may have both conditions. Additionally, young often display different tooth morphologies than adults, and sexual dimorphism is also displayed in the teeth of many sharks (e.g. Scyliorhinus, Carcharhinus, and Galeus) (Cappetta, 1987). The difficulty of using shark teeth is heightened by convergence in tooth morphology among different species of sharks specialized in similar feeding habits. Researchers studying fossil shark teeth often described new genera and species based on small samples of teeth. This practice, coupled with the scarcity of comparative material among the living sharks, has led to many invalid shark taxa (Cappetta, 1987; Compagno, 1988). In one example, Case and Cappetta (1997) described 44 fossil shark species from a single Late Cretaceous locality; to put this into context, there are only approximately 350 total extant species of sharks worldwide. Teeth are not the only shark skeletal element to be preserved, however. Neoselachians secondarily calcify regions of their vertebral centra with a bone—like tissue (e.g., Applegate, 1967), and the centra are commonly found as fossils. Hasse (1879— 1885) and Ridewood (1921) both recognized the potential for study of the calcification patterns within shark vertebral centra. Shark centra are morphologically distinct between different orders, but are also distinct among different genera within a single family. Because of the well—documented problems with fossil shark teeth, these additional fossil elements are potentially an important source for additional data, but fossilized shark centra are rarely studied or even identified. This study describes the morphology of the vertebral centra from four carcharhiniform families, and discusses their usefulness for identification and systematics. The Order Carcharhiniformes, sistergroup to the Order Lamniformes, is the focal clade for this project (Figure la). Carcharhiniformes includes eight families and over sixty fossil and extant genera, and represents about 55% of the approximately 350 living shark species and 25% of all living elasmobranchs (Compagno, 1988; Naylor, 1992). Carcharhiniform sharks are the most numerous in species, and also likely in individuals, of all sharks. They predominate in warmer seas over continental shelves and slopes, but are distributed worldwide, ranging fi'om tropical to polar seas, and inhabit both benthic middle zone (double cone) " . iagonal calcification Figure l. A. Ordinal relationships of the Recent Elasmobranchii. Modified fiom Carvalho (1996). B. Cross section of Sphynra centrum. Modified from Ridewood (1921). and pelagic regions (Compagno, 1984). Comparative Recent material is relatively easy to obtain as a result of their abundance. Carcharhiniform sharks are also an old group, with their first appearance in the Jurassic (Cappetta, 1987). More advanced carcharhinids first appear in the Paleogene, but did not achieve their high diversity until the Neogene (Maisey, 1984). Carcharhiniform fossils, including vertebral centra, are common in deposits along both the East and West Coasts of North America, and are also found worldwide in other shallow marine deposits. Despite their high taxonomic diversity, carcharhiniform sharks are far less diverse morphologically than other shark orders, and have few highly specialized sharks (Compagno, 1988). With the exception of the lmmmerheads, members of this order are less distinct from each other than are lamniform or orectolobiform sharks. Carcharhiniform sharks form a morphological gradient, with scyliorhinids at one extreme, and highly derived carcharhinids and sphyrnids at the other, with all the other members between these extremes (Compagno, 1988). This “gradient” makes additional techniques for clarifying carcharhiniform taxonomy essential. This study focuses on ten genera among four families of carcharhiniform sharks, including the Triakidae (houndsharks), Hemigaleidae (weasel and snaggletoothed sharks), Carcharhinidae (requiem sharks), and Sphymidae (hammerhead sharks). These four families are the most derived of the order, are common as fossils, and (with the exception of Hemigaleidae) have readily obtainable Recent comparative specimens. The four families not included are the Scyliorhinidae (catsharks), Proscylliidae (finback catsharks), Pseudotriakidae (false catsharks), and Leptochariidae (barbeled houndsharks). Members of these four families are rare as fossils and Recent comparative material is difficult to obtain. Scyliorhinids, proscylliids, and pseudotriakids are all deepwater sharks or are found on continental slopes, environments that offer a low probability of fossilization. In addition, proscylliids, pseudotriakids, and leptochariids have low diversity, with only six genera among the three families. SHARK CENTRA The vertebral column of neoselachian sharks, including the Order Carcharhiniformes, consists of a central axis of centra, with a series of neural arches on the dorsal surface, and ventral ribs (anteriorly) or haemal arches (posteriorly) on the ventral surfaces. The neural arch encloses the dorsal nerve chord, and the haemal arches protect the dorsal aorta, and the posterior cardinal vein in the caudal region. The vertebral centrum is the major unit of the vertebral column, housing the notochord or its remnants and providing structural support for the shark. Shark centra are amphicoelous to varying degrees, with concave anterior and posterior faces, giving them a bobbin— or spool—like shape. The notochord passes through the center of the centra which may be constricted and reduced to intervertebral lenses through ontogeny (Cappetta, 1987). Consecutive centra are firmly united by fibrous rings, or intervertebral ligaments, and by white fibrous tissue that occur between the cartilage of the neural and haemal arches. Unlike teleosts, neoselachians do not have zygapophyseal processes for articulation of the vertebrae (Ridewood, 1921). Neoselachian sharks have a cartilaginous skeleton, virtually lacking true bone. The vertebral column, likewise, is composed of cartilage. Neoselachians secondarily calcify regions of their vertebral centra, adding rigidity and strength to the column as an adaptation to the compressive forces experienced during locomotion through the dense aquatic environment (Ridewood, 1921; Daniel, 1934; Budker, 1971). This secondary calcification often forms distinct antero—posterior cross—sectional patterns within different lineages. Much discussion has occurred on the systematic significance and usefulness of the calcification patterns for elasmobranch classification (e.g., Gill, 1893; Jordan and Evermann, 1896; Jordan, 1923; White, 1938; Fowler, 1941; and Smith, 1949), which will be revisited in this dissertation. Development Information on development of the elasmobranch vertebral column is derived mainly from Hasse (1879—1885) and Ridewood (1921). In the early stages of development, a chordal cell—produced sheath invests the notochord. This sheath differentiates into two regions; forming externally is the membrana elastica externa, and on the interior a thick, fibrous sheath forms called the membrana elastica interna. This layer of cells forms the innermost portion of the notochordal sheath. Skeletogenous tissue is applied to the lateral surfaces of the notochordal sheath as the sclerotomes differentiate fi'om the myotomes. This Skeletogenous tissue is not continuous on the dorsal and ventral surface of the notochordal sheath. The arch—cartilages differentiate later in the upper and lower parts of the tracts while the middle part of each tract becomes reduced to a layer roughly two cells thick (Figure lb). During early development, the membrana elastica interna differentiates into three regions, known as the inner, middle, and outer zones (Figure lb). The middle zone is composed of fibrous cartilage, which calcifies and forms the double cone of the centra. The inner zone is composed of hyaline cartilage, and is mostly found at the apices of the double cone. The outer zone initially consists of hyaline cartilage, and comprises a large volume of the centrum. This outer zone may become secondarily calcified in adults by the development of an investing layer immediately external to the double cone. The outer zone may also secondarily develop longitudinal calcified laminae, either in a radiating star—like pattern (e.g., Lamniformes) or in a series of concentric tubes (e.g., Squatiniformes). These secondary outer zone calcifications are independent of the primary double cone calcification. Skeletogenous cells next invade the inner fibrous sheath of the notochord through foramina in the membrana elastica externa, especially from the arch—bases, and produce a layer of cartilage between the membrana elastica extema and the inner fibrous sheath (or membrana elastica interna). The once continuous elastica interna differentiates into rings of hyaline cartilage set end to end. These rings continue to grow and develop into centra. The remaining areas of the chordal sheath ultimately develop into the fibrous intervertebral ligaments articulating adjacent centra together. As the centra continue to rapidly grow and develop, the notochord becomes constricted at the apices of the double cone by the hyaline cartilage of the inner zone, but continues to increase in size in the intervertebral spaces. Any calcification outside of the notochordal sheath layers is termed peripheral calcification (Ridewood, 1921). The neural and haemal arches develop independently of the centra, with the exception of the Skeletogenous cells immigrating to the inner fibrous sheath through foramina at the arch—bases. The neural and haemal arches are, for the most part, morphologically distinct and external to the centra. The arches do become continuous with the sheath cartilage at the areas where the arches attach to the centra as a result of the foramina in the membrana elastica externa. Each centrum has a pair of dorsal and a pair of ventral foramina for the attachment of arch cartilages. A complete neural arch in elasmobranch fishes is composed mainly of the basidorsal and interdorsal cartilages, and to a lesser degree, small, unpaired supradorsals and suprainterdorsals (Figure 2). The basidorsal cartilage is typically immediately dorsal to the centrum, with the exception of the transition zone (i.e., the transition from monospondylous to diplospondylous vertebrae). The interdorsal cartilages are immediately dorsal to the intervertebral ligaments. The haemal arches are composed mainly of basiventral cartilages. The interventral cartilages appear with less regularity than do their dorsal counterparts. While the haemal arches are complete and closed in the caudal region, those vertebrae anterior to the cloaca are incomplete, and the basiventral cartilages in this more anterior region are known as transverse processes and form ventral ribs (Cappetta, 1987). Centrum Morphology As mentioned earlier, a single vertebral centrum has a cylindrical shape that varies fi'om a bobbin- to spool-like shape. The lateral sides of a centrum may be concave, somewhat resembling an hourglass, straight, or slightly convex (Kozuch and Fitzgerald, 1989). Kozuch and Fitzgerald (1989) classify four basic shapes for shark centra. The centrum may be a cylinder, where the sides are nearly flat, with perhaps slight concavity or convexity to the outer wall. The modified cylinder is like the cylinder, except it has concave sides that recurve slightly near the rim. A fluted cylinder is a long, Supradorsal Basidorsal \ \ Interventral Interdorsal Vertebral centrum Basiventral \ Figure 2. Lateral view of a section of a shark vertebral column. Modified from Cappetta (1987). slender centrum that flares at each end. Finally, some centra have a strong hourglass shape with a pinched middle area and sides that are strongly recurved at the rims (Figure 3). As mentioned earlier, shark centra vary in length regionally (i.e., MP centra tend to be much longer than DP or DC centra), but also vary among different genera. Fluted cylinder—shaped centra tend to be relatively the longest centra, along with hourglass— shaped centra. Cylinder— and modified—cylinder shaped centra tend to be of moderate to short length. In articular view, shark centra are most often round. The centra of many genera are ovoid in articular view, with a medio—lateral breadth that is greater than dorso—ventral height. Less commonly, some centra are larger in dorso—ventral height than medio—lateral breadth. An obvious feature on both Recent and fossil carcharhiniform centra is the foramina on the dorsal and ventral surfaces of the centra In life, these foramina housed the cartilage of the basidorsal and basiventral cartilages. These foramina are typically oval, rectangular, or square, and have varying width and length proportions depending on genus, or, to a lesser degree, the region within the column. In some genera, the foramina extend fully into the rims, while in others, the foramina are not sufficiently long. The interforaminal wall, that is, the space between the two dorsal or two ventral foramina, also varies in width. The width of the dorsal interforaminal wall tends to be more consistent within the column than the width of the ventral interforaminal wall, which widens dramatically towards the caudal regions of the column. 10 Figure 3. Four common shapes of carcharhiniform centra. A. Fluted cylinder (UCMP 148061, 'l'riakis, dorsal view), B. Modified cylinder (CAS 65084, Galeocerdo, dorsal view), C. Cylinder (AN SP LCM 27, Can-harhinus, dorsal view), D. Hourglass (AMNH 93 843, Sphyma, dorsal View). Scale bars = 10 mm. Afler Kozuch and Fitzgerald (1989). ll Another physical characteristic of carcharhiniform centra is the size and spacing of the small pores found between the foramina. These pores vary in size from species to species, and follow a variety of patterns (Kozuch and Fitzgerald, 1989). These pores may be scattered along the external surface, may follow the outline of the rim and the foramina, may be concentrated in irregular clusters at the rim, or may be found midway between the rims along the sides of the centra (Kozuch and Fitzgerald, 1989) (Figure 4). Hoenig and Walsh (1982) identify these pores as cartilage canals, rod—shaped or branching structures penetrating the intermedialia, external to the notochordal sheath. These canals are especially well developed in the carcharhiniform families Carcharhinidae and Sphyrnidae. These structures are found throughout the intermedialia in mature individuals, though the presence of newer cartilage surrounding and filling some canals suggests destruction by chondrogenesis. These cartilage canals may serve nutritive roles, providing blood to the cartilage, or as part of the mechanism that regulates the levels of calcium in the serum and tissues, though the function is not clearly understood (Hoenig and Walsh, 1982). Hoenig and Walsh (1982) did, however, rule out any bone—forming role of these canals. In some shark centra, a notochordal canal remnant is still present at the very center of the centrum. This canal remnant is typically very small in carcharhiniform sharks, and is more often completely closed over with calcified cartilage and reduced to intervertebral lenses (Cappetta, 1987). It is most often present in less derived neoselachians that have less calcified skeletons (Cappetta, 1987). 12 Figure 4. Variation in presence and distribution of cartilage canals (pores) in carcharhiniform centra A. Pores absent (CAS 25825.13, Triakis, dorsal view), B. Scattered (G. Hubbell Collection Sphyrna mokarran, dorsal view), C. Dense at rim (G. Hubbell Collection Prionace glauca #1, lateral view), D. Encircling (AMNH 99058.52, Sphyma, dorsal view). Scale bars = 10 mm. Due to the secondary calcification of shark vertebral centra, their internal structure is complex and varied. In the past, the easiest method of studying these structures in a vertebra was to make a cross—section through the narrowest portion of the notochordal canal of a centrum and at right angles to the length of the vertebral column (e.g., Hasse, 1879, 1882). More recent advances in x—radiography provides a superior and non-destructive method of studying these same patterns in shark centra (e.g., Ridewood, 1921; Compagno, 1988; Gottfried, 1999). In most carcharhiniform sharks, hyaline cartilage fills the areas between the basidorsals and basiventrals, the latter of which may be displaced by calcifications from the periphery of the centrum or from the double cone. These sharks likewise have an area of poorly calcified cartilage that extends inwards to the double cone from the basidorsals and basiventrals. There are also four calcified areas, found between the each of the arch cartilages; one between the two basidorsals on the dorsal surface of the centrum, one between the two basiventrals on the ventral surface, and one on each side of the centrum found between a basidorsal and basiventral. These four calcified areas are known as intermedialia, and are found in nearly all carcharhiniform sharks as well as in a number of other shark clades. Intermedialia can vary in their expansion into the chondrified body of the centrum, but do broaden with age (Ridewood, 1921). Intermedialia grow radially and in a centrifugal manner, so that the newest calcifications in these wedges are on the most external surface (Ridewood, 1921). Carcharhiniform intermedialia vary greatly in their shape and size both ontogenetically and between taxa, but Compagno (1988) was able to classify them into 14 three basic types. The first, rudimentary intermedialia, form thickened calcified pads but are not expanded as wedges into the centrum. This type of intermedialium is found in some scyliorhinids and all proscylliids. The second type, hollow intermedialia, are strong wedges of calcification that extend into the centrum, but have hyaline cartilage cores. These hollow intermedialia are found in the scyliorhinids Atelomycterus and Aulohalaelurus. The third type, solid intermedialia, likewise forms strong wedges or trapezoids that extend varying distances into the centrum (Figure lb). The solid intermedialia are the most important elements in the “Maltese cross” or “carcharhinoid” vertebral calcification type of Applegate (1967) (see “Nomenclature” discussion below). This Maltese cross pattern is present in all members of the Triakidae, Hemigaleidae, Carcharhinidae, and Sphyrnidae and Leptocharias (Compagno, 1988). The rudimentary intermedialia are present in many members of Scyliorhinidae and in all members of Proscylliidae. Some scyliorhinids have short, wedge—shaped intermedialia that extend only a part of the way into the vertebral body, and are transitional between the rudimentary and solid types. In addition to the calcified intermedialia, carcharhiniform shark centra also have diagonal calcifications. These calcification are a set of four extensions of the calcified double cone into the basal cartilages (Compagno, 1988). These are absent in many scyliorhinids, some proscylliids, a few triakids and carcharhinids, and in Pseudotriakis, but are otherwise widespread in the order. In some scyliorhinids and proscylliids, these take the form of thick, rounded diagonal calcified knobs. In most carcharhiniform sharks, these diagonal calcifications are thin plates, termed diagonal calcified lamellae (Ridewood, 1921) (Figure 1b). 15 Massare and Sharkey (2003) studied the effect drying had on the morphology of vertebral centra. When comparing a radiograph of a fresh skeleton with the dried skeleton of Carcharhinus Iimbatus (Black—tip sharks), they found that the overall morphology was essentially the same. Some shrinkage did occur in the dried skeleton, but no other distortion in centrum proportions was observed. Drying the skeleton actually allowed regional variations to become more apparent. Age Determination using Centrum Growth Rings Ridewood (1921) noted the presence of concentric growth rings in shark vertebral centra, which were later predicted to have a potential use as age indicators (e.g., Haskell, 1949; Urist 1961; Applegate, 1967). These growth rings have subsequently been used in successful shark age determination (e.g., Tanaka and Mizue, 1979; Thorson and Lacy, 1982; Parsons, 1985; Branstetter and McEachran, 1986; Officer et al, 1996; Wintner, 2000). The nature of the systemtic physiological disturbances that forms the grth rings is not well understood, however (Clement, 1992). A growth ring is usually defined as a pair of bands, an opaque, calcified band and a translucent, less—calcified band (e.g., Cailliet et al., 1985; Wintner, 2000). The annual deposition of a band pair has been verified (Smith, 1984; Parsons, 1993), with the opaque bands deposited in the summer, and the translucent bands in the winter (Cailliet et al., 1985). Brown and Gruber (1988), however, demonstrate that band pairs are not necessarily annual in all genera, and the need exists to determine the timing of band pair forrmtion for each species. 16 Using grth rings for shark age determination has other difficulties. Different increment readers may obtain significantly different counts for the same vertebrae (Officer et al., 1996). Fortunately, many different techniques exist for enlumcing the growth rings for the increment readers, such as alcohol or oil immersion, xylene impregnation, alizarin red staining, x—radiography, and x—ray spectroscopy (Cailliet et al., 1985). No significant variation in growth counts is apparent between at least some of the different techniques (e.g., using alizarin red staining versus x—radiography) (Officer of al., 1996). The variation in counts also decreases with more experienced increment readers (Officer et al., 1996). Officer et al. (1996) discovered significant variation of increment counts from different regions of an individual vertebral colurrm. Counts fiom centra within a single region ofthe column, however, were not significantly different. Natanson and Cailliet (1990) concluded that in Pacific angel sharks, these regional increment count differences were due to differences in vertebral development. Officer et al. (1996) found that it was not possible to determine whether the regional count variations in the triakids they studied were due to vertebral development or due to the methodology used to display the increments. Because sharks lack the scales, otoliths, and bones used in age determination in bony fish, growth ring counts in vertebral centra remains the best widely applicable method for age determination in sharks. Continued validation of the timing of band development and evaluation of the methods employed are necessary to improve the results, however. 17 Regions of the Verteme Column Krefit (1968) and Compagno (1970) have divided the shark vertebral column into three regions. The first region is made up of monospondylous precaudal centra (MP). These include those vertebrae fi'om the occipital centrum to the transition between monospondylous and diplospondylous centra This transition is typically at or just posterior to the pelvic girdle in carcharhiniform sharks (Compagno, 1988). Monospondylous centra gradually increase in length posteriorly in the vertebral column, and are typically longest just anterior to the MP—DP transition (Springer and Garrick, 1964). The first several centra typically have extremely wide ventral foramina, that progressively narrows to the width found in the remaining centra. The second region includes the diplospondylous precaudal centra (DP), beginning at the MP—DP transition to the caudal fin. The transition between the MP—DP zones in many carcharhiniform sharks is marked by an abrupt decrease in centrum length, often with a centrum of intermediate length separating the two zones (Compagno, 1988). Finally, the diplospondylous caudal centra (DC) are the centra of the caudal fin. These vertebrae are readily recognizable by their expanded haemal arches for support of the hypural lobe of the caudal fin (Compagno, 1988). With the exception of the MP—DP transition, centrum proportions gradually change fi'om one centrum to next with the changes cumulative. The formation of diplospondylous vertebrae occurs during embryonic growth of the shark, when vertebral basal cartilages and centra in the postpelvic tail double to produce two vertebrae for every myomere. The centra and basal cartilages transform by elongating, dividing, and reforming into two vertebrae, with new interdorsal cartilage between the basidorsals (Ridewood, 1899b; Secerov, 1911; Goodrich, 1930; and 18 Compagno, 1988). The diplospondylous centra are short, typically three—quarters to one— half the length of the longest monospondylous centra. In larger carcharhiniform slmrks with numerous short centra, such as carcharhinids and sphyrnids, the diplospondylous centra are nearly as long as the monospondylous centra (Compagno, 1988). The multiplication of the vertebrae in the caudal region may increase the flexibility of the tail, usefirl for the caudal propulsion exercised by sharks (Regan, 1906; Daniel, 1934; Goodrich, 1930; Cappetta, 1987). In a number of carcharhiniform species, the diplospondylous region begins with a transition zone of alternating long and short centra, also known as the stutter zone (Compagno and Springer, 1971; Compagno, 1973a, b). This transition zone may extend well into the caudal fin (Compagno, 1988). Isolated long centra may also be found in the diplospondylous precaudal region. The posterior—most region of the vertebral column becomes highly specialized. The basal cartilages become irregular and replaced by fiised blocks of cartilage that form a continuous rod enveloping the posterior extremity of the notochord (Goodrich, 1930). Vertebral counts and ratios of the different regions have been identified for most extant carcharhiniform sharks (e.g. Springer and Garrick, 1964), but do not help when studying isolated fossil centra. While counts can be variable among the different species, a definite increase in total vertebral counts occurs in the succession from Scyliorhinidae to Proscylliidae to Triakidae to Hemigaleidae and finally to Carcharhinidae and Sphyrnidae (Compagno, 1988), suggesting an increase in vertebral count is a derived feature. Compagno (1988) identified the mean total of vertebrae for each of these families: scyliorhinids 128.4, proscylliids 139.4, triakids 145.8, hemigaleids 154.6, 19 carcharhinids 178.0, and sphyrnids 175.7. Among carcharhiniform sharks, Leptocharias and Pseudotriakis have very high vertebral counts and do not fit the above succession, having 198—223 and 180—186 vertebrae respectively. Composition and Function Elasmobranchs typically have two types of cartilage in their vertebral centra: uncalcified hyaline cartilage, which does not have a high preservation potential, and calcified cartilage, which is much sturdier. Hyaline cartilage is found in the inner zone of the centra, and also initially in the outer zone, prior to its secondary calcification. This non—calcified cartilage also extends into the neural and haemal arches. The calcified cartilage that makes up the centra in elasmobranchs is formed as a result of the “impregnation of the intercellular material with a complex mixture of calcium phosphate and carbonates. . .” (Budker, 1971). Urist (1961) and Applegate (1967) examined these calcified deposits using x—ray diffraction and found the resulting pattern was essentially identical to the apatite found in bone in other vertebrates, though Clement (1992) suggests all tissues containing mineralized collagen have essentially indistinguishable x— ray diffiaction patterns. No osteocytes have been found in the calcified cartilage, however, and elasmobranchs cannot definitively be said to have true bone in their centra (Clement, 1992). The only skeletal material that appears to be true bone is found at the bases of teeth and dermal denticles (e.g., Moss, 1970). The presence or absence of bone in shark cartilage is a subject of continuing debate. The similarity of vertebral cartilage to bone in other vertebrates increases the potential for elasmobranch centra preservation, 20 especially when compared to the non—calcified portions of their skeleton (Kozuch and Fitzgerald, 1989). Ridewood (1921) hypothesized that the development of calcified cartilage in the vertebrae of sharks was a response to the stresses associated with locomotion in water. Strengthening of the cartilage skeleton was necessary in certain areas to withstand these rigors while still retaining motility. In addition to being related to activity, calcification is also related to habitat, as discussed above with the deepwater forms, and size (Budker, 1971; Compagno, 1988). Larger sharks would need the calcification due to the strong swimming muscles of their trunk, working in association with the vertebral column (Budker, 1971). A related area for further study would be the mechanical and firnctional differences between sharks with radiating areas of calcification (e.g., Lamniformes and Carcharhiniformes) versus those with concentric bands (e.g., Squatiniformes). Nomenclature Hasse (1879—1885) attempted to classify the Neoselachii into three groups using the patterns seen in the cross—section of their vertebrae. The first group were those with cyclospondylous vertebrae, which are simple cylinders with a ring of calcified cartilage around the notochordal sheath. This group includes the squalids and hexanchids. Tectospondylous vertebrae have calcified cylinders deposited concentrically around the double cone, found in the squatinids. Asterospondylous vertebrae, found in the galeomorph sharks, have strong secondary calcifications formed either as solid intermedialia (the Maltese cross of carcharhiniform sharks) or radii (thin branching plates found in larnniform sharks). Woodward (1889), Regan (1906), Goodrich (1930), 21 Ridewood (1921), Applegate (1967), and Compagno (1988) have discussed the problems associated with the use of Hasse’s terms. Ridewood (1921) felt that though these patterns were “. . . very striking and may prove to be of some taxonomic value, Hasse did not succeed in expressing and coordinating the facts in a scheme that adequately meets the requirements of the case.” (p. 337—338). Ridewood also found that the terms were used inconsistently in the literature. Applegate (1967) suggested that classification based on these vertebral types lumps unrelated groups together. Applegate (1967) saw some value in using vertebral centra for classification, and tried to remedy the problems of Hasse’s system with his own terminology, which reflected the genera in which the different morphologies occurred. He proposed eight morphotypes that could in turn be further subdivided, three of which characterized carcharhiniform sharks. The first of these are the pristiuroid vertebrae found in some members of the Scyliorhinidae (e.g., Pristiurus and Cephaloscyllium). A strongly calcified middle zone with a surrounding weakly calcified outer layer characterizes these vertebrae. The second type, atelomycteroid vertebrae, is also found in some members of the Scyliorhinidae (e. g., Atelomycterus), and Applegate envisions these as transitional between the pristiuroid type and those found in the Carcharhinidae. In this type, both the middle and outer zones are calcified with strongly developed radii into the basalia areas and peripheral calcification of the intermedialia. The final type is the carcharhinoid type, found in the families Triakidae, Carcharhinidae, and Sphyrnidae. Applegate envisioned this vertebral type as the culmination of a trend started in the atelomycteroid type. In the carcharhinoid vertebrae the notochordal sheath may vary in size. The four radii in the basalia are strongly developed, and the intermedialia are well calcified. 22 Hasse (1879) first suggested that there might be an identifiable trend in the development of secondary calcification patterns in the carcharhiniform sharks. Hasse and others (e.g., White, 1937; Nakaya, 1975) assumed that the cyclospondylic (or Applegate’s pristiuroid) centra are primitive in carcharhiniform sharks, and the Maltese cross (or carcharhinoid) centra are the cuhnination of the trend. Nakaya (1975) especially envisioned a distinct trend starting with primitive centra with only thin, rudimentary intermedialia and no diagonal calcifications. As this trend continues, diagonal knobs and intermedialia form and eventually develop into large solid intermedialia and diagonal lamellae (Figure 5). Dissenters (e.g., Regan, 1906; Ridewood, 1921) suggested that some cyclospondylic centra, such as in Galeus, may have secondarily developed from more heavily calcified centra similar to the type seen in Atelomycterus. Applegate (1967) suggested that cyclospondyly in scyliorhinids and squaloids evolved in parallel through reduction in a deepwater habitat. Weak rudimentary intermedialia exist today in sharks more specialized for deepwater environments, while shallow—water forms typically have strong rudimentary, solid, or hollow intermedialia (Compagno, 1988). Compagno (1988) used neither Hasse nor Applegate’s system of vertebral classification, and instead discussed only the different types of elements of vertebral centra. Compagno (1988) suggests that the primitive carcharhiniform calcification pattern might consist of low but strong solid or thick rudimentary intermedialia and diagonal calcification (either lamellae or knobs), and that extreme cyclospondylic centra are not necessarily primitive. Compagno (1988) also notes that intermedialia size can vary and sometimes overlap between families, such as between some triakids and carcharhinids. Diagonal calcified 23 0 IE] I] Figure 5. Trend in cross sectional calcification pattern of carcharhiniform vertebral centra according to Nakaya, (1975). A. Cephaloscyllium, Scyliorhinus; B. Apristurus, Galeus; C. Halaelurus, Proscyllium; D. Mustelus, Triakis; E. Scoliodon, Rhizoprionodon; F. Pterolamiops. Modified from Nakaya (1975). 