A Test of the Multiregional Hypothesis of Modern Human Origins Using Basicranial Evidence from Indonesia and Australia

Home , Franz Weidenreich, Solo Man, Teuku Jacob

University of Tennessee, Knoxville TRACE: Tennessee Research and Creative Exchange

Doctoral Dissertations Graduate School

12-2004

Authur C. Durband University of Tennessee - Knoxville

Follow this and additional works at: https://trace.tennessee.edu/utk_graddiss

Part of the Anthropology Commons

Recommended Citation Durband, Authur C., "A Test of the Multiregional Hypothesis of Modern Human Origins Using Basicranial Evidence from Indonesia and Australia. " PhD diss., University of Tennessee, 2004. https://trace.tennessee.edu/utk_graddiss/2155

This Dissertation is brought to you for free and open access by the Graduate School at TRACE: Tennessee Research and Creative Exchange. It has been accepted for inclusion in Doctoral Dissertations by an authorized administrator of TRACE: Tennessee Research and Creative Exchange. For more information, please contact [email protected]. To the Graduate Council:

I am submitting herewith a dissertation written by Authur C. Durband entitled "A Test of the Multiregional Hypothesis of Modern Human Origins Using Basicranial Evidence from Indonesia and Australia." I have examined the final electronic copy of this dissertation for form and content and recommend that it be accepted in partial fulfillment of the equirr ements for the degree of Doctor of Philosophy, with a major in Anthropology.

Andrew Kramer, Major Professor

We have read this dissertation and recommend its acceptance:

Lyle Konigsberg, Murray Marks, Gary McCracken

Accepted for the Council:

Carolyn R. Hodges

Vice Provost and Dean of the Graduate School

(Original signatures are on file with official studentecor r ds.)

To the Graduate Council:

I am submitting herewith a dissertation written by Arthur C. Durband entitled “A Test of the Multiregional Hypothesis of Modern Human Origins Using Basicranial Evidence from Indonesia and Australia.” I have examined the final electronic copy of this dissertation for form and content and recommend that it be accepted in partial fulfillment of the requirements for the degree of Doctor of Philosophy, with a major in Anthropology.

______Andrew Kramer______Major Professor

We have read this dissertation and recommend its acceptance:

Lyle Konigsberg______

Murray Marks______

Gary McCracken______

Accepted for the Council:

____Anne Mayhew_____ Vice Chancellor and Dean of Graduate Studies

(Original signatures are on file with official student records.)

A Test of the Multiregional Hypothesis of Modern Human Origins Using Basicranial Evidence from Indonesia and Australia

A Dissertation Presented for the Doctor of Philosophy Degree The University of Tennessee

Arthur C. Durband December, 2004

Acknowledgements

Of course, an undertaking of this magnitude cannot be accomplished without a lot of help. Many different people have provided many different things for me every step of the way to get me this far.

First I must thank my main advisor, Andrew Kramer. Throughout my many long years at the University of Tennessee he has been instrumental in providing me with many different opportunities for growth as an aspiring paleoanthropologist. In particular, he brought me to Java during the field season at Rancah, Central Java, in the summer of

1999. During this trip I got to study my first original fossils, and those observations form the basis for this dissertation. He has always encouraged me and pushed me to succeed as a scientist, even though my conclusions about the fossil record and modern human origins in particular have often differed from his own. This work is a direct reflection of this collaboration.

The other members of my doctoral committee, Lyle Konigsberg, Murray Marks, and Gary McCracken have all provided significant help and insights for this work. Their patience and dedication to this project is very much appreciated.

A number of individuals provided me with access to materials in their care, and also made my many travels to collect data both enjoyable and successful.

In Indonesia Fachroel Aziz and Teuku Jacob both provided me with access to original fossils in their care. Tony Djubiantono allowed me to view a cast of the Ngawi calvarium in his laboratory.

iii

During my trip to Europe John de Vos gave me access to the Wajak 1 cranium and also went out of his way to be helpful during my stay. He gave me a ride from

Frankfurt to The Netherlands, taking me to the Neandertal site museum along the way.

He also graciously provided me with several copies of original reprints by Dubois that were cited in this dissertation. In Frankfurt, Friedemann Schrenk allowed me to view the original Sangiran fossil material in his care. Christine Hertler was also a great help during my studies in Germany.

My work in Australia was greatly assisted by Michael Westaway, who provided me with places to stay and too many helpful contacts to mention. Without his help my trip would not have been anywhere near as successful or enjoyable as it was. At the

Shellshear Museum Denise Donlon allowed me to study the impressive collection of

Pleistocene Australian casts in her care, and also took the time to show me around the

Darling Fish Market. Ann Macintosh supplied me with a stack of original reprints of her late husband’s work that were a real treat to receive, and also entertained me with many great stories. In Canberra, my day of work and conversation with Alan Thorne was one of the most interesting and rewarding experiences I’ve had as an anthropologist. David

Bulbeck was also kind enough to take an afternoon to discuss my work during my stay.

Harvey Johnston included me in his guided tour of Lake Mungo during my last weekend in Australia, and his efforts are also appreciated.

At the American Museum of Natural History, Ian Tattersall, Ken Mowbray, and

Gary Schwartz provided me with access to materials in their care. Ken Mowbray and

Sam Márquez also took the time to discuss their views on the Ngandong TMJ with me.

At the Smithsonian Institution, Rick Potts and Jennifer Clark allowed me to borrow casts

in their collection to be scanned for the chapter on TMJ morphology. Rebecca Snyder taught me how to use their impressive 3-D scanner to collect my data, and lent her expertise when necessary. David Hunt allowed me access to skeletal material in his care.

During my stay in Washington Lisa Nevell provided me with a place to stay as well as a lot of help making TMJ scans. Kristen Yoder spent many days working with me on the color-coded TMJ maps in Chapter 6. Mike O’Neil of the statistical consulting center at the University of Tennessee assisted me with the Manova tests for Chapter 5.

Fred Smith has also been instrumental in my development as an anthropologist.

He taught the first class I took on the human fossil record and also served as the main advisor on my Masters committee. His influence on this project is considerable. He also allowed me to borrow some of his personal casts, most notably Sambungmacan 3, to complete this work.

This work would not have been possible without the support of my friends and family. Their patience and confidence in me has really made this journey possible, and a lot more enjoyable than it could have been.

Jim Kidder has collaborated with me on a number of projects in the past several years, and we’ve attended a lot of AAPA meetings together. He has been a great person to work with as well as a good friend, and I hope our work can continue.

Dan and Laura Weinand are like family and have kept me sane during this long and often tortuous process. Without their help I would have given it all up and moved to

Guatemala to grow coffee beans a long time ago.

Meredith Tarczynski has also put up with a lot from me these past few years, and

I appreciate her patience and sense of humor. Altogether way too much fun, indeed.

And last, but certainly not least, my family has been all I could have asked for and more. My brother Rob continues to amaze me with everything he knows how to do (he really did get all of the talent in our family), and I continue to hope that I can grow up to be more like him. Most of all, however, I need to thank my Mom, who has made all of this possible. Because of all her sacrifices I get to do what I love every day of my life, and it just doesn’t get any better than that.

Funding for this project was provided by University of Tennessee, Northern Illinois

University, Sigma Xi, and the American Museum of Natural History.

Abstract

Proponents of the Multiregional Hypothesis of modern human origins have consistently stated that the material from Australasia provides one of the most compelling examples of regional continuity in the human fossil record. According to these workers, features found in the earliest Homo erectus fossils from Java can be traced through more advanced hominids from Ngandong and are found in both fossil and recent Australian

Aborigines.

For this study, non-metric observations will be used to determine the degree of similarity between earlier Homo erectus from Sangiran, the Ngandong fossils (including the Sambungmacan hominids), and fossil/modern Australian Aborigines in the cranial base. This study will examine the hypothesis that a number of non-metric features will show an overall similarity between these samples, and will reject this hypothesis if it can be shown that significant dissimilarity exists between these groups.

The results of this project highlight a suite of features on the cranial base in the

Ngandong sample that appear to be unique to that group. These morphologies include a dual foramen ovale, the location of the squamotympanic fissure, the small size and parallel orientation of the occipital condyles, and the marked expression of the postcondyloid tuberosities. The presence of these autapomorphic characters in the

Ngandong population, in conjunction with previous work on the Pleistocene paleoecology of Java, suggests that multiple hominid species inhabited that island during the Pleistocene. This work also provides strong evidence of discontinuity between

Indonesian Homo erectus and the earliest Homo sapiens in the Australasian fossil record.

vii

Table of Contents

Chapter 1: Previous research on modern human origins in Australasia ...... 1 Chronology of the Indonesian fossil sample ...... 8 Previous interpretations of the Indonesian fossil record ...... 18 Chronology of the Australian fossil sample ...... 26 Previous interpretations of the Australian fossil record ...... 29 Chapter 2: Materials and Methods ...... 40 Materials ...... 40 Methods ...... 42 Notes about possible sources of error ...... 52 Chapter 3: Descriptions of Indonesian cranial bases ...... 53 Sangiran 2 ...... 53 Sangiran 4 ...... 54 Sangiran 12 ...... 55 Sangiran 14 ...... 56 Sangiran 17 ...... 57 Sambungmacan 1 ...... 58 Sambungmacan 3 ...... 58 Sambungmacan 4 ...... 60 Ngandong ...... 61 Ngawi 1 ...... 69 Wajak 1 ...... 70 Chapter 4: Descriptions of Australian cranial bases ...... 72 Keilor ...... 72 Talgai ...... 73 Kow Swamp ...... 74 Lake Mungo ...... 76 Nacurrie 1 ...... 78 Nacurrie 2 ...... 79 Cossack ...... 80 Mossgiel ...... 81 Cohuna ...... 81 Lake Nitchie ...... 83 Chapter 5: Results of non-metric examinations...... 86 Individual character comparisons ...... 88 Discussion ...... 103 Chapter 6: Temporomandibular joint (TMJ) morphology ...... 113 Introduction ...... 113 Results ...... 116 Discussion ...... 133 Chapter 7: Discussion and Conclusions ...... 135 References Cited ...... 150 Appendix A: Raw non-metric data for the modern human sample ...... 174 Appendix B: Raw non-metric data for the fossil sample ...... 182

viii

Appendix C: Raw metric data for the comparative samples ...... 184 Appendix D: Results of Manova tests on the modern comparative samples .....191 Appendix E: Cranial base photographs of the Indonesian fossil sample ...... 198 Appendix F: Cranial base photographs of the Australian fossil sample ...... 222 Vita ...... 237