24 lamellae likewise are quite variable in size in these two families, from long and well developed to completely absent. MATERIALS AND METHODS Radiographs Radiographs were prepared at the MSU Department of Radiology in a triple phase general radiographic room produced by General Electric, using DuPont Quanta Detail screens, with exposures at 52—58 kilivolts and 2.0—10.0 milliamperes. Isolated fossil and Recent vertebrae were radiographed in articular view. Photographs Photographs of vertebral centra were taken on an Olympus C3000 digital camera, and manipulated in Adobe Illustrator 9.0 and Adobe Photoshop 5.0. Centrum measurements Eleven measurements were taken on each fossil and Recent centrum. Centra were measured to a hundredth of a millimeter using digital calipers. The measurements include (Figure 6): 1. Diameter 1: Diameter of centrum measured from lateral side to side at the rim. Because anterior and posterior are impossible to determine in isolated centra, the larger of the two rim measurements was recorded for centrum width 1. 2. Diameter 2: Diameter of centrum measured from lateral side to side at the rim. The smaller of the two diameter measurements was recorded for centrum diameter 2. 25 4——— 1 ——> _ l I / xi : f g): i T I ' i l i? l 5 3 is ,. 7 .‘___ 2 > Figure 6. Carcharhiniform cenrum measurements. A. Dorsal view; 1) Diameter 1, 2) Diameter 2, 3) Length, 4) Width at apices of double cone, 5) Dorsal foramen length, 6) Dorsal foramen width, 7) Dorsal interforaminal width. Note: measurements 5-7 repeated for ventral foramina. B. Lateral view; 8) Height. After Kozuch and Fitzgerald (1989). 26 3. Centrum length: The length of the centrum measured from anterior to posterior rim. 4. Centrum width at apices of the double cone: Width of the centrum at the apices of the double cone. Centrum walls at this location are concave to convex in varying degrees. 5. Centrum height: Height of centrum measured at apices of the double cone. 6. Dorsal foramen length: Length measured from anterior to posterior margin of a single dorsal foramen. Because fossil centra are sometimes imperfectly preserved, the length was measured on whichever of the two foramina that was preserved best. 7. Dorsal foramen width: Maximum width of the same dorsal foramen that was measured for length. 8. Dorsal interforaminal width: Width of the wall separating the two dorsal foramina. 9. Ventral foramen length: Length measured from anterior to posterior margin of a single ventral foramen. Because fossil centra are sometimes imperfectly preserved, the length was measured on whichever of the two foramina that was preserved best. 10. Ventral foramen width: Maximum width of the same ventral foramen that was measured for length. 11. Ventral interforaminal width: Width of the wall separating the two ventral foramina. These measurements were recorded into Microsoft Excel, and later transferred to SYSTAT for analysis (see Discriminant Analysis). 27 Specimens available for study Fossil and Recent specimens were studied at or borrowed from eleven institutions and individuals. The list of institutions and their abbreviations are listed below. Lists of examined specimens are found in the Systematic Description and Appendix A. Specimen cataloging system Institutional catalog numbers are used for specimens when available. When multiple centra are assigned a single catalog number, the number is expanded to include decimal numbers to identify individual centra. For example, five associated Negaprion specimens with the catalog number UF 3245 become UF 3245.1, 3245.2.. .3245.5. Unofficial numbers are assigned to fossil specimens that are uncataloged, as follows: AN SP GM 1—8: Academy of Natural Sciences specimens collected from Gardinier Mine, Bone Valley, Florida. MSP AF 1—18: Academy of Natural Sciences specimens collected fi'om Agrico Fort Green Mine, Bone Valley, Florida. AN SP LCM 1-104: Academy of Natural Sciences specimens collected from Lee Creek Mine, Aurora, North Carolina. AN SP BVA 1—5: Academy of Natural Sciences specimens collected from Bethany, Virginia. 28 Recent specimens without catalog numbers are distinguished by the identification of the shark and the individual who loaned or donated the specimens, and then individually numbered, as follows: Gordon Hubbell, private collector in Gainesville, Florida. For example: G. Hubbell Collection, Carcharhinus perezi 1—3. Lisa Whitenack, student at University of Southern Florida. For example: L. Whitenack donation, Sphyrna mokarran, 1—8. Abbreviations of Institutions AMNH: American Museum of Natural History, New York, New York. AN SP: Academy of Natural Sciences, Philadelphia, Pennsylvania. BIVINH: The Natural History Museum, London, England. CAS: California Academy of Sciences, San Francisco, California. CWV: Calvert Marine Museum, Solomons, Maryland. SDNHM: San Diego Natural History Museum, San Diego, California UCMP: University of California Museum of Paleontology, Berkeley, California. UF: Florida Museum of Natural History, Gainesville, Florida. USNM: Smithsonian Institution, National Museum of Natural History, Washington, DC G. Hubbell Collection: private collection of Gordon Hubbell, Gainesville, Florida L. Whitenack Collection: donation from Lisa Whitenack, University of Southem Florida, Tampa, Florida. 29 SYSTEMATIC DESCRIPTION The following section is formatted into “systematic paleontology” style, but it combines information for fossil and Recent centra of a single genus under one heading. 'This information was combined to make the comparisons more manageable. In addition, a description of Hemipristis based solely on fossil specimens and descriptions of two additional morphotypes of unknown identification are included. Definition and description of the morphological terms employed can be found in the Background on Shark Centra and Materials and Methods. Class CHONDRICTHYES Huxley, 1880 Subclass ELASMOBRANCHII Bonaparte, 1838 Order CARCHARHINIFORMES Compagno, 1973c (Carcharhinida White, 1936) TRIAKIDAE: Gray, 1851 T riakr's Miiller and Henle 1838A Figure 7 Triakis Muller and Henle 1838A, p. 36; 1838c, p. 84 (no species mentioned); 1839, p. 63. Description of modern Triakr's centra Recent Referred Specimens. — CAS 25825 — T. semifasciata, entire skeleton; UCMP 136058 — T. semifasciata, entire skeleton. Centrum Proportions. — Centra of Triakis include both cylinders and fluted cylinders, depending on location within the vertebral column. The largest centra, those 30 Figure 7. Recent and fossil centra of Diakz‘s. A-C: Recent centra A. CAS 25825. 13, dorsal view, B. CAS 25825.21, dorsal view, C. CAS 25825.13 (left), ventral view, and CAS 25825.37 (right), dorsal view. D-F: Fossil centra D. SDNHM 71142.05, dorsal view, E. UCMP 148061, dorsal view, F. UCMP 148078, dorsal view. G-I: X-radiographs. G. CAS 25825.14, articular view, H. CAS 25825.36, articular view, I. UCMP 148059, articular view. Scale bars = 5 mm. 31 found in the mid-trunk region before the MP—DP transition, are of the fluted variety. The length of these centra is greater than the width, which is partly responsible for the fluted cylinder classification. T riakis has proportionately the longest centra of any carcharhiniform observed. These fluted cylinder centra have strongly concave lateral walls at the apices of the double cone. The smaller centra are cylinders, and are found either in the anterior—most region or in the posterior diplospondylous regions. The cylinders have greater width than length. These centra likewise display strongly concave lateral walls at the apices of the double cone, though to a lesser degree than the fluted cylinders. This concavity decreases with decreasing centrum length. All centra lack any recurve at the rims. The medic—lateral breadth is greater than the dorso—ventral height. Foraminal Proportions. —— The dorsal foramina are typically straight, elongate ovals, while the ventral foramina are straight rectangles. All foramina are wide, dominating the either the entire dorsal or ventral surfaces of the centra. As is normal for Carcharhiniform centra, the ventral foramina on anterior centra are especially wide. The dorsal and ventral interforaminal walls are typically narrower than the width of a single foramen, though the anterior centra have ventral interforaminal walls wider than the Width of a single, ventral foramen. With a few exceptions, the foramina do not extend into the rims of the centra. Pore Characteristics. — No pores are present on these centra. Notochord Canal Characteristics. — The canal for the notochord at the apices of the double cone varies fi'om centrum to centrum, but is mostly open and large. Calcification Pattern. ———- The radiographs of CAS 25 825 (Triakis semifasciata) reveal four strongly calcified intermedialia, extending nearly completely to the calcified 32 double cone. The intermedialia have the typical carcharhiniform wedge shape, coming to a point at the center of the centrum. The angle of this interior point in the lateral intermedialia makes approximately a right angle, or a slightly obtuse angle. The external surfaces of the lateral intermedialia are slightly convex, though straighter than the intermedialia of other genera, especially those with little concavity of the lateral walls at the apices of the double cone. The dorsal intermedialia (the intermedialia between the two dorsal foramina) and ventral intermedialia (the internwdialia between the two ventral foramina) are very narrow, and the outside surface is strongly concave. The foraminal areas are long and straight in these centra, narrowest next to the calcified double cone and increasing in width to the outside surface of the centra. The ventral foraminal areas are Wider than dorsal in these centra and are spaced further apart. The calcified double cone is clear in centra on these radiographs, but is delicate and small. It is, however, comparatively larger than the calcified double cone in carcharhinids and sphyrnids. Extending from the calcified double cone are four diagonal calcified lamellae. These have an unusual form in T riakis. Instead of four narrow lamellae of approximately equal length, T riakis has dorsal and ventral lamellae of dramatically unequal length. The two dorsal lamellae are very short, extending less than a quarter of the length of the foramina, and are very narrow. The two ventral lamellae are much longer, extending almost half the distance of the foramina. These lamellae are thicker at the base than the dorsal lamellae, and end with an extreme increase in thickness. In cross—sectional view, these almost appear as knobs on the end of the thinner base. 33 Description of fossil Triakr’s centra Referred specimens. — SDNHM 61933.12, 61933.13, 71142.05, 71142.10, 71142.20; UCMP 148039, 148058, 148059, 148061, 148066, and 148078. Age and distribution. —— Paleocene to Recent in the West Indies, Europe; Recent in East South Atlantic, Eastern Pacific, Western North Pacific, South Pacific, and Western Indian Oceans (Compagno, 1984, 1988; Cappetta, 1987). Discussion. — F ossilized Triakis centra were rare among the specimens examined for this study. All fossil Triakis centra were from California localities, which is consistent with their modern and fossil North American distribution. The eleven fossil Triakis centra were very similar to the Recent comparative material. All were long, fluted cylinders with strongly concave lateral walls. Cylinder—shaped Triakis centra were not identified among the fossil centra on loan to the MSU Museum, though identification of these cylinder—shaped centra to the generic level can be difficult due to a lack of diagnostic characters. Foraminal aspects of the fossil centra were likewise similar to Recent centra. The foramina were wide, typically long rectangles or ovals, with narrow interforaminal walls. In a few specimens, a single foramen extended completely into the rim, though this was not the normal case. The notochordal canal was visibly open only in SDNHM 71142.20. The canal was obscured with sediment in all other Triakis specimens. The medic—lateral breadth was larger than the dorso—ventral height in all specimens except UCMP 148061, which was round in articular view. UCMP 148043 was crushed during preservation, though it possesses a fluted cylinder shape and triakid—like foramina. UCMP 148059 and 148061 were the only fossil Triakis centra with x— radiograph images available for study. The radiographs of these specimens are very 34 similar to the x—radiograph of Recent specimen CAS 25825. The lateral intermedialia have interior angles that are slightly obtuse, and have slightly convex external surfaces. The centra of CAS 25 825 have relatively straight external surfaces compared to other triakid centra, but those of UCMP 148061 are straighter. The dorsal and ventral intermedialia are also very narrow with concave external surfaces. The foraminal dimensions of the specimens are also similar. The calcified double cone of UCMP 148061 is visible, but is small and delicate. 1? our short and slender diagonal lamellae extend fi'om this double cone. Unlike CAS 25825, these lamellae are all of equal length, and extend roughly a quarter of the length of the foramina. UCMP 148061 lacks any evidence of thickened lamellae in the ventral foramina UCMP 148059, however, has four robust diagonal lamellae of approximately equal length, extending approximately one—quarter to one—third of the distance to the Surface of the centrum. Mustelus Linck, 1790 Figure 8 Mustelus Linck, 1730, p. 31. Description of modern Mustelus centra Material Examined. — CAS 53006 - M. califomicus entire skeleton; G. Hubbell Collection — M. canis, 3 centra; UCMP 136060 - M. califomicus or lunulatus, entire skeleton; UCMP 128660 — M. henIei, entire skeleton. 35 Figure 8. Recent and fossil centra of Musrelus. A-C: Recent centra. A. CAS 53006.26, dorsal view, B. CAS 53006.28, dorsal View, C. G. Hubbell Collection Mustelus canis, dorsal view. D-F: Fossil centra. D. SDNHM 71142.02, dorsal view, E. UCMP 148041, dorsal view, F. UCMP 148053, dorsal view. G—H: X-radiographs. G. G. Hubbell Collection Mustelus canis 1, articular view, H. G. Hubbell Collection Mustelus canr's 2, articular view. Scale bars = 5 mm. 36 Centrum Proportions. — Centra of Mustelus include both cylinders and fluted cylinders, depending on location within the vertebral column. The largest centra, those found in the mid—trunk region before the MP—DP transition, are fluted cylinders. The length of these centra is greater than the width, partly responsible for the fluted cylinder classification. These fluted cylinder centra also have strongly concave lateral walls at the apices of the double cone, with a tendency to recurve towards their prominent rims. The smaller centra are cylinders, and are found either in the anterior—most region or in the posterior diplospondylous regions. These cylinders have greater width than length. These centra likewise display strongly concave lateral walls at the apices of the double cone, though to a lesser degree than the fluted cylinders and with less recurve at the rims. This concavity decreases with decreasing centrum length. The medic—lateral breadth is much greater than dorso—ventral height, giving these centra a very ovoid shape in articular view. F oraminal Proportions. — Dorsal and ventral foramina on centra of Mustelus are usually straight, elongate ovals. Ventral foramina on the large, fluted cylinder centra are typically slightly bowed medially. Foramina vary in width. Dorsal foramina are typically moderate in width, while ventral are wider. Both have an interforaminal wall that is wider than the width of a single dorsal foramen. Both the sets of foramina are comparatively narrower than in other triakids. The foramina are long, and extend fully into the rims. Pore Characteristics. — No pores are present on these centra. Notochord Canal Characteristics. — The size of the canal for the notochord at the apices of the double cone varies fi'om centrum to centrum, though is only rarely closed. 37 Calcification Pattern. — Radiographs of G. Hubbell Collection Mustelus centra reveal four strongly calcified intermedialia, as in T riakis. The interior angles of the two lateral intermedialia are either approximately right angles or slightly acute. The outside surface of these intermedialia are very convex and rounded, adding to the overall round appearance of the cross sectional view of the centra. The dorsal and ventral intermedialia are narrow as in Triakis, but much shorter. This shortness is related to the ovoid shape of the centra in articular view. The surfaces of the dorsal and ventral intermedialia are sharply concave, almost appearing gouged. The non—calcified areas previously housing the arch cartilage are not long and straight as is commonly seen in carcharhiniform centra. Instead, they are short and wide. These foraminal areas flare out rapidly until reaching a maximum width near the surface of the centrum and then recurve heavily, giving the foraminal area a bulbous shape. Calcified ridges, connected to the intermedialia, follow and define the foraminal areas past the surface of the centrum. The calcified double cone is small and delicate in this genus, as in Triakis, with four very short and thin diagonal lamellae extending fi'om its surface. These lamellae extend about a quarter to a third of the distance to the surface of the centra. Description of fossil Mustelus centra Referred specimens. — SDNHM 28495, 61933.05, 61933.07, 61933.14, 71142.01—71142.03, 71142.11, 71142.14, 71142.15, 71142.21, 71142.26; UCMP 148041, 148053, and 148056. 38 Age and distribution. — Lower Eocene through Recent in Europe and North America; Recent in all tropical and temperate seas (Compagno, 1984, 1988; Cappetta, 1987; Purdy et al., 2001). Discussion. — The fossil centra identified as Mustelus vary between the cylinder and fluted cylinder shapes, but are recognizable based on other characters. All of the fossil Mustelus centra have strongly concave lateral walls that recurve at the rims. The medic—lateral breadth is also usually greater than the dorso—ventral height, especially in the longer, fluted cylinders. The foramina are elongate oval, still tend to be narrower than in other triakid centra, and are often bowed medially in the fluted cylinders. The canal for the notochord remnant is open in the Recent specimens, though only open in three fossil Mustelus centra. The remaining centra either have a closed notochordal canal or are obstructed with sediment. No x—radiographs of fossil Mustelus centra were available for study. Galeorhinus Blainville, 1816 Figure 9 Galeorhinus Blainville, 1816, p. 121. Description of modern Galeorhinus centra Material Examined. — CAS 25822 - G. zyopterus, entire skeleton; G. Hubbell Collection - G. galeus, 1 centrum Centrum Proportions. — Galeorhinus centra are mostly cylinders, though once again the largest few may be classified as fluted cylinders. The fluted cylinders in 39 Figure 9. Recent and fossil centra of Galeorhinus. A-C: Recent centra A. CAS 25822.46, dorsal view, B. CAS 25822.41, dorsal view, C. G. Hubbell Collection Galeorhinus galeus, dorsal view. D-F: Fossil Centra D. SDNHM 61933.03, dorsal view, E. SDNHM 71142.09, dorsal view, F. UCMP 148048, dorsal view. G-I: X-radiographs. G. CAS 25822.1, articular view, H. CAS 25822.2, articular view, I. UCMP 148045, articular view. Scale bars = 5 mm. 40 Galeorhinus have concave walls at the apices of the double cones, but not to the same degree as in other triakids. The smaller cylinders likewise have concave walls, but in decreasing measure as the size decreases posteriorly. These centra lack any recurve at the rims. Galeorhinus centra are also shorter than other triakids. Length of the centra never exceeds width, giving them a stocky appearance. The proportions as seen fiom the articular view exhibit considerable variation. The Galeorhinus specimen from the G. Hubbell Collection has a medic-lateral breadth that is quite a bit larger than dorso— ventral height, as do many of the anterior centra from CAS 25822. A few centra in CAS 25822, however, have a dorso—ventral height that is larger than medic—lateral breadth. The bulk of these specimens are round in articular view. Foraminal Proportions. — The dorsal and ventral foramina are either rectangular (on the fluted centra) or square (on the cylinders). The squared—edges of all the foramina are due to the fact that they extend fully into the rims, so much so that the rims make the dorsal and ventral boundaries of the foramina. The foramina are very wide in Galeorhinus centra, more so than observed in any other triakid, and dominate the dorsal and ventral surfaces. The dorsal interforaminal walls are always narrower than the width of an adjacent dorsal foramen, sometimes less than 1 mm in width. These walls are so narrow in some that they are barely present. The ventral interforaminal walls are likewise narrower than the width of a ventral foramen, except in the anterior-most centra where the wall is quite wide, even exceeding the width of an adjacent foramen. Pore Characteristics. — No pores are present on these centra Notochord Canal Characteristics. —— The canal of the notochord remnant is Open in the available specimens, in most cases quite large compared to the size of the centrum. 41 Calcification Characteristics. — Centra fi'om CAS 25822 and G. Hubbell Collection G. galeus were radiographed for this study. The intermedialia of Galeorhinus are well—calcified, as in the previous two triakid genera, but are unique in that this calcification does not appear connect with the calcified double cone. Instead of coming to a sharp angle at the very center of the centrum, these intermedialia stop short, and have a very blunt or rounded medial surface. The outside surface of the two lateral intermedialia are only slightly convex, giving the centra laterally compressed appearances. The dorsal and ventral intermedialia are extremely narrow and rod—shaped. The outside surfaces of these intermedialia are slightly concave. The non—calcified foraminal areas are narrow near the center of the centrum and flare out towards the surface, where they are quite wide. Like other triakids, the calcified double cone is small and delicate, with four clear diagonal lamellae projecting from it. These lamellae are likewise narrow and delicate. The two lamellae projecting into the dorsal foraminal areas are the shorter pair, extending roughly a quarter of the distance to the surface of the centrum. The ventral lamellae are longer, extending two—thirds of the distance to the surface of the centrum. Description of fossil Galeorhinus centra Referred specimens. — ANSP 3331, 15415.07; SDNHM 61933.01—61933.04, 61933.06, 61933.07, 61933.09—61933.1 1, 61933.15, 71142.06, 71142.08, 71142.09, 71142.12, 71142.13, 71142.16, 71142.17, 71142.22—71142.25, 71142.28, 71142.29; UCMP 148020, 148021, 148028, 148029, 148031, 148033, 148045, 148047, 148048, 42 148050, 148054, 148062, 148064, 148065, 148069, 148070, 148072-148074, and 148080. Age and distribution. — Upper Cretaceous through Recent in Europe, North America, and North and West Afi'ica; Recent in all seas (Compagno, 1984, 1988; Cappetta, 1987; Purdy et al., 2001). Discussion. — Many more fossil centra have been identified as Galeorhinus than the other two triakid genera described in this study (Triakis, Mustelus). Centra of this genus are distinguished by their relatively short, stocky appearance and very wide foramina. It must be acknowledged, however, that the cylinder—shaped centra of T riakis and Mustelus may appear very similar to those of Galeorhinus, and possibly some centra that were identified as Galeorhinus may, in fact, be diplospondylous centra from one of these genera. In those cylinder—shaped centra, the wide foramina may be the only character useful for identification. The above centra all have cylinder to fluted cylinder—shaped centra, and appear relatively stocky compared to other triakid centra The foramina are also wide in these centra, though some are relatively narrower than those observed on Recent Galeorhinus centra, making identification less certain. In the two Recent specimens, the foramina were always rectangular or square and extended into the rims. In the some of the fossil centra identified as Galeorhinus, many of the foramina were oval, and often did not extend into the rims (e.g. ANSP 3331, UCMP 148021). Only seven of the forty—five fossil Galeorhinus centra had an open notochordal canal. The remaining centra were either completely closed or obscured by sediment. In the centra where the opening remained, the canal remnants were wide and obvious. 43 X—radiographs of UCMP 148028, 148031, and 148045 were available for study. The cross—sectional views of these centra are consistent with the Recent specimens, CAS 25822 and G. Hubbell Collection G. galeus. One of the major differences is the medial surface of the intermedialia. On the Recent specimens, the intermedialia have a rounded and blunt medial surface that does not extend to the calcified double cone. The medial surfaces of the intermedialia on the three fossil centra form a sharp corner that extends to the double cone. The dorsal and ventral intermedialia are somewhat wider than on the Recent specimens, but do have concave external surfaces. The non—calcified foraminal areas are wide, and flare towards the surfaces of the centra. The calcified double cone and four diagonal lamellae are visible on all fossil Galeorhinus specimens. These calcifications on UCMP 148028 very closely resemble those of CAS 25822, but UCMP 148031 and 148045 have much thicker diagonal lamellae. The lengths of the lamellae are comparable to CAS 25822. Summary of Triakidae Centra of these three genera have many similarities. All three include both cylinder and fluted cylinder shapes. The smaller centra are usually the cylinder shape, and are short to medium in length. These cylinders appear both in the anterior—most region of the column, and in the diplospondylous region. Their shorter lengths give them a more disk—like appearance, though all do have at least some constriction at the apices of the double cones. The largest, fluted cylinder centra are extremely long for their diameter, and have strongly concave lateral walls. Of the three, T riakis centra are relatively the longest and narrowest, while Galeorhinus have the thickest, most robust centra. Mustelus centra are the only to show any consistent recurve at the rims. The foramina vary depending on location of the centra within the column, but typically are very wide, dominating the dorsal and ventral surfaces of the centra. The ventral foramina are especially wide in the anterior—most region of the column, separated by a wide ventral interforaminal wall. F oraminal shape is not consistent in triakid centra. Elongate oval, rectangular, and square shaped foramina are found in centra of this family, sometimes multiple shapes in a single genus. Overall, the Mustelus foramina appear to be narrower for their length than those of Triakis. The ventral foramina in the fluted cylinders of Mustelus how so that the apices of the curves face each other medially, while those in Triakis are straight. Foramina on Triakis centra do not extend into the rims, while foramina on Mustelus and Galeorhinus centra do. Galeorhinus centra have extremely wide foramina compared to the other two genera. The cross—sectional patterns of secondary calcification in Recent triakid centra are distinct fi'om one another. The only arguable similarity, beyond those similarities common to all carcharhiniform sharks, is the amount of widening of the non—calcified foraminal areas towards the surface of the centra. The shapes of the intermedialia and diagonal lamellae are unique to each genus. Unlike carcharhinids or sphyrnids, triakid centra completely lack pores. This feature, along with the presence of fluted cylinders and wide foramina are the most reliable characters for distinguishing triakid centra from centra of other carcharhiniform sharks. Great care must be taken, however, when identifying isolated fossil triakid centra to the generic level. The fluted cylinder—shaped centra are generally distinct from one 45 another. These distinctions diminish in more posterior regions of the column, as the centra become relatively shorter. HEMIGALEIDAE: Hasse 1879 (18858) Hemipnstis Agassiz 1843B Figure 10 Hemipristis Agassiz, 1843B, p. 237. Referred specimens (fossil). — ANSP 7055—7061, LCM 3, 9, 18, 51, 91, 96; USNM 464, 288017, 288020, 288023, 288039, 288042, 288045, 290319, 467529, 467530; UCMP 148001, and 148008. Age and distribution. — Middle Eocene through Pleistocene in Europe, North and South America, North and West Africa, India, and Indonesia; Recent in Indian and western Pacific Oceans (Cappetta, 1987; Compagno, 1984). Description of fossil Hemipristis centra Centrum Proportions. — Hemipristis centra are very short, disk—like cylinders. They are, along with Prionace centra, consistently the shortest of carcharhiniform centra. The lateral walls are typically straight, though are sometimes slightly concave. All Hemipristis centra have wide, distinct rims. These centra are usually round in articular view, though the largest centra are often ovoid due to a larger medic—lateral breadth than dorso—ventral height. F oraminal Proportions. — Foramina are the most unique feature of Hemipristis centra. Each foramen is perfectly bisected by a thick diagonal lamella that extends 46 completely to the surface of the centrum. As a result, each foramen appears as two adjacent elongate oval openings. Each of these adjacent openings appears narrow by themselves, but the foramina, as a whole, tend to be quite wide. These foramina generally do not extend into the rims. Dorsal interforaminal walls are narrower than the foramina. Ventral foramina are usually narrower than the foramina, though on some of the larger centra these can be slightly wider than the foramina. Pore Characteristics. — All Hemipristis centra have medium to large pores. These pores typically encircle the rims and foramina closely, though a small percentage of the pores are scattered over the entire external surface. Notochordal Canal Characteristics. — The notochordal canal is filled with sediment in these centra. The only exception is USNM 288023, where the canal is very large and open. Calcification Pattern. — USNM 467529, AN SP LCM 3, and LCM 51 were radiographed for this study. The pattern of secondary calcification in Hemipristis centra is easily recognizable, for reasons mentioned above. The intermedialia are unremarkable. The lateral intermedialia have obtuse interior angles and very convex exterior surfaces. The narrower dorsal and ventral intermedialia have straight to slightly convex exterior surfaces. The non—calcified foraminal areas are quite wide, and are completely bisected by thick diagonal lamellae that extend completely to the surface. This is the only genus to consistently have lamellae that are so fully formed. The calcified double cone is not visible in these radiographs. Discussion. — Recent Hemipristis skeletal material was not available for comparative purposes. Fossil centra of this genus were recognized by Purdy et al. (2001) 47 Figure 10. Fossil centra of Hemipristis. A. ANSP 7060, dorsal view, B. UCMP 148001, dorsal view, C. USNM 288042, dorsal view, D. USNM 288045, dorsal view, E. X—radiograph of USNM 467529, articular view. Scale bars = 10 mm. 48 in association with fossil teeth at Lee Creek Mine, North Carolina, and were used for comparison. Hemipristis centra are difficult to confirse with centra from other genera, because of the distinct characteristics described above. While Carcharhinus centra are also short cylinders with straight walls and encircling pores, Hemipristis centra have thicker rims and very distinct foramina with diagonal lamellae. CARCHARHINIDAE: Jordan and Evermann 1896 Galeocerdo Miiller and Henle 1837 Figure 11 Galeocerdo Muller and Henle, 1837, p. 115. Description of modern Galeocerdo centra Material Examined. — CAS 65084 — G. cuvier, 3 centra; AMNH 99048 — G. cuvier, entire skeleton. Centrum Proportions. — All of the observed Galeocerdo centra have a distinct modified cylinder shape, that is, a cylinder with concave sides that recurve at the rims. Even the small caudal centra have this shape. The lateral walls are very concave, though never to the degree seen in triakids. The centra are medium in length, though the length is less than the width. When compared to length, Galeocerdo centra are comparatively longer than most Carcharhinus centra, and slightly shorter than Sphyrna centra. Galeocerdo centra have robust anterior and posterior rims. The centra are round from an articular view. 49 Figure 11. Recent and fossil centra of Galeocerdo. A. Recent centrum CAS 65084.1, dorsal view. B-D: Fossil centra B. USNM 494467, dorsal view, C. ANSP LCM 52, dorsal view, D. ANSP LCM 53, dorsal view. E-F: X-radiographs. E. CAS 65084.3, articular view, F. UCMP 148068, articular view. Scale bars = 10 mm. 50 F oraminal Proportions. — Dorsal and ventral foramina are mostly rectangular, approaching a square shape in the shorter centra of the anterior—most and caudal regions. Oval foramina were not observed. (Kozuch and Fitzgerald (1989) observed oval foramina exclusively in Galeocerdo, though this author disagrees with their classification based on the photographed specimen they included.) Foramina are medium width. The interforaminal walls are usually narrower than the foramina, though may be wider dorsally close to the MP—DP transition. The foramina never reach the rims. Pore Characteristics. —— Galeocerdo centra have many extremely large pores, the largest of any specimen studied. The pores encircle the rim and foramina in a regular pattern, with a few scattered on the lateral walls and interforaminal spaces. Notochord Canal Characteristics. — The canal for the notochord is completely sealed in all observed specimens. Calcification Characteristics. — Centra of CAS 65084 were radiographed for this study. The four intermedialia in Galeocerdo centra are fairly typical in their overall appearance. They all originate at the very center of the centrum next to the opening for the notochord. The two lateral intermedialia have an interior angle of approximately 90° with slightly convex outside surfaces. The dorsal and ventral intermedialia are thin wedge shapes with straight outer surfaces. The non—calcified foraminal areas are straight, and widen rapidly towards the surface. The ventral foraminal areas are wider and spaced