List of Tables

Table Page

2.1 Fossil specimens included in this project ...... 41 2.2 Modern human sample used for this project ...... 42 2.3 Cranial base features examined for this project ...... 43 5.1 Summary statistics for the relative area of occipital condyle to foramen magnum in the modern sample ...... 102 5.2 Results of t-test for equality of means for the relative area of occipital condyle to foramen magnum area in the modern sample ...... 102 5.3 Non-metric traits used in this study and their frequencies in each segment of the proposed regional continuity sequence for Australasia ...... 104 5.4 Results of Fisher’s Exact tests comparing data for Sangiran and Ngandong/Sambungmacan ...... 105 5.5 Results of Fisher’s Exact tests comparing data for Sangiran and the Pleistocene Australians...... 105 5.6 Results of Fisher’s Exact tests comparing data for Sangiran and modern Australians ...... 106 5.7 Results of Fisher’s Exact tests comparing data for Ngandong/ Sambungmacan and the Pleistocene Australians...... 106 5.8 Results of Fisher’s Exact tests comparing data for Ngandong/ Sambungmacan and the modern Australians...... 107 5.9 Results of Fisher’s Exact tests comparing data for the Pleistocene Australians and the modern Australians ...... 107 5.10 Polychoric correlation results for the potential Ngandong and Sambungmacan autapomorphies ...... 110 D.1 Results of Manova test for expression of the pharyngeal tubercle in the modern samples ...... 192 D.2 Results of Manova test for expression of tympanic plate contact with mastoid in the modern samples ...... 192 D.3 Results of Manova test for expression of the alar tubercles in the modern samples ...... 193 D.4 Results of Manova test for expression of the orientation of the occipital condyles in the modern samples ...... 193 D.5 Results of Manova test for expression of the opisthionic recess in the modern samples ...... 194 D.6 Results of Manova test for expression of the postcondyloid tuberosities in the modern samples ...... 194 D.7 Results of Manova test for expression of the foramen lacerum in the modern samples ...... 195 D.8 Results of Manova test for expression of the juxtamastoid process in the modern samples ...... 195 D.9 Results of Manova test for expression of the occipitomastoid crest in the modern samples ...... 196

D.10 Results of Manova test for expression of the location of the carotid foramen relative to the squamotympanic fissure in the modern samples .....196 D.11 Results of Manova test for the size of the postglenoid tubercle in the modern samples ...... 197 D.12 Results of Manova test for expression of the orientation of the squamotympanic fissure in the modern samples ...... 197

List of Figures

Figure Page

1.1 Site map of relevant localities in Indonesia and Australia...... 7 1.2 Approximate correlations for dating methods used at Sangiran...... 17 2.1 Map of the cranial base features to be examined ...... 44 6.1 3-D scans of the right TMJ of Ngandong 12, inferior view ...... 118 6.2 Right TMJ of Ngandong 12 (cast) ...... 119 6.3 3-D scans of the right TMJ of Ngandong 7, inferior view ...... 120 6.4 3-D scan of the left TMJ of Ngandong 7, lateral view ...... 121 6.5 3-D scan of the right TMJ of Ngandong 10, lateral view ...... 122 6.6 3-D scan of the left TMJ of Sambungmacan 3, inferior view ...... 123 6.7 Sangiran 4 left mandibular fossa (cast) ...... 125 6.8 3-D scan of the right TMJ of Sangiran 17 ...... 127 6.9 Right TMJ of OH 9 (cast) ...... 129 6.10 3-D scans of the right TMJ of OH 9 ...... 130 6.11 3-D scans of the left TMJ of WT 15000 ...... 131 6.12 Right TMJ of WT 15000 (cast) ...... 132 E.1 Sangiran 2 cranial base ...... 199 E.2 Sangiran 2 posterior foramen magnum (cast) ...... 200 E.3 Sangiran 2 left TMJ (cast) ...... 201 E.4 Sangiran 4 cranial base ...... 202 E.5 Sangiran 4 left TMJ (cast) ...... 203 E.6 Sangiran 4 foramina ovale ...... 204 E.7 Sangiran 4 foramen magnum and occipital condyles ...... 204 E.8 Sangiran 12 occipital ...... 205 E.9 Sangiran 14 basioccipital ...... 206 E.10 Sangiran 17 cranial base ...... 207 E.11 Sangiran 17 right TMJ ...... 208 E.12 Sangiran 17 foramen magnum ...... 208 E.13 Ngandong 7 cranial base ...... 209 E.14 Ngandong 11 cranial base ...... 210 E.15 Ngandong 12 cranial base ...... 211 E.16 Ngandong 12 right posterior cranial base ...... 212 E.17 Ngandong 12 foramen magnum ...... 213 E.18 Ngandong 7 foramen magnum ...... 214 E.19 Ngandong 11 posterior cranial base ...... 215 E.20 Ngandong 12 left foramen ovale ...... 216 E.21 Ngandong 7 left foramen ovale ...... 217 E.22 Ngandong 12 left tympanomastoid fissure in lateral view ...... 217 E.23 Sambungmacan 1 left TMJ ...... 218 E.24 Sambungmacan 3 posterior foramen magnum (cast) ...... 219 E.25 Sambungmacan 3 left TMJ (cast) ...... 219 E.26 Sambungmacan 4 cranial base ...... 220

xii

E.27 Wajak 1 cranial base ...... 221 F.1 Lake Mungo 1 occipital bone near (or at) opisthion (cast) ...... 223 F.2 Lake Mungo III occipital bone (cast) ...... 223 F.3 Kow Swamp 5 cranial base (cast) ...... 224 F.4 Kow Swamp 5 left TMJ, mastoid, and occipital (cast) ...... 225 F.5 Kow Swamp 5 midline occipital bone (cast) ...... 225 F.6 Keilor cranial base (cast) ...... 226 F.7 Keilor left TMJ and tympanic bone (cast) ...... 227 F.8 Keilor foramen magnum (cast) ...... 227 F.9 Nacurrie 1 left TMJ (cast) ...... 228 F.10 Nacurrie 2 cranial base from Westaway (2002b) ...... 229 F.11 Mossgiel cranial base (cast) ...... 230 F.12 Mossgiel left TMJ (cast) ...... 231 F.13 Mossgiel foramen magnum (cast) ...... 231 F.14 Cossack right TMJ (cast) ...... 232 F.15 Cohuna cranial base (cast) ...... 233 F.16 Cohuna left TMJ and tympanic bone (cast) ...... 234 F.17 Lake Nitchie cranial base (cast) ...... 235 F.18 Lake Nitchie posterior cranial base (cast) ...... 236 F.19 Lake Nitchie right posterior cranial base (cast) ...... 236

xiii

Glossary of Indonesian fossil nomenclature

During the past 102 years of paleoanthropological research in Indonesia a number of different conventions have been used to name/number the fossil discoveries. This has often made it quite confusing when consulting different references on these specimens. Below are guides to help the reader sort out various synonyms for the different Indonesian crania that will be discussed in this dissertation.

Fossil name Pithecanthropus # Trinil 2 Pithecanthropus I Sangiran 2 Pithecanthropus II Sangiran 3 Pithecanthropus III Sangiran 4 Pithecanthropus IV Modjokerto Pithecanthropus V Sangiran 10 Pithecanthropus VI Sangiran 12 Pithecanthropus VII Sangiran 17 Pithecanthropus VIII

Tjg-1993.05 Skull IX

Original Modern Jacob Solo I Ngandong 1 Ngandong 1 Solo II Ngandong 2 Ngandong 2 Solo III Ngandong 3 Ngandong 3 Solo IIIa Ngandong 4 Ngandong 4 Solo IV Ngandong 5 Ngandong 5 Solo V Ngandong 6 Ngandong 6 Solo VI Ngandong 7 Ngandong 7 Solo VII Ngandong 8 Ngandong 8 Tibia A Tibia A Ngandong 9 Tibia B Tibia B Ngandong 10 Solo VIII Ngandong 9 Ngandong 11 Solo IX Ngandong 10 Ngandong 12 Solo X Ngandong 11 Ngandong 13 Solo XI Ngandong 12 Ngandong 14

The “original” numbering system for the Ngandong crania was used by Weidenreich (1951) and Santa Luca (1977, 1980) in their work on these specimens. The “modern” system is the scheme typically used in current publications on these fossils, and it is the one employed in this dissertation. The “Jacob” system is that preferred by Teuku Jacob, one of the leading Indonesian paleoanthropologists and author of a number of papers that include the Ngandong crania.

xiv

Chapter 1: Previous research on modern human origins in Australasia

The notion of regional continuity linking the ancient and modern populations of

Southeast Asia and Australia has a long and storied history in paleoanthropology. These

ideas were primarily formulated by Franz Weidenreich (1943, 1945a, b, c, 1951) after his

extensive studies of fossil humans from throughout these regions. Weidenreich (1943:

276) contended that “[t]here is an almost continuous line leading from Pithecanthropus

through Homo soloensis and fossil Australian forms to certain modern primitive

Australian types.” Later workers such as Thorne (1975, 1976, 1977; Thorne and Wolpoff,

1981), Macintosh (1963, 1965, 1967a), Kramer (1989, 1991), and Wolpoff (1989, 1992,

1999) have expanded upon this work, contributing to the perception that the Australasian fossil assemblage represents one of the best cases for regional continuity in the origin of modern humans.

Much of the interest currently focused on these regional characteristics in

Australasia can be credited to the “Center and Edge” model of Thorne and Wolpoff

(1981). In this paper the authors propose a theory to explain the mechanism behind regional continuity as well as providing one of its more famous examples: the facial similarities between the Indonesian Homo erectus fossil Sangiran 17 and the modern

Kow Swamp material from Australia (Thorne and Wolpoff, 1981). Other work has linked the WLH 50 partial cranium from Australia with the Ngandong crania (Hawks et al.,

2000; Wolpoff et al., 2001), and the mandibular sample from Sangiran with modern

Australian Aborigines (Kramer, 1989, 1991). This work emphasizes the persistence of

several morphological characters in the skulls, jaws, and teeth of these specimens that potentially provide evidence for the maintenance of genetic cohesiveness in the region for well over a million years.

Recent work has highlighted several difficulties with this theory of regional continuity, however, and often provides evidence for replacement of the archaic populations. Baba and colleagues (1998, 2000) reconstructed the face of Sangiran 17 and failed to find support for the earlier conclusions of Thorne and Wolpoff (1981) regarding regional continuity with the Kow Swamp specimens. Studies by Brown (1981, 1989) and

Antón & Weinstein (1999) have pointed out the presence of artificial cranial deformation in Pleistocene Australians, calling into question the utility of flat frontal bones as indicators of continuity with Indonesian H. erectus. WLH 50 has also had its alleged transitional nature assailed in work by Neves et al. (1999) and Stringer (1998), and has even been suggested to be pathological (Webb, 1989, 1990). WLH 50 has also been recently redated to approximately 14,000 years by Simpson and Grün (1998), a date that would make the specimen too young to serve as an intermediary between the Ngandong fossils and modern Australians. Likewise, Ngandong has been redated by Swisher et al.

(1996) to be as recent as 27,000 years. This surprisingly recent date would negate these specimens as potential ancestors to the Australians.

While Rogers (1995: 676) suggests that “regional continuity is unlikely to be reflected in fossils,” Relethford (2001a: 203) has suggested that some fossil evidence is

“perfectly consistent with a multiregional interpretation.” Recent work on genetic evidence relevant to this question (Relethford, 2001a, b; Templeton, 2002) has provided support for the Assimilation model of Smith and colleagues (1989), which posits that

humans migrating out of Africa interbred with local populations of archaic humans. This research has shown that while a majority of the DNA regions studied indicate that Africa has played a dominant role in the formation of the modern human genome, a small percentage of these data provide evidence for gene flow with non-African populations

(Templeton, 2002). These studies cast doubt on a total replacement scenario. However, even with this type of model not all archaic populations necessarily contributed to the modern human gene pool. As Relethford (2001a: 206) explains, “[l]ocal populations frequently became extinct and were replaced by migrants from neighboring populations…” From this we can infer that it is possible, if not probable, that many fossil populations are not ancestral to modern humans.

With this in mind, this dissertation will be focused upon this question of continuity vs. replacement in the human fossil sample from Indonesia and Australia. My previous work on the Ngandong specimens has contributed to the arguments against regional continuity in the Australasian fossil record. These studies have pointed to several features on the Ngandong basicrania that were potentially unique to that sample, thereby strongly indicating discontinuity between them and the later modern populations in the region (Durband, 1998, 2002a, b, 2004). These results echoed previous work by

Macintosh and Larnach (1972, 1976; Larnach and Macintosh, 1974) that indicated the presence of autapomorphic features in the Ngandong specimens. This body of work, in conjunction with other studies conducted on the Ngandong cranial vault (Grimaud-Hervé,

1986; Santa Luca, 1980), suggests that these hominids have progressed beyond the typical morphological pattern seen in Homo erectus but not in the direction of more modern humans.

It is my intention to build upon these important early results with this doctoral dissertation. While my initial work provided significant evidence for unique features in the Ngandong specimens, these results need to be tested against a large sample of human crania that represents broad temporal and geographical areas. This expanded sample will include both older H. erectus material from Indonesia as well as fossil and modern specimens that were not included in my thesis work. My analysis will focus on the cranial base, an area of the cranium that Howells (1969) contends is largely ignored in analyses of this type. Most comparative studies tend to focus on the cranial vault or facial structures to determine phylogenetic relationships. While this approach is often necessary due to the fragility of the skull base, the fossils that do preserve this portion of the cranium have been underexploited in the literature. This is a trend that should be reversed, as research has demonstrated that the cranial base is an area of the skull that is controlled primarily by genetic factors (Van Limborgh, 1970; Burdi, 1976). The basicranium appears early in the development of the skull, formed initially from mesenchyme which then develops into a chondral template that provides the framework for ossification of the chondrocranium (Burdi, 1976; Sperber, 1989). Due to the early formation of these structures, they are able to develop with little influence from other cranial structures (Burdi, 1976; Smahel and Skvarilova, 1988; Sperber, 1989). It has been suggested that the cranial base may be evolutionarily conservative and could therefore provide valuable phylogenetic information. Further study is needed to confirm this supposition (Lieberman et al., 2000).

Cranial non-metrics have been used extensively by physical anthropologists to reconstruct population relationships (e.g. Richtsmeier et al., 1984; Frayer, 1992; Ishida

and Dodo, 1993; Konigsberg et al., 1993). While there has been some debate about their efficacy (e.g. Corruccini, 1974; Sjøvold, 1984; Van Vark et al., 1989; and see discussion in Hauser and De Stefano, 1989), many of the difficulties encountered in their use are due in part to a lack of standardization, and hence comparability with other similar studies

(Hauser and De Stefano, 1989). Previous studies using basicranial characters have found them to show similar levels of heritability as other neurocranial or facial characters

(Richtsmeier et al., 1984; Cheverud, 1995). Strait (1998) also demonstrated that phylogenetic analyses using basicranial characters produced results that were quite similar to those obtained from facial or neurocranial characters. Konigsberg and colleagues (1993) found that cranial non-metrics are less influenced by artificial cranial deformation than craniometrics, and as a result discrete traits are more useful for determining population relationships for affected individuals. As noted earlier, several scholars believe that some Pleistocene Australian crania have been affected by artificial deformation (e.g. Brown, 1981, 1987; Antón and Weinstein, 1999).

Unfortunately, little work has been done beyond a simple description of many of the pertinent fossils from Southeast Asia and Australasia, and this is particularly true of much of the late Pleistocene/early Holocene material from Australia. The few works that have been published focus primarily on the cranial vault and facial skeleton and mention little about the base (e.g. Brown, 1989; Freedman, 1985; Freedman and Lofgren, 1979;

Santa Luca, 1980; Thorne, 1975; Weidenreich, 1945c). Weidenreich (1943, 1951) provided the first detailed descriptions of the cranial base of the Ngandong fossils. His comparisons between Ngandong and other specimens were limited by the small comparative sample available at that time, however, and as a result he was able to draw

few conclusions about the possible role of the Ngandong people in the later evolution of modern humans in the region. Santa Luca (1980) sought only to establish the place of the fossils relative to other Homo erectus and Neandertal fossils, and not their relationship to modern humans. The lack of consideration for the morphology of the cranial bases in these samples precludes the use of the literature to study this important region of the skull. Many specimens from Australasia retain this fragile section of the cranium, including specimens from Sangiran (skulls 4, 14, and 17 with portions on others),

Ngandong (skulls 7 and 12 with portions preserved on several others), and many late

Pleistocene/early Holocene specimens from Australia. This group of fossils encompasses the entire lineage proposed by Wolpoff (1989, 1992, 1999) to represent the evidence for regional continuity in Australasia.

The scholarship surrounding the peopling of Australia is varied and complex.

While I cannot hope to present a complete picture of this dynamic research, the following sections will allow the reader to become familiar with the background and history of this work. To achieve this aim, a brief recounting of the current understandings of the chronology of the fossil specimens of both Indonesia and Australia will be provided, followed by a synopsis of the interpretations of each region’s fossil records. These sections will be specifically focused on the arguments for continuity that have so long been a part of paleoanthropological inquiry in this part of the world (Figure 1.1).

Figure 1.1: Site map of relevant localities in Indonesia and Australia. Modified from a map created by Peter Brown.

Chronology of the Indonesian fossil sample

The dating of the Indonesian hominid sample has always been problematic at best, and these issues began with the activities surrounding the first discoveries at Sangiran. After

Dubois had prospected at Sangiran and failed to find any fossils, GHR von Koenigswald moved to that site in the early 1930’s (Larick et al., 1999). Finding no hominids on his own, von Koenigswald turned his attentions to the local farmers in the region, offering to pay them for fossils they located in their fields. As one would expect, this brought huge numbers of fossils to his doorstep. It was because of this offer to purchase fossils that the first hominid specimen from the site, the Sangiran 1 jaw, was brought to him in 1936

(Koenigswald, 1956).

While this business was a boon to von Koenigswald’s scientific endeavors and quite profitable for the farmers, it had a very negative and lasting impact on later efforts to both collect and understand the fossil sequence at Sangiran. By placing a monetary value on the fossils brought to him, von Koenigswald unwittingly unleashed a profit motive in the local population. This had the immediate effect of larger, more complete fossils being broken up into smaller pieces in order to be sold many different times, as happened with the Sangiran 2 skullcap (Koenigswald, 1956). And while that problem was remedied by higher prices being offered for larger, more complete specimens, a more serious problem resulted from farmers hiding the find sites of their fossils to protect their future moneymaking potential. Clearly, it was in the best fiscal interests of the farmers to keep the location of their best finds a secret, because if the scientists knew where they came from they could just prospect there themselves and cut out the middlemen.

Provenience matters little to a farmer just trying to feed his family, so this is

understandable from a humanitarian standpoint, but it is incredibly unfortunate from a scientific one. Thus, the results of this new culture of profit surrounding fossils in the

Sangiran dome mean that very few hominid fossils have a reliable provenance (Larick et al., 1999). It also means that specimens like Sambungmacan 3 end up on the international marketplace for auction to the highest bidder (Laitman and Tattersall, 2001), and there are likely other fossils that have suffered the same fate unbeknownst to scientists.

These problems aside, a number of workers have utilized varying approaches in an effort to develop a chronology of the Indonesian hominids. The earliest efforts were predominantly biostratigraphic and placed the hominids within series of fauna that correlated to shifts in the ecosystem upon Java as well as fluctuating connections with mainland Asia (e.g. Aziz et al., 1995; de Vos, 1985, 1995; de Vos et al., 1982, 1994;

Storm 2000, 2001a, b). More recent efforts have included magnetostratigraphy (e.g.

Hyodo et al., 1993), 40Ar/39Ar dating (e.g. Swisher et al., 1994; Larick et al., 2001),

ESR/EPR dating on associated faunal remains (Swisher et al., 1996), and mineralogical testing on the specimens themselves (e.g. Matsu’ura, 1982, 1986; Matsu’ura et al., 1992,

1995; Sighinolfi et al., 1993). Through these efforts many of the earlier Javan finds without solid provenance from sites such as Mojokerto and Sangiran have been correlated with various dated strata. Thus, the beginnings of a chronological framework for the

Indonesian fossil sequence have emerged.

The development of the Javan biostratigraphic and lithostratigraphic sequences, and the various controversies and debates surrounding them, make for interesting reading by themselves. However, for the purposes of this dissertation the reader will be spared those details and only the basic outline of the chronologies will be presented. The

lithostratigraphic series that has been constructed is based solely on the sediments contained in the Sangiran dome, which has made it difficult to apply outside that site

(though some have tried). The biostratigraphic sequence is an attempt to tie different sites together and place them in sequence by way of various faunal turnovers. While the biostratigraphic methods have the longest history on the island, the various geological and radiometric dating methods have received a great deal of attention in more recent studies. In an effort to keep this discussion in historical perspective, the biostratigraphy of

Java will be addressed first.

The use of various animal fossils to relatively date strata containing human fossils is certainly nothing new. Dubois was the first to use this method on Java when he compared the fossils he recovered from Trinil to the Siwalik fauna from India. Using this strategy he arrived at a Pliocene or early Pleistocene date for his Trinil finds (Shipman,

2001). Von Koenigswald (1937) further described this Trinil fauna and recognized a series of additional faunal collections that could be identified by guide fossils peculiar to each zone. These fauna were, from oldest to youngest, the Tji Sande, Tji Djoelang, Kali

Glagah, Djetis, Trinil, Ngandong, Sampoeng, and Recent, and these were named after the type locality for each collection (Koenigswald, 1937). Von Koenigswald split the site of

Trinil itself into at least two faunal zones, with Dubois’ finds coming from the older

Djetis and the fossils collected by the Selenka expedition coming from the Trinil zone

(Koenigswald, 1937). The early Javan fauna were stocked by a Pliocene invasion from

Asia, the so-called “Sivamalayan” fauna, while a later Pleistocene influx of the

“Sinomalayan” fauna occurred in later Pleistocene times (Koenigswald, 1937).

Later authors have considerably changed these faunal groupings in an attempt to enhance the precision of the sequence. The biostratigraphic series in current use has seven stages, one less than the original succession, and these are (oldest to youngest):

Satir, Cisaat, Trinil Hauptknochenschicht (H.K.), Kedung Brubus, Ngandong, Punung, and Wajak (Aziz, 1990; Aziz et al., 1995; van den Bergh et al., 2001; though note that this sequence is disputed by Bartstra, 1983 and Hooijer, 1983). These stages represent an elaboration of von Koenigswald’s (1937) original faunal succession as well as, in many cases, corrections of perceived errors and miscalculations in the collections (Theunissen et al., 1990). For example, de Vos and colleagues (1982) found that von Koenigswald’s

(1937) original Trinil faunal stage represented a composite fauna. The current Trinil H.K. stage is the result of reanalysis of Dubois’ original notes and fossil collections, and represents a valid paleoecological unit (de Vos, personal communication; de Vos et al.,

1982; Sondaar et al., 1983; Aziz et al., 1995).