firrther apart than the dorsal foraminal areas. The calcified double cone fi‘om which the diagonal lamellae extend is not visible in Galeocerdo centra, and the opening for the notochord is much smaller than in the centra of triakid genera. Two diagonal lamellae extend ventrally about one third of the 51 distance to the surface. These lamellae are thin at the base and widen slightly at the distal end. No dorsal lamellae are present. Description of fossil Galeocerdo centra Referred specimens. — AN SP LCM 52, 59, 60; USNM 494467; UCMP 148002, 148068, 148079; and BMNH 1309. Age and distribution. — Lower Eocene through Recent in Europe, North and South America, North, West, and South Africa, Celebes, India, and Japan; Recent in all warm—temperate and tropical seas (Compagno, 1984, 1988; Cappetta, 1987). Discussion. — Centra of Galeocerdo were rare among the fossils available for study. The eight fossil centra identified as Galeocerdo share the long, modified cylinder— shape and large, encircling pores found in Recent centra of the genus. The main dissimilarity is the presence of oval foramina found in UCMP 148079. The retraining fossil centra had rectangular foramina, as was observed in all Recent Galeocerdo centra. Kozuch and Fitzgerald (1989), however, observed oval foramina in Galeocerdo centra, suggesting that either rectangular or oval foramina may be present. In AN SP LCM 60, the pores are large, but very scarce. USNM 494467, on the other hand, has large pores in abundance, mostly encircling the rims and foramina, but also covering the lateral walls. This specimen is somewhat weathered, however, which may obscure the true nature of pore distribution. UCMP 148068 was the only fossil Galeocerdo centrum for which an x— radiograph was available. The cross—section of UCMP 148068 is similar to that of the Recent specimens, CAS 65084, in most ways, but with a few notable differences. The 52 interior angle of the lateral intermedialia is approximately 90° in the Recent specimen, but is clearly obtuse in the fossil centrum. The dorsal and ventral intermedialia, however, are quite similar. The calcified double cone is not visible in the Recent centrum, and has two slender diagonal lamellae extending into the ventral non—calcified areas. In the fossil centrum, the calcified double cone is visible and robust, as are four diagonal lamellae. The two ventral lamellae are approximately twice the length of the dorsal lamellae. It must be noted that these discrepancies might be better understood with additional x— radiographs of Galeocerdo centra, especially considering both specimens are relatively short and small, and are mostly likely diplospondylous caudal centra. Rhizoprionodon Whitley 1929 Figure 12 Rhizoprionodon Whitley, 1929, p. 354. Description of modern Rhizoprionodon centra Material Examined. — AMNH 22826 — R terraenovae, entire skeleton; G. Hubbell Collection — R. terraenovae 5 centra; L. Whitenack Collection — R. terraenovae, 10 centra. Centrum Proportions. — The centra of Rhizoprionodon are unlike those of any other carcharhinid. Their centra, in fact, more closely resemble triakid centra. The centra range from cylinder to fluted cylinder in overall shape and are longer tlmn typical carcharhinid vertebrae. The size tends to be small compared to other carcharhinids. The lateral walls of the centra are strongly concave. The centra have a medic—lateral breadth 53 Figure 12. Recent and fossil centra of Rhizoprionodon. A-C: Recent centra. A. G. Hubbell Collection Rhizoprionodon terraenovae l, dorsal view, B. G. Hubbell Collection Rhizoprionodon terraenovae 2, dorsal view, C. G. Hubbell Collection Rhizoprionodon terraenovae 5, dorsal view. D-F: Fossil centra. D. ANSPAF 19, dorsal View, E. UF 122234, dorsal View, F. UF 123154, dorsal view. G-H: X-radiographs. G. G. Hubbell Collection Rhizoprionodon terraenovae 2, articular view, B. G. Hubbell Collection Rhizoprionodon terraenovae 4, articular view. Scale bars = 5 mm. 54 larger than dorso—ventral height, sometimes extremely so. This gives the centra an ovoid appearance fi'om articular view. F oraminal Proportions. — Foramina on these centra are also unlike any other carcharhiniform centra observed. They are almost completely closed over by what appears to be calcified cartilage or remnants of the arch—cartilage. The foramina are all elongate ovals, and tend to be quite narrow. Due to this narrowness, the interforaminal walls are all wider than the foramina, sometimes greatly so. Wall width varies with region of column as is normal for carcharhiniform centra. Pore Characteristics. —— No pores were visible on any Rhizoprionodon centra, even with a dissecting microscope. This makes them the only carcharhinid centra observed without pores. Notochord Canal Characteristics. — The canal for the notochord was closed in every centrum of Rhizoprionodon. Calcification Characteristics. — Centra from G. Hubbell Collection R. terraenovae were radio graphed for this study. The cross—sectional pattern of secondary calcification in Rhizoprionodon is almost indistinguishable from the pattern seen in Mustelus. The most obvious difference is the shape of the lateral intermedialia. In Rhizoprionodon, the interior angle of these intermedialia are at right angles to obtuse angles, while they tend to be acute in Mustelus. The external surfaces of these intermedialia in Rhizoprionodon are very convex, giving a very round appearance to the centra in the radiograph. The dorsal and ventral intermedialia are quite similar to those of Mustelus, both are quite narrow and short. In Rhizoprionodon the surface of these two intermedialia are sharply concave. The non—calcified areas previously housing the arch 55 cartilage are not long and straight as is commonly seen in carcharhiniform centra, but are rather short and wide. These foraminal areas flare out rapidly until reaching a maximum width near the surface of the centrum and then recurve heavily, giving the foraminal area a bulbous shape. Rhizoprionodon does not have this shape to the same degree as Mustelus. Calcified ridges, connected to the intermedialia, follow and define the foraminal areas past the surface of the centrum. The calcified double cone is not visible in the radiographs of these centra, though four small diagonal lamellae do extend into the non—calcified foraminal areas. These lamellae are thin at the base and wider distally, and extend a quarter to a third of the distance into the foraminal areas. Description of fossil Rhizoprionodon centra Referred specimens. —— ANSP AF 19; UP 122858, 122234, and 123154. Age and distribution. —— Lower Eocene through Recent in Europe, North Africa, and North America; Recent in warm—temperate to tropical Atlantic and Indo—Pacific Oceans (Compagno, 1984, 1988; Cappetta, 1987; Purdy et al., 2001). Discussion. — These four fossil specimens are poorly preserved. Their identification as Rhizoprionodon, however, is certain. Key morphological features are still visible on these centra that are common to Rhizoprionodon. The centra are all long, fluted cylinders with strongly concave lateral walls. The medic—lateral breadth is much greater than the dorso—ventral height, giving the centra an ovoid shape in articular view. The fossil centra also display the same unusual foramina seen in the Recent centra. The foraminal surfaces are closed over with what appears to be arch—cartilage. On the fossil 56 specimens, this is observed as thin ridges where the foramina are normally located. No pores are visible on the UF specimens, but the poor quality of these fossils would preclude any possibility of observing pores. AN SP AF 19 has what appears to be a few small pores adjacent to the foramina, though preservational quality of this specimen is poor enough to make this uncertain. The presence of pores on this specimen would make it the only Rhizoprionodon specimen with this character. An x—radiograph was available for UF 123154, but no internal detail was visible. The only information that was visible was the ovoid cross—sectional shape, similar to the shape observed in the x—radiographs of the G. Hubbell Collection Rhizoprionodon centra. Prionace Cantor 1849 Figure 13 Prionace Cantor, 1849 p. 1399. Description of modern Prionace centra Material Examined. — G. Hubbell Collection — Prionace glauca, 5 vertebrae. Centrum Proportions. — Prionace centra are disk—like cylinders. For their diameters, they are the shortest carcharhiniform centra. Despite their shortness, they still have strongly concave lateral walls. Their rims are quite pronounced, giving them a very unique appearance compared to other carcharhinids. Prionace centra are round in articular view. These centra also seem more delicate than other carcharhiniform centra. The cartilage has a low density and chalky texture. The walls also have a somewhat fibrous appearance. These centra are from an adult animal that measured 9’6”, so lower 57 Figure 13. Recent and fossil centra ofPrionace. A-B: Recent centra. A. G. Hubbell Collection Prionace glauca l, dorsal view, B. G. Hubbell Collection Prionace glauca 1, lateral view. C-F: Fossil centra. C. CMMV 815, dorsal view, D. ANSP LCM 64, dorsal View, E. ANSP 308.2, dorsal view, F. ANSP 308.2, articular View. G-I: X-radiographs. G. G. Hubbell Collection Prionace glauca 4, articular view, H. G. Hubbell Collection Prionace glauca 5, articular view, I. ANSP 308.2, articular View. Scale bars = 10 mm. 58 density is not a result of the animal being juvenile with little calcification of the cartilage. These centra should be compared with centra from other P. glauca specimens. Foraminal Proportions. — In all of the centra examined, the foramina are short, wide rectangles. Their short length is due to the overall short length of the centra, but the width of the foramina is comparable to other carclmrhinids. The foramina extend firlly into the rims. The interforaminal walls are generally quite narrow, much narrower than the foramina. Pore Characteristics. — Pores are present and abundant on these centra. The pores are miniscule, making magnification necessary to see them clearly. The pores are found surrounding the foramina in fairly low density and in a single, regular line following the rims. This pattern is unique to Prionace centra. Notochord Canal Characteristics. —- The canal for the notochord is not entirely closed, but small enough to be barely noticeable. Like Negaprion, the Canal appears to have been more open recently, but is now closed over by cartilage. Calcification Characteristics. — Centra fi'om G. Hubbell Collection P. glauca were radiographed for this study. As in the external view of the Prionace centra, the cross-sectional calcification patterns are unique. The interior angles of the lateral intermedialia are always obtuse and extend completely to the opening for the notochordal remnant. The external surfaces of these intermedialia all have a unique shape. These surfaces have the normal convex, rounded appearance, but recurve strongly near the foramina, giving them a gentle “W” shape. The dorsal and ventral intermedialia are long, but vary quite a bit in width. The external surfaces on all of these intermedialia are very concave, sometimes drarmtically so, and to a degree more so than in any other observed 59 carcharhiniform shark centra. The four non—calcified foraminal areas are narrow at the base and flare towards the external surfaces. No calcified double cones are visible in Prionace centra, nor are there any diagonal calcified lamellae. Description of fossil Prionace centra Referred specimens. — CMMV 815; ANSP 3081—3084, and AN SP LCM 64. Age and distribution. — Pliocene of Italy; Recent in all tropical and temperate seas (Cappetta, 1987). Prionace is the widest ranging modern cartilaginous fish (Compagno, 1984, 1988). Discussion. — Six fossil centra from the eastern United States have been tentatively identified as Prionace. Of these six, AN SP 3081—3084 are associated and likely fiom the same individual. These six centra have only been tentatively identified, because Prionace centra are not otherwise known from Neogene deposits in North America. Cappetta (1987) lists the only known fossil occurrence for Prionace as the Pliocene of Italy. These six fossil centra closely resemble the Recent Prionace centra that were available for comparison. All were short, disk—like cylinders with strongly concave lateral walls, strong rims, and were round in articular view. The foramina were all short, wide rectangles that extended into the rims. The Recent Prionace centra were observed to be very delicate compared to other carcharhinid centra. They seemed to have a lower density cartilage, and an almost fibrous texture. All six fossil centra likewise seemed much more delicate, despite mineralization during fossilization. 60 The Recent centra have unique pore characteristics compared to other carcharhinid sharks. Their pores are miniscule and closely followed the rims in a single line. The pores on these six fossil centra are larger, and encircle the foramina and rims, though not in the regular, single line observed on the Recent centra. Of special note is the opening for the notochordal canal in ANSP 3081—3084. These openings are enormous, ranging from 3.4 mm to 6.5 mm in diameter. In other carcharhiniform centra, the openings for the notochordal remnants usually resemble tiny pinholes, and are not generally measurable. The reason for the unusually large openings in these four centra is not clear, though it does not appear to be the result of weathering or poor preservation. CMMV 815 and ANSP 308.1—308.4 had x—radiographs available for study. In overall appearance, the cross—sectional views of the fossil and Recent centra appear similar, but differ in a few characteristics. The Recent Prionace centra have lateral intermedialia with a convex outer surface that recurve strongly near the foramina that give them a gentle “W” shape. This feature is only visible on one lateral intermedialium on CMMV 815, and only to a small degree. This centrum is heavily worn, however, and the condition could have been present when the centrum was in better condition. AN SP 308.1—308.4 have only a hint of this shape on the lateral intermedialia. The strongly concave external surfaces of the dorsal and ventral intermedialia seen in the Recent centra are also apparent in ANSP 308.1—308.4, but not in CMMV 815, though the poorer condition of preservation could have obscured this feature. No calcified double cone or diagonal lamellae are visible on the G. Hubbell Collection Prionace specimens. AN SP 308.1—308.3 also lack any evidence for these 61 calcifications. AN SP 308.4, however, has two small diagonal lamellae extending into the dorsal non—calcified areas. CMMV 815 has four thick diagonal lamellae extending over half of the distance towards the surface of the centrum. Negaprion Whitley 1940 Figure 14 Negaprion Whitley, 1940, p. 11. Description of modern Negaprion centra Material Examined. — G. Hubbell Collection — N. brevirostris, 2 centra. Centrum Proportions. — Negaprion centra have a cylinder shape with concave lateral walls. In one specimen, this concavity is minor, with a small amount of recurvature at the rims. The second specimen has strongly concave lateral walls and a large amount of recurve. Negaprion centra are short, though longer than the average Carcharhinus centrum. Both centra are round in articular view. In overall appearance, they are very similar to centra of Carcharhinus. There may, indeed, be no unique character possessed by either that can distinguish them fiom each other. The sample size of two centra is too small, and more centra are needed to draw final conclusions. F oraminal Proportions. — Foramina are all rectangular with medium width and length that do not extend into the rims of the centra. Both the dorsal and ventral interforaminal spaces are wider than the foramina. Pore Characteristics. — These Negaprion centra have numerous pores closely following the rims and foramina. The pores are miniscule, almost too small to be seen 62 Figure 14. Recent and fossil centra of Negaprion. A. Recent centrum G. Hubbell Collection Negaprion brevirostris 1, dorsal view. B-C: Fossil centra B. UF 3245.2, dorsal view, C. UF 3245.5, dorsal view. D-F: X-radiographs. D. G. Hubbell Collection Negaprion brevirostris 1, articular view, E. UF 3245.2, articular view, F. UF 3245.5, articular view. Scale bars = 10 mm 63 without magnification. Kozuch and Fitzgerald (1989) record Negaprion pores as being scattered over the surface, which is clearly not the case in the G. Hubbell Collection specimens. As mentioned earlier, more centra are needed to increase the sample size. Notochord Canal Characteristics. — Both centra still have a small opening for the notochord, but they are miniscule, smaller even than the pores on the surfaces. This opening was clearly larger earlier in the life of the shark, but is now closed over by cartilage. Calcification Characteristics. — Centra from G. Hubbell Collection N. brevirostris were radiographed for this study. The pattern of secondary calcification in Negaprion centra is very similar to that of Carcharhinus. The lateral intermedialia have a very obtuse interior angle with moderately cdnvex outer surfaces. These intermedialia extend completely to the center of the centra where they connect with the openings for the notochordal remnant. The dorsal and ventral intermedialia are long, wide wedges with straight to slightly concave exterior surfaces. The non—calcified foraminal areas are narrow wedge shapes. The dorsal foraminal areas are slightly shorter than the ventral foraminal areas, and are straight. The longer ventral foraminal areas are slightly bowed laterally. The calcified double cone is not visible in the radiographs of these centra. F our diagonal lamellae extend from the center of the centra. These lamellae are comparatively wide and quite short, extending less than a quarter of the distance to the surface of the centrum. Description of fossil Negaprion centra Referred specimens. — UF 3245.1—3245.5. Age and distribution. -— Middle Eocene through Recent in Europe, North and South America, and West Africa; Recent in warm—temperate to tropical Atlantic and Indo—Pacific Oceans (Compagno, 1984, 1988; Cappetta, 1987). Discussion. — The five fossil Negaprion centra are associated with teeth identified as Negaprion brevirostris, and are likely from a single individual. UF 3245.4 and 3245.5 appear to be diplospondylous caudal centra. They are smaller than the other three fossil centra, and are very short. All five of the fossil centra are shorter than the two Recent Negaprion centra. The three larger fossil centra may be diplospondylous precaudal centra. The fossil and Recent centra closely resemble one another in most other ways. The fossil centra have slightly concave later walls and prominent rims, and are round in articular view. The foramina are rectangular with wide interforaminal walls. All the fossil centra have fine pores scattered on the lateral walls and surrounding the foramina. No opening for the notochordal remnant was observed on any of the fossil centra All five fossil Negaprion centra were radio graphed for this study. The cross— sectional view is very similar to that of G. Hubbell Collection Negaprion brevirostris. The intermedialia have the same obtuse interior angle and convex outer surfaces. The dorsal and ventral intermedialia likewise are long, wide wedges with slightly concave exterior surfaces. All four non—calcified were narrow and straight. The calcified double cone is not visible in the fossil specimens, though all have four thick diagonal lamellae. In the three fossil centra that appear to be diplospondylous precaudal, the lamellae extend 65 a quarter to a third of the distance to the external surface of the centra. In the diplospondylous caudal centra, the lamellae extend two—thirds to three—quarters of the distance to the surface, and are thicker than the lamellae in the precaudal centra. The unfortunately small sample size of Recent Negaprion centra greatly limits the identification of fossil centra of the same genus. UF 3245.1—3245.5 would most have most likely been identified as Carcharhinus centra had it not been for the associated Negaprion brevirostris teeth. Hundreds of fossil centra have been identified as Carcharhinus below, though a few Negaprion centra may be present. Carcharhinus is much more abundant and diverse in both modern waters and in the fossil record (e.g., almost 30 extant species of Carcharhinus versus 2 extant species of Negaprion. (Compagno, 1984)). Another difficulty with the comparison of the fossil and Recent Negaprion centra is that the specimens were fi'om different regions within the vertebral column. G. Hubbell Collection Negaprion brevirostris centra are monospondylous centra, while all of the fossil centra appear to be diplospondylous. Variations along the column were most certainly responsible for some of the differences seen between the two sets of specimens. Carcharhinus Blainville 1816 Figure 15 Carcharhinus Blainville, 1816, p. 121 . 66 Description of modern Carcharhinus centra Material Examined. — AMNH 218147 — C. leucas, entire skeleton; G. Hubbell Collection - C. leucas, 3 centra; G. Hubbell Collection — C. leucas, 5 centra; G. Hubbell Collection — C. leucas, 2 centra; AMNH 93846 — C. brevipinna, entire skeleton; AMNH 218150 — C. acronotus, entire skeleton; G. Hubbell Collection — C. altimus, 2 centra; G. Hubbell Collection — C. falciformis, 4 centra; G. Hubbell Collection — C. falciformis, 5 centra; G. Hubbell Collection — C. limbatus, 3 centra; G. Hubbell Collection — C. limbatus, 3 centra; L. Whitenack Collection - C. limbatus, 15 centra; L. Whitenack Collection — C. limbatus, 12 centra; G. Hubbell Collection — C. perezi, 4 centra; G. Hubbell Collection — C. perezi, 3 centra; G. Hubbell Collection — C. perezi, 4 centra; G. Hubbell Collection — C. plumbeus, 5 centra; G. Hubbell Collection — C. signatus, 4 centra. Centrum Proportions. — Specimen availability for this genus was better than for any other carcharhiniform shark. Carcharhinus is also the most diverse carcharhiniform genus. Their centra display some of this diversity, but the morphology of the centra for the different species is similar. All species observed have cylinder-shaped centra. C. brevipinna is unique among the observed specimens in having a few of the largest centra that could be classified as fluted cylinders. C. falciformis, C. altimus, and C. signatus have a few centra with a slight amount of recurve at the rims, approaching a modified cylinder shape. Overall, Carcharhinus centra are very short relative to their diameters, giving them a disk or bobbin—like appearance. Some of the largest centra, found just prior to the MP—DP transition, are longer and more similar in length to those of Sphyrna or Galeocerdo. Carcharhinus centra have less concave lateral walls than any other genus. 67 Figure 15. Recent and fossil centra of Carcharhinus. A-C: Recent centra A. G. Hubbell Collection Carcharhinus perezi 3, dorsal view, B. G. Hubbell Collection Carcharhinus leucas 2.1, dorsal view, C. G. Hubbell Collection Carcharhinus signatus 4, dorsal view. D-H: Fossil centra D. ANSP LCM 27, dorsal view, E. ANSP LCM 46, dorsal view, F. AN SP LCM 47, dorsal view. G-H. Associated specimens. G. USNM 24914.04, dorsal view, H. USNM 24914.12, dorsal view. I-K. X-radiographs. I. G. Hubbell Collection Carcharhinus leucas 2.1, articular view, J. G. Hubbell Collection Carcharhinus signatus 4, articular view, K. BMNH 4546.1, articular view. Scale bars = 10 mm. 68 While this concavity does exist in some species, it is minor, with the exception of C. brevipinna and a few centra of C. acronotus. Centra typically have straight or even convex lateral walls. Carcharhinus centra are usually round in articular view, though may vary between medic—lateral breadth greater than dorso—ventral height or dorso—ventral height greater than medic—lateral breadth F oraminal Proportions. — F oraminal shape is quite variable in Carcharhinus centra. Most foramina are oval or rectangular, though square foramina are found in some of the shortest caudal centra. F oramina are mdim in width, and never extend into the rims. Except on the anterior—most centra, the ventral interforaminal width is narrower than the width of the foramina. Dorsal interforaminal width, however, can be both narrower and wider than the width of the foramina in a single individual. Pore Characteristics. — Pores are always present on Carcharhinus centra, though in varying numbers. The pores are very small to medium in size, never as large as those found on Galeocerdo centra. The pores are commonly scattered over the entire external surface or encircle the rims and foramina. Notochord Canal Characteristics. — The passage for the notochord is rarely intact. When it is open, it is quite small and barely visible without magnification. The area typically looks to have been previously larger and closed over with cartilage. Calcification Characteristics. —- Centra from G. Hubbell’s C. signatus, C. perezi, and C. leucas were radiographed for this study. While centra of Carcharhinus do not have any characters of the calcification pattern that are unique, the overall pattern is identifiable. This pattern, unfortunately, does closely resemble the pattern seen in Negaprion centra. The four intermedialia are unremarkable. The lateral intermedialia 69 each have an obtuse interior angle, and a very rounded, convex exterior surface. The dorsal and ventral intermedialia vary in width, as in all carcharhiniform centra, and have a straight to gently concave exterior margin. The non—calcified foraminal areas are fairly narrow. The dorsal foraminal areas are typically shorter than the ventral, and are straight. The longer ventral foramina are either straight or curve laterally gently. The calcified double—cone is faintly visible in these radio graphs, and is quite narrow, especially considering the medium to large size of these centra. Extending from the double—cone are four delicate, short diagonal lamellae. These four lamellae are equal in length, and extend a fourth or fifth of the distance towards the exterior of the centra. The lamellae are thinnest at the base, increase slightly in width distally, and taper at the end. Description of fossil Carcharhinus centra Referred specimens. -— CMMV 1013, 1131, 1133, 1575, 1577, 1578, 1581, 1582, 165201—1652] 1, 1784, 1887; ANSP GM 1, 4, 5, ANSP 308.5, 6480, 7054, 8111, 8113, 8122, 8125, 8131, 8040, 14698, 15415.02—15415.06, 15415.08, 15433, 20148.1, ANSP AF 5, 11, 13, ANSP LCM 2, 4, 5, 7, 8, 12, 14, 21, 23, 26—30, 44—50, 53—58, 62, 63, 65- 67, 72—75, 79, 81—83, 85—88, 90, 92, 93, 94, 97, 104, ANSP BVA 1—5; USNM 24914.01—24914.13, 288037, 369890, 435324.01, 435324.02, 467532, 494463.01— 494463.11; SDNHM 25782, 61937.01—61937.03, 63154, 65993—65995, 71142.04; UP 92276, 92277, 92280—92282, 92291, 128906; UCMP 148003, 148005—148007, 148009— 148012, 148015, 148017, 148018, 148025—148027, 148032, 148035, 148044; BMNH 4645.1, 4645.2, 4645.4, 4645.6, 5571.1—5571.4, 35611a.1—35611a.8, and 13799. 70 Age and distribution. — Middle Eocene through Recent of Europe, North America, North and West Afiica, Australia, Japan, New Zealand, and India; Recent in all warm—temperate and tropical seas (Compagno, 1984, 1988; Cappetta, 1987). Discussion. — Carcharhinus is abundant and diverse today, as it was during much of the Cenozoic. Fossil centra fi'om Carcharhinus are much more abundant than any other genus among the material examined for this study. While some diversity in their morphology is expected, their centra are still readily identifiable (though Negaprion centra may easily be mistaken for Carcharhinus). All of the fossils examined are short, cylinder—shaped, and have slightly concave to convex lateral walls. Also similar to Recent Carcharhinus centra, the foramina on the fossil centra vary fi'om round to rectangular, though they never consistently extend into the rims. The remmnt for the notochordal canal was closed in every fossil centrum. Pores are always present (when fossil preservation is sufficiently good), either scattered over the entire external surface, or found encircling the rims and foramina. UCMP 148010 is the only observed exception. The pores on this centrum followed the rims in single, regular lines. This pattern is similar to that of Prionace centra, though these pores were much larger. UCMP 148010 did not resemble Prionace centra in any other characteristic. Over 50 fossil Carcharhinus centra were x—radiographed for this study. The cross—sectional patterns on these fossils are very similar to those of Recent Carcharhinus centra. One feature that is especially distinct on the fossil centra is the shape of the ventral non—calcified areas. At times, these non—calcified areas are straight and narrow, but in many, they curve laterally (e.g. CMMV 1013, 1652.02, 1652.07, ANSP 20148). 71 This characteristic was noted on Recent Carcharhinus centra, but not as often, nor to the same degree (with the exception of a single C. leucas centrum). The calcified double—cone is rarely visible in the radiographs. When it is visible, it is faint and delicate, similar to those of Recent Carcharhinus centra. Diagonal lamellae are always present and well developed. The lengths of these lamellae vary between extending one—quarter to three—quarters of the distance to the surface of the centrum. Like those observed on Recent Carcharhinus centra, the lamellae are thinnest at the base, increase slightly in width distally, and taper at the end. Summary of Carcharhinidae The members of Carcharhinidae are numerous and diverse, and this diversity is visible in the morphology of the vertebral centra. Of the five genera examined, only Carcharhinus and Negaprion centra closely resemble each other, and distinguishing the two will be difficult. Galeocerdo centra are among the most easily recognized of the carcharhinids. These centra have distinct modified cylinder shapes with large encircling pores. In radiographs, the presence of only two diagonal lamellae is a unique feature found in only Galeocerdo, though more samples may prove otherwise. Carcharhinus centra are very common as fossils, and are generally not difficult to recognize. The combination of their extremely short, disk—like cylinder shape, typically straight to slightly convex lateral walls, and scattered to encircling distribution of tiny pores is enough to identify them. Those centra that are longer are more difficult to distinguish, as they may be confused with Negaprion or perhaps Galeocerdo if the lateral 72 walls are strongly concave or display any recurve at the rims. Long centra of unknown identification will have to be examined on a case-by—case basis and compared to centra of known identification. The morphology of Rhizoprionodon centra is quite distinct from the morphologies of other carcharhinid centra. Their strongly concave lateral walls and length alone are enough to distinguish them. Their lack of pores is also very distinct among carcharhinids. Distinguishing these centra from triakid centra poses more of a problem. Both have similar shapes, lack pores, and often have elongate oval foramina. The cartilage that closes over the foramina will be the key identifying feature. This cartilage does appear to be preserved in fossils, as in ANSP AF 19, UP 122858, 122234, and 123154. Negaprion centra closely resemble the centra of Carcharhinus. Both have similar cylindrical shapes, and while Negaprion centra have more heavily concave lateral walls, these concave walls are still sometimes found in Carcharhinus centra. The shapes in articular view and foraminal characteristics are likewise similar. The most distinguishing characteristic of Negaprion centra is their greater length. Comparisons of the two genera were hampered by a lack of an appropriate sample size of Negaprion centra. Prionace centra are really quite unique among carcharhiniform centra. Their lengths are among the shortest of any carcharhinid, a family that is has many taxa with short centra. Their strongly concave walls, unique pore distribution, and shape of their intermedialia also make them distinct. Finally, their delicate cartilage is unlike the centra of any other carcharhiniform shark. From a taphonomical sense, these centra seem unlikely to be preserved as fossils. Six fossil centra have been tentatively identified as Prionace because of their similar morphologies and delicate cartilage. 