The Satir fauna does not contain humans, but each of the other stages has yielded hominid fossils (Aziz et al., 1995; though de Vos et al., 1994 contend that the first hominids occurred during the Trinil H.K stage). According to Aziz and colleagues

(1995), the types of species represented in the various faunal stages can be used as an indicator of Java’s relationship to the mainland at a given point in time. For example, the

Satir and Cisaat faunas each have a poor variety of species, a condition which probably indicates that Java was an island without connections to mainland Asia (de Vos et al.,

1994; Aziz et al., 1995; van den Bergh et al., 2001). Trinil H.K. also has a relatively small number of species, which suggests that Java still lacked or had only very limited contact with mainland Asia (Aziz et al., 1995; de Vos 1987; de Vos et al., 1982, 1994;

Long et al., 1996; van den Bergh et al., 2001). The remaining faunal stages, Kedung

Brubus, Ngandong, Punung, and Wajak represent more balanced faunas (de Vos et al.,

1994; Aziz et al., 1995; Long et al., 1996). Van den Bergh and colleagues (2001) note that sea levels dropped considerably during the Kedung Brubus period, and this allowed a major influx of new mammalian species from the mainland. The succeeding Ngandong fauna is quite similar to Kedung Brubus, but also shows signs of isolation and endemism

(van den Bergh, 1999; van den Bergh et al., 2001). While no new species appear in the

Ngandong fauna, the majority of species in that fauna disappear with the arrival of the

Punung fauna in the late Pleistocene (de Vos et al., 1994; van den Bergh, 1999; van den

Bergh et al., 2001).

According to the architects of this faunal succession, these groupings of animal species can be used to glean significant information about the paleoenvironments inhabited by early humans on Java (de Vos, 1983, 1985, 1995; de Vos et al., 1982, 1994;

Aziz et al, 1995; Storm, 2000, 2001a, b; van den Bergh et al., 2001). The Trinil H.K.,

Kedung Brubus, and Ngandong fauna, all containing Homo erectus fossils, suggest an open woodland habitat (de Vos, 1983, 1985, 1987; de Vos et al., 1982, 1994; Storm,

2000, 2001a, b; van den Bergh et al., 2001; but see Pope, 1985, 1995). By contrast, the

Punung fauna “is strongly indicative of a more humid, more densely forested environment” (Storm, 2000: 230). The scant human remains so far recovered with a

Punung association (two upper incisors, one upper and one lower canine, and one upper molar) have been diagnosed as probably belonging to Homo sapiens (de Vos, 1985).

Storm (2000, 2001a, b) has suggested that Homo erectus was driven extinct with other open woodland species when tropical rainforests expanded southward into insular

Southeast Asia. Homo sapiens, who was more technologically advanced and therefore better adapted to survive in a difficult environment like a rainforest, subsequently immigrated to Java during the late Pleistocene and Holocene (Storm, 2000, 2001a, b).

This scenario will be revisited toward the end of my dissertation.

As a compliment to these biostratigraphical studies, the lithostratigraphy of the island has also been examined in an attempt to date the Javan hominid collection. This work has been concentrated almost exclusively on the geology of the Sangiran dome because it has proven to be the most productive site for hominid fossils. Various workers have divided the sediments present at Sangiran into a series of formations, which are

(from oldest to youngest): Kalibeng, Pucangan (Sangiran), Grenzbank, Kabuh (Bapang), and the Notopuro (Larick et al., 2001; Widianto, 2001). The Kalibeng is a marine sediment and the Notopuro consists of breccias and lahars, and neither of these layers has yielded any hominid fossils (Widianto, 2001). All the hominids with known provenance recovered from Sangiran have come from the Pucangan, Grenzbank, and Kabuh layers

(Koenigswald, 1956; Widianto, 2001). These levels have been partially correlated with the biostratigraphic sequence previously mentioned. For example, de Vos and colleagues

(1982) have established that the Trinil H.K. fauna found by Dubois and Selenka correlates with the Grenzbank or the lower Kabuh formation.

The strata at Ngandong are considerably more challenging to interpret. While the

Solo crania were recovered during an actual excavation, and thus have exact provenances, the site geology has proven difficult to understand. Bartstra (1987) first recognized two terraces along the Solo River at Ngandong, a High and a Low, but later this number was increased to three distinct levels (Bartstra et al., 1988). Other workers

have distinguished as many as six (Swisher et al., 1996) or even seven (Sartono, 1976) discrete sedimentary units at the site. At Sambungmacan none of the hominid fossils discovered to date were the result of controlled excavations (Matsu’ura et al., 2000;

Laitman and Tattersall, 2001; Widianto, 2001; Baba et al., 2003). Concretions found inside Sambungmacan 1 match a layer of conglomerate at the site, and above this layer lie other fossiliferous deposits (Widianto, 2001). To my knowledge nothing has been published regarding the exact find locations of the other two Sambungmacan crania, and the Sambungmacan 2 tibia was a surface find (Matsu’ura et al., 2000). Likewise, the

Ngawi skull was found underwater in the Solo River when it was kicked by a swimmer and clearly lacks any provenance (F. Aziz, personal communication).

In an attempt to interpret the strata of central and east Java a number of different dating techniques have been used. As noted earlier, these have included chemical, mineralogical, and radiological methods. Fluorine dating, easily the least “absolute” of these methods, has been used on a number of specimens from Sangiran (Matsu’ura, 1982;

Matsu’ura et al, 1992, 1995; Kondo et al., 1995), Trinil (Matsu’ura, 1986), and the

Sambungmacan 2 tibia (Matsu’ura et al., 1990, 2000). While fluorine and other types of element analysis do not provide a numerical age for the fossils studied, oftentimes they do allow specimens to be assigned to a particular stratum at a site. This can be valuable in the event of a surface find (e.g. Matsu’ura et al., 1990, 2000), or when a specimen is purchased rather than excavated firsthand. As a result of this work, a number of fossils previously without solid provenance have been correlated with particular strata. In this way it is possible to begin constructing a relative timeline for the fossil assemblages.

Absolute dates for the Javan hominids have been obtained through radiometric methods, including K-Ar, 40Ar/39Ar, and ESR/EPR with highly controversial results.

Dates on the volcanics from Sangiran have produced widely varying ages for the sediments containing hominid fossils. Curtis (1981) produced a K-Ar date of 1.2 ± 0.02 mya for the base of the Pucangan formation, and ages of 0.9-1.3 mya for the upper part of that formation. The lower Pucangan was later dated by Swisher and colleagues (1994) to

1.66 ± 0.04 mya using K-Ar, but this date has been roundly criticized for lacking context with fossil remains (de Vos and Sondaar, 1994) and utilizing pumice that could not have derived from the Pucangan level (Larick et al., 2001). Larick and co-workers (2001) prefer an age of approximately 1.5 mya for the base of the Kabuh (Bapang) formation, and allege that the Pucangan hominids must therefore all be older than 1.5 mya. Those same authors date the top of the Kabuh formation to 1.02 ± 0.06 mya (Larick et al.,

2001). This series of dates collected at the site provides ages for a number of important crania, including Sangiran 10 (1.2 ± 0.2 mya) and Sangiran 17 (1.02 ± 0.13 mya) (Larick et al., 2001).

Electron Spin Resonance (ESR), also known as Electron Paramagnetic Resonance

(EPR) has also been used on Java at the site of Ngandong. Swisher and colleagues (1996) revisited the site and recovered fossil bovid teeth, which they compared to teeth collected by Oppenoorth during the original excavations. ESR/EPR dating on those teeth provided startlingly young ages of between 27.3-46.4 kyr for Ngandong and 27.3-53.4 kyr for

Sambungmacan (Swisher et al., 1996, 1997; Rink et al., 1997). However, these dates have also met with a great deal of resistance from the scientific community. Grün and

Thorne (1997) contend that the taphonomy of the site indicates that the hominids are not

contemporaneous with the sediments and/or the fauna present. Instead, they argue that the hominids have been re-deposited from older layers, echoing a similar theory espoused by

Santa Luca (1977, 1980). Westaway (2002a) also supports the notion that the Ngandong hominids are older than the associated faunal remains. Storm (2000, 2001a, b) and de

Vos (personal communication) point out that the recent ESR dates do not fit with the biostratigraphic sequence that has been determined for the island, and assert that the sediments can be no younger than 126 kyr. In addition, Westaway and colleagues (2003;

Westaway, personal communication) feel that it is probable that Swisher and colleagues

(1996) excavated Dutch backfill during their re-examination of the site and did not find faunal elements in an undisturbed state. Finally, Weeks and others (2003) question the very assumptions behind the EPR technique, and note that chemical changes in the depositional environment might be responsible for creating changes in the enamel matrix that are not necessarily age-dependant. These changes “would destroy the basis of this dating procedure, that is, the dose-rate invariability with the age of the fossil” (Weeks et al., 2003: 9888). Thus, the dispute over these dates remains far from resolved.

The end result of all this work does provide a framework upon which to begin sorting the hominid finds of Java (Figure 1.2). This framework is by no means complete, nor even necessarily sound in all aspects, but by using this panoply of methods as checks and balances on each other a working chronology has been obtained that allows us to make some order of the Indonesian hominid sample.

Absolute dates Lithostratigraphy Hominids Biostratigraphy

Notopuro

1.02 ± 0.06 Sangiran 3 (Larick et al., 2001) Upper Tuff

Sangiran 2, 10, 12, 14 Kedung Brubus Middle Tuf f Sangiran 17, Skull X Kabuh (Bapang)

Low er Tuff

1.51 ± 0.08 (Larick et al., 2001) Trinil H.K

Gre nzbank

Sangiran 4

Cisaat

Pucangan (Sangiran)

1.66 ± 0.04 Satir (Sw isher et al., 1994)

Kalibeng

Figure 1.2: Approximate correlations for dating methods used at Sangiran.

Previous interpretations of the Indonesian fossil record

Of course, any history of the interpretations surrounding the Indonesian hominids must

begin with the opinions of Eugene Dubois, the man who discovered the first Pleistocene

human fossils from the region. As is well known, Dubois (1935, 1936, 1937a, b, 1938a,

b, 1940a, b, c) felt that his famous specimens from Trinil represented a true human

ancestor while other fossils from Sangiran and Zhoukoudian in China were too derived to

have been ancestral to later humans. His theories, however, were widely misunderstood

at the time because Dubois insisted on emphasizing the human characteristics of the

fossil assemblages from Ngandong, Zhoukoudian, and Sangiran while seemingly

denigrating his own find by interpreting it as a giant gibbon (Dubois, 1935, 1937a, b,

1940a, b, c). This somewhat bizarre reading of the evidence was roundly chastised by

scientists of the time (e.g. Le Gros Clark, 1937; Weidenreich, 1951) and generally

dismissed by the scientific community as a whole. Only decades later, through the efforts

of Theunissen (1989) and later Shipman (2001) were Dubois’ motives for this stance

made clear. They involved his work on the evolution of the brain, which he undertook in

the years following his discoveries at Trinil (Theunissen, 1989). As a result of his pre-

conceived notions, Dubois (1923, 1924, 1928) formulated a hypothesis that the brain

evolved through a series of spontaneous doublings in size from one species to another.

While his Pithecanthropus had a brain too large to have given rise to modern humans

through simple doubling, by assuming that Pithecanthropus had body proportions similar to that of a gibbon Dubois was able to manipulate the formula to make it work in terms of relative body size (Dubois, 1935; Theunissen, 1989). Thus, Pithecanthropus had double the encephalization of a modern gibbon and half that of a modern human, and was

therefore the true human ancestor (Dubois, 1935, 1937b). On the other hand, none of the other purported human ancestors from China or Java could truly be ancestral to later humans because their brains were too large to evolve into a modern form through

Dubois’ proposed mechanism (Dubois, 1938b, 1940a, b, c; Theunissen, 1989).

Though interesting as a historical footnote, Dubois’ interpretations lend little of substance to scholars in their search for relationships between the fossil humans from

Java. Other individuals responsible for many of the important finds pre-dating the Second

World War, however, provided more lasting contributions. In particular, Oppenoorth,

Weidenreich, and von Koenigswald exploited comparative studies with the limited fossil sample available at the time in an effort to elucidate the position of Java man in the human phylogeny. Oppenoorth (1937) opined that Ngandong shares a number of similarities with Rhodesian man. After extensive comparison with Neandertals,

Oppenoorth (1937: 352) states that similarities between the Ngandong fossils and

European Neandertals “are more seeming than real as far as proof of identity.” Instead, details of the supraorbital region and particularly the occiput were thought to ally

Ngandong much more closely with the Broken Hill skull and placed them as the oldest known representatives of “Homo sapiens fossillis” (Oppenoorth, 1937). According to

Weidenreich (1951), however, the views expressed by Oppenoorth (1937) represent a considerable change from an earlier stance (Oppenoorth, 1932a, 1932b, 1932c) that

Ngandong represented a Neandertal type. This about face was likely due to the influence of Dubois, who, as mentioned earlier, was vehemently opposed to the idea that any of the fossils from Sangiran or Ngandong represented anything but ancient Homo sapiens

(Weidenreich, 1951).

An alternative viewpoint held that the Chinese and Indonesian fossil material that

had been recovered to that point (minus the Wajak fossils and the Upper Cave material

from Zhoukoudian) represented a single species, Homo (Pithecanthropus) erectus. This idea was first put forth by von Koenigswald and Weidenreich (1939) regarding the material from Sangiran and Zhoukoudian, and later work (Weidenreich, 1943, 1951) also included the Ngandong hominids in this grouping. While he used many different names, such as Sinanthropus and Pithecanthropus, to refer to these various fossil specimens,

Weidenreich (1951: 227) explained that he used these names only to “assign a given hominid specimen to a place in the phylogenetic morphological sequence” and that he did not consider them taxonomic designations in the strict sense. In fact, the Chinese and

Indonesian forms were identical in 57 out of 74 character states that could be examined, and the two samples differed in only four characters according to a summary by

Weidenreich (1943).

Weidenreich (1943, 1945b) found that Sangiran 2 represented the same type of hominid as the Trinil skullcap, despite the previously mentioned objections by Dubois

(1936, 1937b, 1938b). Sangiran 3 was diagnosed as a juvenile, but nonetheless exhibited characteristics typical of Trinil and Sangiran 2 (Weidenreich, 1943, 1945b). Sangiran 4, on the other hand, was more difficult to interpret due to its larger size as well as the retention of some more primitive characteristics, such as a maxillary diastema

(Weidenreich, 1943). Weidenreich (1943, 1945a, b, 1946) felt that Sangiran 4 might represent a link between more robust older forms, represented by the massive Sangiran 6 mandibular fragment, and the more lightly built Pithecanthropus skulls. Or, sexual dimorphism might also be invoked to explain the diversity in size and robusticity seen at

Sangiran (Weidenreich, 1943). The Ngandong hominids were seen as “an enlarged

Pithecanthropus type on the way to an advanced form” (Weidenreich, 1943: 274).

Weidenreich (1943) regarded the Solo specimens as more primitive than the Neandertals and morphologically very similar to the preceding pithecanthropines, yet somewhat closer to modern humans. It was on this basis that he made his famous pronouncement that “[t]here is an almost continuous line leading from Pithecanthropus through Homo soloensis and fossil Australian forms to certain modern primitive Australian types”

(Weidenreich, 1943: 276). As mentioned earlier, this assertion reverberates to the present day.

GHR von Koenigswald (1956) generally agreed with Weidenreich in his interpretation of the growing Javan fossil sample, but differed on a few key points and taxonomic designations. Koenigswald (1956) erected the species Pithecanthropus modjokertensis after discovery of the Modjokerto child in 1936, and placed Sangiran 4 in that species as an adult example. In his opinion the Sangiran 1 jaw, which was larger and more robust than the other pithecanthropines discovered to date, also belonged with this species and felt that this assemblage likely represented a robust ancestor of P. erectus

(Koenigswald, 1956). Weidenreich (1943, 1945, 1946) agreed in principle with this interpretation, but placed Sangiran 4 under a different species name, Pithecanthropus robustus, which would lead to some confusion in later publications. Koenigswald (1956) also differed with Weidenreich (1943, 1951) in considering the Ngandong fossils to be a tropical Neandertal. This view presupposes a worldwide “Neandertal phase” of hominid development prior to the development of more modern features. Koenigswald (1956) felt

that the Ngandong specimens were too recent and advanced to represent a pithecanthropine and instead provided examples of this Neandertal phase on Java.

When large-scale excavations resumed on Java in the 1960’s Indonesian scientists began publishing their own interpretations of the fossils, both new and old, and attempted to reconcile newer finds with previous ideas regarding the fossil sequence. Sartono (1964,

1967, 1968, 1971, 1972, 1975, 1990) was one of the more prolific Indonesian workers and participated in a number of fossil discoveries. His views of human evolution in

Southeast Asia were very similar to Weidenreich’s (1943, 1951) in that he perceived a lineal progression from Javan Pithecanthropus to later modern humans in Australia

(Sartono, 1975). The discovery of Sangiran 17 (which he referred to as Pithecanthropus

VIII) in 1969 yielded the most complete cranium recovered to date, and it also had a relatively large cranial capacity. This skull provided valuable new insight into the level of variation present at Sangiran. Sartono (1975) envisioned two different scenarios to explain the progression of forms on the island. The more speciose hypothesis contained five different species or subspecies (Meganthropus, P. dubius, P. modjokertensis, P. erectus, and P. soloensis) while his preferred model utilized two subspecies, a small- brained group and a large-brained group (Sartono, 1975). The large-brained group was considered chronologically younger and more advanced, and contained the Zhoukoudian skulls, Sangiran 17 (Pithecanthropus VIII), and the Ngandong crania. Sartono (1975) posited that the small-brained group evolved into the large-brained group, and that

Sangiran 12 (Pithecanthropus VII) could have served as an intermediary form between the two groups.

Jacob (1969, 1972a, b, 1975, 1976, 1978a, b, 1979, 1981, 1984, 2001) has also written extensively on the Indonesian hominid sample. He has argued convincingly that the Sangiran and Ngandong crania were probably not victims of cannibalism (Jacob,

1969, 1972b, 1978a, 1981) and also corrected what he perceived as errors by

Weidenreich (1951) in his descriptions of the Solo fossils (Jacob, 1969, 1978a). Like

Sartono, Jacob has identified a number of different groups in the Javan fossil assemblage.

These groups include robust and gracile lines, with the robust group including P. modjokertensis (represented by the child’s skull and Sangiran 4) and P. soloensis

(Ngandong, Sambungmacan 1, and Sangiran 17), and the gracile group inhabited by P. erectus (the remaining Trinil and Sangiran specimens) (Jacob, 1975, 1978b, 1979, 1984).

The robust group appears during the early Pleistocene and survives until the end of that period, while the gracile group is known only from the middle Pleistocene (Jacob, 1975,

1979). Jacob (1975, 1976, 1978a, b, 1979) suggests that P. modjokertensis evolved into

P. soloensis during the middle or late Pleistocene, and that the latter species is differentiated by a number of unique features on the cranial base. P. erectus also evolved from P. modjokertensis, and “in turn evolved into late progressive pithecanthropines and early primitive Homo whose remains have not yet been discovered” (Jacob, 1979: 9). P. soloensis may also have evolved into Homo and contributed genes to later H. sapiens in the region (Jacob, 1976).

Santa Luca (1980) included most of the important fossils from Java in his oft- cited analysis of the Ngandong crania. Through his study of the craniometrics of this group he identified several characteristics of this sequence that he found noteworthy. For example, Santa Luca (1977, 1980) differed from Sartono (1975) in that he found the

Trinil and Sangiran 2 calvaria probably represented the most primitive forms on the island while Sangiran 4 is actually more advanced morphologically. In his opinion it was more likely that Sangiran 17 was closely related to Trinil/Sangiran 2 while the Ngandong crania might have evolved from an ancestor similar to Sangiran 4 (Santa Luca, 1977,

1980). Santa Luca (1977, 1980) also discounted Jacob’s (1975, 1976, 1979) placement of

Sangiran 17 and the Sambungmacan 1 cranium into a Pithecanthropus soloensis group with the Ngandong skulls.

During the mid-1980s the Javan finds played a role in a growing debate over the validity of the species H. erectus. A number of authors, including Stringer (1984),

Andrews (1984), and Wood (1984, 1991) proposed splitting the African and Asian specimens that were currently subsumed under H. erectus into two species. The Asian specimens would retain that designation, while the African fossils, which lacked certain autapomorphies deemed peculiar to the Asian forms, would be placed in Homo ergaster

(Andrews, 1984; Stringer, 1984; Wood, 1984). This hypothesis was countered by a number of studies which pronounced that the so-called autapomorphies were not only variable within the African and Asian samples but could also be found on non-erectine specimens (Rightmire, 1990; Kennedy, 1991; Bräuer and Mbua, 1992). In response,

Wolpoff and colleagues (1994) have called for the sinking of H. erectus altogether, claiming that the evidence for regional continuity is so convincing that these specimens should instead be re-classified as early Homo sapiens. This solution has also been met with a great deal of skepticism.

Rightmire (1984, 1990, 1992, 1994) is one of the leading skeptics of this extreme lumping viewpoint. He asserts that Homo erectus is clearly a diagnosable taxon separate

from later H. sapiens (Rightmire, 1992), and feels that splitting the sample into two or more species is unwarranted (Rightmire, 1984, 1990). In fact, Rightmire (1981, 1990) contends that there is little evidence for significant change over the lifetime of the species. This apparent stasis would indicate that Homo erectus was not evolving in the

direction of modern humans and was likely replaced by more advanced hominids in the

late Pleistocene (Rightmire, 1990, 1992, 1994). Rightmire (1990) does not see any

justification for more than one species of Homo in the Pleistocene of Java. He feels that

the specimens share a number of similarities, and any differences in size can probably be

explained by sexual dimorphism (Rightmire, 1990). Even the Ngandong hominids show a

typical Homo erectus pattern that is not intermediate in form (Rightmire, 1994).

Kramer (1989, 1991, 1993, 1994; Kramer and Konigsberg, 1994) has also

contributed substantially to this debate. His work on the Sangiran mandibles led to the

dual conclusions that the sample could be accommodated within a single species, and that

that species (H. erectus) shared a number of similarities with modern Australian aborigines (Kramer, 1989, 1991). Further, the robust Sangiran mandibles do not show any australopithecine affinities (Kramer, 1989, 1994) as has been suggested by some authors (e.g. Tyler, 1991, 1994). Kramer (1993) also supports a single-species scenario for the cranial specimens allocated to H. erectus, and does not see any justification for splitting the sample into African and Asian taxa.