73 SPHYRNTDAE: Gill 1872 Sphyrna Rafinesque 1810 Figure 16 Sphyrna Rafinesque, 1810A p. 46, 60. Description of modern Sphyrna centra Material Examined. — AMNH 93 843 - S. tiburo, entire skeleton; AMNH 99058 - S. zygaena, entire skeleton; AMNH 99064 — S. Iewini, entire skeleton; G. Hubbell Collection — S. mokarran, 1 centrum; L. Whitenack donation — S. mokarran, 8 centra; UCMP 136052 — S. Iewini, entire skeleton. Centrum Proportions. — Sphyrna centra are usually cylinders to modified cylinders, often exhibiting strong recurve at the rims. The centra of S. tiburo are the sole examples of the hourglass shaped centra. These centra are pinched in the middle and have sides recurving strongly at the rims. All centra of Sphyrna have concave walls at the apices of the double cones, though in many centra this is fairly minor. The concavity is never as pronounced as in triakids. Centra are medium in length, typically longer than carcharhinids but shorter than triakids. Centrum length never exceeds centrum width except for a few centra from UCMP 136052, a juvenile S. lewini. The centra are usually round in articular view to slightly greater in dorso—ventral height than medic—lateral breadth, with the exception of S. tiburo. Most of S. tiburo centra around round, though the largest centra located near the MP—DP transition are much larger in medic—lateral 74 Figure 16. Recent and fossil centra of Sphyrna. A-D: Recent centra. A. G. Hubbell Collection Sphyrna mokarran, dorsal view, B. AMNH 9905 8.52, dorsal view, C. AMNH 99064.65, dorsal View, D. AMNH 93843.37, dorsal view. E-I: Fossil centra. E. USNM 288057, dorsal view, E ANSP LCM 68, dorsal view, G. SDNHM 71143, dorsal View, H. UCMP 148049, dorsal View. I. USNM 494466, dorsal view. J-K: X-radiographs. J. G. Hubbell Collection Sphyma mokarran, articular View, K. X-radiograph of SDNHM 71143, articular view. Scale bars = 10 mm. 75 breadth than dorso—ventral height. With the proportions discussed, Sphyrna centra appear stocky and robust. Foraminal Proportions. — Foramina are almost always rectangular. The one exception is found on S. tiburo, where the few foramina that were not covered with arch cartilage displayed strongly bowed, elongate oval foramina. Sphyrna foramina are wide in all cases except S. tiburo. The dorsal interforaminal wall is wider than the width of the dorsal foramina, while the ventral interforaminal wall is narrower than the ventral foramina. The foramina do not extend into the rims. Pore Characteristics. — All Sphyrna centra have abundant pores, and are usually very small, though may reach medium to large sizes in rare cases. The pores are scattered, and do not form a recognizable pattern over the outer wall. The pores on S. tiburo are sparse, and are small enough that mgnification is required to see them. Notochord Canal Characteristics. — In every centrum, the notochord canal is completely closed. Calcification Characteristics. — G. Hubbell Collection S. mokarran centrum was radiographed for this study. The pattern of secondary calcification in Sphyrna is similar to most carcharhinids, but does have a unique enough morphology to be distinguishable. The lateral intermedialia have obtuse interior angles, with well rounded, obtuse exterior margins. The dorsal and ventral intermedialia are narrow wedge—shapes with straight to slightly concave exterior surfaces. The dorsal intermedialium is long, adding a distinct hump—shape to the dorsal margin in cross—sectional view, and is partly responsible for the dorso—ventral height being larger than the medic—lateral breadth The four non— 76 calcified foraminal spaces are very long, narrow, and straight. With the dorsal and ventral intermedialia being so narrow, the foraminal spaces are closely spaced. The calcified double—cone is very small in Sphyrna, similar to that of Carcharhinus. The opening for the notochordal remnant likewise has a small diameter. Four very tiny diagonal lamellae extend fiom the calcified double—cone. These diagonal calcifications are extremely short, projecting approximately 1 mm into the non—calcified areas. Description of fossil Sphyrna centra Referred specimens. — CMMV 1139; ANSP GM 2, 3, ANSP AF 3, 6, 7, ANSP LCM 1, 6, 10, 13, 17, 19, 20, 25, 31, 68—71, 76, 80, 84, 89; USNM 288026, 288027, 288041, 288043, 288049, 288051, 288055, 288056.], 2880562, 288057, 494464— 494466, 495870; SDNHM 71143; UP 92289; UCMP 148049; and BMNH 13795. Age and distribution. — Lower Miocene through Recent in Europe, Asia, and North America; Recent in all warm—temperate and tropical seas (Compagno, 1984, 1988; Cappetta, 1987; Purdy et al., 2001). Discussion. — Identification of fossil Sphyrna centra has been very difficult, as there are few characteristics common in centra of all members of this genus. Many of the specimens on loan arrived with the identification of Sphyrna, though none were associated with teeth. These centra were always compared with Recent Sphyrna centra, and always were similar in most of the characteristics. In the process of identifying these fossil Sphyrna centra, several inconsistencies did arise. Sphyrna centra are always medium to long, and almost always have recurve at the rims. The amount of recurve, 77 h6wever, varies dramatically among the centra. Some of the centra have only slight recurve (to almost none whatsoever (e.g. CMMV 1139, AN SP AF 6, 7)), while others have very strong recurve (e.g. NMNH 288026, 494465). An hourglass sluped centrum appears to be diagnostic of Sphyrna, but not all members of this genus have centra with this shape. Kozuch and Fitzgerald (1989) found this shape only in Sphyrna tiburo, and was only found in this species among the Recent comparative material examined for this study. A few fossil centra possess the long, hourglass shape consistent with the morphology of Sphyrna tiburo (e.g. AN SP LCM 19, UCMP 148049). The recent Sphyrna centra on loan typically have rectangular foramina, while oval were more common on the fossil centra. The ventral interforaminal wall is very narrow on many fossil centra, similar to those of Recent centra. Some of the largest specimens resemble one another very closely, but are very distinct when compared to other fossil Sphyrna centra. These have very heavily concave lateral walls, heavy recurve at the rims, and a medic—lateral breadth that is much greater than dorso—ventral height. No fossil specimen were observed with an opening for the notochordal canal. The most stable characteristics of Sphyrna centra are the pore size and distribution. The pores are almost always minute, enough so that magnification is often necessary for proper observation. The pores are also scattered over the entire surface of the centrum. Fossil Sphyrna centra are most similar to Galeocerdo centra, as both have modified cylinder shapes. Galeocerdo centra have large, encircling pores and generally more strongly concave lateral walls (with the exception of the largest specimens). 78 CMMV 1139, ANSP GM 2, 3, ANSP AF 3, 6, 7, USNM 288056, 494466, SDNHM 71143, and UCMP 148049 were all radiographed for this study. The overall morphology of these cross—sectional views varied to a small degree, which is not surprising considering the diversity of shapes observed in centra from Sphyrna. Each centrum has one interforaminal wall that is very narrow, which is also observable on the external surface. The most diagnostic characteristic shared by every fossil centrum is the complete lack of diagonal calcifications, and no visible calcified double cone. The only instance of diagonal lamellae being observed on a Sphyrna specimen was the Recent G. Hubbell Collection S. mokarran centrum. In this centrum, four very small and delicate lamellae are slightly visible on the x—radiograph. Every other carcharhiniform genus observed have well—developed diagonal lamellae in their centra, with the exception of Prionace. Ridewood (1921), however, notes that in Sphyrna, along with Carcharhinus and Galeocerdo, the diagonal lamellae are the most developed. While well—developed diagonal lamellae were expected, they were never observed. Indeterminate Carcharhiniform 1 Figure 17 Fossil Material Examined. — BMNH 1965, 5752.1, 5752.2, 13794, 13796, 13797. Centrum Proportions. — The six centra that comprise indeterminate carcharhiniform 1 are very distinct from any other carcharhiniform centrum examined in this study. They are best classified as robust cylinders with a medium length. Their lateral walls tend to be somewhat irregular, without showing any concavity. The rims are 79 Figure 17. Fossil centra of Indeterminate Carcharhiniformes 1. A. BMNH 5752.1, dorsal view, B. BMNH 5752.1, ventral view. C. Cross section of BMNH 5752.2, articular view. Scale bars = 10 mm. 80 visible, but not conspicuous. The medic—lateral breadth is much greater than the dorso— ventral height, giving these centra a pronounced ovoid shape in articular view. These centra resemble flattened Carcharhinus centra. All of these centra are very large, with the smallest having a width of about 50mm. Foraminal Proportions. —— The foramina are long, straight ovals in all of these centra. The dorsal interforaminal wall is typically narrower than the width of a single foramen, while the ventral interforaminal wall is wider. Pore Characteristics. — Pores are difficult to see on most of the centra due to poor preservation of the external surface. Small pores are visible encircling the rims and foramina on BMNH 5752.1, however. Notochordal Canal Characteristics. — The notochordal canal is closed in every centrum, though this may be in part due to poor preservational quality. Calcification Pattern. — All of these centra are heavily mineralized, and the foramina are filled with matrix, so no radiographs are available. BWH 5752.2 has been sectioned in the same plane as the radiographs, however. The lateral intermedialia are large, have an obtuse interior angle, and a very convex external surface. The dorsal and ventral intermedialia are very narrow, with rounded external surfaces. The non—calcified areas are narrow and straight. No calcified double cone is visible, nor are any diagonal lamellae. The lack of these structures may be a result of poor preservations. Discussion. -— These six specimens arrived from The Natural History Museum with the identification of Galeus. This identification is unlikely. Galeus is a small scyliorhinid, rarely reaching even 90cm (Compagno, 1984). These centra are far too large for this genus. Galeus was a synonym for Galeorhinus (Cuvier, 1817), though this 81 identification would also be unlikely. Fossil and Recent Galeorhinus centra have been examined, and are not morphologically similar to these unknown centra. These centra are far too large to be consistent with the size of Galeorhinus, which reach a maximum size of about 190cm (Compagno, 1984). The unknown centra have pores on the external surface, which are lacking in all observed triakid centra, including Galeorhinus. Indeterminate Carcharhiniform 2 Figure 18 Fossil Material Examined. — ANSP AF 1, 2, AN SP LCM 16, 22, 24, 32—43, 61, 78, 102, 103; USNM 288014. Centrum Proportions. — The centra of this indeterminate carcharhiniform are quite distinct. The centra are very long cylinders, and with gently concave lateral walls. Strong recurve at the rims is present in some of the centra, while in others recurve is completely lacking. The rims are narrow but distinct. All of the centra are markedly greater in medic—lateral breadth than dorso—ventral height, giving the centra a wide ovoid shape in articular view. F oraminal Proportions. — The foramina are also unique on these centra. The centra have elongate oval foramina that never extend into the rims. The foramina are usually straight on the shorter centra of this morphotype, and strongly bowed on the longer centra. These bowed foramina are positioned with the apices facing medially. The foramina often have what appear to be remnants of cartilage closing over the openings, similar to the condition seen in Rhizoprionodon centra. 82 Figure 18. Fossil centra of Indeterminate Carcharhiniformes 2. A. ANSP LCM 37, dorsal view, B. ANSP LCM 39, dorsal view, C. ANSP LCM 32, dorsal view, D. ANSP LCM 34, dorsal View, E. ANSP LCM 36, dorsal view. F-H: X-radiographs. F. ANSP LCM 16, articular View, G. ANSP LCM 37, articular view, H. ANSP LCM 39, articular view. Scale bars = 10 mm. 83 Pore Characteristics. — Pores are present on all of these centra, though in varying numbers. The pores are small, and are usually scattered over the surface of the centrum. In a few centra, the pores are more heavily concentrated in the interforaminal areas. Notochordal Canal Characteristics. — The notochordal canal is closed in every example of this morphotype. Calcification Pattern. — ANSP AF 1, 2, and AN SP LCM 16, 33, 37, and 39 were radiographed for this study. The intermedialia are well—developed and unremarkable in cross—sectional view. The lateral intermedialia are wide and well—developed, with obtuse interior angles, and strongly convex external surfaces. The dorsal and ventral intermedialia are narrow with either straight or gently concave external surfaces. The non—calcified areas are more distinct in these images. Instead of straight or flaring, as is usually observed in carcharhiniform centra, these narrow non—calcified areas have irregular, undulating margins. No calcified double cone is visible in any of the images, nor are any diagonal lamellae. AN SP LCM 37 is unusual among these six centra in that is has what appear to be calcified growths along the anterior and posterior margins of the non—calcified areas. In cross—sectional View, this is seen as a thick, knobby projection in each non—calcified area. Discussion. — These indeterminate centra are most similar to centra of Rhizoprionodon. Both are much longer than other carcharhinid centra, and both have a greater medic—lateral breadth than dorso—ventral height. Both also have long, oval foramina that are often covered or encrusted with additional cartilage. These unknown fossil centra differ fi'om Rhizoprionodon centra in a number of ways, though. 84 Rhizoprionodon centra have heavily concave lateral walls and completely lack any recurve at the rims. The unknown fossil centra, however, generally have only slightly concave lateral walls, and often have heavy recurve at the rims. These fossil centra also have distinct pores scattered on the external surfaces, while no pores were ever observed in Rhizoprionodon, even with magnification. These fossil centra also resemble centra fiom Sphyrna in a few minor ways. Both often have long centra, though these unknown centra tend to be far longer. Sphyrna centra also have recurve at the rims, though again not usually to the degree observed in the unknown centra. Both appear to usually lack diagonal lamellae in the radiographs, though the Recent Sphyrna mokarran centrum has minute lamellae, and AN SP LCM 37 has some extra calcification in the non—calcified areas. These unknown fossil centra most likely do not belong to Sphyrna. All of these centra are fi'om the Lee Creek Mine, and would most likely be from a genus represented by teeth in the Lee Creek Mine deposits, but matching the centra to a genus has not yet been possible. While the centra appear to be carcharhiniform, they do not resemble Hemipristis, Galeocerdo, Carcharhinus, or Sphyrna centra, genera that are well represented by teeth at Lee Creek, and of sufficient size. The indeterminate centra do not resemble scyliorhinid or triakid centra. About 50 Paragaleus (Hemigaleidae) teeth are present in these deposits, from sharks between 1.5 and 1.7 m in total length (Purdy et al., 2001). Recent Paragaleus are usually smaller and reach a maximum size of only 1.4 m, (Compagno, 1984). Sharks of these sizes would be small for the size of the indeterminate centra. Five teeth of T riaenodon are also present in the Lee Creek Mine deposits. These sharks are somewhat larger than Paragaleus, reaching maximum sizes of 85 about 2.1 m, but are rarely larger than 1.6 m (Compagno, 1984). Because teeth are sturdy and durable, more T riaenodon teeth would be expected if these twenty—two indeterminate centra were fiom T riaenodon. DISCRIMINANT ANALYSIS TESTING CLASSIFICATION OF FOSSIL CENTRA Introduction to Discriminant Analysis Discriminant analyses are generally employed for one of two reasons. The first relates to interpretation when studying the ways in which groups differ. The discriminant analysis will identify which variables contribute most to discriminating cases among groups. The second purpose involves classification: that is, predicting the group membership in naturally occurring groups for cases of unknown membership (Klecka, 1980; Tabachnick and Fidell, 1983). The analysis involves two types of variables. The grouping variable defines the groups into which the cases are being classified, and the discriminating variables are used to distinguish between groups. In discriminant analyses the variables are not defined as dependent or independent. If causation were indicated, the analysis would be analogous to a multiple regression analysis (Klecka, 1980). Several assumptions exist when conducting this type of analysis (Klecka, 1980): I The grouping variable must involve two or more mutually exclusive groups, with at least two cases per group. I The analysis must involve at least two more cases than discriminating variables. I The discriminating variables must be interval or ratio scale variables. 86 I Discriminating variables may not be linear combinations of other discriminating variables, and may not be perfectly correlated. To include this type of variable would involve redundant information in the analysis. I The covariance matrices for each group must be approximately equal. I The discriminating variables must have a normal distribution. The data used in these analyses meet all of the above assumptions. When the analysis is run, a classification matrix is produced that indicates the number of cases that were predicted correctly. The classification matrix also indicates the number of incorrectly classified cases, and the groups into which they were predicted. The classification matrix is especially useful when using discriminant analysis as a means of predicting the group membership for cases with unknown membership. The classification matrix can also be used to test the ability of the discriminating variables to classify cases of known grouping. An analysis can be conducted on the known cases, and the percent correctly classified indicates the accuracy of the procedure (Klecka, 1980). The more successfirl the discriminating variables are for distinguishing between groups, the higher the percentages in the classification matrix. In the following analyses, morphological measurements of various shark vertebral centra characteristics serve as the discriminating variables and identification at the family or genus level is used as the grouping variable. The ability of the independent variables to successfully discriminate for identification was first tested by performing analyses using only Recent centra where identification was already known. Once established, fossil centra with hypothetical 87 identifications were added to test whether these centra were identified as predicted in the analysis. A value called Wilks’ Lambda is produced with the output of a discriminant analysis, which is a measure of the level of discrimination in the analysis. Values of lambda that are near zero denote high discrimination. When lambda approaches a maximum value of 1.0, it is denoting less discrimination (Klecka, 1980). To test significance, Wilks’ Lambda is converted to an F value (provided in the output of the analysis) and compared to a standard statistical table. A jackknifed classification matrix is also included with the output of the analysis. Bias enters the discriminant analysis when the data fi'om the case being classified is included when developing the classification equation (Tabachnick and F idell, 1983). When all variables are included in the jackknifed classification, each case is then classified on the basis of all data except the specific case being classified, and biased is removed (Tabachnick and Fidell, 1983). The results of a jackknifed classification can be viewed as a more realistic estimate of the ability of the discriminating variables to distinguish between groups (Tabachnick and F idell, 1983). The following analyses were conducted to test the identification and significance of identification given to individual fossil centra, predict identification for unknown centra, and provide a testable model to the identification of sometimes morphologically similar fossils. 88 Methods The data used in the statistical analyses include eleven interval scale variables, measuring external centrum and foraminal proportions of the shark vertebrae. These eleven measurements were taken from all accessible centra from thirty-eight Recent individuals and 507 fossil centra that were either on loan to the MSU Museum or were examined during collection visitat ions. The eleven variables, listed by the variable names used in SYSTAT, include: 1. DIAMETER_I: The larger of the two diameter measurements, measured fi'om lateral side to side at the rim. 2. DIAMETER_2: The smaller of the two diameter measurements, measured from lateral side to side at the rim. 3. MAX_LENGTH: The length of the centrum measured fi'om anterior to posterior rim. 4. WAIST_WIDTH: Width of the centrum at the apices of the double cone. The centrum walls at this location are concave in varying degrees. 5. WAIST_HEIGHT: Height of centrum, measured at the apices of the double cone. 6. D_FORAM_L: Length of one dorsal foramen 7. D_FORAM_W: Width of one dorsal foramen, measured at its maximum. 8. D_WALL_W: Width of the dorsal interforaminal space. 9. V_FORAM_L: Length of one ventral foramen. 10. V_FORAM_W: Width of one ventral foramen, measured at its maximum. 11. V_WALL_W: Width of the ventral interforaminal space. 89 These measurements were recorded into Microsoft Excel, and later transferred to SYSTAT. The analyses of these centra were designed to examine the overall morphology of centra fiom different genera. Centra may dramatically differ in size, however, even within a single genus. The differences in size mask the overall morphology in a discriminant analysis, and centra would be classified based primarily on their size. Because of this, the interval scale variables have been transformed into ratio scale variables. Variables 2—11 were divided by DIAMETER_I. DIAMETER_l was chosen over MAX_LENGTH because the latter can vary dramatically depending on the location of the centrum within the vertebral column. Centrum length is relatively the longest just anterior to the monospondylous—diplospondylous transition. Length decreases rapidly in the caudal region. The diameter of the centra is more consistent with overall size. Following are the names of the ten transformed variables used in SYSTAT. 12. T_DIAMETER_2 13. T_MAX_LENGTH 14. T_WAIST_WIDT 15. T_WAIST_HEIG l6. T_D_FORAM_L 17. T_D_FORAM_W 18. T_D_WALL_W 19. T_V_FORAM_L 20. T_V__FORAM_W 21. T_V_WALL_W 90 Two nominal variables were included for use as grouping variables in the discriminant analyses. 22. GENUS 23. FAMILY Nine genera were measured and included in the data matrix, and were given the following nominal classification in the GENUS variable: 1. Galeorhinus 2. Mustelus 3. T riakis 4. Carcharhinus 5. Galeocerdo 6. Negaprion 7. Prionace 8. Rhizoprionodon 9. Sphyrna Three families were measured and included in the data matrix, and were given the following nominal classification in the FAMILY variable: 91 1. Triakidae 2. Carcharhinidae 3. Sphyrnidae The raw data used in these analyses is found in Appendix A. Transformed data is available upon request. SYSTAT output, including Wilks’ Lambda, Approximate F values, and classification matrices, is found in Appendix C. Discriminant analyses testing identification of modern centra: The first discriminant analysis performed was on data that included only measurements of Recent centra of known identification. This was designed to test the ability of the analysis to successfully discriminate centra using the morphological variables listed above. Many of the cases were not included in the final discriminant analyses for several reasons. First, varying numbers of centra were available from each individual shark, ranging from the entire column of over one hundred centra to as few as one or two centra fiom an individual. In order to prevent a few specimens with hundreds of centra fiom overwhelming the data set, a maximum of ten centra with complete measurements were randomly included fi'om a single individual. Second, only centra from the monospondylous region were included in the data set. The MP—DP transition is marked by an abrupt decrease in centrum length Diplospondylous centra converge in morphology posteriorly, even among different families. To prevent these morphological similarities fi'om obscuring the differences in the larger, more anterior centra, they were removed from the data matrix. Two genera were not included in the analysis because of an exceedingly small sample size. Only two monospondylous centra were available for 92 the genus Negaprion and, excluding fluid—preserved specimens, three monospondylous centra of Rhizoprionodon (two being the minimum number of cases in a group allowed by a discriminant analysis (Klecka, 1980)). Finally, any centrum that had missing data was not included. Remnants of the neural and haemal arches were sometimes present on the vertebral columns of the Recent skeletons. These remnants often prevented measurements from being recorded, especially WAIST_HEIGHT, D_FORAM_W, and D_WALL_W. The final data set of Recent centra includes 116 centra from 28 individuals. A correlation table of the transformed variables demonstrates that these variables are not perfectly correlated, supporting the assumption necessary for discriminant analyses (Appendix B). The most highly correlated variables are T_D_FORAM_L and T_MAX_LENGTH with correlation coefficients of 0.95, T_V_FORAM_L and T_MAX_LENGTH at 0.96, and T_D_FORAM_L and T_V_FORAM_L at 0.97. All other variables were correlated with coefficients less than 0.82. Discriminant analyses were generally performed using the complete estimation option, which includes all variables in the solution. Stepwise Options are available, which test the contribution of individual variables to the discriminating process. Those that do not discriminate the cases well are removed. This option is particularly usefirl for exploratory purposes, or when the goal is to interpret the ways the groups differ. The first discriminant analysis performed on the data from the Recent centra used all ten transformed ratio variables, with FAMILY as the grouping variable. The analysis used 0.01 tolerance and the complete estimation option, and the distances were saved. The Wilks’ Lambda was 0.0729, and Approximate F was 28.1148, which is significant 93 below the 0.001 level. 100% of triakid centra were predicted correctly (31 total), as were 100% of sphyrnid centra (16 total). Of the sixty—nine carcharhinid centra, 96% were discriminated properly, with three incorrectly predicted as sphyrnid centra. Overall, 97% of the centra were correctly classified, with 94% correctly classified in the jackknifed classification matrix. For exploratory purposes, a backward stepwise discriminant analysis was performed on the same ratio variables. This analysis begins with all variables used in the discriminant function, and removes variables when the F score drops below the identified threshold. FAMILY was once again the grouping variable in this analysis, and 0.15 as the to Alpha—to—enter and Alpha—to—remove. Wilks’ Lambda was 0.0617, and the F score was 54.4438 (significant below the 0.001 level) for this analysis. In this analysis one Carcharhinus centrum was classified as a Sphyrna centrum. The total correct predictions ' was 99%, with 96% correctly predicted in the jackknifed classification matrix. Of the ten variables, T_WAI ST_HEIG, T_D_WALL__W, T_D_FORAM_L, T_V_FORAM_L, T_V_FORAM_W, and T_V_WALL_W were useful in discriminating the families. T_DIAMETER__2, T_MAX_LENGTH, T_WAIST_WIDT, and T_D_FORAM_W were not entered into the solution by SYSTAT. To firrther test the use of the transformed ratio variables for classification, all ten variables were included in a discriminant analysis with GENUS as the grouping variable. The analysis used 0.01 tolerance and the complete estimation option, and the distances were saved. The Wilks’ Lambda was 0.0120, and Approximate F was 11.6830 (significant below the 0.001 level). The included variables discriminated for genus very well, though not with quite the same success as for family. Among the triakids, only 94 Mustelus centra (5 total) were discriminated properly 100% of the time. Galeorhinus centra (11 total) were predicted correctly 91% of the time, with one centrum predicted as a Galeocerdo centrum. T riakis centra had the lowest number of correctly classified centra, with 80% being properly predicted. Of the 15 centra, 12 were predicted correctly, with two predicted as Galeorhinus and one as Mustelus. The centra from triakid genera tend to be very similar to each other, and identification is difficult. The discriminant analyses consistently had the most difficulty correctly predicting the classification of triakid genera when fossil centra were added to the data matrix. Of the three carcharhinid genera, Galeocerdo (12 total) and Prionace (5 total) centra were both predicted correctly 100% of the time. Carcharhinus centra (52 total) were classified correctly 85% of the time, with two centra classified as Galeorhinus centra, three as Galeocerdo, and three as Sphyrna. Sphyrna centra (16 total) were classified correctly 100% of the time. Overall, 90% of the centra were correctly classified. The Jackknifed classification matrix scores for triakids were somewhat lower, as they were in the family discriminant analysis. The percent classified correctly remained the same for Galeorhinus and Mustelus (91%, 100%, respectively), but decreased to 67% for T riakis centra. Five of these centra were misclassified, with two centra predicted as Galeorhinus, two as Mustelus, and one as Galeocerdo. All three carcharhinid genera had lower percentages of correctly classified centra. Carcharhinus centra were predicted correctly at 83%, as one additional centrum was misclassified as Sphyrna. Galeocerdo centra dropped to 92%, with one centrum predicted as Carcharhinus. Prionace centra were predicted correctly at only 60%, with one 95 centrum classified as Mustelus and one as Carcharhinus. Sphyrna centra were correctly predicted 88% of the time, with one centrum predicted as Carcharhinus and one as Galeocerdo. The total correct predictions were 83% in the Jackknifed classification matrix for genera. For exploratory purposes, a backward stepwise discriminant analysis was also performed on the same ratio variables, with genus once again as the grouping variable, and 0.15 to enter and remove. Wilks’ Lambda was 0.0076, and the Approximate F score was 17.8044 (significant below the 0.001 level) for this analysis. This new analysis successfully classified 93% of the centra, with 88% correctly predicted in the jackknifed classification nuttrix. Of the ten variables, T_MAX_LENGTH, T_WAIST_WIDT, T_WAIST_HEIG, T_D_FORAM_L, T_D_WALL__W, T_V_FORAM_L, T_V_FORAM_W, and T_V_WALL_W, were useful in discriminating the families. T_DIAMETER_Z and T_D_FORAM_W were not entered into the solution by SYSTAT. After the above analyses were run, Negaprion and Rhizoprionodon centrum measurements were included into the data set. Some fossil centra were available that were consistent with the morphology observed in those two genera (especially fossil centra fi'om FMNH). Despite having a small sample size of centra from the Recent genera, they proved necessary for comparison. Two discriminant analyses were performed on the file including the Negaprion and Rhizoprionodon centrum measurements. The first amlysis used FAMILY as the grouping variable. The Wilks’ Lambda was 0.0726, and the Approximate F was 29.5631 (significant below the 0.001 level). The overall percentage of the classification matrix was 98%, and the Jackknifed classification 96 matrix was 96% predicted correctly. Three Carcharhinus centra were misc lassified as sphyrnids. The second analysis used GENUS as the grouping variable. The Wilks’ Lambda was 0.0067, and Approximate F was 9.9247 (significant below the 0.001 level). 85% of the centra were classified correctly, with 74% correctly classified in the Jackknifed matrix. The new genera, with their small sample sizes increased the error in the predictions. One of the two Negaprion centra was misclassified as a Carcharhinus centrum and two of the three Rhizoprionodon centra were misclassified as Prionace. The addition of these two genera also lowered the percentage of correct predictions for Carcharhinus and Prionace, as some of these centra were predicted as the added genera. Discriminant analyses including fossil centra The results of the discriminant analysis on modern shark centra justify the use of the transformed ratio variables as a test for identification, with correct classifications being in the ninety—percent range for both family and genus classification. When adding fossil centra measurements to the Recent centra data set, the existing classification problems continued, and therefore lowered the overall percent predicted correctly for the following analyses. For example, centra from Carcharhinus acronotus and C. brevipinna were often misclassified in the above analyses. Their centra are unusually long compared to other Carcharhinus centra, and resemble triakid centra based on measurements alone. The analyses continued to misclassify these centra after fossil centra data was added. Measurements of fossil centra were recorded in the same manner as for Recent centra. These centra were also transformed into ratio scale variables by dividing 97 DIAMETER_I into each variable. Before the fossil centra were included in the data matrices for discriminant analyses, they were first given tentative identification based on characteristics compared with the modern centra. As above, centra that were missing data (i.e., incomplete centra) or centra that were from the diplospondylous regions were not included in the analyses. Calvert Marine Museum fossil centra Added to the data set of modern shark centra (not including Negaprion or Rhizoprionodon) were twenty—two fossil centra from the Calvert Marine Museum. Once constructed, the data set was subjected to discriminant analyses. The first analysis involved all transformed ratio variables with FAMILY used as the grouping variable. Wilks’ Lambda for this analysis was 0.1027, and the Approximate F was 26.7246 (significant below the 0.001 level). The classification matrix had an overall score of 96% predicted correctly, and the jackknifed classification matrix had 92% correct predictions. For the fossil centra, only one was predicted contrary to the tentative identifications given them. CMMV—1139 was identified as a sphyrnid in the nominal variable FAMILY, but was predicted as a triakid. This centrum is a fluted cylinder, that is, a long cylinder with strongly concave lateral walls and no recurve at the rims. This centrum is large, however, with a diameter of almost 40 mm, and has very small, scattered pores similar to those present on recent Sphyrna centra. The same data set was also subjected to a discriminant analysis with GENUS as the grouping variable. The Wilks’ Lambda for this analysis was 0.0183, and the Approximate F was 12.2994 (significant below the 0.001 level). The classification matrix 98 had an overall score of 91%, and the jackknifed matrix was 83% predicted correctly. As for the discriminant analysis with FAMILY as the grouping variable, only CMMV-1139 was predicted differently than its initial identification. While identified as a Sphyrna centrum, it was predicted as Triakis. Again, the fluted cylinder morphology of this centrum probably led to this classification. Overall size and pore distribution are not being analyzed during the discriminant analyses. Of interest is CMMV-815. This centrum is large, with a short, disk-like cylinder shape, has a large amount of concavity at the apices of the double cone, and small to medium size pores that closely follow the rims. This centrum seems less dense and more delicate than other fossil centra, and is not preserved as well. These characteristics closely match those of modern Prionace centra. Because of these similarities, this centrum was classified as Prionace in the GENUS variable, and was predicted as the same genus in the discriminant analysis. Cappetta (1987) lists the only known fossil Prionace specimens from the Pliocene of Italy. CMMV—815 was found as float at the Calvert Cliffs in Maryland, known for its Miocene shark fossils. Florida Museum of Natural History fossil centra Fossils fi'om the Florida Museum of Natural History include sixteen centra from the Late Miocene through Pleistocene of Florida. Five of the centra were associated with three teeth and were identified as Negaprion brevirostris. Of the seventeen centra fi'om the University of Florida collection, six were caudal centra and not included. Three other centra of poor quality were excluded due to a lack of complete measurements. Eight centra remained, of which three were associated. 99 The University of Florida centra measurements were combined with the data set that included Negaprion and Rhizoprionodon Recent centra. This new data set was subjected to a discriminant analysis using the complete estimation option and included all transformed ratio variables and used FAMILY as the grouping variable. The Wilks’ Lambda for this analysis was 0.123 and the approximate F was 21.6587 (significant below the 0.001 level). The classification matrix had 92% of centra predicted correctly, and a jackknifed matrix with 88% predicted correctly. Of the eight fossil centra added to the matrix, five were discriminated correctly. Of the three misclassified, UP 92281, identified as Carcharhinus, was predicted as a sphyrnid. UF 92289, identified as Sphyrna, was predicted as a carcharhinid. Finally, UF 123154, identified as Rhizoprionodon, was predicted as a triakid. The same data were subjected to a discriminant analysis using the complete estimation option that included all transformed ratio variables and used GENUS as the grouping variable. The Wilks’ Lambda for this analysis was 0.0155 and the approximate F was 8.2754 (significant below the 0.001 level). The classification matrix had 83% of centra predicted correctly, and a jackknifed matrix with 71% predicted correctly. Four of the eight fossil centra were discriminated incorrectly for genus. UF 3245.2, one of the Negaprion brevirostris centra, was classified as Carcharhinus, while the other two were predicted correctly. This is puzzling, as the three associated centra closely resemble each other, and have similar dimensions. The unfortunately small sample size for Negaprion centra (five total) likely played a role. UF 123154 may have had similar problems. In this case, the sample size for Rhizoprionodon, including UF 123154, was only four centra. While identified as Rhizoprionodon, this centrum was 100 classified as Mustelus. Both genera have long, fluted cylinder shaped centra and a medic— lateral breadth that is greater than dorso—ventral height, which could have led to the misclassification. UF 123154 is clearly a Rhizoprionodon centrum based on examination of the foramina. UF 128906, identified as Carcharhinus, was predicted as Rhizoprionodon, which again may be attributed in part to the confusion introduced due to the small sample of Rhizoprionodon centra. UP 92289 was tentatively identified as Sphyma, but was classified as Galeocerdo. This centrum is quite small, and does resemble Galeocerdo in some aspects, but is missing the characteristic large pores and modified cylinder shape. Neither of these characteristics would be apparent in the measurements taken from the centra and included in this data set. The Natural History Museum (London) fmil centra Sixteen fossil centra fi'om The Natural History Museum were analyzed using discriminant analyses. Four centra from this collection were excluded because they are unlike any recent comparative material available for this study. These centra were previously identified as Galeus sp. Other centra were excluded due to incomplete specimens or because they were caudal centra. The measurements from the sixteen centra were added to the data set of Recent centra that excluded the Negaprion and Rhizoprionodon centra, as there were no fossil centra in this set tint resembled those two genera. The new data set was subjected to a discriminant analysis using the complete estimation option and using all transformed ratio variables and FAMILY as the grouping variable. The Wilks’ Lambda was 0.0860, and the approximate F was 28.9251 for this analysis (significant below the 0.001 level). 101 The classification matrix had 96% of the centra predicted correctly, with a jackknifed matrix with 92% predicted correctly. The only fossil centrum misclassified was BMNH 13795, which was identified as Sphyrnidae and classified as Carcharhinidae. The same data were subjected to a discriminant analysis using the complete estimation option, and again using all transformed ratio variables and GENUS as the grouping variable. The Wilks’ Lambda was 0.0146, and the approximate F was 12.6585 for this analysis (significant below the 0.001 level). The classification matrix had 91% of the centra predicted to genera correctly, with a jackknifed matrix with 85% accuracy. Once again BMNH 13795 was misclassified. While identified as a centrum from Sphyrna, the analysis predicted it as a Carcharhinus centrum. While the measurements may suggest the latter classification, simple observation suggests this is actually a centrum from Sphyrna. The centrum is relatively much longer than Carcharhinus centra tend to be, and is closer to the relative length of Sphyma. The centrum did arrive from The Natural History Museum with a carcharhinid identification, however. BMNH 1309 was also misclassified in the analysis. This centrum was entered as a Galeocerdo centrum, and was predicted as Carcharhinus. This centrum is not entirely complete, which will affect measurements to some degree. The centrum has a somewhat shorter relative length than most Galeocerdo centra, but has a slightly modified cylinder shape and large pores. The centrum arrived from The Natural History Museum with the identification Galeocerdo. 102 San Diego Natural History Museum fossil centra The vertebral centra on loan from the San Diego Natural History are unusual by having a heavy concentration of triakid centra, and little else. Carcharhinidae, and especially Carcharhinus centra, are usually more abundant. Triakid centra tend to be difficult to identify to genus due to similarities on overall morphology, and lack of any distinct characters. Measurements fi'om thirty—nine complete, non—caudal centra were included in the data set of modern carcharhiniform centra measurements. The transformed ratio variables were subjected to a discriminant analysis with the complete estimation option using FAMILY as the grouping variable. The Wilks’ Lambda for this analysis was 0.1375, and the approximate F was 24.2635 (significant below the 0.001 level). The classification matrix shows 95% of centra classified correctly, with a jackknifed matrix with 91% correct classification. Only two fossil centra were misclassified in this analysis. One was SDNHM 61933.06, classified as a carcharhinid and predicted as a triakid. The other was SDNHM 61933.09 (not associated with SDNHM 61933.06), classified as a triakid, but predicted as a carcharhinid. The same transformed ratio variables were subjected to a discriminant analysis with the complete estimation option, using GENUS as the grouping variable. The Wilks’ Lambda was 0.0229, and the approximate F was 12.9108 (significant below the 0.001 level). The classification matrix had 86% predicted correctly, with a jackknifed classification matrix with 81% correct predictions. Nine of the thirty—nine fossil centra were misclassified. SDNHM 61933.06 was once again misclassified. It was identified as Carcharhinus and was predicted as a Galeorhinus centrum. It is very possible that the 103 initial classification of this specimen was incorrect, and the discriminant analysis has predicted it correctly. The specimen is a small, short cylinder with heavily concave lateral walls. It also lacks pores, which is a triakid character. This specimen may be a centra fi'om the diplospondylous region, misleading the initial classification. SDNHM 61933.08 (not associated with the other centra in lot SDNHM 61933) was initially classified as Carcharhinus, but was predicted as Galeocerdo. The small size of this specimen makes classification difficult, especially if it is a carcharhinid, as it was likely from a juvenile. This centrum has a cylinder shape that is approaching a modified cylinder. The lateral walls are very concave, and the length is relatively long for a carcharhinid centrum. Pores are present, but due to the small size of the vertebra, are difficult to observe clearly. While clearly a carcharhinid, the identification to the generic level is uncertain. SDNHM 61933.09 (not associated with other centra in lot SDNHM 61933) was identified as Galeorhinus and predicted in the analysis as Galeocerdo. The initial identification of Galeorhinus is more likely in this case. The centrum is a small cylinder with strongly concave lateral walls and no hint of recurve at the rims. The relative length compared to diameter would be similar to that of Galeocerdo but is also consistent with Galeorhinus. This specimen also completely lacks pores on its lateral surface and interforaminal areas, once again suggesting the Galeorhinus prediction is correct. SDNHM 61933.14 (not associated with other centra in lot SDNHM 61933) was initially identified as Galeorhinus, and was predicted as Mustelus during the analysis. While this specimen is clearly a triakid centrum, generic classification is not as clear. The centrum has foramina that are narrower than is usually seen in Galeorhinus centra, and 104 there is some recurve at the rims. Both of these characteristics are common to Mustelus. In this instance, the predicted classification of Mustelus is most likely correct. Specimens SDNHM 61933.15 (not associated with other centra in lot SDNHM 61933), 71142.08, and 71142.28 (specimens from lot SDNHM 71142 are not associated) were initially described as Galeorhinus and were predicted as Galeocerdo. These specimens are approaching a fluted cylinder classification, lack recurve at the rims, have wide foramina that extend into the rims, and completely lack pores. These specimens are clearly not Galeocerdo centra. The reason for this prediction during the analysis is not clear. SDNHM 61933.16 (not associated with other centra in lot SDNHM 61933) was initially classified as Triakis, and was predicted to be Galeorhinus as a result of the analysis. The ventral interforaminal area on this centrum is extremely wide, suggesting this centrum was from the anterior-most region of the vertebral column. Centra fi'om this region tend to converge in morphology among different genera, and are difficult to identify. This centrum is clearly from a triakid, but generic identification is not certain. SDNHM 71142.14 was initially identified as Mustelus, but was predicted as T riakis during the analysis. The reason for the initial identification of Mustelus was the presence of slight recurve at the rims. The centrum has foramina that are wider than normal for a Mustelus centrum, but they do extend into the rims. Triakid foramina do not appear to extend into the rims, nor do they have any recurve at the rims, suggesting this centrum actually is fiom Mustelus. 105 Academy of Natural Sciences of Philadelphia fossil centra Measurements from ninety-seven fossil centra were included in the following analyses. Two hundred and thirty—four specimens are on loan from the Academy of Natural Science, but many of them are incomplete or are clearly caudal centra. Also, several centra that have been identified as Hemipristis sp. were not included because of a lack of recent centra of that genus for comparison. Finally, one morphotype was present in this sample that did not match any other centra from modern sharks, and any centrum with this morphology was not included in the discriminant analysis. Measurements fiom the ninety-seven complete, non—caudal centra were included in the data set of modern carcharhiniform centra measurements. The transformed ratio variables were subjected to a discriminant analysis with the complete estimation option using FAMILY as the grouping variable. The Wilks’ Lambda for this analysis was 0.2039, and the approximate F was 24.4166 (significant below the 0.001 level). The classification matrix had 88% of centra classified correctly, with a jackknifed matrix with 85% correct classification. The percent of correctly classified centra is lower for this analysis than in other analyses. Several centra that were identified as Sphyrnidae had long, fluted cylinder shapes and were predicted as triakid centra. These centra had small, scattered pores on the lateral walls and interforaminal areas, however, which are never present on triakid centra. Other centra were commonly confused between Carcharhinidae and Sphyrnidae. The same transformed ratio variables were subjected to a discriminant analysis with the complete estimation option, using GENUS as the grouping variable. The Wilks’ Lambda was 0.0660, and the approximate F was 11.7652 (significant below the 0.001 106 level). The classification matrix had 86% predicted correctly, with a jackknifed classification matrix with 80% correct predictions. Twelve of the ninety—seven fossil centra were misclassified in this analysis. The same problems that occurred in the first analysis were also present in this analysis. LCM 19, 70, 71, and 80 were identified as Sphyrna centra, but predicted as a variety of triakid genera. These centra had long, fluted cylinder shapes, which is sometimes present in Sphyrna centra, but more common in triakid centra. These centra had small, scattered pores covering their surfaces, and the identification of Sphyrna is more likely correct. One centrum of Sphyrna (LCM 25) was predicted as Galeocerdo, which is understandable. The two genera both have modified cylinder shapes. One of the main observable differences between them is the difference in pore size and distribution. Sphyrna centra typically have small pores scattered over their surfaces, while Galeocerdo centra have large, encircling pores. This character was not tested during the discriminant analyses. AN SP 308.1 and 308.2 were identified as Carcharhinus and were predicted as Prionace. These centra do superficially seem very similar to Prionace, but lack the single line of pores along each rim, and are more likely Carcharhinus centra. LCM 64, 66, 86, 97 was tentatively identified as Prionace for the analysis. Its morphology and pore distribution closely resembles that of Prionace. This centrum was predicted as Prionace in the analysis. Agrico Fort #4 was identified as Galeorhinus and predicted as T riala‘s. While this centrum is triakid, it is possible that it could belong to either genus. It resembles 107 Galeorhinus in its short, stocky shape, but it has heavily concave lateral walls and rather narrow foramina, similar to that seen in T riakis. Generic classification may not be confidently made in this case. ANSP 15415.06 and 15415.08 (not associated) and Lee Creek 15 were all identified as Carcharhinus and predicted as Sphyrna in the analysis. AN SP 15415.08 is clearly a centrum of Carcharhinus, with its short, cylinder shape. The other two are more problematic. Both AN SP 15415.06 and Lee Creek 15 are cylinder shape, though are slightly longer than most Carcharhinus centra. Their foramina are relatively large for their diameter. It is possible these are actually Sphyrna centra, but identification is not certain. Finally, LCM 50 was identified as a Galeocerdo specimen, but was predicted to be Carcharhinus. Upon further examination, the initial identification is still likely correct. This centrum is a large, modified cylinder with noticeable recurve at both rims. The length of this centrum exceeds that which is normally seen in Carcharhinus, though is shorter than most Galeocerdo centra The pores are encircling, but smaller than is usually observed in Galeocerdo. National Museum of Natural History fossil centra A total of forty—one fossil centra on loan from the National Museum of Natural History were included in the data matrix with modern shark centra These fossils are unusually high in Carcharhinus and Sphyrna centra. No triakid centra were observed. A discriminant analysis of the new data matrix was conducted with the complete estimation option using all transformed ratio variables and FAMILY as the grouping 108 variable. The Wilks’ lambda for this analysis was 0.1469 and the approximate F was 23.3286 (significant below the 0.001 level). 87% of the centra were predicted according to the FAMILY variable, with 83% correct predictions in the jackknifed classification matrix. The centra misclassified in this discriminant analysis were similar to that of the following discriminant analysis, and will be discussed more thoroughly below. The same data matrix was subjected to a discriminant analysis, with the complete estimation option, using GENUS as the grouping variable. The Wilks’ lambda was 0.0483, with an approximate F of 9.8783 (significant below the 0.001 level). The classification matrix had 83% of the centra predicted correctly, with 76% correct in the jackknifed matrix. Of the forty—one fossil centra, seven were misclassified. Most of these misclassifications were expected, however. USNM—V—288049, 288055, 459870, 494465, and 494466 were all identified as Sphyrna. These centra are all extremely large, relatively long, and have extremely concave lateral walls at the apices of the double cone. With the effects of overall size eliminated as a result of transformation of the data, the morphology of these centra is very similar to that of triakid centra. These fossil centra are clearly not triakid, however. They all have well—developed pores, and are much too large. It is most likely these centra are actually monospondylous centra from Sphyrna. USNM—V—288041 was identified as Sphyrna, but predicted as Galeocerdo. These two genera have similar centra, but this particular centrum closely matches the five centra listed above. It is long and has extremely concave lateral walls. Galeocerdo centra have large, encircling pores, while this specimen has small, scattered pores. USNM—V494467 was identified as Galeocerdo, but was predicted as Sphyrna. This specimen is not as rehtively long as most Sphyrna centra tend to be, but is longer 109 than Carcharhinus centra. The centrum has a modified cylinder shape, with large, well— developed pores that appear to be mostly encircling, though the outer surfaces of this centra are poorly preserved. The remaining centra all were predicted as identified. USNM—V—24914 has 13 associated centra, and USNM—V494463 has 11 associated centra, and the predictions were consistent among all of these specimens. University of California Museum of Paleontology fossil centra A total of forty—seven fossil centra on loan from the University of California Museum of Paleontology were included in the data matrix of modern shark centra. A discriminant analysis of the new data matrix was conducted with the complete estimation option using all transformed ratio variables and FAMILY as the grouping variable. The Wilks’ lambda for this analysis was 0.1222 and the approximate F was 26.4336 (significant below the 0.001 level). 91% of the centra were predicted according to the FAMILY variable, with 90% correct predictions in the jackknifed classification matrix The centra predicted incorrectly in this discriminant analysis were similar to that of the following discriminant analysis, and will be discussed more thoroughly below. The same data matrix was subjected to a discriminant analysis with the complete estimation option, using GENUS as the grouping variable. The Wilks’ lambda was 0.0264, with an approximate F of 12.9225 (significant below the 0.001 level). The classification matrix had 86% of the centra predicted correctly, with 79% correct in the jackknifed matrix. Of the forty—seven fossil centra included in the analysis, seven were misclassified. 110 UCMP 14802 is one of the largest specimens on loan fi'om the University of California Museum of Paleontology. It was identified as Galeocerdo due to its large encircling pores and modified cylinder morphology. This centrum was predicted to be a Galeorhinus centrum as a result of the analysis, which it clearly is not. This centrum appears to have some damage that occurred during preparation, which could alter the specimen enough to result in a misclassification. UCMP 148003 is a short, cylinder—shaped centrum with straight lateral walls, and resembles Carcharhinus centra in all observable characteristics. It was, however, predicted as a Sphyrna centrum during the analysis. The initial identification of Carcharhinus, however, is more likely. UCMP 148041 was initially identified as Mustelus, but was classified as Galeorhinus during the analysis. This centrum is a small cylinder to modified cylinder with strongly concave lateral walls that recurve towards the rims. The medic—lateral breadth is larger than the dorso—ventral height, giving the centrum an ovoid articular view. These characteristics led to the initial identification of Mustelus, which is more likely than the predicted classification of Galeorhinus. UCMP 148046 is a long, fluted cylinder, with no evidence for pores on the surface. This centrum is poorly preserved, so pores may have been obliterated during fossilization. This centrum was identified as Sphyrna, but was classified as Mustelus. This centrum does have strongly concave lateral walls that recurve at the rims. The dorso—ventral height is larger than medic—lateral breadth. The proper identification of this centrum is unclear. 111 UCMP 148048 is a small, cylinder—shaped centrum with strongly concave lateral walls. This centrum lacks any pores on the lateral walls or in the interforaminal areas. The centrum is shorter than most triakids, but similar to Galeorhinus. These characteristics led to an initial identification of Galeorhinus, though was predicted to be Galeocerdo. While these two genera do share some similarities in overall morphology, the small size of this centrum and lack of pores makes the prediction of Galeocerdo unlikely. UCMP 148049 is a unique specimen. It is a long, fluted cylinder, though the lateral walls are only slightly concave. These walls recurve at the rims, giving it a shape that approaches an hourglass. This centrum has small, encircling pores. This centrum closely resembles the centra of Sphyrna tiburo, and was therefore identified as a Sphyrna centrum. USCMP 148049 was predicted as T riakis during the analysis, which it clearly is not. The long, narrow shape would be misleading in an analysis, especially when recurve at the rims and pore information are not represented in the data matrix. The final centrum to be misclassified in this analysis was UCMP 148056. This centrum is a long, fluted cylinder. The medic—lateral breadth is greater than dorso—ventral height. Overall, it is a narrow, delicate specimen. The centrum was initially identified as Mustelus, but was predicted to be a Galeorhinus centrum. Galeorhinus centra tend to be shorter and stockier in appearance than is observed in this centrum. Mustelus is a more likely classification. 112 PHYLOGENETIC ANALYSIS To test the hypothesis that the centrum characters discussed earlier reflect phylogeny, a data matrix of twenty-three unweighted clnracters was scored and subjected to a maximum parsimony analysis. Included in the data matrix were all genera discussed in the Systematic Description (with the exception of the two outgroups), and also the two indeterminate taxa. Information on ingroup taxa was obtained through examination of specimens during collection visitations and those available by loan at the Michigan State University Museum. Outgroup characters were obtained either through examination of specimens available by loan or from Hasse (1879—1885), Ridewood (1921), Nakaya (1975), and Compagno (1988). The list of centrum characters can be found in Appendix D. To assess the relationships within the ingroup, two closely related taxa were chosen as the outgroup (Maddison et al., 1984). The two outgroup taxa chosen for this analysis are Scyliorhinus and Haploblepharus, which are both basal members of the Order Carcharhiniformes. The heuristic search algorithm with closest addition sequence was used with the PAUP" 4.0 program (Swofford, 2000). This search algorithm selects a representative sample among the possible most parsimonious trees. For comparative purposes, the branch—and—bound search algorithm was also used on the same data matrix. In every analysis, both search algorithms provided identical results. Too many taxa were being analyzed to use the exhaustive search algorithm. Two different analyses were run. In the first analysis, characters 4, 5, 15, 16, 18, and 19 were ordered while the rest were unordered. The six ordered characters were all multistate characters with states that 113 potentially imply progression. In the second analysis, all characters were unordered. Both Strict and Adams consensus trees were obtained from the resulting most parsimonious trees. MacClade 4.03 (Maddison and Maddison, 2001) was used to further manipulate the topology of the trees. After the data matrix with centrum characters was subjected to maximum parsimony analyses, a second data matrix of unweighted characters was constructed and subjected to a maximum parsimony analysis. This second matrix included one hundred and thirteen characters of morphological characters fi'om complete specimens of Recent carcharhiniform sharks, modified fiom Compagno (1988). Taxa included in this analysis were nearly the same as those included in the analyses using centrum data, minus the two indeterminate carcharhiniform sharks, and minus Haploblepharus, due to lack of complete character information. The heuristic search algorithm with closest addition sequence was used with this data matrix, and all characters were unordered. Finally, the original twenty-three centrum characters were added to this matrix, for a new matrix totaling one hundred and thirty—six whole body plus centrum characters. The heuristic search algorithm with closest addition sequence was used with this data rmtrix also. Characters 4, 5, 15, 16, 18, and 19 were analyzed as both ordered and unordered. Both results were identical, and will only be discussed once. Results The first analysis using only centrum characters, with characters 4, 5, 15, 16, 18, and 19 ordered produced three equally parsimonious trees, each 70 steps long with a 114 consistency index of 0.784. The Strict and Adams consensus trees have the same topology (Figure 19a). The second analysis using only unordered centrum characters produced eight equally parsimonious trees, each 66 steps long with a consistency index of 0.600. The Strict and Adams consensus trees differ only by the position of Rhizoprionodon. In the Strict consensus tree (Figure 20a), Rhizoprionodon is part of an unresolved polytomy including the outgroup taxa, Mustelus, the clade containing Triakis and Galeorhinus, and the clade containing all other taxa. In the Adams consensus tree Rhizoprionodon is united in a clade with Mustelus. These consensus trees are similar to those created using ordered characters, with a few notable exceptions. The two sets of trees differ with the position of Prionace. In the analysis using ordered characters, Prionace is positioned as the sister— taxon to the clade containing Hemipristis and the clade containing Galeocerdo, Sphyrna, Negaprion, and Carcharhinus. In the analysis using unordered characters, Prionace is positioned as the sister—taxon only to the clade containing Galeocerdo, Sphyrna, Negaprion, and Carcharhinus. Also, Hemipristis and the two indeterminate taxa no longer form a monophyletic clade as they do in the previous analysis. Finally, in the analysis using unordered characters, Triakis and Galeorhinus form a monophyletic clade, unlike the analysis using ordered characters. The third analysis using 113 morphological characters from complete specimens of carcharhiniform sharks produced two equally parsimonious trees, each 116 steps long. The Strict and Adams consensus trees had a consistency index of 0.750, and produced trees with identical topologies (Figure 21a). The consensus tree resembled the trees obtained using centrum characters in the placement of the triakids and Hemipristis. The 115 9 0 o '\ 'L 9 so of - xv :0 .90 ““0 Q\O 6‘0 @096 (6'90 {$00 6 \Q 0 \ \\ ngQ 1609 906°‘19Q3‘« \Qi‘OQGWbbW‘ng‘xog 3 3,11,12 2.11.13 44. 17 17 $19,” 15 6,12 6, 11,23 7 8 3 8 1,13,15-17, 21 4,9,10,14 7 A 5 0 N ‘5 0 006° . ‘0‘ ’ o . 09 \00 \09v 09 $99 {\00 ° ‘39 0’0“, 096‘? (p ‘6 0 Q{‘00 {Go .\i°\(p9 0..., \oo{°.o‘36§\9; .009 go“ 6080 9(90‘6 09 \h‘) é«6°& C9» Qy‘ %O\ bé‘kobo 61‘0“ WOQ B Figure 19. Cladograms showing relationships among Carcharhiniformes using centrum characters (with characters 4, 5, 15, 16, 18, and 19 ordered). A. Consensus tree of three equally parsimonious trees, showing character changes. Underlined numbers indicate reversal to primitive state. B. Manipulated consensus tree topology for Carcharhiniformes, resulting in 3 additional steps. 116 09 v0 9 0b ‘0. 9 o ’0‘.) 0 0Q ye {06‘ {‘09 006’ o’b‘& ‘5) (P (6 0 {‘00 {96‘ 6‘0 oéoé 0° 0\°0"§3 \0 ‘1' 669 bég. 06\\ 00° \ “‘6‘ ago «$0 9 f ‘f 1‘ 7’ Q9 \ a V‘ Q8 0 9 6 c 8 45,9, 1 1014 11 14 112312 3 9 7 5,8,1], 15,7 12’” 4.10.14 5,9,13 1, 6, 8, 15-17, 21 A 9 c o W N . 9 “Q .69 (96 ’ '-e .909 $0 Q\0 0&0 ° 0 o\00 {it be“ 56‘. 6‘9 00° 6600 v OQOQ ‘6‘ 9°‘<\°\l'\°C9ch\c6°Qc,9Q\soe B Figure 20. Cladograms showing relationships among Carcharhiniformes using centrum characters (with all characters unordered). A. Strict consensus tree of eight equally parsimonious trees, showing character changes. Underlined numbers indicate reversal to primitive state. B. Manipulated consensus tree topology for Carcharhiniformes, resulting in 4 additional steps. 117 00 e e . o6 09 {on 00 (0‘1“) 0'65 95600 09°00 0° {9° {9‘0 i‘o 9‘0\ «Site \°° 0Q \0°° 66‘“ 80 00°) 0690 0° ‘8’ 4‘ 0° x\° 0° 0° 0&9 Q9806 1 9-12 23 78-80 65, 66 77 62-64 27-42, 55 A 9-12,17-20 6 O . 09 09 .9 {@006 \(‘05 \0 0 0 Q's ~1\‘°{o 48° if 006“ «‘9 «5'0 050M ‘69 {0&0 9° 4° 4"" 0° 0° 0° 0° Qfo Q2“) 9°62? 0° Li fl 44.4 .8. 6.1 51 ' 121, 9“ . 6,23 12, 88-89, 101-103 11, 100 85-87 1, 13, 15-17, 44, 45, 47, 50-65, 78 4, 5, 9,10,14, 40-43 7, 8 B 18, 19, 3235 Figure 21. Cladograms showing relationships among Carcharhiniformes using complete morphological characters, showing character changes. Underlined numbers indicate reversal to primitive state. A. Consensus tree of two equally parsimonious trees, centrum characters not included. B. Single most parsimonious tree, combining complete morphological characters with centrum characters. 118 biggest differences were in the position of the various carcharhinid genera and Sphyrna. In the analysis using whole body morphology, Prionace forms an unresolved polytomy with Negaprion and Carcharhinus. Sphyrna is still positioned within the carcharhinids, as is Rhizoprionodon. Galeocerdo is the sister—taxon to all other carcharhinids and Sphyrna in this analysis. The final analysis combining centrum characters with other morphological characters produced a single most parsimonious tree with 166 steps (Figure 21b). This tree is nearly identical to the tree using whole body characters minus the centrum characters, with the only difference being that the polytomies are now resolved. Relationships of carcharhiniform sharks using centrum data In the analysis using ordered characters (Figure 193), the three equally most parsimonious trees vary little fiom one another. The topologies of the trees are nearly identical; the only variation in topology is the placement of Rhizoprionodon and Mustelus. Rhizoprionodon shares some plesiomorphic characters with triakids that are not present in other carcharhinids, including the fluted cylinder shape (character 1), absence of pores (character 15, 16) and bulbous basidorsal and basiventral cartilage shape in cross—sectional view (character 21). Rhizoprionodon share some apomorphic characters with other carcharhinids, including dorsal and ventral foramina that do not extend into rims (character 8), the ratio of dorsal interforaminal wall width/width at apices of the double cone (character 12), interior angle of the intermedialia (character 17), and presence of thin diagonal lamellae (character 19). In the consensus trees, Rhizoprionodon is the sistergroup of Mustelus, though other morphological (e.g., Compagno, 1988) and 119 molecular (e.g. Lavery, 1992) data do not support this relationship. Mustelus is normally considered to be a less derived member of Triakidae, while Rhizoprionodon is united with the Carcharhinidae, suggesting convergence of centrum morphology among these taxa. Genera that are typically united under the Triakidae (Mustelus, Triakis, and Galeorhinus) do not form a single clade in the consensus trees, and are therefore paraphyletic in this analysis. Maisey (1984) likewise reports the triakids to be paraphyletic. The relative placement of these genera in the trees compared to the other carcharhiniform sharks is consistent with previous hypotheses (e.g., Maisey, 1984; Compagno, 1988). The Emily Carcharhinidae, as usually defined (e.g., Compagno, 1988), is polyphyletic because Sphyrna is embedded in the clade that is normally classified as Carcharhinidae, and is the sistergroup to the clade containing Carcharhinus and Negaprion. Compagno (1988) suggested a reclassification of the Carcharhinidae where the harnmerheads (Sphyrna and Eusphyra) were assigned to the tribe Sphyrnini within the Carcharhinidae. Naylor (1992) conducted a distance Wagner analysis using protein variations among carcharhiniform sharks, the results of which support the suggestion of Compagno (1988) to combine the sphyrnids with the carcharhinids (Figure 22c), though their exact relationship has yet to be determined. Prionace is usually considered a derived carcharhinid (e.g., Naylor, 1992), but is excluded fi'om the clade that contains other carcharhinids when only centrum characters were used in the analysis. As discussed previously, centra of Prionace are distinct, and lack many of the derived characters present in other carcharhinids. Prionace retains many 120 ,s 4“" iffy 6°93“? ”\9 .o‘} ‘booé‘ 006° (9 9 ‘oQ 49$ 90010999 «’0 fifeéf Q“? of @0‘0" B «@046 919960‘0:;’°‘60 $0 90¢ 0°00 60°64’60“ 968009 0°Q§M0°0°Q°9° C Figure 22. Cladograms showing relationships among Carcharhiniformes using molecular data. A. Majority rule consensus tree 1. B. Majority rule consensus tree 2. C. Distance Wagner tree. Modified from Naylor, 1992. 