Other recent contributions to this debate include work by Antón (2001, 2002,

2003; Antón et al., 2002), Kidder and Durband (2000, 2004), and Durband (1998, 2002a, b, 2004). Metric work on the African and Asian samples by Kidder and Durband (2000,

2004) found that the African and Indonesian fossils share a similar metric pattern to the

exclusion of the Zhoukoudian specimens, and this conclusion has since been supported

by Antón (2001, 2002, 2003). These studies show general homogeneity in the overall

cranial shape in the Indonesian sample, and provide support for a single lineage of

hominids on Java. Work with the non-metrics of the Southeast Asian hominids, however,

may indicate that a multiple-species solution is required. Durband (1998, 2002a, b, 2004)

has identified a number of potential autapomorphies on the cranial bases of the

Ngandong, Sambungmacan, and Ngawi samples that may indicate evolution beyond the

condition seen in the Sangiran hominids. These features are also not found in any modern

human samples examined to date, which would suggest that that group of specimens

represented a population that went extinct after a period of differentiation from earlier

hominids found on the island (Durband 2002b, 2004).

Chronology of the Australian fossil sample

Currently the oldest accepted archaeological evidence for the presence of humans on

Australia comes from the Malakunanja II site in the Northern Territory, which has been dated to approximately 50 kyr (Roberts et al., 1990). In 1996 claims of dates as old as

176 kyr were made for the Jinmium rockshelter, Northern Territory (Fullagar et al.,

1996), but more recent work has pointed out methodological flaws in this study (Roberts et al., 1998). Some workers have pointed to changes in the vegetational regime and pollen composition in Australia starting as early as 120 kyr as the earliest signs of human activity (Singh et al., 1981; Wright, 1986), but these conclusions are based on very limited studies and have no archaeological evidence to support them (Bowdler, 1993).

Further circumstantial evidence for the early appearance of humans on Australia is

supplied by the disappearance of Genyornis newtoni, a flightless bird, at approximately

50 kyr (Miller et al., 2003). Miller and colleagues (2003) theorize that centuries of intensified burning in the interior of Australia by human colonizers profoundly altered the ecosystem and contributed to the extinction of Australia’s megafauna. Miller and his team (2003) used amino acid racemization to pinpoint the last appearance of Genyornis and point the finger at a potential human causation for this extinction event (though see

Stone and Cupper, 2003).

The earliest human skeletal evidence from Australia is from Lake Mungo, New

South Wales (Thorne et al., 1999; Bowler et al., 2003). J.M Bowler and his team excavated two reasonably complete skeletons, Mungo I and III, found eroding out of calcrete blocks. Mungo I was discovered in 1968 (Bowler et al., 1970) while Mungo III was not found until 1974 (Bowler and Thorne, 1976). There is general agreement among scientists that these two skeletons represent the oldest securely dated human fossils on the continent, however there has been a great deal of recent debate concerning the exact age of Mungo III, the older of the two. The original radiocarbon dates on the site placed

Mungo I at approximately 32 kyr (Bowler et al., 1970) while Mungo III was dated to between 28-32 kyr (Bowler and Thorne, 1976). These dates were later revised by Bowler

(1998) to between 42-45 kyr after detailed work on the site, and this agreed with an additional published Thermoluminescence date of 43 kyr on Mungo III sediments by

Oyston (1996). In 1999, Thorne and colleagues redated the Mungo III burial using ESR and U-series techniques and arrived at an age estimate of 62 ± 6 kyr for the bones themselves and 61 ± 2 kyr for the sediments of the gravesite. This new set of dates was met with a great deal of skepticism by a number of authors (Gillespie and Roberts, 2000; 27

Bowler and Magee; 2000; Brown and Gillespie, 2000; O’Connell and Allen, 2004) who pointed out that these new ages were quite incongruous with the series of over 200 dates sampled throughout the stratigraphic sequence at the site. In an effort to correct these anomalies, the Lake Mungo burials have been recently redated (again) to 40 ± 2 kyr through a series of 25 optical ages (Bowler et al., 2003). This set of dates agrees more closely with earlier work (Bowler, 1998; Bowler and Price, 1998; Gillespie, 1998;

Oyston, 1996) on the ages of the Mungo skeletons (O’Connell and Allen, 2004). Further, these dates place the initial occupation of the Willandra Lakes region at between 40-50 kyr (Bowler et al., 2003) during a period of increased aridity. As will be discussed in the next section, some workers attribute the high levels of robusticity present in Pleistocene

Australian populations to living in a drier, harsher climate (Bulbeck, 2001).

The remainder of the archaeological and skeletal evidence found on the continent is 40 kyr or younger, and a number of sites date to between 30-40 kyr (Bowdler, 1993).

The majority of these older sites are purely archaeological, however, and do not contain any appreciable human skeletal remains. After the Mungo remains, several other skeletons in the Willandra Lakes series have been dated to between 11-18.6 kyr (Webb,

1989), but these dates are now questioned by Gillespie (1997, 1998). Further south, the

King Island skeleton dates to approximately 14.2 kyr (Sim and Thorne, 1990). The Kow

Swamp skeletons have recently been redated to between 19-22 kyr through single aliquot

OSL dating (Stone and Cupper, 2003), which is significantly older than the 6.5-13 kyr dates obtained through radiocarbon analysis (Thorne, 1975; Brown 1992). However, since the OSL dates are on sediments and not the skeletons themselves they have not been generally accepted (Thorne, personal communication; Bulbeck, personal

communication). Some of the Coobool Creek specimens are thought to be as old as 14-15 kyr, with one U-series date taken from cranium 65 providing an age of 14.3 ± 1 kyr

(Brown, 1989). The Nacurrie 1 skeleton has an AMS date of 11.4 kyr (Brown, 1994b) and is considered to be morphologically similar to both the Kow Swamp and Coobool

Creek samples (Brown, 1992). The Keilor cranium is approximately 12 kyr (Macintosh and Larnach, 1976), while the soil horizon from which Talgai was excavated has been dated to 11.6 kyr (Oakley et al., 1975). A number of other skeletons (Mossgiel, Green

Gully, Cossack, etc.) are dated between 5-9 kyr (Pardoe, 1993). The Cohuna cranium is undated, but a reappraisal of the site by Macumber and Thorne (1975) led them to posit a probable date of 9-13 kyr for the cranium. Brown (1987, 1989) considers Cohuna to date to the terminal Pleistocene based on its morphological similarity to Kow Swamp and

Coobool Creek.

Thus, based on the best available evidence Australia has been occupied by humans for approximately 50 kyr and perhaps slightly longer.

Previous interpretations of the Australian fossil record

There has likewise been considerable debate during the past 50 years surrounding the composition of the first permanent human groups to reach Australia. The earliest theories, from workers like Birdsell (1949, 1950, 1967) and Morrison (1967), sought to explain the variation present in living populations of Aboriginals though the genetic contributions of multiple founding populations. This initial work profoundly influenced the history of this debate into the late 20th century and surely contributed to later attempts to explain the variation seen in skeletal remains of Pleistocene inhabitants though a population

hybridization model (Thorne, 1971, 1976, 1977, 1981, 1984, 1989; Thorne and Wilson,

1977; Thorne and Wolpoff, 1981).

Of these early theories, Birdsell’s (1949, 1967) trihybrid model was clearly the most influential. While many early authors, including Keith (1925) and Jones (1934) maintained that the Australian Aboriginals represented a homogenous population,

Birdsell (1967) criticized this work on the basis that it was done only on cranial samples and did not incorporate any measurements or observations from living subjects. Birdsell

(1949, 1967), on the other hand, formulated his theory of three successive waves of migrants after work involving anthropometric measurements, skin and eye pigmentation, hair color and type, dental morphology, and various blood groups. Using these characters,

Birdsell (1949, 1967) hypothesized that there were three ancestral sources for the gene pool of modern Australians that arrived in successive waves of immigration. The first was the Oceanic Negrito, which was characterized by short stature, dark skin, woolly hair form, and a short narrow face (Birdsell, 1950). The second wave brought the Murrayian people, who were characterized by short stature, relatively light pigmentation, wavy to straight hair form, and a massive face with large brow ridges (Birdsell, 1950). Finally, the third major influx of genes was brought by the Carpentarians. This group was characterized by tall stature, very dark skin, wavy to straight hair form, and a high and narrow skull with large brows (Birdsell, 1950). The distribution of features in modern

Australians could be explained through interactions between these groups. Birdsell (1949,

1950, 1967) found that much of the Negrito contribution to the gene pool had been swamped by the subsequent waves of invaders. Present day descendants of the Negritos had been marginalized to only a very small percentage of the landmass, while the

Murrayians had settled in the southern part of the continent and the Carpentarians took the northern areas (Birdsell, 1967). Birdsell (1967) felt that the archaeological record of both Australia as well as mainland Asia supported this theory and provided evidence for each of his three types in the distant past. In fact, he categorized some of the known

Pleistocene Australian skulls, calling Keilor “classic Murrayian in type” (Birdsell, 1967:

148) and also claimed that the Wajak 1 skull from Java represented this group. It is interesting to note that Birdsell (1967) also speculated that the more primitive traits seen in his Murrayian and Carpentarian types could potentially be attributed to genetic exchange with Homo soloensis or other non-sapiens populations.

Another hybridization model for the origins of the indigenous Australians was put forth by Morrison (1967: 1056), who felt that the Aboriginals “are derived from at least two successive waves of immigrants, who were genetically dissimilar.” He based his theory on a number of different genetic markers found in the blood, including ABO,

MNS, Gc serum, Gm serum, Haptoglobins, and Transferrins (Morrison, 1967). These markers were found to show significant differences between those Australians who lived in the interior and those inhabiting the peripheral areas of the continent (Morrison, 1967).

Certain genetic markers were restricted to the center of Australia, and Morrison (1967) felt that the best explanation for the lack of these markers in other areas was best explained by the presence of at least two founding populations.

A number of other researchers arrived at the more economical conclusion that the data do not warrant hybridization events and instead represent a single, morphologically variable founding population (Abbie, 1951, 1963, 1966; Howells 1973, 1977; Larnach,

1974; Macintosh and Larnach, 1976). As Howells (1977) points out, before 1950

anthropologists were concerned mainly with the idea of “pure racial strains” and tended to approach problems of populational history through hybridization or migrations to explain variation. These obsolete ideas clearly influenced Birdsell (1949, 1967) and his theories of large-scale population amalgamations in Australia. Howells (1973a, b, 1977) found that the cranial samples from Australia were quite uniform and that there was no basis for subdividing the sample. Larnach (1974) likewise found the Australian sample to be fairly invariable and homogenous, and colorfully adds that “images of Negritos,

Murrayians, and Carpentarians are ghost images which disappear as we trace them back towards Aboriginal beginnings” (213). Instead of wholesale movements of diverse

“racial” stock, more complex interactions involving selection and adaptation to localized environments could be invoked to explain the diversity seen in the cranial, anthropometric, and serological data sets (Howells, 1977; Pardoe, 1991a).

Just as the Birdsellian ideas regarding multiple founding stocks were falling from favor, however, a new brand of hybridization theory was proposed by Alan Thorne

(1971, 1976, 1977, 1980, 1981, 1984, 1989; Thorne and Wilson, 1977). As a doctoral student Thorne took part in the excavations of both the Lake Mungo and Kow Swamp burials, and he was intrigued by the differences between these sets of individuals.