121 plesiomorphic centrum characters, including dorsal and ventral foramina that extend into the rims (character 8), dorsal foramina width/dorsal interforaminal wall width ratio less greater than 1.0 (character 1 1), dorsal interforaminal wall width/width at apices of double cone ratio less than 0.28 (character 12), and lack of diagonal calcified lamellae (characters l8, 19). The last two characters dealing with diagonal calcified lamellae were coded as absent because no lamellae were present in any modern Prionace centra. Some of the fossil centra tentatively identified as Prionace have diagonal calcifications, but due to uncertainties in the identification, were not included in the analysis. The two unknown taxa, Carcharhiniformes indeterminate A and B, are united with Hemipristis in the analyses, suggesting that they may be either members of the family Hemigaleidae or sister—taxa to that family. These indeterminate taxa are less derived than most carcharhinids based on centra characters. The relationships of Galeocerdo, Negaprion, and Carcharhinus were consistent with other hypotheses (Compagno, 1988; Naylor, 1992) (Figure 22). The consensus trees were mnipulated using MacClade to explore how many additional steps were required to alter the tree topology to one that was more consistent with other morphological and molecular data. Rhizoprionodon was separated from the Mustelus, and Mustelus, T riakis, and Galeorhinus were united in a single, monophyletic clade. This manipulation required two additional steps. Placing Prionace as the sistergroup to the clade containing Galeocerdo, Sphyrna, Negaprion and Carcharhinus added one additional step to the tree (Figure 19b). Finally, moving Sphyrna outside the clade containing Galeocerdo, Negaprion, and Carcharhinus required two additional 122 steps, but was not shown in the figure, as it is not consistent with more recent hypotheses (e.g., Compagno, 1988; Naylor, 1992). The analysis using unordered centrum characters produced eight equally most parsimonious trees, and their consensus trees were very similar to those discussed above. The Strict and Adam’s consensus trees differed only in the position of Rhizoprionodon. In the Adam’s consensus tree, Rhizoprionodon was grouped with Mustelus, while it was isolated fiom Mustelus and included in a polytomy in the Strict consensus tree. The unordered characters produced a few other, minor differences fi'om the analysis using ordered characters. In this second analysis, Triakis and Galeorhinus were united. Hemipristis was no longer united with the two indeterminate carcharhiniform sharks. The most notable difference was the placement of Prionace in this analysis. Instead of being placed as the first outgroup to the clade containing Hemipristis and the indeterminate carcharhiniform sharks, Prionace was placed within the clade containing Galeocerdo, Sphyrna, Negaprion, and Carcharhinus, more consistent with the hypotheses of Compagno (1988) and Naylor (1992). Because the resulting trees fi'om the two analyses using centrum characters are so similar, it is difficult to indicate whether ordered or unordered characters performed better. These consensus trees were manipulated in the same manner as the previous analysis. Removing the polytomy containing the outgroups Scyliorhinus and Haploblepharus required three additional steps. Placing Mustelus in the clade containing Triakis and Galeorhinus required no additional steps. Removing Rhizoprionodon from the polytomy required one additional step (Figure 20b). Placing Sphyrna outside the 123 carcharhinids added two steps but was not shown in the figure, as it is not consistent with more recent hypotheses (e.g., Compagno, 1988; Naylor, 1992). Relationships of carcharhiniform sharks using all available morphological characters The results of the analysis using whole specimen carcharhiniform sharks, minus centrum characters (Figure 213), produced two equally parsimonious trees that are very similar to one another. The only difference between the two trees is with the placement of Mustelus. In the first tree, Mustelus is part of an unresolved polytomy with Scyliorhinus, T riakis, and the clade containing the remaining taxa. In the second tree, the polytomy contains Scyliorhinus, Triakis, and the clade containing the remaining taxa, with Mustelus as the sistergroup to all taxa minus Scyliorhinus and Triakis. The Strict and Adams consensus trees are identical to one another, and are also identical to the first tree discussed above (Figure 21a). This tree topology is similar to the topologies of trees using centrum characters in rmny ways. The three genera typically grouped together as Triakidae (Mustelus, Triakis, and Galeorhinus) do not form a distinct clade, making Triakidae paraphyletic. Hemipristis is the sistergroup to all of the carcharhinids and Sphyrna. RhiZOprionodon and Prionace are now included with the other carcharhinids, and Sphyrna is the sister group to the clade containing all of the carcharhinids except Galeocerdo. Of interest is the crown polytomy that includes Prionace, Negaprion, and Carcharhinus, indicating less clarity in the relationships of these taxa based on the whole specimen clmracters than analyses including centrum 124 characters, but consistent with the hypothesis of Naylor (1992), who also hypothesized that these three taxa form an unresolved polytomy. Tree topology was manipulated using MacClade for exploratory purposes. Grouping Triakis and Mustelus together required no additional steps. Adding Galeorhinus to the other triakids required four additional steps. Finally, moving Sphyrna outside the other carcharhinids (outside Galeocerdo) added one step to the total length of the tree. The analysis tlmt included the combination of centrum characters with all morphological characters resulted in a very distinct, perfectly pectinate tree (Figure 21b). As in previous analyses, the three genera typically normally grouped together as Triakidae (Mustelus, Triakis, Galeorhinus) do not form a distinct clade, making Triakidae paraphyletic. Hemipristis, being the only representative of the Hemigaleidae, is not grouped with any other taxa. This analysis once again suggests that Carcharhinidae, as normally defined, is paraphyletic. Sphyrna was included within the cluster of genera normally attributed to Carcharhinidae, supporting the suggestion of Compagno (1988) and Naylor (1992) to unite the Carcharhinidae and Sphyrnidae into a single family. The inclusion of the other morphological characters in the analysis produced a more customary placement for Rhizoprionodon, as it was once again included within the clade that includes Carcharhinidae and Sphyrna. The previous Prionace, Negaprion, and Carcharhinus polytomy was resolved when centrum characters were included in the data matrix. The tree was manipulated using MacClade to explore how many additional steps were required to group the three triakid genera together. One additional step was required 125 to group Mustelus and Triakis, but seven additional steps were required to also include Galeorhinus. Galeorhinus is generally considered to be the most derived triakid of the three included in the analysis (Compagno, 1988), and has several centrum apomorphies that distinguish it from other triakid genera. These apomorphies include a medic—lateral breadth approximately equal to dorso—ventral height (character 4), a centrum length/diameter ratio between 0.6 and 0.95 (character 5), dorsal foramina length/diameter ratio less than 0.57 (character 9), dorsal foramina length/dorsal foramina width ratio less than 3.0 (character 10), and ventral foramina length/ventral foramina width ratio less than 2.5 (character 14). Sphyrna was placed as the sistergroup to the carcharhinids, a move that required three additional steps. CONCLUSIONS Identification of carcharhiniform verteme centra Vertebral centra fiom the ten genera discussed in this study have distinct morphologies that allow them to be identified to the generic level. In most cases, examination of external morphology alone provides enough information for identification. When examination alone is not sufficient, measurements can be taken and applied to a discriminant analysis to predict identification. Discriminant analysis is useful for predicting the group membership in naturally occurring groups for cases of unknown membership (Klecka, 1980; Tabachnick and F idell, 1983). The present study corroborates that of Kozuch and Fitzgerald (1989) who successfully used vertebral centra to identify sharks to the specific level at archaeological sites. 126 The morphology of the centra was not, however, distinct enough to identify specimens to genus in every instance. The centra of Negaprion and Carcharhinus have nearly identical morphologies. Most centra tint fit the description of these genera are labeled as Carcharhinus simply because of the high diversity and common occurrence of this genus. Without teeth to aid in identification (e.g. UF 32451—32455), the identification of a centrum as Negaprion or Carcharhinus cannot be entirely certain. Because only one specimen of Negaprion with two centra was sampled in the study, a larger sample size may reveal additional differences useful for distinguishing the two genera. Diplospondylous caudal centra are also very difficult to identify. These centra are proportionately much shorter than MP or DP centra, and tend to converge towards a cylindrical shape and lose morphological complexity as they become smaller among all carcharhiniform taxa. The pore distribution is still distinct in these centra, but is not enough for identification of the centra without other characters. The six fossil centra tentatively identified as Prionace are significant because this genus is not currently known from Neogene deposits in North America. Currently, this genus is only known in the fossil record fiom the Pliocene of Italy (Landini, 1977). If these fossil centra are indeed fiom Prionace, it would indicate a significant geographic range extension. Prionace is found in oceans worldwide today (Compagno, 1984), and their presence in the Neogene of North American should not come as a surprise. Because Prionace glauca is a pelagic species, they are not as likely to be recovered as fossils. Unfortunately, Prionace centra were available fi'om only one individual for the present study, and more comparative material is needed before any conclusions can be drawn. 127 Taxonomic value of shark vertebral centra Carcharhiniform vertebral centra are also useful in a phylogenetic context. The results of the first two cladistic analyses based on centrum characters alone (Figures 19, 20) were very similar to published accounts of carcharhiniform phylogeny (Figure 22). A few taxa had positions within the Cladograms that were contrary to previously published hypotheses, but in all of these cases, when the trees were manipulated to resemble the other hypotheses, the resulting trees were only a few steps longer. Despite the paraphyletic nature of Triakidae, their relationship to the other carcharhiniform sharks was similar to the hypotheses of Maisey (1984), Compagno (1977), and Compagno (1988). Triakids are considered to be the sistergroup of the clade containing Hemigaleidae, Carcharhinidae, and Sphyrnidae. The paraphyletic nature of the three triakid genera studied in this analysis may not be problematic, as it required very few additional steps to combine them into a monophyletic clade (2 steps in the analysis using ordered centrum characters, and no additional steps in the analysis with all centrum characters unordered). Rhizoprionodon is considered a carcharhinid (Compagno, 1988), but has usually been grouped with Mustelus or as part of an unresolved polytomy with the outgroups, triakids, and the clade containing the remaining taxa. This relationship was not entirely a surprise, however, as Rhizoprionodon has centra very similar to triakids in a number of characters, as discussed above. The habitat and distribution of Rhizoprionodon is not significantly different from that of other carcharhinids (Conrpagno, 1984), so it is unlikely that the vertebral differences from other carcharhinids are the result of 128 environmental efi‘ects. It is, however, unclear whether the differences are the retention of plesiomorphic characters or reversals. The usefulness of the diagonal calcifications as characters in a phylogenetic analysis needs to be fiirther examined. While some authors have discussed apparent trends in these calcifications (e.g. White, 1938; Nakaya, 1975) (Figure 5), it is apparent fi'om this study that the diagonal calcifications vary within a single genus or even an individual. A separate analysis of characters derived fi'om whole specimens, minus centrum data, was conducted to compare with the analyses based on centrum data alone. The resulting consensus tree (Figure 21a) had Rhizoprionodon and Prionace placed in a manner consistent with other hypotheses (e. g. Naylor, 1992). Triakidae was once again paraphyletic. The consensus tree resulted in two unresolved polytomies. The first was at the base of the tree that included Scyliorhinus, Triakis, Mustelus, and the clade containing the remaining taxa. The second polytomy included the three most derived taxa, Prionace, Negaprion, and Carcharhinus. When centrum data were added to the matrix containing characters from complete specimens, the resulting tree changed very little. The addition did result in clarity, however, as the polytomies became resolved in the final solution. Compagno (1988) and Naylor (1992) suggest a re—classification of the hammerhead sharks (genus Sphyrna) as the tribe Sphymini within the Carcharhinidae based on both morphological and molecular data. The hammerheads have traditionally been assigned to their own family (Sphyrnidae), but the results of the phylogenetic analyses presented here create a polyphyletic Carcharhinidae. Based on the addition of centrum data, I suggest the cladistic re—classification of the hammerheads be adopted. 129 This re—classification would not affect the discriminant analyses using GENUS as the grouping variable, but would reduce the number of groups from three to two in the analyses using FAMILY as the grouping variable. Vertebral centra do carry a phylogenetic signal. This hypothesis is supported strongly by the results of the several phylogenetic amlyses performed above. In the two analyses using only centrum characters (Figures 19, 20) the topologies of the trees are very similar to other published hypotheses (e.g. Compagno, 1988; Naylor, 1992), (Figure 22), including the inclusion of Sphyrna within the clade containing carcharhinid genera. The hypothesis that vertebral centra carry a phylogenetic signal is further supported by the added clarity centrum data supplied to the analysis using whole body morphology (Figures 21a, b). In the analysis using whole body morphology without centrum data, several unresolved polytomies appeared. By adding centrum data, these polytomies became resolved, and resulted in a single most parsimonious tree. While Hasse (1879—1885) and Applegate (1967) tried to create a series of classifications based on the cross sectional pattern of vertebral centra, no evidence was found for a distinct morphotype that could readily identify a particular genus or family in this study. Hasse (1879—1885) and Applegate (1967) may have been too optimistic about the value of the calcification patterns, but their work, along with Ridewood (1921), has laid an important foundation for studying shark centra. When distinct elements within the centra are examined separately (similar to Compagno, 1988), and coded as individual characters, the use of centrum morphology becomes more apparent. The use of centrum morphology becomes even more apparent when coupled with all possible shark ' morphological characters. 130 It is obvious that more Recent comparative specimens are needed for future analyses. The results presented in this study only include ten genera fiom four families, a small fraction of the nearly fifty genera of carcharhiniform sharks. In addition to incorporating additional taxa, centra fi'om more individuals and from different regions are needed to insure an appropriately large sample size is available. Additionally, a more comprehensive selection of outgroup taxa, perhaps including basal lamniform sharks, would benefit future analyses. Future work Many possible areas for firture research logically follow the research presented here. Fossil shark teeth have been studied extensively, but associated centra are often overlooked. Purdy et al. (2001) studied the shark fauna of the Lee Creek Mine. They did identify the centra fi'om three different genera of carcharhiniform sharks, but there are many more left unidentified. Twelve carcharhiniform genera fi'om the Lee Creek Mine were identified based on teeth alone. Additioml centra may be identified to genus upon a more thorough examination of these centra, thus increasing the known diversity in the fossil record. Three regions have been identified in the vertebral column of sharks, the monospondylous precaudal zone, diplospondylous precaudal zone, and the diplospondylous caudal zone (Compagno, 1970). The three identified regions have never been studied in a quantitative fi'amework, and the nature of the variation between and within the zones is poorly understood. In addition to variation of centrum proportions, the internal calcification patterns also vary with position. The proportional variations within 131 the column of a single individual can be better understood using geometric landmark morphometrics. X—radiographs of centra from the same individual will allow the nature of variation in calcification along the column to be studied. The minor variations between adjacent centra will be revealed as well as broader patterns along the entire column. These data will be compared to other sharks of the same species to compare consistency of the positional variations. Similar studies quantifying shape change using geometric rnorphometrics can be conducted on a single species at different growth stages. Calcification of the vertebral centra increases throughout the lifetime of sharks, though the nature of this increase is poorly understood. A better understanding of the changing calcification patterns will improve the use of fossil centra for phylogenetic analysis. The data gathered may also be important for firture studies to interpret the relationship between centrum morphology and swimming characteristics in extant taxa. Bucholtz (1998, 2001) has identified a correlation in regional variation of centrum morphology and swimming mode in cetaceans. Massare and Sharkey (2003) and Kajiura et al. (2003) have conducted some preliminary work on the role of shark vertebral centra in swimming styles. Massare and Sharkey (2003) examined the variation in length, width, and height of vertebral centra with their position in the vertebral column of five sharks (four carcharhiniforms and one orectolobiform). They found considerable variation in the pattern of centrum shape, without identifying what the differences imply for swimming styles. Kajiura et al. (2003) show that there is no relationship between the number of vertebrae and the ability to bend the body laterally in juvenile Sphyrna tiburo, S. Iewini, and Carcharhinus plumbeus. Because carcharhiniform sharks increasingly calcify their 132 vertebral centra, the relationship between number of vertebrae and flexibility may be different for adult sharks. Vertebral proportions, regional variations, and degree and pattern of calcification, joined with counts of total vertebral number, can be used to evaluate the vertebral contribution to body flexibility and its role in swimming styles for carcharhiniform sharks. This research can serve as a model for future studies among other clades. Fossil shark centra are regularly preserved fi'om other groups, including Lamniformes, the sistergroup to Carcharhiniformes. The research presented here can serve to establish criteria and a methodological approach for including centrum morphology in future morphofunctional analyses of lamniforms and other chondrichthyans. The results of this study support the hypothesis that carcharhiniform verteme centra are morphologically distinct, identifiable, and useful in interpretation of phylogenetic relationships. Because of the well—documented difficulties with identifying isolated fossil shark teeth, shark vertebral centra should be utilized whenever they are available. The combination of fossil shark teeth and vertebral centra will provide a more rigorous means for studying fossil sharks than teeth alone. 133 2 .m nod 2.2 2 no.2 an.2 2 «.2 20m 2 0020:0200 002.220.2020..» ovom 22m 22222.42 SN mod 3.2 mnd 22.2 2.2 2.2 320220.200 352220220200 anon 2222a :22)? on.~ 2.o 2.2 od n.n R2 2 om.212 30:02.00 3220:0209 2m.o22~ =22< Nm.m A062 2 on.n~ 2.2 22.2 monm ooom 2.2.2.2302 o vnéoooo =22< 2 .m 2N.n2 noon nn.om 2.~N 222m omdm 2.5.2.02 0.2.3 o 2.083 2.2232 o3. 2~.n2 moon anon oodm anon nmdm 2.25202 0.2.3 o voéoooo 2.22%.? So no.2 mm. 2 m 3.2 2.2m ondm oodu 2.25.30 x: R... . 8.083 :22)? mod 22 m 2.2” 2.o~ 2.2 ooom 222m 2.25302 05.3 o ooéoooo 2.22—>2 voé 2.2 vodm 2 2.2 no.2 onolm. 2n.om 2.25.302 3.5.... . mméoooo 2222)? v.2. 2 2.22 moon 3.: 2.2 2.8 Nn.2 0200 0.2 o ooooooo $222.2 $2. non oou no.2 2.2 2.o~ on.2 000% 0.2.3 o 2.223% 2.22—>222. 22. mod no.2m mn.n2 2.2 nmom vow 000020 3...... . undoooo 2.222)? 2 .2. on.” anon 2.: 2 2.2... 2 vn.2 oo.oN 0220020 05.3 o 2833 2.222)? and non mmom oo.2 2 222.2 on.2 0:002» 3:2,... . 2.2233 2.222“. no.2” ovn «N2 mo.n2 22Io.2 2 2.2 om.2 0.200% 05.3 o 22.2233 2.22222. nm.m 2 .m no.2 3.2 2.22 20m 2 ood 2 20.2.50 0220000206 «223$ :23? mm.m ooé 2.2 $2 no.» 8.2 32.2 20.2.50 022000025 3:253 :23? mod 26...“ no.2 on.2 on.» 2.2 2n.2 .20.:50 02020000225 mmovooo 2222)?. ad woo 2.2 2n.2 no.2 vn2 2.2 20.050 02020000206 omdvooo 2.222)? nw.m moo 2.2 em .2 no.2 2N.2 nm.2 20.200 0220000225 2.253 2.22220. nn.m no 2 .2 32 mod no .2 on.2 20.2.50 020000206 202.com 222.4. n06 mmo 22.». 2 on .2 no.2 52 no.2 20.2.50 020000206 mmdvooo 2222.0. no.2. nmo 222.2” 2 22 mod mn.m 2 no.2 20.2.50 00.2000020D 22.263 2.22—>24... on.m own oo.m 2 $2 2 mod on.2 Ed 2 20.2.50 0220000209 22.25% 2.27232 2n.m nod 2.2 no.2 2 clad wn.2 2M2 20.2.50 0220000200 omdwooo g 20.2 8.». own 36 m 2 .w 2..» Na $230.22 3232920200 nmovwmo 2.2% 0.2 ow.m non moo woo no.» won 022.. 50.22 32220220200 2.2.26 :23? 2N.2 no.2“ v.n 2.6 omo no.» 2n 0522050222 35.202200 2.2.26 2.222222. no .2 2o.m vno Woo 2N.o 2.n vmn 022.2 50.3 32.222.820.200 govwmo 2.2% no.2 mod on.2 no.2 3.2 2.2 3.2 0.322 gmflmm. 2.2.23 E 82 on.w 22.2 no.2 no.2 no.2 vm.2 0.522.: 022.3%. 3:2.me 2.2% 2.2 m.n olo.o2 22.2 2.22 mh2 onr.2 03% 053.22%. 3.2.23 222 3 EEO.”— Q .2 EEO“— Q 52.22 2903 E28 8.08 Emcou x05 n 2020805 2 Evan—D 28582252205 m EVE—gm— 300805802 05000 232m 2:082 - E < 52522.2(. 134 $22082:ng 0.222200 232m 2:832 - $2250. non non nod 22.2. ov.o22n E man an no.» woo. om.o22n E and end nm.o m 2m.o22n 2.22—>3. no on.2. on.2 2 .o vnéoooo :25 nn.m 20.2. on.2 0n.» 3.2.250 :22)? on.2” n22. on.2 and 3.253 EZE< on.2” n22. om.n2 no 8.1228 2.22232 2nd o2. 2&2 ooo ooéoooo =22Z< non o 2. m2 No.» moéoooo 2.222)? no.2 n m nn.» no.2. oodoooo 2.2% on.2 n.m 2n.» no.2. modoooo 2.222)? 2 .n o m ood ooo nmdoooo 2223?. SN L 22 m no.» 2.6 2.233 222.0. an ood on.» on.2. nvfioooo 22$? 2o.m nm.m omd on.2. nmdmcoo 2.2232 3.2” om.m $.20 nm.m «magma 2.2% 2.2. Wm moo wm.m 2.2233 2223?. o2. 2....” mod wmd 3.253 2.2% no.2. and no.o mn.m 3.233 2.2722)?V on.2. on m 36 22.2” ovdvooo IZE< 2.2. 2 .v om.o on.m 20.26% 2222.0. o2. n.m 2o.o wnn 3.233 2279?. on.2. no.2. and 2o.n 3.25% E 2.2. no m 2.o non 3.263 2.2232 2.20 2o m nloo 2n.n ondvoma 2.2233. :2 mn.2 no.2. on noovwmo 2222.4. o.2 mon mo.2q mod 2.633 2222‘ on.2 2.2 22. no.2 ovovwmo 52S? non no.n on.m onn 3.2.26 :22)? o.2 3.2 3.2 22.2. ndewmo 2.22.53. on.2 2.n no.2 2 22.20 22.2223 2.22222. on.2 onn. 26 3.2. 8.2.23 2.223? B :03 o B Egon o u Egon D 8 :03 D mcogmfl < 2222.52.30 135 no a; 3% 3.9. 3.: 33 2.3. 3.. Egg £§§G 383 m5 2; m; USE a? 3mm 3% 2.9. a? .530 QESNNS «.358 m5 «3 :2 2:: 02: on.» NZ: 3: 3.: 9.qu :8 323: 8.82m 95 is. an. 2.: and mm.» 3: a: 8.: 335%?» 323% ”382 m5 3.: m3 as 2.2 § 2: 3.: 2.: 258: .Eaafié 8.2me m5 3 3 a: m; 3 3.: m2: 2.: 555:5” 33: 3.58 20 mm a.“ m; n: 3.2 «.2 8.: a: 33.38522 $3: 8.2% 36 8a 3..” v; 8.: 2:: a: 3.: 8.: 23.38522 33: 2.3%” m5 m2 N2 8.” a? a.» 3.: 2.: 8.: 2938an £35 2.2me m5 3 2d 8.” m; an.» 3.: x: 3.: 2938552 33: 5.3%” m5 «3 3a 3 3.2 8.2 8.: 9.: 5.: 23.83528 $3: 3.3me m5 :3 m3 3.” 9.: 8.2 3.: 2.: 2. 85.585525: 2683 m5 and :3 3.2 an: 3.2 3.: <2 8.: «3.38558 3%: 2.3% m5 :3 :3 Wow: 3.: :2: $2 8.: 3.: ssuésafist 2.3me m5 a: 3m 2% 8.0 no a.» as as 2:2 o 3.3338 2&an m5 .2 :..N 3. 8.0 Rd ”E :k NE 3239 33828 $.2me m5 m2 2 n2 3 2.0 .3 as as 58399 35.185 ”~8me m5 2 N.“ E K. S 3.0 2 8s 3239 $5.283 835 m5 2 E 2:. b 3.0 3 3 3 55:39 323338 885 m5 2: E 3 8s 3 86 3 3 353% 33338 3.2% m5 3 N 8... 8s 3 m3 3 as 25399 3.3638 83% m5 .1 3 3m 2 .5 m8 2: 3.” 3 25399 33338 835 m5 2 m: S E S n: m; 2.” 32% 35:85 8.8%” m5 2: m3 3 S “no he 2. 2 222:3“ 3:338 5.8%” m5 3 :35 a a see“: a A SE a 5%: as? 523 as? ems: E: N 53855 : $.55 8:82:52 3: sammm mufiuaogoz “.5600 Mada—m «convex n m»> 3895m> 48a8m> #5 . m 35825802 «:50 zap—m Booed - m<0 137 2.9 2.: 2.2 N2 m2 :2 2.2 B. 5 K .298. N 8.38 .38: .0 8... 8... 2.2 R2! 2. : MN FNIN .228. 222328. 2 8.3.8 .32.: .0 2... 8.2 2.2 2.2 2... 8.2 8.2 2 8.3.8 .38: .0 2... 2.... 8.2 2.2 m... 2.2 8.2 8. o2 .2. .298. N 8.38 .32.: .0 8... Bo 2.2 P2: P2... m2 P22 .23 3:888 F 8.3.8 .33: .0 2.2 8.2 2.2 8.2 2.2 2.2 8.2 2 8.3.8 .32.: .0 8.2 m... 2.2 9.2 2.2 2.2 8.2 8. 8 8.... .298. N 8.3.8 .32.: .0 8.2 m... 2.2 8.2 82 ON 2|.PN 3.3:- 35888 2 8.3.8 .32.: .0 2.2 2.2 2.2 2.2 2.2 8.2 8.2 2 8.3.8 .32.: .0 2.2 2... 8.2 8.2 29. : 2.2 3.2 8. 2 6.... .298. N 8.3.8 .32.: .0 2.2 2... :2 m2 Nla... E as... 3.3.5 228580 F 8.3.8 .32.: .0 n... 2.2 2.2 2.2 2.2 8.2 2.2 8. SN .22 .22... N 8.3.8 =3..: .0 2... al.: BR 8.8 2.2 E 2|...N 32.... 32.588 2 889.8 .32: .0 2.2 to 9.2 3.2 2.: 2.2 3.2 2 8.3.8 32.: .0 2.2 2.2 8.2 2.2 2.2 ... 2 2. 2 2.. 8... .5... 8.8. N 8.3.8 .32.: .0 8.2 2.2 2. 2 NI... 2 4.2.2 aim. 2 N.NI2 880. 3:22.28 . 8.3.8 3%: .0 8.2 2... 3. 2 8.2 2.2 2. 2 2.2 m 8.38 .32.: .0 2... 2.2 2. 2 2. 2 2.2 2. 2 8. 2 9 8.3.8 .32: .0 N3 8... 2.2 8.2 2.2 2.2 2.2 2 £38 3%: .0 2.2 2... a. 2 o. 2 2.2 8. 2 8. 2 3. 22. 6.2 .998. N 8.3.8 .32: .0 8.2 m... 2.:I2 E B: 2.2 2. 2 32.2. 2:522:28 , P 8.3.8 .32: .0 2.2 :2 2.2 2.2 22 N2 2N. 2 .. 8.3.8 .32.: .0 2.2 8... 2.2 2.2 9.2 B. 2 :. 2 B. 2: .8 N 8.3.8 .32: .0 2.2 9.2 2W2 9.8 2.2 92m E 85.22 22:38 , F 8.3.8 .32.: .0 2.... 2. : 3N 8.2 2.2 2.2 2.2 v 8.3.8 .32.: .0 5.2 8. : 2.2 2.2 9.2 2 2.2 2 8.3.8 .38: .0 2.2 2.2 2.8 2.2 8.2 2.2 8.2 B. 22 .22 .22. N 5.8.8 .32: .0 9N2 JN|N2 2. 2 2k 2'2 NI...8 SM. 253.... 32:38.8 . 8.3.8 3.2.: .0 22 2.2 2.2 2.2 2.2 2.2 3.2 53 8 N 8828 .32.: .0 8.2 om. Bab 8.2 NI2.2 2.2 3N. 3......» 3:52.98. 2 8.3.8 .38: .0 SN 2... 8.2 2.2 2.2 2. : 2. : 2 8.3. .33: . BN 2... 2.2 R... 2.: 2.: 8.: 382 N538_.o2...: .0 L22 Rm. km 85 MN: mm: o. : 2:8 3.22.. .. . 2,... 2... 8.2 .2 8. : W2 9.2 2.... 85238 2 8.3.8 .38... .0 3 2.2.... o 4 228 a 52.04. .223 22.3 2.23 59.3 .32 N 5.25.0 . .2255 8285.8... .. 82.8 m 220805302 .2800 xbfim «2000M - 20:00:00 .3325 .O 138 8.... 5... 5. 3 m: B. 5 K .88. w 8.3.8 .32.: .0 29 P2. «In... B... .38. 2588.8 . 8.82.8 .32.: .0 a... a... 8.0 8... n 8.82.8 .32.: .0 a... B... 9.0 no... 3. O9 6... .228. N 8.3.8 .32.: .0 8.. Po... al.... elm... .83 35.888 . 8.3.8 .38: .0 8... N... 3 .3 n 8.3.8 .38: .0 no. a...” 9.0 8.0 B. 8 .8... .88. u 8.3.8 .38: .0 9.9 9.” olox Rm 8.3.... 2.55.88 .8828 82...: .0 9.0 8.” on... on... m 8.3.8 3%: .0 8... a... 9... 9.0 8. m. 6.... .228. u 8.3.8 .32.: .0 o... m... a: ale... 3.3:... 2.25298 .8828 .32.: .0 8.2 mu... 1. x: B. 8... .32 .22.... N 8.3.8 .32.: .0 9.: W3 3 B... «82.. 3:858 .8828 .32.: .0 9.8 8... 9.9 «3 n 8.3.8 .32.: .0 2.3 3.0 8.9 no.“ 8. 8n .8.» .298. a 8..8..8 .32.: .0 9.9 E KP. a... 83... 8288.8 .8818 .32.: .0 3.9 «no 88 o: m 882.8 3%: .0 8.9 .80 8... «2 v 8.3.8 .32.: .0 8.8 no 38 9.» o 8.3.8 .38: .0 2.9 N... 8.8 an... 3. B... .8... .22... a 8.3.8 822.: .0 8.9 a... 9.... h. "8.3 85888 .8828 .32.: .0 3.9 3+ 3... 8.... v8.uo..oo..o£..: .0 z. z 2... mm... :6 3. o: .8 N 8.3.8 .32.: .0 9.9 5... 5.9 «be 888.... 322828 . 8.3.8 .32.: .0 2.9 9.... 8: 6... «8.3.8 3%: .0 3.9 5+ 89 88 o 8.3.8 .38: .0 8. : «an 9.9 6.... 3. 8m .8 .22. m 8.3.8 .32.: .0 8... mm 9.9 Ba E88858 .8828 .32.: .0 8.» S 9... 8... 232 u 8.3.8 82...: .0 8.8 ole... «Imps v3 35.... 82:88.8. . 8.3.8 3%: .0 «3 5m 8... 8a m 882.8 .32.: no... 3a 89 9a 22. 8 a 8.3.8 .38: 0 Ba 38 RE 3 8.8.. 28.2.: .r... . 8.. E... mm... 9... 95... 85808 g 3 .35 > B 5.8 > 1. see. > 3 .33 o 8..8..|..82 .. 8.... .m 320895082 «.550 #58 9.80% - 5.80:0“. =anm .O 139 000 P9 MIN... Jaw E 90H... 0h}! 88.3. 9 808.8 .32.: .0 3.0 00.0 00.00 00 00 00.09 3.00 0.00 0 8.80.8 .32.: .0 00.0 00.0 90.00 x. 00 00. 9 .000 00.00 0 808.8 .300: .0 00.0 00.9 00.00 0 00 90.9 9.00 00.00 0 803.8 .32.: .0 00.0 00.0 00.00 90 00 00.9 00.00 00.00 0.0 0.0... 0 5.8.8 .300: .0 00.0 0I0.0 9B0 0:0..0M 9. 9 W00 0I0.00 88.0 805m 9 808.8 .32.: .0 00.9 00.0 00.0 00 0 00.0 8.9 9.0.9 0 803.8 .32.: .0 00.. 00.0 00.0 90 0 00.0 00.0 00.0 0.00 8 0 808.8 .32.: .0 9... 0.0 0I0.0 N 0 00.0 «w. E 80282.0. 0888805.. 9 808.8 .32.: .0 00.0 00.0 00. .0 00 9 9. 9 00.00 00.00 0.00 8 0 8.3.8 .300: .0 00.0 $0 00k 0|00|0 9.9 00. 90 00. .0 0.002.520 5.0302 9 808.8 .32.: .0 00.0 00.0 00.00 00.00 90. 9 .000 00.00 0 8.3.8 .300: .0 9.0 0.0 00.00 00 00 00. 9 .060 9.9.0 0 8.80.8 .300: .0 00.0 90.9 00.00 9. 00 90.9 00.00 00.00 B. 9. 0.9.0 0.0:. 0 8.80.8 .32.: .0 00... 0I0.0 9+8 0R0 00.9 § 00.00 3.80.0 00.020028 . 808.8 .300: .0 00... 0.9 00.00 00 .0 9.9 0.00 .090 0 8.0.2.8 .32.: .0 9.0 00.9 00.00 00 00 00.9 0... 90 0. .0 0 808.8 .300: .0 00.0 8. 9 00.00 .0 00 00.9 9. .0 990 0 8.0.0.8 .300: .0 00.0 00.0 00.9 9 9 00. 9 00.0. 00.9 .00 0.0... 0 808.8 .300: .0 .0... 0I0.0 E0 0E0 0.9 0&0 mm. .0 00028.0 00.0.0088 . 8.0.0.8 .300: .0 00.0 00.0 00.9 00 9 00.0 9.9 00.9 v 8.3.8 .300: .0 00... 9.0 5.9 00 9 0.0 00.9 00.9 0 8.0.0.8 .32.: .0 00.0 00.0 00.9 90.9 00.0 00.9 00.9 ,9 9.0 0.0... 0 8.00.8 .32.: .0 00.0 0.0 00.9 00.9 W00 0I0.0. 0I0.0. .980 00588.8 9 8.3.8 .300: .0 00.0 00. 9 9.00 90. 90 00.9 00. 90 00. 90 0 8.80.8 .300: .0 00... 0.0 RN v0.m.0 0I0. 9 00. 90 00..0..0 0 8.3.8 .300: .0 9 .850 a 0 50.00 a .00.... 0.03 20.9 0.03 .0080 .32 0 8.28.0. .9085 80800.8... 0 88.800 000080050002 09000 383m 0.008% - 00:00:00 zonnam .0 140 00.0 0.00 0N0. 0I0.0 00..0..0... .. 00... 00.0 00.0 00... 00.0 00.0 .00 00.0 00.0 90.0 3.9 00.0 0.0 00.0 00.0 9... 00 0 00.0 00.0 9.... 00.0 00.0 00.0 00.0 .00 00.9 00.0 00.0 00.0 E 9.0 mm 800808. 0880.30.00. 00.0 00.0 00.0 00.0 9.0 9.0 0.....0 9.0 0.530020 8.80002 00.0 0... 00.0 00.0 00.0 00.0 00.0 00.0 00.0 00... 00.9 9.0 3. 09 0.9.0 0.0... 00.0 00!.0 WW0 £0.01 20000.0 0020.058 00.0 00.0 00.9 00.0 9.0 90.0 9.0 00... 00.0 00.0 00.0 00... 00.0 .00 00.0 00.0 $0 0.0... 9.9 3.. 00.0 0.0.0 0088.8. 2.2.0.088, 9.0 9... 00.0 00.0. 00.0 02. 0.0 00.0 00.0 00.0 00.0 90... .9 9.0 0.2. 9.0 mm... 0I0.0 0I0.0 .003 005.8028 00 0 00... 00. 9 00... 9 0I0.0 09.0 00.0 3 :03 > B 0.0.0.. > 1. .0900. > 3 :03 n. 00.08%..80. 000900050002 09000 madam 9.000% - 0098:00 =3n=m .0 141 00.0 00.0 00 9 9.0 90.0 00.0 00.0 33% 00.0 00.0 3.9 0.0 00.0 v0.0 00.0 0.900830: 2:ka m 22.959 0.002823 .4 00.. 09.0 9.9 9.0 00.0 00.0 .00 20:0on again . 8.3200. 2008...; .4 00.. 90.. 0.0 00.0 00.0 N0 0.0 330% «0.9 0.9 5.0 00.0 00.0 «0.0 0.0 2.202.. 0:30:50 .0 : 22.080 0.002233 .4 $9 00.. 00.0 00.0 00.0 00.0 00.0 2.002. 02.2.8... .0 0. 00.380 20022.23 .4 00.9 00.0 00.0 00.0 00.0 0.0 00.0 0:202. 6:60.000... .0 0 22.0200 0.002233 .4 00.9 00.9 00.0 00.0 00.0 90.0 .00 2.002.. 02.9.8... .0 0 00.3200 20028.23 .4 00.9 0.9 00.0 00.0 0.0 00.0 00.0 2.52. 02.0202... .0 0 20.380 2008......» .4 v.0 00.0 00.0 00.0 00.0 00.0 00.0 0:202. 09.0202... .0 0 20400200 04002823 .4 0... 00.9 00.0 00.0 00.0 00.0 00.0 o=co>a 53.8202... .0 m 2090209 04002233 .4 9.. 00.. 90.0 00.0 90.0 00.0 00.0 0:55 02.02:... .0 0 20.00200 2008...; .4 v0.9 5.. 00.0 00.0 8.0 8.0 8.0 0:202. 69.00.02... .0 n 22.260 0.002823 .4 00.. 00.. 00.0 00.0 0.0 .00 00.0 2.202.. 02.02:... .0 0 09.080 2008...; .4 00.9 00... 8.0 0.0 «0 8.0 2.0 0:52. .0325... .0 4 2090009 04002233 .4 00.. 000 0I0.0 400.0 00.0 00.0 00.0 [0.33.05.00.28 42% 00.. [r 9.0 00.0 00.0 00.0 00.0 00.0 02820:... 00.05.020.00 3 22.0200 8009.23 .4 00.9 00.0 00.0 09.0 00.0 00.0 00.0 02.02:... 0:550:28 0. 00.00200 4820......» .4 00.9 00.0 0.0 9.0 .0... 00.0 9.0 02.02:... 005.20.00.00 0. 20.8200. 808......» .4 90.. 0.0 00.0 00.0 00.0 00.0 00.0 02.02:... 0.2.2.2288 : 20.2.80. 0.008223 .4 00.9 3.0 09.0 09.0 «0... 90.0 0.0 052.000.: 03205020000 2 202980 04002323 .4 00.. 3.0 9.0 9.0 00... 0.0 00.0 0902:... 092.29.00.00 0 22.0200. 20020.23 .4 00.. 00.0 0.0 9.0 00... 00.0 0.0 0302:... 03.2.3200 0 20.00209 2820......» .4 00.9 0.0 00.0 0.0 00... 00.0 00.0 03.02:... 0.02.55.09.00 0. 2090200 0.002833 .4 90.9 00.0 00.0 00.0 9... 00.0 00.0 00.02:... 002.220.002.00 0 2020209 20029.23 .4 00.. 00.0 9.0 00.0 9.... 00.0 00.0 00.02:... 00.05.020.00 0 29.00004 0008...; .4 90.9 9.0 90.0 00.0 906 0.0.0 00.0 0:200:00... 02:...00000200 v 20.8200 0.00223? .4 00.. 00.0 00.0 00.0 0... 00.0 0.0 05.02:... 0223.05.00 0 2020200. 20022.23 00.. 00.0 .00 00.0 00.0 00.0 0.0 03020:... 002.00.020.00 0 20.2.80. 20028.3 00.. 00.0 00.0 00.0 .0... 00.0 9.0 0902:... 005.06.00.00 _ 2020009 20029.23 .4 3 2.0.0.. o 4 2.0.0... n. 50.010003 2.0.3 .003 50:04 x02 0 8.02.05 . 8.0.20.0 00.080.02.00. 0 222% 002080020002 05200 imam 020034 - 29:0qu 48:83? .4 142 00.. «N 0.0 3.0 :20on 0.0. .. w 2 22.0200 20023.23 .4 00.. 0.0 0.0 00.0. 200.0on 0E2...” 0 820.52 2008.23 .4 .00 00.0 0.0 8.0 20:0on 02.0220 . 820.62 200% 00.. 8%.. mm. 0.0 2.202.. 00.02:... .0 N. 22.0.80 20028.23 .4 00... .... «0.. 00.0 2.202.. 00.02:... .0 .. 22.0.80 20028.23 .4 00.. v... 00.. 0.0 0:202. 00.02:... .0 2 22.0200 20022.23 .4 S... «0.. 00.. 0.0 0:55 .00.02.:.. .0 0 22.0.80 2820...; .4 0.... u... .0. 00.. 2.2020 00.02:... .0 0 22.0200 20028.23 .4 00.. 0... 0.. 00.. 0:202. 00.02:... .0 0 22.0220 20020.23 .4 0.... «0.2 00.. 0..~ 2.205 00.02:... .0 0 22.0220 20020223 .4 0.... 00.2 0.. .00 0:202. 00.02:... .0 0 22.0.80 20020223 .4 0.... .0... 0.... 0.0 0:202. 00.02:... .0 0 22.0220 200202....» .4 «0.. 00.0 0..... 8.0 0:203 00.02:... .0 m 22.0.80 0.002022? .4 00.. .00 00.. 8.0 0:52. 00.02:... .0 N 22.0.80 2820.51.514 «0.. 00.. 00.. 0.0 0:202. 00.02:... .0 . 22.0200 20028.23 .4 00.. .0. .00. 8.0 00.02:... mg 0. 22.0200 20023.23 .4 .0. .0. 00.0 .00 00.02:... 002502800 .2 22.0.80 2002022... .4 00.. 00.. 00am 00.0 00.02:... 0055020000 2 22.0.80 20020223 .4 8.. 00.. 00.0. 0.0 00.02:... 002502800 N. 22.0200 20022.23 .4 0.. 00.. RN 00.0 00.025... 0025020000 2 22.0200 20020.23 .4 00.. 00.. 00.0 00.0 00.02:... 0022.05.00 o. 22.0.50 20022.2... .4 00.. 8.. 00M 00M 00.02:... 002505.000 0 22.0200 20020223 .4 00.. 0.0.. 0“.“ and 00.025... 005.50.00.00 0 22.0200 20020.33 .4 00.. 0.0.. 3.0 «0.0 00.02:... 002500.200 .. 22.0200 20020223 .4 00.. «0.. 00M 00M 00.02:... 002500.200 0 22.0200 20020223 .4 v0... 00.. med 00.0 00.02:... 00:.502900 0 22.0220 20020.23 .4 00.. 00.. mm.“ 00.0 00.095. 0025020000 2 22.0220 20028.23 .4 00.. 00.. 0...“ 00.0 00.02:... 002.50.00.00 0 22.0220 20020223 .4 0.. 00.. 00.0 00.0 00.02:... 00.02.020.00 N 22.0.80. 2005......» .4 00.. 00.. 00.0 00.0 00.02:... 002502800 4 22.0.80 20022.23 .4 3 .23 > 3 $80 > 4 55... > 3 .23 2 828008... 0 Seam 0.20.20.03.02 0.2200 0.0020 .2003. - 22.0.80 20022.23 .4 143 mud mam 3.9 8.0 $6 9d via :26on mExcqm w cognac x0833? :— 5.~ mm.» and 8.0 8v 3a 3d :26on mEchw R. 8:38 x0383? .4 «a 3 8d «3 Be «ta 9d :95on 2:qu e 8228 x8853 .1— 3.. 3 8a 8d 86 3 5a :26on mEEQm n 8:38 xoacozés .1— 3 88o“. a 4 58m a a :5 Eu? 523 ES.» $93 .82 a 56:85 _ 5.255 8:85:62 1% 8.58am 95080330: 95:00 Mgm “€000“ u "nomad—HOG MONGUHEB :H 144 145 84 3d 9.» 9m :85on 2.2% m a 5:580 €30.33 .4 a: and Ed and cgmon mExtqm .4. cosmcco 48:83? .4 5 and QM mod 28.6on mExcqm o cognac 48:35? .4 t 34. mm.» mod :36on mExcqm m cognac 385:; .4 3 :53 > 3 898m > 4 830m > 3 ~33 O 8:85:89 “a anamomm madcap—5302 EufiDU xazm «QDODM u fiomugog £0564ng {H 24 36 >4. 5.” and $6 on.» bad 8398x458 ”39$ mdmoom— EDD wm4 RN no.5 3; mad m6 3.x 5.x uBBQbEmm @335 1382 $203 Nm4 EN No.5 2.6 and 5d and o.» Sutgbfimm “.3ch mdmoofl $203 35805302 «550 imam €834 - EDD 146 mm4 WON mms mamoofl EDD 34 SN w—H inflow: EDD 3.4 «Na mud magma EDD $588882 EEoD 4.52m 80on - ADEDD 147 NS 2.8 28 8.8 8.5 $5 8.: 8.: d .83 o .5. .52< 8.8 2.5 8.5 8.2 8.: 8.2 n: 5.: d .83 2. 2.. .52< :2 :1. 8.2 8.: 8.2 8.2 8.2 8.: .2 .83 m 5. .52... 2... 8+ 5.2 5.2 3.8 5.2 8.8 8.8 .2 .83 N .._< .52< 8.: 2.: 3.2 8.8 8.8 8.8 d .83 _ .2 .52... 2 .5 2+ 8... .8 2.8 5.2 8.8 2.8 d .20: 3:8 .52< S... 8.8 85 2.8 8.8 8.: 8.8 8.8 3 .83 3.2 .52... :8 5.8 2.8 8.2 8.2 8.5 8.2 5.2 .2 .55 8.2.2 .55. 8.8 8.5 8.8 8.: 8.2 8.8 2.2 8.: .2 .2; 8.22 .52... 28 8.5 82 8.2 5.2 2.5 5.2 85.2 22 .59 8.22.2 .52< a.“ 8.:. 8.8 22 8.2 2.8 2.: 2.: 22 .59 8.22 .52... 2.8 5... 8.8 5.5 8.8 5.2 8.8 8.8 22 .28 8.2.2 .52< 8.5 5.5 R5 58 3.8 5.2 8.8 8.8 22 .58 8.22 .52< 8... 8.. $5 8.2 5.2 8.: 8.8 8.8 2.38m .28 82.2 82¢. 5... 5... 88 8.2 8.8 5.: 8.8 8.8 <> .5: 55 .52... 5.5 8.8 8.8 8.8 8.8 85.2 5.8 8.8 d. 8m 9.8 .52... 5: 8 2.5 2.8 8.8 :2 8.8 8.8 22 .83 $8 .525. 8.5 2 .8 2.8 85.8 8.: 2.8 5.8 .2 .83 88 .52.. 3.: 22 .83 88 .52... E 8.5 8.5 8.8 8.2 22 .83 88 .52... 2.2 .2 .83 $8 .52... NE :8 8.5 8.8 8.8 8.2 2.8 8.8 22 .83 88 82¢. 8.5 2.8 I 2.8 :8 8.: 8.8 8.8 22 .83 88 .52... 8.8 8.8 8.2 :8 8.8 8.2 2 .8 8.8 2522 .52 88 82¢. 8.8 88 5.: 5.8 8.8 8.2 8.8 2.8 .8052: 88 .52... 3 2.8 E. 8.5 2.2 8.5 2: 5: 8 .25 58 .52... S... 2... 2.8 8.: 8.: :8 2 .2 8.2 m: .28 38 .52... E. 8.8 2.5 3.8 8.8 8.2 2.8 8.8 mm 22. 5.58 .52.. 2.“ 8.... 2 .5 2.8 :2 5.: 2.8 8.8 81:2. 8.58 .52< 8... 8.5 Rs 8.8 2 .8 8.: 8.8 5.8 ma .5: 2.58 .52... 3 :53 o B see: a 2 .582 o 520: 2.3 523 253 £83 83 8 5255 2 .2255 8:83 5 8.5082 mufloaosmmoz 95:00 Mgw ammom u mmz< 148 «in VTN 2.0 0 n2 mmZ< :6 _o.m 5w vm 3 Ego..— > A 896m > *2. 8860mm.— mwfluaosgoz 6.5300 u—Hfln—m ammom a mmz< 149 1..» 8.» 8.“ 2 8.». 2.8 8.2 8.2 oz .85 8 2o; 85» :8 8.8 8.2 8.2 2 .2 8.2 8.2 8.2 oz .85 8 2o; .82... 8.» 3. 2.» 2.2 8.8 8.: 2.8 8.8 oz .85 8 2o; 5% 8.» 8.8 8.8 8.: 8.: »».2 8.2 0.2 oz .85 8 2o; 5% 2.8 8.8 8.: 8.2 8.2 8.2 8.2 2.2 oz .85 2 2o; .52... 2: 2.8 a.“ 8.8 2.8 2.2 :8 8.8 oz .85 2 2o; mmz... 8... 8.». 8.2 8.8 8.8 8.2 8 :8 oz .85 2 2o; .82... 8.» 8.8 2.2 8.2 8.: ».2 8.2 8.2 oz .85 2 2o; .82». 8.8 8.“ 2.8 8.2 8.8 8.2 8.8 8.8 oz .85 2 2o; 5mz< 2..» 8.8 88 2.8 8.8 8.: 2 .8 8.8 oz .85 2 So; 824.. 8.8 .2 2.2 8.8 8.8 8.2 _.8 2.8 oz .85 2 2o; .82». 8.8 8.». 2..» 3.8 ».8 8.2 2.8 8.8 oz .85 2 2o; 5mz< 8.». 8.8 22 8.8 8.2 8.: 2.8 8.8 oz .85 223 5% 8... 8.8 »E 8.8 2.8 8.2 8.8 8.8 oz .85 8 2o; 5mz< 8.» 8.» 8.: 8.8 8.8 8.: 2.8 8.8 oz .85 » 2o; 85. 3. 8.8 8.2 8.8 8.8 8.2 8.8 58 oz .85 8 2o; .82». 8.8 8.». 2.8 8.2 8.2 2:: 8.8 2.8 oz .85 0 2o]. 82». 82. 2.2. 2.» :..8 8.8 8.2 2.8 8.8 oz .85 8 2o; 82¢. 8.» »3. 2.8 8.8 2.8 8.8 8.8 _.8 oz .85 2. 23 .55. $2. 8.... 8.». 8.8 8.8 8.2 8.8 3.8 oz .85 8 2o; .82». 88 8.». 8.8 2.8 8.8 2.2 8.8 8.8 oz .85 8 2o; .85. 88 8.». :2 8.8 8.8 8.8 8.8 $8 oz .85 _ 2o; 82¢. 8.8 2.8 8.8 8.: 2.2 2..» 8.2 8.2 5 .Ee » 20 82... 2 .m a» 8.: ».8 _».8 8.: 2.8 8.8 5 .39 8 2o mmz... 8.». 8.8 8.8 2.2 8.2 8.8 3.2 2.2 <> .8; » <>m fizz... 8.» 2.8 8.» 8.8 3.8 8.: 8.8 8.8 <> .8; 8 <>m 5mz< 8.» 3. 8.» 2 .8 8.8 ».: 2 .8 8.8 <> .EH _ <>m 5mz< 8.8 8.8 8.8 8.2 8.2 8.8 8.: 8.: 5 .85 2 2» 8z< 2..» .8 2.“ _.2 2.2 8.» 8.2 2.2 .5 .85 : m< 5:»le 8.8 $8 8.». 8.2 8.2 :.2 8.2 »I».2 5 .85 h 2» 82». 3 83 o 3 5.85 a ._ seen. a 52»: :33 523 EB .88.. .32 8 2255 2 5.25.5 .2382 .2 8.58am— muCOEP—Swwoz aha—MOD Madam ammom I mmz< 150 8.» »».» 2.» 8 2o; 5% »o.» 8.» 8.» 8 2o; 5% 8.» »».» :.» 8 2o; 82¢. 2.» »».~ 8.» 8 23 E »».8 8.8 2.» 2 2o; 82¢. 8.» 8.» 8.» 2 :3 82¢. 8.» 8.» 8.2 2 2o; mmz< 2..» 2 .8 8.: 2 2o; 82¢. 2.» »o.» 8.» 2 2o; 82¢. 8.» 8.» 2.» »_ 2o; 82¢. 8.» 8.» »».2 2 2o; 82¢. »o.» 2.» 8.» 2 2o; 5% 8.» 8.» 8.2 22o; 82¢. 8.» 8.» 8.» 8 2o; 82¢. 82 ».» »».2 » 2o; 82¢. 88.» »o.» 8.2 » 2o; 82¢. »».» 8.» 8.» » 2o; 5% 8.2 ».» 8.» » 2o; 82¢. ».» 8.» 8.» » 2o; E 8.» 8.» 8.» » 2o; mwzk ».» 8.» 2.» 8 2o; .82.» 8.» »o.» 8.2 _ 2o; .82,» 2.» »o.» 8.» » :6 82¢. »».» »».» 8.2 » .20 82¢. 8.» 8.» 8.» » ¢.>m 82¢. 8.» 8.» a.» 8 ¢.>m mg 8.2 »».8 8.» _ $5 .82,» »».» 8.» »».» 2 .2 82¢. 8.» 2.» 8.» : .8. 82¢. :.» mm J8.» » 2¢. 82¢. 3 83 > 3 .58» > 2 .55» > .2 EEQM 320805802 2.5200 Madam ammom - mmz< 151 »».» 8.» »8.» » ”.8 .».»8 2 .2 »».8 2.8 oz fig 8 2o: »% 2.» »».» 8.... »».8 »».»8 »».». 8.8 2.8 oz SE 8» 2o: 82¢. 2.» »».» 8».» »88 »».8 8»... 8.8 2.8 oz 8:» .» so: »E »».» 8».» »».» »».»8 »».»8 .».8. ».»8 »».»8 oz .8:» a» so» 82¢. »».» ».» »».» 8».8 ». .88 ».8. »».8 »».8 oz 8:» 8» 2 o: 82¢. .8.» 8.» »».» 8. .8 »».2 8... 8.2 »».2 oz 8:» »» 2 o: 82.» »».» :.» »».» »».»8 »».»8 8.. . 8.8 »».»8 oz 8:» 2. 2o: 82¢. » .».» ».» 8.2 »».8 ».» .».8 »».8 oz 8:» »» 2o: 82... :.» »8.» »».» »».8 »».8 »8.» 8.8 8.8 oz 8:» »» 2o: 82< 8.» »8.» 2.». »..»8 8.8 »..2 8.8 88.8 oz 8:» »» 2o: 82¢. 8».» »».» ».». ».». »».». 8.2 8.». 2.». oz 8:» 8» 2o: 82¢. »».8 »».8 »».8. ».». 8.». »8.». »».». 8.». oz 8:» 8» 2o: 8%.“ 8.» »».» 8.». »».2 8.». 8.8 ».». ».». o2 8:» .» 2o: 82¢. »».» »».» 8.». »».2 8.». 8.2 2.8 »8. 8 oz 8:» »» 2o: 82¢. »».» 8.» »».2 8.». 8.». 8.8 »».8 8. 8 o2 8:» 8 2o: 82< .».» 8 »».». »8.2 »».8 »».8 .»..8 »8.88 oz 8:» 8 2o» 82¢. »».» »».» 8.2 »».2 8.2 2.8 »».8 8. 8 oz 8:» »8 2o» 82¢. »».» ».» 88.2 ».». 8.8 »».8 88 »».88 oz 8:» 8 2o: 82< 8».» »».» »».». 8.2 »».2 2.8 8.8 »».8 o2 8:» 8 2o: 82¢. »».» 8 »».». 88.8 »».8 »».8 8.8 »».8 oz 8:» 8 2o: 82< »».» »».» »8.». 2.88 »».»8 »».2 »».»8 »».»8 oz 8:» 8 2o: 82¢. 8.» »».8 8.». »».2 »».8 2.88 »».8 8.8 oz 8:» 8» 2o: 82¢. »».8 »».» »».». 8.8 »».8 8».». 88.8 »».8 oz 8:» 8 2o: E 8»... .».» »».» 8 »».2 ».». 8.88 .».88 oz 8:» 8 2o; 82¢. »».» »».» »».» 8.2 »».2 8.» 2.8 88.8 oz 8:» 8 2o: 82¢. .».» 8.» »».» »».8 »».88 8.2 8.88 8.8 oz 8:» 8 2o: 82¢. »».» »8.» 8.» ».88 .».8 8.: »».8 »».8 oz 8:» »8 2o: »E 8.» »».» ».» 8.8 88.8 »».». 8 .8 oz 8:» »8 2o: 82¢. »».8 »».» »».» »».2 8.». ».2 8.2 »».2 oz 8:» »8 2o: 82¢. »».8 2.8 ».».» 8.». »».». 8.2 8.». 8.». oz 8:» 8 2o: 82¢. 3 :3... o 3 Sec» 8. 8 .55... a :.»»... .83 523 EB 58%. .32 8 .3258. . 5.28.». 8:83 » say. mufloaogmaoz «Hun—00 Madam ammom I mmz< 152 mm.” 0:. 8.0 mm 20.— me< 3.5 2 .m 3.3 mm 20.— mmZ< We v0.0 36 E 20.. mmz< No.5 86 8.x. cm 20.. mg 3.2 or... on.” 3 20.. mmz< 2.0 9mm mmfi «v 20.. me< No.3 36 mm.” 3. 20.. mmz< hNS 3N Vmfi av 201 .373. 36 mvd 3.». 3. 20.. mmz< 36 N56. ma.” 3. 20.. me< omé On wad— ? 20.. LE 26 afim v2: NV 204 mmZ< 36 mod 3.: 2. 20..— me< end _v.m wad. o». 20.— mmZ< end. end QM: an 204 mmz< no.» mud :.m— an 204 me< and En m6. mm 204 mmz< Rd :..m 8.2 cm 20..— E a... Nm.m 3.2 mm 20.— mmZ< ems cod VGS Em 20A mmz< 36 :.v v52 mm 20..— mmz< Rd Vmfi 22M: mm 204 me< 3.0 E .n Nd ; 20..— me< 3.6 34. no.” on 20..— mmZ< mud mi 9.6 on 204 mmz< no.0 moé .2: mm 20..— LE no.» :3. 8.x um 20.— .579. 3.3 _o.m 3 .w em 20..— mmz< 84. m6 Nwd mm 204 mmz< nod hvd v6 3.. 20.— .572 3 :36 > 3 880m > 1— Each > n Goa—mm— mwfloaogoz aha—HOD v—uflSm ammom I mmz< 153 8.5 on t .22 8.8 88 3.2 8.2.8 :8 oz .82... 8 2o; .52< 2.8 8.. 2... 8.2 3.2 2... 2.2 25.2 oz .82... 22 2.6.2 .52< a... 3.8 .2... 8.2 3.2 8.: 3.2 2.2.. oz .82: 3 2o; .52< 8... $8 .5... 8.2 $2 8.3 8.2 5.2 oz 622.2 22 2o; .52< 2.8 5.8 5... 2.2 2.... 2.... 2 2.2 oz .82: 22 2o; 2% 2.8 5.2 3... .23. 2.2.22 2.» 2.3 8... oz .82: 5 2o; 2% 8.2 3.5 2% 2.2 8.2. 8.3 3.2 8.: oz .82... 8 2o; 2% 2 .m 8.8 8... 25.2 3.3 22.5 2:: 2.:: oz .82... 8 2o; .52... .. 8.8 S. 8.2 8.... 8.2 2.2 2.2 oz .82... 8 2o; .52... :.2 8.5 3.22 8.... 8.2 2.2 :2 8.2 oz .82... 2. 2o; .52... 2.5 ..2 .8 a... 3.2 9...: 8.. 8.2 8.2 oz .82... 8 2o; .52... 3.8 28 8... 2.2.. $2.. $8 .52 8.2 oz .82; 8 2o; .52... .2... 32.... 8.. 228 3.8 5.2 8.8 8.8 oz .8... 2. 2o; .52< $2 .. 8.2 8.2 2.2.2 3.2 2...: 28 oz .02: 22 2o; .52... 5.8 :2. 2.2 8.2 5.... 8.8 3.8 8.8 oz .82: 8 2o; .52< 23 8.5 5.8 28 22.5 5.8 2.5 3.8 oz .82: 2 2o; .52... :8 2.2. 5.: 8.: 2.2 .25.... 32.2 3.2 oz .82... 2. 228 .52... 3.8 8.2. 2.5 8.8 2 .2 2.3 3.2 2.2 oz .852 S 2o; .52... 8.8 2 one 3.2 8.2 o 8.2 3.2 oz .82... 3 2o; .52... 22.5 85.5 3.3 3.2 3.2.2 8.3 $2 3.2 oz .82... 3 28 2E 28 8.5 8... 5.2 2.2 3. 2 .2 5.2 oz .82: 3 23 .52... 3... 8.5 8.3 8.2 2.2 3.3 8...: 3.: oz .82: 2. 2o; .52... :8 2.8 2.3 8.... 5.2 8.2 8.2 8.2 oz .82: s 23 .52< 8... 8.5 8.2 3.: 5...: 2.: 8.2 8.2 oz .822 8 2o; .52< 3.8 2.8 8.3 :.22 3. : 8.3 3.2 8.2 oz .82... an 23 2%.... :5 2 .5 2.3 8.2 8.8 8.2 2.2.8 8. 28 oz .82... .5 23 E 2.... 2... 2.... 2.8 88 8.2 2.8 3.2.8 oz .82... 5 23 2g 8... 2 2. .5... 8.8 3.8 8.2 88 2.22.8 oz .82... 5 So; .521... 23 3... $22 88 8.2.8 8.2 8.2.8 2......” oz .82: 5 2o; .52... 8.2 S... m... 38 38 E2 258 :8 oz .82... 3 2.6.2 .52... 3 22.3 o B as... a .2 .552 a 52...: 2.3 52.23 2.3 5228.2 .322 N .2255 2 .3255 2.2283 2 8.58.25 5220805232 82200 282m ammom - .52.... 154 3.6 5... cod an 204 mmz< 32m N_.m fine on EU.— mmZ< 3N SN 2.6 v» .204 .522... mad. fin 2.4. mm .20.— .372 v 8.2.” 3.6 mm .204 mmZ< QWN 3.. 2.4. 3 .204 mmZ< ad 3N no.5 cm 20.— mmz< and o. .N 3.3. an 204 .572 8.». fin 3.” ms 204 mmZ< _N.m mmd mm.» on 204 mmZ< 5.2.” Hum 56 mm ED.— 226% and mud No... Q. 204 mmZ< :25 56 no.» N5 204 mmz< 3d mmé 3.0— :. EDA mmZ< no.2.” Que 86 _ on 204 mmZ< was mod No.5 00 204 mmZ< mud N23. 2 .2 no 204 mmZ< mod 8.». 9.6 no .20.— mmZ< $6 and mmfi no 204 mmZ< wad. m _ .m 3% no 204 mmZ< ”v.2. and m6 we .204 mmZ< mad and 36 me .204 mmz< no.2. 38.. 5d 3 .204 mmZ< and mmé fix an 204 mmZ< mad and Wm am .204 mmZ< NS _m.m mmd an EU..— mmZ< mé Wm 3d hm 204 mmZ< mad mod 2%.” on .204 mmZ< end— vmé 5a mm 204 mmZ< and. ..m :6 Vm 204 mmZ< 3 =a3 > 3 28.52 > A 822.5“— > a 558mm 5220822302 8250 2.82m ammom - .579. 155 8.8 5.8 2.2.2 8.2. 5.85 5.2 8.8 2.8.2. oz .822 3. 2o; .52< 85.2. 3.5 58.2 8.: 2.5.2 522.58 58.2 58 oz .82: 52 2.6.2 .522. 85.5 55.5 85.2 2.: 5.2 8.2.8 8.8 85.88 oz .82: 82 2o; .52< 2 .5 52.5 22.5 88 8.2.8 8.2 8.8 8.8 oz .822 8 2o; .52< 2.5.5 2.8 :8 8.8 88.8 2.8.: 2.5.8 5.8 oz .822 8 2o; .52< 8.5 822.5 522.8 55.88 5.58 3.: 8.8 8.8 oz .82: 3 2o; .52< 8.5 5.5 5.8 8.2 58.2 88.2 55.2 8.2 oz .82: 8 2o; .52< 522.5 8.2. 8.2 5.2.8 5.2.88 85. 552.8 55.58 oz .82: 8.2 So; .52< 8.8 8.8 82. 2.5.2.8 558 8.22 888 8.58 oz .822 5 2o; .52< 522.5 822.5 55.8 8.58 22.8 8.2 88.8 8.8 oz .82: 8 2o; 2% 8.2. 8.5 5.2 8.8 2.8 5.2 3.8 2 8.8 oz .82: 8 2o: .52< 3 :83 a 3 52.2 o 2 22.2 a .2820: 2.23 .2223 252.3 558.2 .32 8 .2255 2 .2255 22:83 a 5220823532 82200 282w ammom - 852.22.. 156 55.5 55.5 55.2 32 56.2 .52... 55.5 58.2. 52.52 52 2.26.2 .2522. 85.5 85.5 2.2 82 2.6.2 .52.... 55.5 5.5 3.82 85 $23 .52< 55.2 55.5 52.5 55 56.2 .52< 8.5 8.5 55.5 3 2.6.2 .525. 8.2. 52.5 25.5 55 5,62 .52... 85.5 85.5 58.82 85 23 .52< 85.2. 55.5 85.2. 25 562 .52< 2.5 55.5 52.5 55 56.2 522.. 85.5 52.5 85.2.2 55 2.62 .52< 3 2253 > 3 88.2.2 > .2 58.2.2 > 22 8828.25 5282082252502 5550 282m 2550.22 - .5775 157 55.5 55.5 2 2.52 85.82 55.5 25.82 85.82 .2... .852 5.52 2555 222252 55.5 2.5 58.2 2 55.2 2 85.5 55.2 2 58.2 2 .2... .852 8.52 2555 225,252 55.5 25.5 88.52 25.52 5.5 2.52 2 .2... .852 5.52 2555 222252 85.2. 28.5 55.52 55.2 55. 22 58.52 5.52 .2... .852 5.52 2555 222252 85.5 55.5 55.2 2 55.52 82.8 55.52 85.52 .25. .852 5.52 2555 222252 55.5 55.5 5.82 25.52 55.5 55.52 5.52 .22. .852 5.52 2555 522252 85.2. 55.5 85.58 5.58 55.82 22.2.8 52.58 .2... .852 8.52 2555 222252 5.5 82 .5 5.8 28.52 55.82 55.8 28.8 .22. .852 2.52 2555 222252 8.5 85.52 85.88 25.58 2.52 55.88 2.58 22522 .8252 55852 222252 25.5 55.5 25.22 85.52 52.5 88.52 52.52 522522 .8222 8.55852 225,252 5.5 22.52 55.58 85.25 5.52 8.55 55.55 2252 .822 85852 22.22.2252 55.5 52.82 88.55 55.55 85.58 55.55 55.55 52252 .822 55852 222252 55.5 55.2 82.58 55.58 85.2.8 55.55 3.85 52252 .8222 5582 222252 52.8 82.28 55.55 55.85 55.58 55.55 85.55 522522 .822 3852 225,252 52 .5 58.58 85.55 55.85 55.88 55.55 85.52. P... .852 2.8585 222252 55.5 55.2 58.8 88.58 55.82 55.58 55.58 2.522252 .55 52855 222252 58.5 55.2 58.28 55.58 5.52 88.58 38 5:22.52 .2258 52855 222252 58.5 85.52 88.28 55.58 55.82 5.58 58.58 5525.52 .2258 2.2855 222252 58.5 85.5 55.82 55.82 85.5 55.82 55.82 .25. .852 5.5555 2222 2.5 82.5 85.52 25.52 5.5 88.52 55.52 .2... .852 5.5555 522252 58.5 58.5 8.82 85.82 88.5 85.52 55.52 .2 .852 5.5555 222252 25.5 55.5 58.8 28.88 55.2 58.88 85.88 .22. .852 8.5555 222.2252 85.5 52 .5 52 .58 55.88 55.82 58.58 58.58 .2... .852 2.5555 222252 5.8 25.55 58.85 55.55 3.55 55.55 85.55 522522 .822 5552 225,252 55.5 5.5 1 85.58 55.58 1% 2 .52 85.58 55.58 o5 .852 552 225,252 3 .5252 52 2 825.2 52 2252552 25253 .2223 2253 2258.2 552.22 8 225.252 2 2225252 5222582 5 55.28.25 52522220225552 25.222200 282m ammom - $235 158 522.5 55.5 88.8 5.5 5.52 2555 22252 58.8 58.8 85.5 2.2 8.52 2555 22252 55.5 85.5 55.5 85.5 5.52 2555 222252 55.5 58.5 5.5 88.5 5.52 2555 522252 55.5 85.5 85.5 85.8 5.52 2555 222252 5.5 55.5 85.5 85.5 5.52 2555 2222522 52.: 85.5 55.5 85.5 8.52 2555 222252 25.5 58.5 58.2 85.5 2.2 2555 22252 58.5 55.5 55.82 5.5 5582 222252 55.5 55.5 55.5 85.5 55852 22252 55.5 5.5 58.52 58.8 85852 222252 28.5 58.5 52.82 52.5 55852 2.2252 85.5 55.5 58.82 25.5 55852 522252 55.8 55.5 85.52 85.5 3852 22252 85.5 88.5 85.8 2.5 28585 522252 2.2 55.5 88.2 55.5 5.2855 2222 85.5 55.8 55.52 55.2 5.2855 2222 5 55.8 5.5 55.5 2.2855 2222 2.5 55.8 55.5 58.8 5.5555 2222 5.5 55.5 58.5 522.5 5.5555 2222 55.5 88.5 55.5 85.2 5.5555 2222 55.8 58.5 5.8 55.5 8.5555 2222 58.52 55.5 8.5 52.5 2.5555 2222 55.5 85.8 55.55 55.5 552 222522 55.5 525.5 55.2 58.5 5552 22252 3 2253 > B 525.2 > .2 E5252 > 3 2253 52 .2 5.2225225 525208022582 5222250 285m ammom - $235 159 55.5 85.8 55.52 58.55 85.55 55.28 88.55 55.55 522 .822 8552 >22o 55.8 52.5 85.52 88.25 88.55 85.52 55.55 55.25 oz .822 5582 >22o 85.5 5.5 5.5 85.55 58.55 8.52 55.55 55.55 522 .822 2.8552 >22o 55.5 5.5 55.5 2.55 88.55 82.52 55.55 5.55 522 .822 558552 >22o 88.5 8.5 2.52 52.55 25.55 58.52 82.55 88.55 522 .82 55.8552 >22o 55.5 58.5 5552 55.2.5 25.55 55.52 55.55 5.55 522 .822 85.8552 >22o 85.8 25.5 2.52 2.2.5 85.55 58.52 5.55 55.55 522 .82 55.8552 >22o 2.5 55.8 55.52 2 .55 55.55 55.52 55.55 55.55 522 .822 558552 >22o 85.5 85.5 2.52 5.55 52.55 52.52 55.55 82.55 522 .822 3.8552 >22o 55.5 55.5 85.52 52.55 85.55 88.52 85.55 55.55 522 .822 55.8552 >22o 55.5 22.5 5.52 55.55 55.55 52.52 55.55 28.52 22 .822 858552 >22o 88.8 52.5 55.52 52.55 22.55 5.52 55.55 55.55 522 .822 25.8552 >22o 85.5 55.5 55.5 5.82 58.2 25.5 55.2 58.2 522 .822 8552 >22o 55.8 55.8 2.5 58.2 58.52 55.8 55.52 52.52 522 .822 2552 >22o 55.5 2.2. 58.5 55.2 55.2 88.8 25.2 85.52 522 .822 5852 >22o 55.5 85.8 85.2 85.25 52.55 58.52 55.85 55.85 522 .822 8852 >22o 55.5 52.8 55.2 55 28.55 58.52 52.55 58.52. 522 .822 5852 >22o 5.8 58.5 2.58 58.85 55.58 85.55 55.55 55.55 522 .822 552 2 >22o 88.5 5.2. 55.52 55.55 52.55 85.52 58.55 52.55 522 .822 5522 >22o 55.5 5.5 25.82 55.55 52.85 25.52 55.85 28.55 522 .822 222 >22o 55.5 28.5 85.52 55.55 5225 55.2.2 55.25 5.85 522 .822 252 >22o 55.5 m: 55.52 155.55 55.85 5.52 582.251 E225 522 .822 525 > 22o 3 2253 52 >2 525.2 52 2 .5252 52 2252...: 25253 .5522 25253 .5582 552 8 5225252 2 2225252 5222582 22 828mg 52582222582 5.222200 282m 2550a - >220 160 52.82 55.8 55.52 8552 >22o 28.52 55.5 52.52 5582 >22o 55.22 5.5 2.22 52.8552 >22o 52.22 55.5 55.52 55.8552 >22o 5.52 55.5 55.5 558552 >22o 58.52 5.5 88.5 858552 >22o 58.82 52.5 55.52 55.8552 >22o 55.52 85.5 55.52 55.8552 >22o 55.22 55.5 25.52 3.8552 >22o 85.5 88.5 55.52 55.8552 >22o 25.82 58.5 52.52 85.8552 >22o 85.52 52.5 58.52 25.8552 >22o 52.8 55.5 85.5 8552 >22o 8.5 55.8 55.5 2552 >22o 55.5 52.5 55.5 5852 >22o 28.8 5.5 55.52 8852 >22o 52.82 85.5 55.22 5852 >22o 55.5 5.5 25.58 5522 >22o 28.22 25.5 2.22 5522 >22o 8.22 8.5 85.52 2522 >22o 55.52 55.5 85.: 5252 >22o 5.8 52.82 25.22 525>22o 3 2253 > 3 8525.2 > 2 25.5.2 > 5 5525225 3:089—3302 QHHGQU MEW mmmom .- EU 161 8.5 25.5 55.52 58.55 55.55 88.28 55.85 8.55 2.22.2.2 55552 .5 52 .5 85.8 8.52 55.2 2 85.82 55.52 55.52 55.52 2.2 2522.2 552 582 .5 55.5 58.5 5.22 58.52 52.2 2 2.2 222.2 55582 .5 55.5 88.82 55.82 2.52 55.52 2.2 222.2 55882 .5 88.2 52.2 5.2 55.5 5.5 58.8 5.5 55.5 2.2 .822 55885 .5 25.2 85.8 88.5 55.2 58.52 2.52 58.82 55.82 2.2 .822 55885 .2: 52.2 85.8 25.5 8.2 5.52 25.5 28.2 55.52 2.2 .822 85885 .5 55.5 55.5 55.52 5.85 25.85 55.88 58.55 5.55 2.2 .822 25885 .5 52 .8 58.8 58.5 85.82 8.2 2 55.5 55.2 2 55.2 2 2.2 .822 55885 2: 58.2 55.5 88.5 85.82 55.2 2 2.8 85.2 2.52 2.2 .822 88885 2: 55.8 28.5 55.8 58 58.58 55.2 2 85.58 25.58 2.2 .822 58885 .5 88.8 82.5 85.5 85.52 55.52 58.8 58.52 85.52 2.2 .822 5.5585 .5 25.8 85.5 85.5 55.82 55.52 58.8 58.52 55.52 2.2 .822 5.5585 2: 58.5 55.5 55.5 5.58 85.28 2.52 88.88 85.88 22 .822 5.5585 8: 2.5 85.5 5.5 55.58 85.28 82 .52 88 25.88 22 .822 8.5585 252 88.5 55.5 55.5 5.58 58.28 58.52 55.88 2.2.88 22 .822 2.5585 2: 3 2253 52 a 525.2 52 2 25.5.2 52 225222 25253 .5523 2253 222582 .32 8 222.5252 2 2225252 5222852 5 22528.22 mHQDEOHH—mmoz “.5650 Jim ammom I PD 162 :62 and :.w— ocawN— m3 nmé 3.5” 05.2 5522mm— "5 wnwmfl 85 emNNQ 85 :.N wad m 2 .2 ammo 85 a0...” and 2.5 ammo 85 5. SN med mama “—3 mod— vad @56— Emma 85 cod mod 3.5” cwmmm «.5 Wm and Sam Emma 55 and. wmé 3..” wanna 85 3.25 36 35.». “mean .85 8.45 mvfi 8.5. v.22 m3 mus 56 «We «.mvmm 55 3.8. 3.5. mad «.33 "5 was 3.5. mad 2655mm 5 3 :53 > 3 85.8.2— > .— 5952 > 2222 :25 m 5252202252502 25.222200 imam 255082 .. 55 163 2 .8 8.88 8.8 8.8 8.8 2.2. :.8 oz .88 .8888 228 8.8 8.2.. 8.88 8.88 8.8 8.:. 8.:. oz .82. 8888 22m: 8: 8.8 8.8 8.8 88.2. 8.:. 2 .2. oz .88 8888 22m: 8.; 8.8 8.8 8. _ 8 8.8 8.2. 8.2. oz .88 2.88 223 8.2 8.2 8.:. 8.2. 8.8 8.2. 8.2. oz .85 2888 22m: :8 8.8 8.8 8.8 8.88 8.2. 8.2. oz .022 2.88 223 8.2 8.2 8.8 8.88 8.8 8.8 2.8 oz .88 8888 225 8.8 8.8 2.8 8.8 8.8 88.8 8.8 555.5 8888 22m: . 8.8 8.8 8.:. 8.:. 8.2 8. 2. 8.2. oz .88 8888 225 8.8 «.2 8.8 8.8 :.2 8.8 88.8 .5555 8888 22m: 3. 8.: 88.8 8.8 R. .8 8.8 8.8 .5555 8888 22m: 8... 8.2 8.8 8.8 2.8 8.8 8.8 565.5 888 22m: «.8 8.8 8.8 8.8 8.2 8.88 8.88 oz .022 8888 22m: 8.» 8.8 2.88 8. _ m 8.2 8.8 .8. _ 8 oz .88 8888 22m: 888 8.8 8.8 :8 8.2 8.8 8.8 oz .88 :88 228 2.8 8.2 8...: 82 8.2 8.2 8.2 oz .88 288 .223 8.8 8.... 8.2 8.2 8.2 8.8 .8 a: .822 2 .228 22.0.: 8.8 8.8 8.2 8.2 8.2 8.8 8.8 a: .822 8. .228 223 8.8 88.8 8.2 8.2 V2 8.8 8. .8 a: .83 2228 22.0.: 8.8 8.8 8.2 8.8 8.2 8.8 8.8 a: .022 2 .288 228 2.8 88 8.2 2.8 8.2 88.8 8.8 a: .88 8.228 228 8.8 8.8 8.2 :.8 8.2 88.8 8.8 a: .022 8.228 22mm 8.8 8.8 8.2 8.2 8.2 8.2 8.8 a: .822 8228 .225 8.8 8.8 8.8 8.8 2.2 8.8 2.8 a: .85 8.228 223 8.8 8.8 2.8 8.2 02 8.2 8.2 a: .85 8.228 228 2.8 8.8 8.8 8.2 8.2 8 8.8 a: .85 8.228 223 8.8 2.8 8.8 8.2 2.2 8.8 8.8 92 .822 8.228 223 8.8 8.8 8.8 8.2 88.2 8.2 8.2 a: .822 8.228 225 8.8 8.8 8.8 2.2 88.2 8.2 8.2 a: .822 8.228 225 88.8 . 8.8 8.8 8.8 2.2 8.8. 8.8 oz .822 2.2. :23 3 888 a .2 55.2 a £2»: .83 523 28> ewes v82 8 53:35 _ 53:35 2:83 8 888m 9:289:30.)— EEGO £95 :80”. - 22w: 164 8.8 8.8 8.8 8.8 _.8888 22m: 28 8.8 8.8 2 8888 22m: 8.: 2.8 82 8 8.8 .888 228 m 8.8 8.8 8... 2.88 228 8.2 8.2 2.2 «.8 88885-2sz 2 .8 8.“ 8.8 8.8 2.88 223 8.2 8.2 8.2 8.8 88885228 2.2 8.8 888 2..» 2.88 22% 8.2 8.2 8.2 8.8 8885-2sz 8.: 8.2 8.2 2.8 888 228 :.8 8... 8.2 8... 8888 223 2.8 83. 8.2 28 8888 22m: $8 8 8.8 8.2. 8885-228 v.8 8.8 8.8 2.2.... 8888->-22m: 8.“ 8.8 88.8 8.8 2885-228 8... 8.8 8.2 8.8 2888 22m: 82 8.8 8.8 8.8 2.228 223 2..: 2.8 8.8 8.8 2.228 22.8: 8.: 8.8 8.8 8.8 :.228 228. 8.8. 8.8 8.8 8.8 2.228 .228 2.2 2.8 2.8 8.8 8.228 .2sz 8.2 8.8 :.8 :8 8.228 225 8.2 3 8.8 8.8 8.228 228 8.2 2.8 8.8 8.8 8.228 223 8.2 2.8 E 28 8.228 222. 8.2 2.8 8.8 8.“ 8.228 228: 8.2 8.8 2.8 8.8 8.228 228: 8.2 8.8 8: 88.8 8.228 225 8.2 2.8 3 8.8 8.228 228 8.8 PE ml: 8.8 82?:sz 3 :25 > 3 888 > a 528 > a 2.3% 8 8&6me mEmEmsmmwS. 8E8 {usw =80". - 22m: 165 88.82 88.88 8.88 2 .88 8.:. s. .8 88.8.8 oz .82: 882.2. .228: 88.2 88.88 88.88 :28 8.88. 88.88 8.88 oz .828 882.2. 228: 88.8 8.88 2 .88 82 .88 8888 88.88 82.2. oz .888 882.2. .228: 2.8 28.8 88.88 2.88 2.88 88.8.. 88.88. oz .888 882.2. 228: 88.8 8...: 8.88 8.828 88.82 88.88 .888 92 .822 2882.2. 228: 88.8 88.2 8.88 88.88 88.82 88.88 88.8.8 92 .822 2882.2. 228: 88.8 8.2 :28 88.88 28.82 88.8.8 88.88 92 .8882 88882.2. 588: 88.8 88.82 88.88 88.88 88.8. 88.8.8 888.8 92 .8822 88882.2. 228: 88.8 88.2 88.88 88.88 88.82 8.88 83.8 92 .8822 88882.2. .228: 88.8 88.: 5.8 88.88 8_ 88.88 82.8 a: .83 88882.2. 228: 88.8 38. 88.8.8 8.88 8.: 22.8 88.8.8 92 .8882 88882.2. 228: 88.8 88.: 88.8.8 8.88 2.82 2.88 8.88 92 .822 3882.2. 228: 88.8 82 2 .88 88.88 8_ .82 88.88 88.88 a: .8882 88882.2. 228: _.8 88.82 88.8.8 8.88 88.82 88.88 8.88 a: .82 88882.2. 228: 88.8 2.8.: 2.8.8.8 8.88 28.82 8.88 88.88 92 .882 8882.2. 228: 8.8.: 2.8. .82. 28.8.. 88.88 88.2. 8.2. oz .828 888888 .228: 8.8.2 88.: 88.88 88.8.. :88 88.88. 88.88. oz .888 888882. 228: 88.8 82 .88 88.88 88.2 .888 2.88 88.88 oz .888 88882. .228: 8.8 88.8 88.2 8:: 2.8 88.2 88.2 oz .828 8.8882. 228: 28.8 88.8. 28.88 88.88 8.8. 88.88 88.88 82 .882 888888 .228: :.8 8.8 88.88 88.88 8.2 2.88 8.88 oz .888 2 8888 228: 88.8 88.88 2.88 8. :. 88.2. 2.8.. 88.8.. oz .828 88888 228: B 8888 a 2 88.88 a .8888: .8283 £23 283 5.8.3 882 8 588885 _ 58885 88:83 8 8.5888 3:089:32). 98:00 {mam =80”. .. 22w: 166 8.8.2 2 .8_ 88.88 88.: 882.2. 228: 88.88 88.: 28.88 8. .8 882.8 228: 8.8.8 8.8 88.88 88.8 882.2. 228: 88.8 88.8 8.8.8.8 88.8 882.2. 228: 8_ 2 .8 8.82 8.8.8 2882.88 .228: 88.82 8.88 28.82 88.8 2882.2. .228: 8.2 8.8 8.8_ 88.8 88882.2. .228: 2.2 88.8 88.2 88.8 88882.2. 228: 8.8_ 8.8 28.2 88.8 8.8832. 228: 88.2 88.8 88.2 8.8 88882.2. 228: 88.82 8.8 88.2 88.8 888822. 228: 88.2 88.8 8.2 88.8 8882.2. 228: 88.8. 88.8 88.2 88.8 8882.2. 228: 8.8. 8_ .8 88.2 8.8 88882.88 :28: 8.: 88.8 82.2 28.8 8.8832. 228: 88.82 2 .2 88.2 88.8 888882228: 88.82 88.8 .882 .8 8888822228: 88.8. 8_ .8 88.2 88.8 888888. 228: 88.8 8.8 88.8 8.8 8.8882. .228: : 88.8 88.82 88.8 888888 228: 8.8 8.8 8.8 8.8 2 8888->-228: 88.8 8.8 88.88 8.8 88888 228: 88 :83 > 3 8858 > .8 88.88 > 3 :83 o 8 888888 9:98.938.)— EEwO {mam =80". - 22w: 167 88... 88.8 88.8. 88.: 8.8 88.8. 2.8. <0 .88.... 88.8... .8 82:28.8 88.8 88... 88.8 88.8 8...8. 8.8. 88.8. <0 .88.... 88.82 .8 82:28.8 2.8 ..8.8 88... 88... 88.8. 88.8. 88.8. <0 .88.. 88.82 .8 2:28 8.8 88.8. ..... 2.8 8.2 8.2 8.... <0 .88.... 88.8... .8 82:28.8 88.8 8...8 8.8. 88.8. 88.8 ..8.8. 2 <0 .88.... .882 .8 82:28.8 ..8 88.8. .8... ..8.8 88.8. 8.8. 88.8. <0 .88.. 88.8... .8 82:28.8 88.8 2.8. 88.8. 88.8 8.... 88.2 8.... <0 .88.. 88.8... .8 2228.8 88.8 88.8. .8... 2.8 ..8.8. 88.8. 88.8. <0 .88.. .88... .8 82:28.8 88.8 88.8. 88.8 88.88 88.... 8.88 .8 <0 .8... 88888 82:28.8 88... 8.8 8. 8.8. 88.8. 88.88 8.8 <0 .88.... 88888 8.8.28.8 8.8 88.2 88.88 8.88 ..8..8 .888 2.88 <0 .88.... .8. 88 58.7.8.8 8.... :.8 .888 8. .88 8...8. .....8 88.8 <0 .88.... 88.8828 2228 8.8 8.... .888 .888 88.88 88.88 88.88 <0 .88.... 88828 2:28 88.8 88.8. 88.8. 88.8. 2.8. 88.8. 88.... <0 .88.... 8.8828 82:28.8 88.8 .88 8...: ..8.8 88.8 88.8. 8.8. <0 .88.... 8.8828 82:28.8 88.8 88.8 88.8 88.8 88.8. 88... ....... <0 .88.. 2 .8828 2:28 .88 8.8 8.8. 8...8 .88 ..8.8. 88.8. <0 .88.. 8.8828 2228.8 88.8 8.8 88.8 88.8 88... 88.8. 88.8. <0 .88.. 8.8828 3.8.28.8 .88 8.8 88.8 88.8 88.8 2.... .8... <0 .88.... ...888.8 25.8 88.8 88.8 8.8 ...8 8.8 88... 8... <0 .88.... 8.8828 82:28.8 .88 :.8 88.8 .88 88.8 .88. 88.8. <0 .88.... 88.8828 82:28.8 ....8 88... .8. 88.8 88.8 .88. 88.8. <0 .88.. 88.8828 88.7.8.8 88.8 88... 88.8 88.8 88.8 88.8 88.8 <0 .88.... 88.8828 2228.8 88.. 88.8 ..8.8 ..8.8 2.8 88.8 8.8 <0 .8... 88.8828 2:28 88.. 88... 88.8 2.8 8.8 88.8 2.8 <0 .88.... 88.8828 82:28.8 8.8 2.... 88.8 ..8.8 ...8 .88 88.8 <0 .8... 88.8828 2228.8 88.. 88.8 88.8 88.8 .88 ...8 2.8 <0 .88.... 88.8828 2:208 88.. 88... 88.8 88.8 88... 88.8 88.8 <0 .88.... .88828 3.8.28.8 88.8 88.8. .88. .8... 88.88 88.8. 88.8. <0 .8882 82.88 2:28 88... 88.8 . .888 88.88 J88... 88.8 88.88 <0 .88.. 88888 2:208 88 sec... 8. .. .888... 8. 58.8.. 8.83 8.8.3 8.8.8 585.. 88.2 8 3.25.8. . 5.25.8. 88.88.. 8 =85.8....8 888.080.38.802 8888.00 0.8.8.5 @8808. - SHE/Em 168 N00 0a...” 2.0 Ewm 8.9.— : EZQm and and 8.2: and mad: F EZn—m v.0 36 E. mod 8&3 E EZQm mwN no...” 5.: 0o.N 3N3 :. 5.90m _n.m mod N. .0 SN 3&3 K 20:35 no...” and 0nd— 0N.N 8.9. K gflw VEN 8mm 8.2 mm. mod: I. 237mm omfi 3.8” $0. an. 5&3 E. v mm.» Nm.m 055 0.0 0300 2375M an... 8.8” via 3.8. $30 2:205 mmfi m0.0 00.0. 00.0 8.0— m0 22205 S... no.0 .md 00.. 8.53.0 232nm 3.0. 005 3.3 8N8 3.8.3.0 SEZDm no.0 _Nm 8d 00d 0. .2830 EZDm de .mfi 0.8. mm. 2.320 EZn—m and «Na mod .06 3 6.3.0 SEZDm mod mvd 3.0 3.. 268.0 £20m Nmfi mmd mmd mm. 2.230 glam vmd Sam 3... mod 2.330 ZEZQm hmd 8.0.m 050 3d 2.26.0 Sizam fin :.m mmfi .NN 3.98.203 wmd 2d .8... mm. “5.230 SEZn—m 86.— ».md 0N.m 0m. Edna—0 SEZQm SN 50.. Kd aod 00.9.30 EZQw «8... 8m. 2. 3“.— wodmo; 232nm E..— VN.N mm... :.o 8.220 ZZZDm ma. :.N a...” 0nd 8.320 SIZDm 8.. m»... :.m a; 8.330 SEZQm VON no... Edm mod movwm SEZQm 08.... mm... :.d 0m... 9wa SEZQm 3 :83 > 3 88.5.2 > A EEO... > BE D u 58E m mun—0805302 MHHGDU Jim ammonm .. 2.5/Ham 169 a... 2 a; No.3 3.3 an... 0... <0 .8I..m .22 2:2@ 2.0 Sn 2.9. .82 2.: 3.3 on: <0 .8... n... E 2:23 a... :.m n 9.2 $2 a... 2.0 <0 .8... an... K 2:23 3.. ten 83 m... .3. 8.“ .3 <0 .20.... ”an... E 2:23 2.. as and .0 9..“ 0% 2: <0 .8... Rd... E 2:23 a... 0. .m 3.0 a... 9.... a... .2 <0 .8... an... :. 2:23 a; 3..” 2.0 2.0 :.n a: 8.. <0 .8... <2... 2 2:23 «3 on... N... $2 a. .m 8.... 2.. <0 .8... m2... .5 2:23 2 a... on... E 2: 2.2 .2 <0 .8... 8.9.. E 2:23 .3 m...“ a... 2.... 3.. .3. 8... <0 .8... .3... E 2:23 ”3 as m: 3.0 am... am... «2 <0 .8... .03... :. 2:23 0: .3... 3 .3 3.... a... 8.... <0 .8... 2.9. 2 2:23 33 23 3.... 2: m2 3.... of: <0 .8... t .N... 2 2:23 a: a. .0 8.0 82 8.. .6... 2° <0 .85. 0. .3. E 2:23 EN v... 8... a... 8.... 8.... 8.... <0 .8... 2.9.. 2 2:23 32 2.0 x... N. a .0... 8.... 8.... <0 .8... 2...... 2 2:23 2... «3 3 m2. 2.... 2 .N. ”3. <0 .8... N. .N... :. 2:23 .32 3. a; .3. .2. 2.... 2.2 <0 .SE 22.”... E 2:23 3 sea“. a 2 ago... a £90: 2.3 £23 3.53 .35.. 5.2 N .3255 . .2255 5:83 2 55.8% EGO—H5330: ”5:00 Mgm mwwom u EQW 170 K2 2 Ka a: 8K 2:28 8.” 3K 3.8 a.» m: K 2:28 3.: 3.. :3 no 8.”: K 2:28 v.2 3.. 38 3° 822 K 2:28 «.2 8: :.m a: K22 K 2:28 a: 3: K8 33 2.9.. K 2:28 3; SN SM 26 as: K 2:28 2.. am :3 m2 2.9.. K 2:28 8: :.N .2 8: ~22 K 2:28 88 SN 8.“ N2 :32 K 2:28 28 8: 3K «3 8.9: K 2:28 3: K2 8.” 3 2.9; K 2:28 gm «3 on.“ K: 2.2: K 2:28 2: mom 8 and 2.“: K 2:28 3.: 5m 8.” a: 28: K 2:28 Ea 2% 33 2:. 2.9; K 2:28 a? a: 2.0 gm 2.9: K 2:28 3.. mam 8.: a: cam: K 2:28 3 :55 > 3 see: > A .56.: > 3 :35 a 2 5.58:8 35805302 «:50 xanm ammom .. ZZZQm 171 ::.: $2: 2.: :2. 8.2 K:.:_ 3.: <0 2:2: :32 :20: :2: :2 :2 :2 2K 8.: :2: <0 .22: :32 :20: 2.:: :1: K:.: 2.: :2. :22 :2: <0 .22: 2.:: :20: :2 K: 8.: ::.:2 :22 8.: K22 <0 .22: 83:: :20: :2 8.: 2.2 :3: 8.:: 8:82 8.: <0 .22: :32 :20: :2. SK K.: K2: 2 3.: :2:: 2 .82 2:: :20: :K: :2 22 K2: 2:: :2:: :.:2 <0 .82 23:: :20: :2 v2 :2: K22 :2 ::.:2 2.: <0 .82 :82 :20: 8.: :2 8 3.: ::.: :2 ::.:2 2 .2 <0 .82 :82 :20: 8.: :2 :2: :22: 8.: 2:2: 2:: 2: .82 :82 :20: 2.: ::K K2: 2:2 :2 2.:: K.:2 <0 .22: 28:: :20: S2 :.: :2 ::.: :2 v2 :2 <0 .22: :82 :20: 8:: :3. 2.: :2 22 :2 :2 <0 .22: :82 :20: :2 2:2: :2: :2: :22 3:: :2: <0 .22: 2:: :20: 2.:: 3.2: 2.:: ::.:: 8.2 2: K2: 2: .82 2:2 :20: 2.: 2:2: :2:: 2: :22 :2: :2:: 2: .82 :82 :20: 2.: :2 ::.: :22 :2. :2: 2:.: <0 .82 28:: :20: :2. 8.: 8.:: 2.:: 3.: :22 :2: <0 .82 2:8: :20: 3.: K: .22 :1: 2.2: 2: 8.:: <0 .82 :2:: :20: :2 2.: :3: :2: :22 22: :2: <0 .82 :28: :20: ::.: :.K :2: :22: :22 :2: ::2: <0 .82 28:: :20: :2 :22 :2: :2:: 2.:. 2.:: :2: <0 .82 2:8: :20: 2.: :2 22:2 2: :2: :2: :2: <0 .82 :82 :20: :2: :2. :22 2...: 2:: :22 :2:: <0 .82 :82 :20: :2: SK 8.:: :.o: ::.: :2: :2: <0 .82 :82 :20: :2 :.K :2: :2: :22 K2: :2: <0 .82 88:: :20: :2 :2 :2: :22 ::2: 2.:: :2: <0 .82 :82 :20: K2 :2 :2: 2:2 2.: :2: :2: <0 .82 :82 :20: :2 8.:: K2: :2: :2:: :.:: :2: <0 .82 :82 :20: on: :2: ::.:I: :I:.:: :Im.:2 :2: :I2: <0 .82 28:: :20: >2 8:8: a 2 8:8: 2 5:20: 2:3 523 2:3 5:52 :2 : 2:52: 2 85:22: 22:82 :2 52H: mammogram—Ag: dun—GOO Mém ammom I EDD 172 ::.: ::.: ::.:2 ::.2 :82 :20: :.: 2.: 8.: 82 :82 :2:: ::.: ::.: ::.: ::2 :82 :20: ::.: :2. 