According to Thorne (1976: 105), “The Kow Swamp crania are large by more recent

Aboriginal standards. They indicate a greater robusticity or ruggedness.” By contrast, the

Lake Mungo I individual “is lightly constructed and has weakly developed muscle insertion sites” and “is striking” in its differences with the Kow Swamp material (Thorne,

1976: 109). Likewise the Mungo III skeleton, which Thorne (1977) diagnosed as male, was classified as gracile in overall form. These two skeletal samples formed the basis for

a new “robust” and “gracile” dichotomy proposed by Thorne (1971, 1976, 1977, 1980,

1981, 1984, 1989) to encompass what he felt was an extreme amount of variation in the

Pleistocene fossil sample from Australia. Thorne later expanded his classifications to include other well-known fossils, placing Keilor, Green Gully, King Island, and Lake

Tandou in the gracile group and specimens such as Cohuna, Cossack, Mossgiel, WLH

50, and the Coobool Creek sample in the robust group (Freedman and Lofgren, 1979;

Habgood, 1986; Sim and Thorne, 1990; Thorne and Wolpoff, 1992). Work by

Pietrusewsky (1979), Macintosh (1963, 1967a, b) and Freedman and Lofgren (1979) supported the notion that two differing morphologies were present in Australia during the

Pleistocene.

As with Birdsell (1967), migration from elsewhere was invoked to explain these radically different morphologies. Thorne (1980: 40) suggested that the robust group had

“the mark of ancient Java” while the gracile group bore the “stamp of ancient China,” and thus both mainland Asia as well as insular Southeast Asia contributed to the modern

Australian gene pool. Features such as a low sloping frontal, large brow ridges, facial prognathism, and thick cranial bone linked the Australian robust group with Indonesian

Homo erectus specimens to their immediate north, particularly the Ngandong fossils

(Thorne, 1977, 1980). Meanwhile, the predecessors of the gracile Australians could be found in the Upper Cave folk from Zhoukoudian and the Liujang cranium (Thorne,

1980). Examples of these gracile peoples could be found closer to Australia in the Niah

Cave deep skull in Borneo, Tabon in the Philippines, and Wadjak 1 in Java (Thorne,

1980). These ideas contributed substantially to the formation of the Multiregional Theory of modern human origins (Thorne and Wolpoff, 1981; Wolpoff, 1989, 1992, 1999;

Wolpoff et al., 1984), and continue to be espoused by Thorne (2002; Thorne and Sim,

1994; Thorne et al., 1999; Thorne, personal communication).

Several workers, among them many Australian scholars, fail to find support for

Thorne’s dihybrid scenario. As stated earlier, both Howells (1973a, 1977) and Larnach

(1974) found the Australian cranial samples to be homogenous, though they also found that Australians as a group retained an unusually high number of primitive characteristics. These plesiomorphies include thicker cranial bone, larger mean size, and greater development of some cranial superstructures like the occipital torus (Larnach,

1974). Multivariate work by Habgood (1986) also cast doubt on Thorne’s robust vs. gracile dichotomy, finding that when compared to a world sample of late

Pleistocene/early Holocene crania the Australian specimens clustered with one another and away from other groups. Pardoe (1991a) concurred, and attributed the variation present in the Pleistocene Australian sample to simple sexual dimorphism. The most vociferous critic of Thorne’s dihybrid vision has been Peter Brown (1981, 1987, 1989,

1992, 2000). Brown has questioned the legitimacy of Thorne’s robust and gracile morphs on several fronts, including the accuracy of sex diagnoses (Brown, 1994a, 1995, 2002;

Brown and Gillespie, 2000), the confounding effects of cultural practices on cranial morphology (Brown, 1981, 1989), and the rationale behind the diagnosis of robustness itself (Brown, 1987, 1992, 2002; Brown and Gillespie, 2000). For example, Brown

(1987, 1989, 1994a) has criticized Thorne’s (1977; Thorne and Wilson, 1977) diagnosis of the Keilor specimen as gracile despite its large size and general robusticity. Likewise,

Brown (1994a, 1995, 2000; Brown and Gillespie, 2002) takes issue with Thorne’s (1980;

Thorne et al., 1999) assignment of male sex to the Mungo 3 specimen, stating that

“[w]hile [Mungo 3] is certainly tall and robustly built in comparison with late

Holocene female Aborigines, outside the Holocene female range for some

postcranial dimensions, this is not enough to indicate male status for a Pleistocene

Australian. Supraorbital morphology, as well as frontal curvature and absence of a

median frontal ridge, is decidedly feminine and contrasts with all of the Coobool

Creek, Kow Swamp, and Nacurrie males.” (Brown, 2000: 748)

A lengthy debate was also waged over the sex of the King Island skeleton (Brown,

1994a, 1995; Sim and Thorne, 1995; Thorne and Sim, 1994), which Brown felt had been too hastily categorized as male by Thorne despite several feminine characteristics. These battles highlight the fundamental concern with the dihybrid scenario in the minds of workers like Brown (1987, 1989) and Pardoe (1991a): a lack of appreciation for variation in the Pleistocene and Holocene record of Australia. Brown (1987:61) found that the

“combination of the craniometric, tooth size and vault thickness results suggests a single, homogenous Pleistocene population” that shows “a consistent Australian Pleistocene morphology.” Artificial cranial deformation practiced by the populations represented by

Kow Swamp and Coobool Creek (Brown 1981, 1987, 1989; Antón and Weinstein, 1999) has also created the false appearance of more extreme cranial variation in the sample. In short, all of the variation present in the Pleistocene Australian cranial sample could be accommodated in a single variable population exhibiting sexual dimorphism, and explanations requiring multiple founding populations were unsupported by the evidence.

While the single population explanation for the peopling of Australia would appear to be the most parsimonious, it must still be able to explain why the inhabitants of

Australia become increasingly robust during the end of the Pleistocene. This observed

increase in robusticity runs counter to the reduction in robusticity seen elsewhere in the world (Wolpoff, 1999) and has been difficult to explain without invoking migrations of skeletally robust people (e.g. Thorne, 1977, 1980). Proponents of the single population theory would argue that selection, and not hybridity, can explain the transition from a relatively gracile early Pleistocene sample to more robust late Pleistocene populations and then back to more delicately built modern groups. Brown (1987) feels that increased

Holocene air temperatures on Australia might have influenced body proportions though a reduction in overall body size. This reduction could have affected cranial dimensions and tooth size and concurrently reduced prognathism as well. Bulbeck (2001) hypothesizes that increasingly harsh climatic conditions during the last glacial maximum on Australia could have led to an increase in robusticity like that seen at Kow Swamp, and that amelioration of those conditions after would have allowed the reduction in robusticity seen into the Holocene. This scenario would be consistent with the relative gracility seen in the early Lake Mungo skeletons, an increase in robusticity throughout the continent around 19-21 kyr, and then the subsequent reduction in robusticity expected in fully modern populations (Bulbeck, 2001; Stone and Cupper, 2003). Better understanding of the nature and timing of climatic changes in Australia during the late Pleistocene (e.g.

Bowler et al., 2003; Stone and Cupper, 2003) may help clarify the selective pressures at work in those populations.

On a final note, recent DNA studies have also provided information germane to the question of the founding population(s) of Australia. Analysis of 101 complete mitochondrial DNA (mtDNA) lineages, including 41 from Australian and New Guinean aborigines, indicates that modern humans arrived on the Australian continent less than

70,000 years ago (Ingman and Gyllensten, 2003). These dates are in general agreement with the archaeological data discussed above. Using this mtDNA data, Ingman and

Gyllensten (2003: 1605) posit “either a common colonization of [Australia and New

Guinea] from a heterogenous source population or independent colonization events and subsequent gene flow.” On average, the mtDNA sequences from New Guinea and

Australia are more closely related than either of those populations is to Asian sequences

(Ingman and Gyllensten, 2003). Other work focusing on the Y-chromosome has identified an Australia-specific haplotype (Kayser et al., 2001; Redd et al., 2002), and suggests a population expansion that may have started from only a few hundred individuals.

Adcock and colleagues (2001) extracted ancient mtDNA from a number of

Australian Pleistocene fossil specimens, including Mungo III, and found no significant distinction between anatomically robust and gracile specimens. While the Mungo III sequence was found to differ substantially from the other fossil and living Australians, suggesting a divergence of the Mungo III sequence before the last common ancestor of contemporary human populations, the remainder of the fossil Australians tested formed a clade with living Australian aboriginal mtDNA sequences (Adcock et al., 2001). These findings cast further doubt on a dihybrid explanation for Pleistocene Australian morphological variation (Adcock et al., 2001; Relethford, 2001b). However, see Smith and colleagues (2003) for an alternative view on the potential hazards of sequencing ancient DNA from Lake Mungo.

Thus, the controversies surrounding the origins of modern humans in Australia have predominantly followed, and in some cases (i.e. Weidenreich, 1943) are responsible for founding, the larger debate of Multiregional Evolution vs. a Recent Replacement model.

From these two predominant schools of thought very clear predictions can be made regarding the fossil sequences of Indonesia and Australia and their contribution to later modern human populations in the region.

The predictions of the Multiregional Hypothesis of modern human origins as they would be applied to the Australasian fossil record are as follows: (1) that modern humans will share at least some diagnosable features in common with the regional archaic populations which preceded them in that region, (2) there will not be only one “modern” morphological pattern, as each regional transition would have proceeded at different rates and upon different regional characteristics, and (3) modern humans in a particular region will most resemble the archaic humans from that same region.

Conversely, a Replacement Hypothesis of modern human origins would posit the following predictions for the fossil record: (1) a morphological break between the archaic hominids in the region (i.e. the Sangiran/Ngandong group) and the earliest modern humans in the region (i.e. Pleistocene Australian populations), and (2) that the modern humans would share one “modern” morphological pattern similar to that found in other modern populations and unlike that pattern found in archaic hominids.

Less extreme variants of the models mentioned above can be found in the Afro-

European Sapiens Hypothesis of Bräuer (1984a, b, 1989, 1992), and the Assimilation

Hypothesis of Smith and colleagues (1989). Both models accept varying degrees of hybridization between modern humans and more archaic forms. Brauer (1984b, 1989,

1992) would emphasize replacement over hybridization and allow for trivial gene flow between the invading modern humans and regional archaic populations. Thus, the predictions for his Afro-European model would largely mirror those of the strict

Replacement model with the possibility of some small regional similarities. The

Assimilation model (Smith et al., 1989; Smith, 1992), on the other hand, stresses regional continuity with archaic populations as a significant component in the transition to fully modern populations. This model, essentially the inverse of the Afro-European Sapiens

Hypothesis (Shreeve, 1995), begins with a similar assumption that early modern humans probably arose in Africa and migrated throughout the Old World. However, those invading early modern populations then “assimilated” the archaic populations in each region into their gene pools resulting in a very high degree of regional continuity.

Examination of the cranial bases present in the fossil records of Indonesia and

Australia would provide an ideal test of the predictions for each of these models. As mentioned earlier, the cranial base is at least as useful as other areas of the skull for determining phylogenetic relationships. This evidence has not been previously examined in this region, and in fact even published descriptions of the relevant specimens are quite incomplete for this purpose. Thus, a study of the cranial base in the context of modern human origins in Australasia would be a valuable contribution to paleoanthropology.

Chapter 2: Materials and Methods

Materials

The fossil sample chosen for this study is shown in Table 2.1. The original fossil specimens from Ngandong, Sangiran 14, and Sambungmacan 1 were examined at Gadjah

Mada University, Yogyakarta, Indonesia on July 14-15, 1999. Sangiran 12 and 17 were examined at the Geological Research and Development Centre, Bandung, Indonesia on

June 18 and July 22, 1999. Wajak 1 was examined at the National Natuurhistorisch

Museum, Leiden, The Netherlands, on May 12, 2003. Sangiran 2 and 4 were studied at the Senckenberg Museum, Frankfurt, Germany on May 13 and 15, 2003.