2:22 8.: :82 :20: :.: :2: ::.: ::.: :82 :20: ::.: ::.: ::.: :.: 3:2 :20: 2.: ::.: ::.: :2 282 :20: ::.: :.: ::.: :2: :82 :20: ::.: ::.: ::.: ::2 :82 :20: 8.: :5. ::.: ::.: :82 :20: 2:: ::.: ::.: ::.: 282 :20: ::.2 :2 :K.: ::2 :82 :20: ::2 ::2 ::.: 82 :82 :20: :2 :.: 2.22 ::.K 82 :20: 8.: 8.: ::.: ::.: :82 :20: 8.: :.: K.:2 :.: :82 :20: 8.: :.: 3.: 8.: 282 :20: ::.: ::.: ::.:2 ::2 :82 :20: ::.: ::.: ::.: ::.: :82 :20: 2.: 8.: ::.:2 :2: :82 :20: 8.: ::.: ::.: ::.: ::2 :20: ::.: :2: :K.: 2:: :82 :20: 8.: ::.: ::.: ::.: :82 :20: ::.: :.: ::.: 2:2. :82 :20: :2: :.: :.: ::.: :82 :20: :.: ::.: 8.: ::.: :82 :20: 8.: 8.: 8.: ::.: :82 :20: :.: ::.: ::.: 8.: 8:2 :20: ::.:2 2.: ::.:2 3.: :82 :20: ::.: ::.:2 ::.: ::.: 8:2 :20: 3 22:3 > 3 5:8: > 2 5:8: > 3 22:3 2 2 8.28:: 35808222 28:80 282m :80: .. EDD 173 ::.: 8.: 8.: :.: :.: 8.: ::.: <0 2'2: :82 :20: 2.: ::.: 8.: 8.: 8.: ::.: 8.: <0 .22: :82 :20: 82 ::.: ::.: :_.: ::.: 8.: 8.: <0 .22: :82 :20: ::.: ::.: ::.: 8.: t .: 2.: 8.: <0 .22: :82 :20: 82 :.: 8.: ::.: ::.: ::.: ::.: <0 .22: :82 :20: :2 8.: ::.: ::.: ::.: ::.: 8.: <0 .22: :82 :20: 82 ::.: 8.: ::.: ::.: ::.: ::.: <0 .22: 8:2 :20: :.: 8.: ::.: ::.: ::.: 8.:_ 8.2 <0 .22: :82 :20: ::.: :.: 8.: 8.: ::.: ::.: ::.: <0 .22: :82 :20: :.: 8.: 8.: : 8.: 8.2 8.8 <0 .22: :82 :20: 8.: 8.: ::.: ::.: ::.S 8.2 8.: <0 .22: :82 :20: ::.: ::.: ::.: ::.2 2 ::.: 8.:. <0 .22: :82 :20: ::.: ::.: :.:. ::.:: 8.: ::.:: 8.: <0 .22: :82 :20: 8.: ::.: :2 .2 ::.e 8.2 2 .2 8.: <0 .22: 8:2 :20: ::.: 8.2 ::.: 8.: 2.:: 8.: :2 <0 .22: .82 :20: ::.: ::.: :2 8.2 ::.2 8.: 8.:. <0 .22: :82 :20: 8.: ::.: 8.2 8.: ::.: 2 .2 8.: <0 .22: :82 :20: :2 ::.: 8.: ::.: 8.: 8.: :.: <0 .22: :82 :20: 8.: ::.: ::.: ::.: 2.: ::.: ::.: <0 .22: :82 :20: ::2 ::.: 8.: ::.: ::.: 8.: ::.: <0 .22: :82 :20: ::.: :.: 2.2 m2 :.: ::.2 ::.:. <0 .22: :82 :20: 3 see: a 2 58:: o 22:: 2:3 523 2:3 8:82 :2 m 222: 2 55.55 2.83 : 8.28:: $5805:on 8:80 22m ammom - EDD 174 ::.: ::.: ::.: ::.: :82 :20: ::.: :2: ::.: 82 :82 :20: ::.: 82 ::.: ::.: :82 :20: 8.: 8.: 2.: 82 :82 :20: :22 ::2 ::.: ::.: :82 :20: ::2 2:2 :.: ::.: :82 :20: :.: ::2 2.: 2:2 :82 :20: 2:2 ::.: 8.: 2:2 :82 :20: :2 ::.: 8.: 82 :82 :20: ::.: 8.: ::.: ::2 :82 :20: 82 ::.: ::.: 82 :82 :20: :.: ::.: 8.: ::.: :82 :20: :.: ::.: 8.: ::.: :82 :20: 8.: 8.: :.: 82 8:2 :20: 8.: 8.: ::.: ::.: 282 :20: 8.: ::.: ::.: ::.: :82 :20: ::.: 8.: ::.:2 ::.: :82 :20: 8.: 82 ::.: ::.: :82 :20: 82 ::2 8.: :.: :82 :20: 82 2:2 8.: 2:: :82 :20: 8.: 2:: ::.: 8.: :82 :20: 3 22:3 > 3 see: > 2 8:8: > 3 22:3 9 : 828:: 8:285:32 8800 22m ammom .. @403 175 oo.H OH.OI oo.H om.on wm.o oo.H vH.o om.on om.o oo.H 3HH03>B 38mm0h>8 HEmuom>B SHHmSQH ma.on on.o NN.o wv.os oo.H BEMHOmQB mm.o| Hm.o hm.o NN.OI om.o oo.H mN.o om.on 00.0: bN.o mH.os mo.ou oo.H 25:20:29 mfiwmumHmmza 00.0 mm.on mn.on mN.o Hm.OI mn.0l Hr.o oo.H mHH mm.on om.o mm.o mH.oa mm.o mm.o vm.on mm.OI oo.H oo.o Ho.OI HH.OI om.o ma.o: mH.OI mm.o mN.o vH.0I oo.H $122<3H 3 2:20: 2ImI >1 fl 2 :H a :mcoflum>nwmno mo HmnEsz B B B B B B onmHBmH:: 3:§::>:: ._:Eeo::>:: 3:. _:3:o:: 3:220::o:: 2:588:02. :_m::2m3:: 83:23:: :.::o..:xm§:: «22820:: 177 vm mH v0 mm Hmuoe vm mH H O m Hm v mm N N OOH O O Hm H uomuuoow m N H xflnume coHumonHmmMHo UmchxxUMO hm mH mm Hm HMMOB OOH OH O O m mm m we O N OOH O O Hm H uomuuoow m N H AmCEDHoo OuCH UmeHmmmHo mmHuoompmo 30: CH mmmmov xHuumE coHuMOHMHmmMHU 0000.0 u no:0 00m 00 u 00 00HH.0N :0 .xouaam mHH N OH H mu mN:0.0 u munEmH mUnEMH .mxHHz "oHnuaua> ucwmiouo an HAHZKh :90: .I910 «Magoo vacuum co nauaaaad undawaduoowa .H «unpack uwahaddd undawafluonwn .0 anzummd 178 Om OH OOH OH vm m OO O uumnnoow m mm 0H OOH OH mm H OOH O uomuuoom m OO Hm HMHOB o O m mO H N H om H N H xHuumE GOHumonHmmmHo UmuchxUMO mO Hm Hmuoe O O m OO O N O Hm H N H AmcESHoo oucH UmHOHmmMHU mmHuoomumo 30: CH mwmmov xHuumE coHumonHmmmHU 0000.0 : noun 0H0 mHH N NH O u On 0000.00 um .xouaam : On 0H00.0 u mnnqu mugEmH .mxHHz "cand«u¢> oaamdouo on MAHZIH £90: .quu auuaoo vacuum no nwuhaacu undcfladuoadu ouwtminm uvuuxxonm .N 179 Om VH o H H O o O O OO o m o H o H o O Nm 0 0 HH H O o O m mm v o m Mv O o N O OO o O H o OH N N m OOH o O o o o O O N HO O o H o O o OH H uowuuoow m h m w m N H xHHumE COHUMUHMHOOMHU UwMHCxxumh. m om OH O OH vv NH O «H HMCOB OOH OH o o o O o o m OOH o m o o o o o h OOH o o NH 0 O o o O mm m o m vv 0 o N v OO O O O O NH H N m OOH O O o o O O O N HO O O H o o O OH H uomuuoow O O O O m N H AmCESHoo ouCH UmeHmmMHo mmHuommumo 30H CH mmmmov xHuumE COHUMOHmemmHU my 0m0O.HH um .xoumam On 0NH0.0 u mnnqu mnnEmH .mxHHz 0000.0 : Comm 000 00 00H 0 0H "candaua> unamiouu on maxim and: .Iulv duvacu uncoom C0 nwuhdaad »CdCdaduonan .m Om OH O OH Ov VH O HH HmuOB OOH OH O O O O O O O OOH O O O O O O O O NO O O HH H O O O O OO O H m vv O O H O OO O O O O OH O N m OOH O O O O O O O N OO O O N O H O O H uumuuoow O O O O m N H xHuumE COHumoHMHOOMHU UmeCxxomO mO OH O OH Ov OH O NH HmuOB OOH OH O O O O O O O OOH O O O O O O O O NO O O HH H O O O O NO H O O OO O O O O OO O O O O OH O N m OOH O O O O O O O N HO O O O O H O OH H pomuuoow O b O O m N H AmCECHoo ouCH UmHOHmmmHo mmHHOOmumo 30H CH mmmmuv xHHumE COHumoHOHmmmHO 0000.0 u noun 000 00 : On 0000.:H um .xouaam 00H 0 0 n ma 0000.0 : munsmq mnnsmH .mxHHz "anuHuu> oaHmaouu nu quzsu Ath .uuuu auuauu vacuum ao uHuMHuau uauaHaHHuuHu uthmunm uvuutxuam .v 181 OO «O OO OOH uomuuoow OO OOH OO OOH uowuuoow MOMKD HO HO HMuOB H O O OO O N O HO H N H xHuumE COHumonHmmmHo UmmHCxxumO HO Hm HMuOB O O m HO O N O HO H N H AmCECHoo ouCH UmHOHmmMHo mmHuoomumo 30H CH mwmmov xHHumE CoHumonHmmmHU 0000.0 u noum 0HH OHN N ON OH u On HmOm.0N um .xouaam mu ON:0.0 u munamq mnnEmH .mxHHz "anuHHu> oaHmaouo nu Nqnzifi aqu .Aauuoauwumunwnm_uau auwumrujz oaHquuaHv Ivan uuuauu uauuum ao «Huhduau uauaHaHuuuHo .n 182 vO OO O Ov O NO OO OO OO NO pumuuoow OO OOH mm OO OO OOH OO OO OOH HO nu mbzuu auHx .HaOanuwdeuwnm_uau auwumtujz oaHuaHuaH. lulu auuauu uauuum ao ONOOOMOOOOQ‘ 03 H H OOOMOOOOKO N ODOOOOOONOO N OOOOOOHHO KO (\OOOHOONMO OOOOOOVNO KO KO \OOOOVOOOOO L0 OOOVOHOOO O OOO0.0 WHOHMHOOOH l‘ H KO H HOOMNOOOO H v noun NHH VOOOONHNHOH v m v H V OOOOOHOOO m HOO O OO OH on H MHHHOOOOOO \0 NH HmuOB NOVNOOOOOO HOOHNOOOOO HNMVLDKDFCDON xHuumE COHumonHmmmHu UmOchxUMO V H OOVOOOOOO N AmCECHoo ouCH UmeHmmMHu mmHuoomumo 30H CH L0 0H Hmpo: OLOOOOOOOO OOHNOOOOO UHNMVLOKOOmm uumnuo O mummov xHuumE COHumoHOHmmMHU MU MU OvNO.O um .xouaad OOO0.0 u MUQEOH MUQEMH .mxHHz "anuHHab oaHanuo «Huaduau uauaHaHuuan .O 183 NO NO NO OO uownnoow OO VO vO OOH uowuuoow AmCSsHoo ouCH UmHOHmmMHo mmHHOOmumo 30H CH mummov 0000.0 u Doug 00H MHKOV' MOLDKD NON N ON OH OO mm Hmuoe H N O OO H N O om H xHuumE COHumonHmmMHo UmmHCxxomO OO Nm HmuOB O H O OO O N O HO H N H xHuumE CoHumonHmmmHU n my O0N:.ON um .xoumam u On :NOH.0 u mnnsmq mnnEmH .mxHHz uanuHHu> oaHmaouo nu MAHZKh 50H: ~luau uuuauu >220 ao aHuhHuaa uauaHaHHuuHa .O 184 OO OO OO NO OO OO OOH NO uomuHOUO HO OO OOH OOH OO OO OOH OOH uomuuoow \D H Oooomoom (I) H OOOOMOOU) H H O OH «O OH OH OH Hmuoe O H O H N O O H. O H O H O O O HH H O O O O H O NO O O m O O H O OH N N O O O O O O O N O O O N O O H O O O m N H xHHumE COHuMOHMHmmmHo UmuHCxxUMO 0.504 O OH «O OH O OH HmuOB O H O H O O O O O O O O O O O NH O O O O O O O HO O O N v O O O OH O N O O O O O O O N O O O O O HH H O O O m N H AmCECHoo ouCH UmHmHmmmHo mmHuoomumo 30H CH mmmmov xHHumE CoHumonHmmMHU On 0000.NH um .xonmaa mu m0H0.0 u mnnsmg mngmH .mxHHz 0000.0 u noun 000 00 HmH 0 0H ”uHAuHuu> oaHmuouu nu mbzuu auH: .uuuu uuuauu >220 ao uHuhHuau uauaHaHuunHa .O OO OO vO OOH uomuHOOO NO vO OO OOH uowuuoom momm MOKDKO OO Om HmuOB H H O OO O N O HO H N H xHuumE COHumonHmmmHo UmOHCxxomO OO vm HmuOB H O O NO m N O HO H N H AmCECHoo ouCH UmeHmmmHo meuomwumo 30H CH mummov xHume COHumoHOHmmMHU 0000.0 u noun 00H va N ON OH On :0OO.HN um .xouamm mu ommH.0 u mugfimq mnnemH .mxHHz ”anuHuub oaHmaouo nu HAHZGH auH: .uuuu uuuauu an ac nHuhHuau uauaHaHuuuHa .O 186 I‘M“: HO HN v O H OH Ov OH OH O HmuOB NO OH H O O N O O O O O O O O m O O O O H O O OO O H O O O H O O O O O H O O O O v O O O O NO O O O O HH H O O O O OO O H H H m Ov O N H v OO O H O O O o NH N O m OOH O O O O O O O O O N vO O O O O O O H O O H uowuuoow O O O O O v m N H XflHUME COHUMUHHHWWMHU UOWHC¥¥UMU MO ON v O O OH Ov OH O OH HOpOB .o/o OO OH O O O N O O O O O 1: ON O H N O O O O H O O OO O H v O O O O O O O OO O O O O O N O O O O OOH O O O O NH O O O O O OO O H O O m vv O H H O OO O H O O O O OH O O m OOH O O O O O O O O O N NO O O O O H O H O O H pumuuoow O O O O O v m N H AmCECHoo ouCH UmHMHmmmHu mmHuoompmo 30H CH mmmmov xHuumE CoHumoHuHmmmHU Ow 0m:N.0 um .xouamm On mmH0.0 u mnnsmq mUQEmH .mxHHz 0000.0 u noga NH: 00 00H 0 0H "anuHug oaHmaouu nu ODE—o fiHx .36 uuuauu .mb ao uHuaHuau vauaHfiHuuuHa .OH NO OO OO OOH pomHHOUO OO vO OO OOH uumuuoom ! 0N 0: «m Hmuo: mH N 0 m 0 0: m N WM 0 0 H0 H 1. m N H XHHUME COHUMUHMHmmmHO Umwflcxxumh 0H H0 Nm Hmuo: OH H 0 m m 00 H N 0 0 H0 H m N H AmCECHoo ouCH UmHmHmmMHo mmHHOOwumo 30H CH mmmmov xHuumE CoHumoHOHmmmHu 0000.0 u noun 0NH OvN N ON OH On HmN0.mN um .xouaam Ow 0O00.0 u mnnEmH anEMH .mxHHz "anuHuu> oaHmuouo nu MAHZIH auH: .uuau uuuauu zzzm ao uHuOHuau aaaaHaHuuan .HH OO OH O l‘ H 0‘ L0 V r-i \D V‘ H HMHOB OO OO NO OO OO OO HO uomuuoow OOOOMOOID [\OOOHOMO LOOOHMNOu—l H VOOOKOHHH O MHHNOOOO NOVHOOHO HOOHMOOO £59: 8383388 838.580 w 1 HO OH O LO H H KO V H II) OH HmuOB OO OO NO OH O O OO O O O O O H OO OOH OOH uomuuouw [\OOOOOV‘O LOOOOMNOO V'OOOmt—lc—IH L0 MOOQ‘OOOO u—l NomOOOOO r-lr-‘IOHNOOO HmCEDHoo ouCH UmHmHmmmHo mmHHOOmumo 30H CH mummov xHuumE COHumoHMHmmMHU On mmmO.NH um .xouaam On OOH0.0 u mUQEMH mvnsmH .mxHHz 0000.0 u noun NH0 00 mNH O 0H "anuHuub oaHmaouo nu maxim avH: .uuuu uuuauu mzzm a0 uHufiHuaa uauaHaHHuan .NH HO OO OO mO uomunoow OO OOH NO OO uumuuoom MOLOLD MOMP OO HO Hmuoe H H O OO O N O OO H N H xHuumE COHumOHOHmOMHo UwOHCxxomO NO OO Hapoe O O O HO O N H OO H N H HmCECHoo ouCH UmHOHmmmHo mmHuoOwumo 30H CH mummov xHuumE COHpmonHmmmHo OOO0.0 u Dona NOH OON N ON OH NO O0OH.ON um .xonamm mu mHmH.0 u mnnEmH monsmH .mxHHz "anuHHu> oaHmaouo nu HAHZKh 39H: ~uuu—u uuuauu zmzom ao uHuhHuaa vauaHaHuuuHo .OH 190 HO OO OO NO OO OO OO NO uuwuuoow OO OO OO OOH OO OO OO OO uomuuoow ON GOOOVOOLD m H GOOOMOOKD H H H m ON mO OH NH O N O O O O O H O H O HH H O O O O OO O O O O O OH H O O O H O O O H H H O O O m N Hmpo: V N HHONHOOO :4vaum xHHumE COHumonHmmmHu vmeCxxUMO V v N N L!) H N OH [\OOOOOV‘O LDLDOHIDNOH H VOOOHOHO Ln MHHmOOOO NHONOOOOO HmCEsHou ouCH UwHmemmHu mmHHOOwumu 30H CH 0000.0 u noun 00: 0O : 00H 0 0H "anuHHu> oaHmaouo uu ODZHU 30H: .uuufl uuuauu zazam ON HmuOH HNONHOOO mmmmov xHHumE COHumonHmmmHo On On GO m0H0.NH um .xoummm 0NN0.0 u mnnsmq mcnsmH .mxHHz uHuMHuaa uauaHaHHuuHa .OH OO OO OO HO uomuuoow OO OO OO OO uowuuoow OO ON OH O OO HO OH N m :ILIIIIIIIIIIH ONH H0 Hmuo: N O m mNH 0 N O” 0 H0 H 1 N H xHuumE COHumonHmOMHo meHCxxUMO ONH HO HmuOB H O m ONH O N O NO H N H HmCECHoo OuCH UwHmemMHu mmHuommumo 30H CH mummuv xHuumE CoHumoHuHmmMHo OOO0.0 u QOHQ OHN NOO N ON OH n On OOH0.0N um .xouaaa On 000N.0 u mUQEmH mngEmH .mxHHz HanuHuu> uaHmaouo uu MAHZKH CUH: .uuuv auuauu mmZfl ao uHuOHuau uaaaHaHuuuHa .OH OO Om KO 0 (\I OOH KO r—i ON (I) H Hmuo: OO O OO O OO N OO H OO O O H O r-i OO mO pumHHOUO H [\OOOMOMO LOOOOr-‘INHKO VOOOMHHH O H MOHNOOOM NOVHHOHN HMONNOOH :4vaum xHHumE COHumonHmmmHo waHCxxomO mu 1 OO HO KO 1‘ H OOH OH HmuOB I\ H l'\ OO OO OO ON O H OO OH O O H O H OO OOH OO pomunoow [\OOONOVO LOOOONMON V'OOOlDt—ic-CO O H MOOQ‘OOOM NOWOOOON HMOHNOHH 1'4vaum HmCEsHou ouCH UwHOHmmmHu mmHHOOwumu 30H CH mmmmov xHHumE CoHumoHOHmmMHo 0000.0 u noun :moH 0O u mu N00:.HH um .xoummm 00N 0 0H u mu 00O0.0 u munEmH mngEmH .mxHHz "anuHuu> oaHmiouo nu mbznw auHx .uuuu uuuauu Quad ao uHuhHuau uauaHaHuuuHa .OH OO OO HO OO uounnoow OO OO OO OOH powHHOUO Om NN mH MOGN NO Om Hmuoe O O O OO O N O om H N H xHuumE CoHumonHmmMHo UmOHCxeMO OO Om HmuOB O O O OO O N O HO H N H HmCEsHoo ouCH UmHOHmmMHu mmHHOOwumu 30H CH mmmmuv xHume COHumoHHHmmmHU 0000.0 n coma va OON N ON OH u mv omNm.mN um .xonaam u On 0O0H.0 u «unamq munEmH .mxHHz "anuHuu> oaHmaouo uu NQHZKH auHx .uuuv uuuauu zzmb a0 nHuMHuau uauaHaHuuuHa .OH 194 :LIIIIIIIIIIIIIH¥11 0: 00 0 HN m0 0N O mH Hmpo: NO OH O m H O O H O O H O O O O H O O NO H O NH O O O O O HO O O O OO H O N O OO H O O O HH O m m OOH O O O O O O O N NO O O O O N O O H uowuuoom O O O O m N H xHHMME COHuMOHmemmHU UmwflcxxUMO MN MO HO O OH OO OH O OH Hmuoe NO HN O N H O O H O OO H O O O O O O O NO H O NH O O O O O NO O O O OO H O N O OO O O O O OH O N m OOH O O O O O O O N HO O O O O H O OH H uomunoow O O O O m N H HmCEDHou ouCH UmeHmmmHo mmHHOOmumo 30H CH 0000.0 u noun m0: 0O 0mH 0 0H "anuHuub oaHmaouu ua mbznu 39H: .uuuu uuuauu zzmb mwmmov xHHumE COHumUHHHMMMHU mu mm:m.0 um .xouaam mu 0O00.0 u mnnEmH mugEmH .mxHHz ao uHuhHuau vauaHaHHuuHa .OH OO OO OO OO uuwnuouO HO OO OO OO pomHHOOO MOON“) mom“) I 1. OO OO HmuOB O N O OO O N H OO H N H xHHumE CoHumoHOHmmmHo UmeCxxUMO OO NO HmuOB O N O OO O N H OO H N H AmCECHoo ouCH UmeHmmmHo meuoowumo 30H CH mummov xHHumE CoHumoHMHmmMHU 0000.0 u gong 00H NOO N ON OH u mu Omm0.0N um .xouQQO u On NNmH.0 u mugfimH mngfimH .mxHHz "anuHuu> uaHmaouo nu MAHzah ath .uuuu uuuauu AzUD ao «Humauaa uauaHaHuuuHa .OH 196 Hmuoe 1—0 m HVMHHNOO ‘4vaum xHuumE COHumonHmmmHo UmmHCxxomO OO OH O NN OO ON OH OO OH O H O H N OO O O O H O H OO O O OH O O O OO O O O OO O N OO O O O O OH O OO O O O O O O OO O O N O O O uuwuuoow O O O O m N OO ON O ON OO OH HH OO OH O O O H H OO O O O H O O OO O O OH O O O OO O O O OO O H OO O O O O OH N OO O O O O O O OO O O N O N O pomHHOUw O O O O m N AmCESHoo ouCH UwHOHmmmHo mmHHOOwumo 30H CH 0000.0 u noun O:0 0O OOH O 0H u HanuHuu> uaHmaouo nu maxim auHx .uuuu auuauu AZUD Hm Hmpo: HKONHHHOO :4vaum mwmmnuv XHHHME COHfiMUflHHmmmHU On ONNO.NH «O .xouaam On OON0.0 u mnnEOH OOQEOH .mxHHz ao anhHuau uauaHaHuuuHo .ON 197 APPENDIX D: PHYLOGENETIC ANALYSIS CHARACTERS A list of centrum characters and their states coded for the matrix 1. Centra typically fluted cylinder simped (0); centra typically non—fluted cylinder shaped (1 ). 2. All centra non—hourglass shaped (0); at least some centra hourglass simped (1). 3. All centra non-modified cylinder shaped (O); at least some centra modified cylinder shaped (1). 4. Medio—lateral breadth > dorso—ventral height (0); medic—lateral breadth = dorso— ventral height (1); medic—lateral breadth < dorso—ventral height (2). 5. Centrum length/diameter ratio typically > 0.95 (O); centrum length/diameter ratio typically between 0.6 and 0.95 (l); centrum length/diameter ratio typically < 0.6 (2). 6. Width at apices of double cone/diameter ratio < 0.85 (0); width at apices of double cone/diameter ratio typically > 0.85 (l). 7. Basidorsal and basiventral arch—cartilage foramina oval (0); Basidorsal and basiventral arch—cartilage foramina square or rectangular in some centra of the vertebral column (1). 8. Basidorsal and basiventral arch—cartilage foramina extend into centrum rims (0); basidorsal and basiventral arch—cartilage foramina do not extend into rims (1). 9. Basidorsal arch—cartilage foramina length/diameter ratio typically > 0.57 (0); basidorsal arch—cartilage foramina length/diameter ratio typically < 0.57 (1). 10. Basidorsal arch—cartilage foramina length/ basidorsal arch—cartilage foramina width ratio typically >3.0 (O); basidorsal arch—cartilage foramina length/ basidorsal arch— cartilage foramina width ratio typically < 3.0 (1). 198 11. Basidorsal arch—cartilage foramina width/dorsal interforaminal wall width ratio typically > 1.0 (0); basidorsal arch—cartilage foramina width/dorsal interforaminal wall width ratio typically < 1.0 (1). 12. Dorsal interforaminal wall width/width at apices of the double cone ratio typically < 0.28 (O); dorsal interforaminal wall width/width at apices of the double cone ratio typically > 0.28 (1). l3. Basiventral arch—cartilage foramina length/diameter ratio typically > 0.54 (0); basiventral arch-cartilage foramina length/diameter ratio typically < 0.54 (1). 14. Basiventral arch—cartilage foramina length/basiventral arch—cartilage foramina width ratio typically >2.5 (O); basiventral arch—cartilage foramina length/basiventral arch— cartilage foramina width ratio typically < 2.5 (1). 15. Pores absent (0); pores small (1); pores large (2). 16. Pores absent (0); pores dense at rim (1); pores encircling rims and foramina (2); pores scattered (3). 17. Interior angle of intermedialia acute in some centra (0); interior angle of intermdialia at 90 degrees or obtuse (1). l8. Diagonal calcified lamellae absent (0); diagonal calcified lamellae length less than one—half the distance to the surface of the centrum (1); diagonal calcified lamellae length typically more than one-half the distance to the surface of the centrum (2). 19. Diagonal calcified lamellae absent (0); diagonal calcified lamellae typically thick (1); Diagonal calcified lamellae typically thin (2). 20. Calcified projections surrounding the basidorsal and basiventral cartilages absent (0); calcified projections surrounding the basidorsal and basiventral cartilages present (1). 199 21 . Basidorsal and basiventral cartilages bulbous in cross—sectional view (0); basidorsal and basiventral cartilages straight in cross—sectional view (1). 22. Basidorsal and basiventral cartilage margins smooth (0); basidorsal and basiventral cartilage margins irregularly shaped (1). 23. Basiventral cartilages project from the calcified double cone in a straight path (0); basiventral cartilages project from the calcified double cone in a bowed path in some centra (1). 200 A list of whole specimen characters and their states coded for the matrix The following list of apomorphies for Carcharhiniformes was taken directly from Compagno (1988). When present in a taxon, each of these apomorphies is coded as 1. When an apomorphy is absent, it is coded as 0. 1. First dorsal over pelvic bases N . Loss of eyelid depressor muscle 3. Reduction of second dorsal fin & . Reduction of clasper parts 5. Loss of rostral node 0\ . Postorbital processes expanded laterally \l . Occiput expanded 8. Loss of fourth lower extrabranchial cartilages 9. Triakid tooth type with anaulacorhizous roots 10. Strongly developed orbital notches on the cranium 11. Bifurcated (triakids) or homlike (“higher” groups) postorbital processes 12. Possibly loss of a fiised clasper groove 13. Lateral teeth asymmetrical and semibladelike 14. Pleats variably developed on claspers 15. Nasal capsules laterally expanded and ovoid 16. Cranium broader l7. Differentiation of dentition into compressed, broadened anteroposterior teeth differentiated from the smaller, narrower medials 201 18. Ectethmoid condyles present 19. Subethmoid fossa present 20. Posterior part of the subnasal plate expanded anteriorly under ectethmo id chamber 21. A pair of medial ectethmoid condyles present 22. Subnasal plate expanded anteriorly, separating the ectethmoid fossa from the nasal fontanelle 23. Mouth triangular 24. Teeth developed as pavement 25. Lower teeth enlarged at symphysis 26. Teeth expanded onto the underside of the lower jaw 27. Internal NLEs with very deep subocular pouches 28. General reduction of spiracles (except Galeocerdo) 29. Stronger monognathic heterodonty 30. Presence of precaudal pits 31. Plesodic pectoral fins 32. Mesorhipidion on the clasper 33. Rippled dorsal caudal margin 34. Nasal fontanelles separate from nasal apertures or lost 35. Primary supraorbital crests lost 36. Extension of levator palatoquadrati muscles into orbits 37. Loss of oral and gill raker papillae? 38. Increase in fin size 39. Anterior movement of first dorsal base 202 40. Loss of clasper books 41. Increase in size 42. Increase in length of distal pectoral radials 43. Extremely strong mono gnathic and dignathic heterodonty (paralleled in more derived carcharhinids) 44. Aortic and efferent hyoidian arterial canals present 45. Further enlargement of gill openings (three or more times eye length) 46. Lower jaw truncated at symphysis 47. Distal crown edges of upper anterolateral teeth arcuate, not angled 48. Mesial serrations on upper teeth and some lowers in adults and subadults 49. A toothless space at the upper and lower symphysis 50. Many tooth rows with imbricate overlap 5 H . Basal ledge ad groove absent from teeth 52. Shifi to osteodonty (living species) 53. Second dorsal more reduced (less than 0.6 of first dorsal length) 54. Shortened terminal lobe of caudal 55. Reduced nasal fontanelles 56. Greatly elongated and narrowed anterior fontanelle with deflected edges 57. Keel on basal plate very strong 58. Keel present in subethmoid fossa 5 9. Efferent hyoidian canals greatly expanded and opening under hyomandibular facets 60. Postorbital processes reduced, not triangular and broad—based, with cylindrical base and expanded, hastate tip 203 61 62 63 64 65 66 67 68 69. 70. 71. 72. 73. 74. 75. 76. 78. 79. 80. 81. Sphenopterotic ridges reduced Ectethmoid foramina present Scroll intestinal valve present Second dorsal fin more or less reduced relative to the first Uniquely enlarged, modified, inflated otic capsules, that extend in front of orbital fissures Postorbital processes situated about at midlengths of otic capsules Lengthened upper labial furrows, longer than those of hemigaleids, Rhizoprionodon, and triakids Broadly arched tooth crowns with heavy serrations and serrated cusplets, secondarily anaulacorhizous tooth roots Keels on caudal peduncle (also in Prionace) Ectotic processes and precerebral pit and keel on cranium High vertebral counts Unique color pattern Cusplets replaced by blades on teeth (also Scoliodon and Sphyrnidae) Greatly elongated preanal ridges and anal radial segments supporting them Very narrow basal plate at orbital notches Stapediocarotid foramina present of basal plate Posterodorsal medial ectethmoid foramina on the ectethmoid condyles Shortening of labial furrows Posterodorsal medial ectethmoid foramina on the ectethmoid condyle Slender body 204 82. 83. 84. 85. 86. 87. 88. 89. 90. 91. 92. 93. 94. 95. 96. 97. 98. 99. Curved—triangular teeth Enlarged medial upper teeth Gill rakers Long, narrow pectorals (also in some large Carcharhinus species) Weak keels on the caudal peduncle A dorsal flare and large epiphysial foramen on the anterior fontanelle Broad—based preorbital processes Gradually sloping sides on otic capsules Blue coloration Brow ridge above eyes Erect, narrow—cusped teeth in both jaws, with serrations reduced or almost entirely absent. Cephalo foil present Endonarial grooves present Labial furrows vestigial or absent (also lost in some carcharhinids) Greatly depressed medial rostral cartilage Expanded, platelike rostral node Depressed and laterally expanded nasal capsules and intemasal septum Nasal capsules entirely lateral to cerebral cavity 100. Subethmoid fossa a transverse slot 101. Slitlike nasal apertures 102. Ectethmoid chambers lateral to ectethmoid condyles 103. Ventral edge of anterior fontanelle at base of medial rostral cartilage 205 104. Double parietal fossae 105. No paired dorsal aortae 106. A single pair of efferent hyoidian foramina on basal plate (paralleled in the carcharhinid Rhizoprionodontini) 107. Development of sphyrnid secondary supraorbital crest 108. Preorbital process far distal on the nasal capsules 109. Postorbital processes extending through middle of levator palatoquadrati muscles 1 10. Levator nictitans muscles far distal on cephalofoil 1 11. Originating on proximal shaft of postorbital processes 1 12. Labial cartilages absent 1 13. Clasper rhipidion lost 206 N ~85 s 335 OE: maSEchwo :otquoz 839:1 :obocothnEm 22803.06 ”$55.0: @3832 «2.36206 02%: mEmEoEoH «I 355% on FOOOPNFNNONNNOO FOFPFFNFFONN‘.°° soOOOOMNOv-OOMNOO FOODOFOFs-Os-v-I-OOO aOOOOOOOOOrs-s-OO aOOOOOFOOOOOrOr EOOOOOFFOFFPOFF aOOoowwopFwaoo FOVOOOPv-V-v-I-u-Fu-v- :OOOOONNOPPFPFF Q Q N ‘ SECOOOOPv-OFFVPFO ,‘EOOOOOOOOOu-u—OOO =OQOOOOOPOPFPOP FOQOPOPFOFFFFOO GOOOPOFPOPPs-Fx-O OOOOOOFFPOPFFPP NOOFv-OOs-OPv-F'POO OOOOOOPOOOFFOv—P DOOOFONFFNNNPNO QOOOPOPFOPPFNOO GOOOOOOs—OOOFFOO NOOOOOOOOOOOPOO FOOOOOPFOFFPFFF 8832552930 .8 88022.0 5.5.80 3 5.5!: Sun 207 88: 252.800 Ea...— uo>tou .uoaco=:_Ea:2ao .8 23022.0 5.53% 0.2.3 .0 x52: Sun. 208 83: 0:53:30 E0... uo>tou .woaco25522a0 .8 23022.0 c2502.» 0.2.3 “.0 55!: Sun. 209 8am: ocuuQEoo E9: 83.2. 63:025....2080 .8 280220 5.502.» 29.3 no 5..qu Sun 210 88: 0:39:00 Ea... $03.20 .00::8.=.Eu..0..uo .2 28022.0 .5502.» 0.9.3 “.0 .353. Sun. 211 83: 0.52.200 Eat 502.03 6028322030 .2 28022.0 c2502.» 22.3 .0 x30... Sun. 212 2.33m ”05232.00 08.3% 0026.5. 0888802 02000030 0020me 20.0.2.3. 2.0.0.6300 9.0.2.... OOOOOOOOOOO OOOOOOOOOOP OOOOOOOOOOP OOOOOOOOOOF OOOOOOOOOOF OOOOOOOOOOP OOOOOOOOOOP OOOOOOOOOOV OOOOOOOOOOF 00000000001- OOOOOOOOOOF 3550.38 a: N: O: a: OS. 8.. O2. 06—. 2:. 3r 8.. 22022.0 E2200 505.3 OOH O9. 3.. 00.. NOH FOP 8.. ONH ONH ONO ONH 20.09.20 EEE8 5.; 83... 252.200 80.... uo>t80 60.28.522.200 .8 28022.0 080.08.» 0.8.3 .0 x52... 83 213 REFERENCES Agassiz, L., 1833-1843, Reserches sur les Poissons fossiles: Neuchatel and Soleure, 5 volumes, 1420 p. Agassiz, L., 1843b, Contenant l’Histoire de l’Ordre des Plaooides. 390+32 p., atlas. Neuchétel, Suisse. Applegate, S. P., 1967, A survey of shark hard parts, in Gilbert, P. W., Mathewson, R. F., and Rall, D. P., eds., Sharks, skates, and rays: Baltimore, Maryland, The John Hopkins Press, p. 37—67. Blainville, H. M. D. D., 1816, Prodrome d’une distribution systematique du regne animal: Bulletin de la Scientifique Société Philomathique de Paris, v. 8, p. 105— 124. Bonaparte, C. L., 1838, Selachorum tabula analytica: Nuovi Annali Delle Scienze Naturali, ser. 1, 2, p. 195—214. Branstetter, S., and McEachran, J. D., 1986, Age and growth estimates of four carcharhinid sharks common to the Gulf of Mexico: A summary paper, in Proceedings, Second International Conference on Indo—Pacific Fishes: Tokyo, Ichthyological Society of Japan, p. 361—37 1. Brown, C. A., and Gruber, S. H., 1988, Age assessment of the lemon shark, Negaprion brevirostris, using tetracycline validated vertebral centra: Copeia, no. 3, p. 747— 753. Budker, P.,l97l, The life of sharks: New York, Columbia University Press, 222 p. Cailliet, G. M., Radtke, R. L., and Welden, B. A., 1985, Elasmobranch age determination and verification: A review: Indo—Pacific Fish Biology, Second International Conference on Indo—Pacific Fishes, Uyeno, T., Arai, R., Taniuchi, T., and Matsuura, K., (eds), p. 345-360. Cantor, T., 1849, Catalogue of Malayan fishes: Journal of the Asiatic Society of Bengal, v. 18, p. 983—1443. Cappetta, H., 1987, Chondrichthyes H, Mesozoic and Cenozoic Elasmobranchii: Handbook of Paleoichthyo logy, v. 3B. Case, G. R., and Cappetta, H., 1997, A new selachian fauna from the Late Maastrichtian of Texas: Muencher Geowissenschafi Abhandlungen, v. 34, p. 131—189. Clement, J. G., 1992, Re—examination of the fine structure of endoskeletal mineralization in chondrichthyans: Implications for growth, aging and calcium homeostasis: Australian Jourml of Marine and Freshwater Research, v. 43, p. 157—181. 214 Coates, M., and Sequeira, S., 2001, A new stethacanthid chondricthyan fiom the lower Carboniferous of Bearsden, Scotland: Journal of Vertebrate Paleontology, v. 21 , no. 3, p. 438—459. Compagno, L. J. V., 1970, Systematics of the genus Hemin'iakis (Selachii: Carcharhinidae): Proceedings, California Academy of Science, v. 39 p. 257—272. Compagno, L .J. V., 1973a, Ctenacis and Gollum, two new genera of sharks (Selachii: Carcharhinidae), and related genera: Proceedings, California Academy of Science series 4, v. 38, p. 63—98. Compagno, L. J. V., 1973b, Gogoliafilewoodi, a new genus and species of shark fi'om New Guinea (Carcharhiniformes; Triakidae), with a redefinition of the family Triakidae and a key to the genera: Proceedings, California Academy of Science series 4, v. 39, p. 383-410. Compagno, L. J. V., 1973c, Interrelationships of living elasmobranchs, in Greenwood, P. H., Miles, R. S., and Patterson, C., eds., Interrelationships of fishes: Zoological Journal of the Linnean Society, Supplement 1, v. 53, p. 15—61. Compagno, L. J. V., 1984, FAO species catalogue. Vol. 4: Sharks of the World. An annotated and illustrated catalogue of shark species known to date. Part 2. Carcharhiniformes. FAO Fisheries Synopsis, v. 125, p. 251—655. Compagno, L. J. V., 1988, Sharks of the Order Carcharhiniformes: Princeton, New Jersey, Princeton University Press, 486 p., 35 plates. Compagno, L. J. V., and Springer, V. G., 1971, Iago, a new genus of carcharhinid shark, with a redescription of I. Omanensis: Fisheries Bulletin, v. 69, p. 615-626. Daniel, J. F., 1934, The elasmobranch fishes: Berkeley, University of California Press, 332 p. De Carvalho, M. R., 1996, Higher—level elasmobranch phylogeny, basal squaleans, and paraphyly, in Stiassny, M., Parenti, L., and Johnson, G., eds., Interrelationships of fishes: Academic Press, p. 35—62. Fowler, H. W., 1941, The fishes of the groups Elasmobranchii, Holocephali, Isospondyli, and Ostariophysi obtained by United States Bureau of Fisheries Steamer Albatross in 1907, chiefly in the Philippine Islands and adjacent seas: Bulletin of the United States National Museum, 100, v. 13, 879 p. Gill, T., 1872, Arrangement of the families of fishes, or Classes Pisces, Marsupiobranchii, and Leptocardii: Smithsonian Miscellaneous Collection v. 247, 49 p. Gill, T., 1893, Families and subfamilies of fishes: Memoirs, National Academy of Sciences, v. 6, p. 125—138. 215 Goodrich, E. S., 1930, Studies on the Structure and Development of Vertebrates: Chicago, The University of Chicago Press, 837 p. Gottfiied, M. D., 1999, Fossil shark vertebral centra: An overlooked data set?: Journal of Vertebrate Paleontology, Abstracts with Programs, v. 19, no. 3, p. 47A. Gray, J. E., 1851, List of the specimens of fish in the collection of the British Museum. Part I. Chondropterygii. 160 p. British Museum (Natural History), London. Hasse, J. C. R, 1879—1885, Das natfirliche system der Elasmobranchier auf grundlage des baues und der entwicklung ihrer wirbelsaule. Eine Morphologische und Palaontologische Studie. Allgemeiner Theil, 76 p., 1879, Besonderer Theil, 285 p., 1882, Erganzungsheft, 27 p., 1885. Hoenig, J. M., Walsh, A. H., 1982, The occurrence of cartilage canals in shark vertebrae: Canadian Journal of Zoology, v. 60, p. 483-485. Huxley, T. H., 1880, On the application of the laws of evolution to the arrangement of the Vertebrata, and more particularly of the Mammalia, in Proceedings, Zoological Society of London, p. 649—661 . Jordan, D. S., 1923, A classification of fishes including families and genera as far as known: Stanford University Publications, University Series, Biological Sciences, v. 3, p. 77—243. Jordan, D. S., and Evermann, B. W., 1896, The fishes of North and Middle America: United States National Museum, Bulletin, v. 47, Part 1, p. 1—1240. Kajiura, S. M., Fomi, J. B., and Summers, A. P., 2003, Maneuvering in juvenile carcharhinid and sphyrnid sharks: The role of the hammerhead shark cephalofoil: Zoology, v. 106, p. 19—28. Klecka, W. R., 1980, Discriminant Analysis: Beverly Hills, London, Sage Publications, 71 p. Kozuch, L., and Fitzgerald, C., 1989, A guide to identify shark centra fi‘om southeastern archaeological sites: Southeastern Archaeology, v. 8, no. 2, p. 146—157. Kreffi G., 1968, Knorpelfische (Chondrichthyes) aus dem tropischen Ostatlantik: Atlantide Report. (10). Scientific Results of the Danish Expedition to the Coasts of Tropical West Africa, 1945—1946: p. 33—76. Lavery, S., 1992, Electrophoretic analysis of phylogenetic relationships among Australian carcharhinid sharks: Australian Journal of Marine and Freshwater Research, v. 43, p. 97—108. Lund, R., 1985, Stethacanthid remains fi’om the Bear Gulch Limestone (Namurian E2B) of Montana: American Museum Novitates, no. 2828, p. 1—24. 216 Maddison, W. P., Donoghue, M. J., and Maddison, D. R., 1984, Outgroup analysis and parsimony: Systematic Zoology, v. 33, no. 1, p. 83—103. Maddison, D. R., and Maddison, W. P., 2001, MacClade 4: Analysis of phylogeny and character evolution, version 4.03, Sinauser Associates, Sunderland, Massachusettes. Maisey, J. G., 1984, Higher elasmobranch phylogeny and biostratigraphy: Zoological Journal of the Linnean Society, v. 82, p. 33—54. Maisey, J. G., 1989, Hamiltonichthys mapesi, g. & sp. nov. (Chondrichthyes; Elasmobranchii) from the Upper Pennsylvanian of Kansas: American Museum Novitates, no. 2931, p. 1—42. Massare, J. A., and Sharkey, S. J., 2003, Centrum shape in sharks: Not a good analog for ichthyosaurs: Paludicola, v. 4, no. 2, p. 27—36. Moss, M. L., 1970, Enamel and bone in shark teeth: With a note on fibrous enamel in fishes: Acta Anatomica, v. 77, p. 161—187. Muller, J. and Henle, F. G. J., 1837, Gattungen der Haifische und Rochen nach einer von ihm mit Hm Henle unternomenen gemeinschafilichen Arbeit fiber die Naturgeschichte der Knorpelfische: Bericht fiber die zur Bekanntmachung geeigneten Verhandlungen der Akademie Wissenschafien Berlin, p. 111-118. Muller, J. and Henle, F. G. J., 1838A, On the generic characters of cartilaginous fishes, with descriptions of new genera: Magazine of Natural History, v. 2, p. 33—37, 88— 91. Nakaya, K., 1975, Taxonomy, comparative anatomy and phylogeny of Japanese catsharks, Scyliorhinidae: Memoirs of the Faculty of Fisheries, Hokkaido University, v. 23, no. 1, p. 1—94. Natanson, L. J., and Cailliet, G. M., 1990, Vertebral growth zone deposition in Pacific angel sharks: Copeia, no. 4, p. 1133—1145. Naylor, Gavin J. P., 1992, The phylogenetic relationships among requiem and hammerhead sharks: Inferring phylogeny when thousands of equally most parsimonious trees result: Cladistics, v. 8, p. 295—318. Naylor, G. J. P., Martin, A. P., Mattison, E. G., and Brown, W. M., 1997, Interrelationships of lamniform sharks: Testing phylogenetic hypotheses with sequence data, in Kocher, T. D., and Stepien, C. A., eds, Molecular Systematics of Fishes, p. 199—218. Officer, R. A., Gason, A. S., Walker, T. I., and Clement, J. G., 1996, Sources of variation in counts of growth increments in vertebrae fi'om gummy sharks, Mustelus antarcticus, and school shark, Galeorhinus galeus: Implications for age 217 determinationz: Canadian Journal of Fisheries and Aquatic Sciences, v. 53, p. 1765-1 777. Parsons, G. R, 1985, Growth and age estimation of the Atlantic sharpnose shark, Rhizoprionodon terraenovae: A comparison of techniques: Copeia, no. 1, p. 80— 85. Parsons, G. R, 1993, Age determination and growth of the bonnethead shark Sphyrna tiburo: A comparison of two populations: Marine Biology (Berlin), v. 117, p. 23- 3 1 . Purdy, R W., Schneider, V. P., Applegate, S. P., McLellan, J. H., Meyer, R L., and Slaughter, B. H., 2001, The Neogene sharks, rays, and bony fishes from Lee Creek Mine, Aurora, North Carolina, in Ray, C. E., and Bohaska, D. J., eds., Geology and Paleontology of the Lee Creek Mine, North Carolina, III, p. 71—202. Smithsonian Contributions to Paleobiology, no. 90, Smithsonian Institution, Washington, D. C. Rafinesque, C. S., 1810, Caratteri di alcuni nuovi generi e nuevi spece di animali e piante della Sicilia, Palermo, 105 p. Regan, T., 1906, A classification of the selachian fishes, in Proceedings, Zoological Society of London, no. 2, p. 722—758 Ridewood, 1899, Some observations on the caudal diplospondyly of sharks: Journal of the Linnean Society (Zoology), v. 27, p. 46—59. Ridewood, W. G., 1921, On the calcification of the vertebral centra in sharks and rays: London, Harrison and Sons, Ltd., Philosophical Transactions of the Royal Society of London. v. 210, series B. Secerov, S., 1911, Uber die Entstehung der Diplospondylie der Selachier: Arbeiten aus dem Zoologischen Instituten der Universistat Wien und der Zoologischen Station in Triest, v. 29, p. 1—28. Smith, J. L. B., 1949, The sea fishes of southern Africa. South Africa, Central News Agency Ltd., 550 p. Smith, S. E., 1984, Timing of vertebral—band deposition in tetracycyline—injected leopard sharks: Transactions of the American Fisheries Society, v. 113, no. 3, p. 308—313. Springer, V. G., and Garrick, J. A. F., 1964, A survey of vertebral numbers in sharks: Proceedings, US. National Museum, v. 116, no 3496, p. 73—96. Swofl‘ord, D. L., 2000, Phylogenetic analysis using parsimony (PAUP*) version 4.0b4a, Sinauer and Associates, Sunderland, Massachusetts. 218 Tabachnick, B. G., and Fidell, L. S., 1983, Using multivariate statistics: New York, Harper & Row, 509 p. Thorson, T. B., and Lacy, E. J. Jr., 1982, Age, growth rate and longevity of Carcharhinus leucas estimated from tagging and vertebral rings: Copeia, no. 1, p. 110—116. Urist, M. R, 1961, Calcium and phosphorous in the blood and skeleton of the Elasmobranchii: Endocrinology, v. 69, p. 778—801. White, E. G., 1936, A classification and phylogeny of the elasmobranch fishes: American Museum Novitates, no. 837, 16 p. White, B. G., 1937, Interrelationships of the elasmobranchs with a key to the Order Galea: Bulletin of the American Museum of Natural History, v. 74, p. 25—138. Whitley, G. P., 1929, Additions to the check—list of the fishes of New South Wales. No. 2: The Australian Zoologist, v. 5, p. 353—357. Whitley, G. P., 1940, The fishes of Australia. Part I. The sharks, rays, devilfish, and other primitive fishes of Australia and New Zealand: Australian Zoologial Handbook, Proceedings, Royal Zoological Society of New South Wales, Sydney. 280 p. Wintner, S. P., 2000, Preliminary study of vertebral growth rings in he whale shark, Rhincodon typus, from the east coast of South America: Environmental Biology of Fishes, v. 59, p. 441—451. 219 ulllllllllltill"