In Australia, access to the original fossil specimens was not possible because many of the human remains from the late Pleistocene and early Holocene have been repatriated and/or reburied by various Aboriginal communities. In the instances where specimens had not been repatriated permission to access the originals was not granted.

Thus, examination of casts, photographs of the original specimens, and previously published descriptions were the only recourse available for this project. Casts of Cohuna,

Cossack, Keilor, Kow Swamp 5, Lake Mungo I, Lake Mungo III, Lake Nitchie,

Mossgiel, Nacurrie 1 (“Murrabit”), and Talgai were studied at the Shellshear Museum,

Sydney, Australia on May 12-14, 2004. Casts of Cohuna, Keilor, Kow Swamp 5, and

Kow Swamp 8 were examined in the lab of Dr. Alan Thorne, Canberra, Australia on May

17, 2004.

Access to high quality photographs was provided for two additional specimens that are either unpublished or in the process of being published. A photograph of the 40

Table 2.1: Fossil specimens included in this project. Casts are denoted by an *.

Sangiran 2 Keilor* Sangiran 4 Talgai* Sangiran 12 Kow Swamp 5* Sangiran 14 Kow Swamp 8* Sangiran 17 Lake Mungo I* Ngandong 1 Lake Mungo III* Ngandong 6 Nacurrie 1* Ngandong 7 Mossgiel* Ngandong 10 Cossack* Ngandong 11 Cohuna* Ngandong 12 Lake Nitchie* Sambungmacan 1 Sambungmacan 3* Ngawi 1* Wajak 1

Sambungmacan 4 cranial base was graciously provided by Dr. Hisao Baba. Likewise,

Michael Westaway was kind enough to provide me with a large copy of a photograph of the Nacurrie 2 cranial base he had previously used in an unpublished report (Westaway,

2002b).

A sample of 309 modern human crania was also examined for this project (Table

2.2). These specimens were chosen in an effort to represent different regional populations and therefore provide a sampling of modern human cranial base morphology from around the world. Sex was estimated by visual examination of the skull, as postcranial elements were not available for study. In the case of the Schmidt Collection from the University of

Leipzig, which included the majority of the crania studied at the Senckenberg Museum, the sex for each cranium was published (Schmidt, 1887).

Table 2.2: Modern human sample used for this project. Tasmanian crania were grouped with the Australian sample during the non-metric examinations.

Location AMNH NMNH FMNH Senckenberg Australia 25 M, 18 F 20 M, 14 F 2 M 1 M Tasmania 5 M, 2 F 1 M, 1 F W. Africa 13 M, 15 F 1 M S. Africa 16 M, 12 F 1M, 1 F Greece 17 M, 8 F Czech 18 M, 7 F Utah 10 M, 10 F Austria 11 M, 13 F Java 20 M, 6 F Egypt 10 M, 7 F India 11 M, 1 F

Methods

Non-metric observations

Seventeen observations were recorded after visual examination of the cranial base, with

one additional character scored metrically. These traits were typically scored on the right

side of the cranium unless it was obscured or missing, in which case it was scored on the

left side. If the trait was unavailable bilaterally it was scored as “NA”. Figure 2.1

provides a map of the skull base with numbers corresponding to the following scored

traits (Table 2.3):

1. Pharyngeal tubercle

The pharyngeal tubercle is located near the center of the basioccipital, and is the attachment site of the superior constrictor muscle of the pharynx (Gray, 1985). This feature is scored as 0-absent, 1-trace (palpable), or 2-present.

Table 2.3: Cranial base features examined for this project. Numbers correspond to location of feature on Figure 2.1.

Feature 1. Pharyngeal tubercle 2. Tympanic plate contact with mastoid 3. Foramen ovale in pit 4. Foramen ovale accessorium 5. Styloid process development 6. Vaginal process development 7. Development of the alar tubercles 8. Orientation of the occipital condyles 9. Opisthionic recess 10. Postcondyloid tuberosity (PCT) development 11. Foramen lacerum development 12. Juxtamastoid crest development 13. Occipitomastoid crest development 14. Location of the carotid foramen relative to the squamotympanic (S-Q) fissure 15. Size of postglenoid tubercle 16. Orientation of the S-Q fissure 17. Relative area of occipital condyle to foramen magnum

Figure 2.1: Map of the cranial base features to be examined. The numbers correspond to the numbering system used for the features in this section. Photograph adapted from White (1991).

2. Tympanic plate contact with mastoid

The inferior-most portion of the tympanic plate is typically joined to the anterior surface of the mastoid process in modern humans, but in some groups of archaic hominids they do not come in contact. The resulting gap is called the tympanomastoid fissure, and some workers (e.g. Andrews, 1984; Stringer, 1984) have used this feature to separate the

African and Asian specimens of Homo erectus. This feature is scored as 0-no contact or

1-contact.

3. Foramen ovale in pit

The foramen ovale is located toward the inferolateral edges of the greater wing of the sphenoid bone. In modern humans this foramen transmits cranial nerve V3 and the accessory meningeal arteries. In Ngandong 7 and 12 the foramen ovale is recessed from the surface of the skull base and is located at the base of a 4-5 mm deep pit. Weidenreich

(1951) was the first to notice the peculiar morphology of the foramen ovale in the

Ngandong hominids, and later work by Jacob (1969) confirmed his earlier diagnosis. This feature is scored as 0-absent or 1-present.

4. Foramen ovale accessorium

A second aspect of the Ngandong morphology of the foramen ovale is the presence of an accessory foramen that lies medial to the main foramen. Jacob (1969: 17) noted that this accessory foramen “is certainly not the foramen of Vesalius or Hyrtl or Civinini, and is not the innominate canaliculus.” Presence of an accessory f. ovale is scored as 0-absent or

1-present.

5. Styloid process development

The styloid process is a thin rod of bone that is typically located on the posterior edge of the tympanic bone approximately 15-20 mm medial of the external auditory meatus. It serves as the origin for the stylohyoid ligament and several small muscles. The lack of a styloid process of the temporal bone has also been cited as a possible autapomorphy for

Asian H. erectus, though Jacob (1969) notes the absence of a bony styloid in as many as

8-34% of modern humans. This feature is scored as 0-absent or 1-present, and would be scored as “present” if there is any indication that a bony styloid had been fused to the tympanic bone.

6. Vaginal process development

In modern humans the vaginal process is a thin, delicate plate of bone that extends inferiorly from the tympanic bone and envelops the bony styloid process. This morphology is different than that exhibited on several archaic specimens, who instead have a thick, rugose ridge along the inferior edge of the tympanic bone that was dubbed the crista petrosa by Weidenreich (1943). This feature is scored as 0-absent (meaning a crista petrosa is present) or 1-present.

7. Development of the alar tubercles

The alar tubercles are formed on the lateral sides of the basioccipital by the attachment of rectus capitis anterior minor. Their development is scored as 0-absent, 1-trace, 2-present,

3-present +.

8. Orientation of the occipital condyles

The occipital condyles are located bilaterally at the anterior edge of the foramen magnum. They serve as the point of articulation between the skull and the first cervical vertebra. In modern humans the long axis of the occipital condyles is typically angled toward the midline at their anterior end. In the Ngandong hominids the long axis of the occipital condyles is parallel to the midline. This feature is scored as 0-long axis parallel to midline or 1-anterior end of condyles angles towards the midline.

9. Opisthionic recess

Opisthion is the anatomical point located at the midline on the posterior border of the foramen magnum, and is typically the posteriormost point of that feature. In modern humans that border is typically rounded. In the Ngandong hominids the posterior half of the foramen magnum is highly constricted, coming to a point at opisthion. This gives the foramen magnum a distinct teardrop shape. This feature is scored as 0-absent (foramen rounded), 1-trace (foramen slightly constricted at rear), or 2-present (foramen forms recess).

10. Postcondyloid tuberosity (PCT) development

The PCT are raised, rugose, discrete mounds of bone that originate immediately posterior to the occipital condyles and run along the very edges of the foramen magnum. These structures can be over 1cm in both breadth and height as measured from the interior rim of the foramen magnum. These features are scored as 0-absent, 1-slight rugosity, 2-

rugosity with discrete tubercle development <10 mm in height, 3-tubercles >10 mm in height.

11. Foramen lacerum development

The foramen lacerum is not a true foramen, but is instead a gap left by the junction of the petrous temporal, the sphenoid, and the basioccipital. Jacob (1969) stated that this foramen was absent or constricted in the Ngandong fossils, and later (Jacob, 1976) amended this view to propose the absence of this feature as characteristic of H. erectus

(including Ngandong). This feature is scored as 0-absent, 1-constricted, or 2-present.

12. Juxtamastoid crest development

There has been some confusion in the literature regarding this crest. Corner (1898) notes the presence of two processes medial to the mastoid process on the cranial base. He proposes “that the name ‘paroccipital’ should be confined to the downward expansion of the occipital bone, and that the name ‘paramastoid’ should be applied to the process at the inner lip of the digastric groove” (Corner, 1898: 386). Taxman (1963) feels that the use of ‘paramastoid’ is confusing, and recommends calling this structure the juxtamastoid eminence or crest. The definition in use for this study roughly follows that of Taxman

(1963) as a structure that lies at the lateral edge of the digastric groove, and is wholly of temporal origin. It is scored as 0-absent, 1-trace, or 2-present.

13. Occipitomastoid crest development

The occipitomastoid crest is synonymous with the ‘paroccipital’ of Corner (1898), and is applied to an inferior projection of bone at the occipitomastoid suture. This crest typically incorporates this suture, but occasionally lies wholly on the occipital bone, and forms the medial border of the digastric groove. This feature is scored 0-absent, 1-trace, 2-present.

14. Location of the carotid foramen relative to the squamotympanic (S-Q) fissure

The carotid foramen is located in the petrous portion of the temporal bone. In modern humans, who typically have a more obtuse petrous-tympanic angle, this foramen is located more anteriorly and is typically bisected by the plane of the S-Q fissure. In earlier hominids, with more acute petrous-tympanic angles, this foramen is located more posteriorly, typically either just at or posterior to the S-Q plane. This trait is scored as 0- foramen fully posterior to fissure, 1-leading edge of foramen even with plane of fissure, or 2-foramen fully bisected by fissure.

15. Size of postglenoid tubercle

The postglenoid tubercle is formed by the inferior projection of the temporal squama at the posterior edge of the mandibular fossa anterior to the S-Q fissure, which separates the temporal squama from the tympanic bone. This trait was scored as 0-tubercle size of 2 mm or less, or 1-tubercle size of ≥3 mm.

16. Orientation of the S-Q fissure

The orientation of the S-Q fissure relative to the midline has also been used by various workers to diagnose specimens. Macintosh and Larnach (1972) include a coronally oriented S-Q fissure as one of the characteristics they use to define H. erectus. They later studied this feature in their analysis of Ngandong traits in modern human populations

(Larnach and Macintosh, 1974). This trait is scored as 0-perpendicular to midline or 1- oblique to midline.

17. Relative area of occipital condyle to foramen magnum

Durband (1998, 2002b, 2004) noted the extremely small size of the occipital condyles in the Ngandong population relative to the area of the foramen magnum. This ratio is computed using the formula (OC L x OC W)/([FM L/2] x [FM W/2] x 3.14159) obtained from Wood (1991). The right condyle was typically measured unless it was damaged or missing.

These non-metric features will be tested for statistically significant differences in the degree of manifestation between the modern populations with a Manova test. For the data on the relative areas of the occipital condyles/foramen magnum, statistical significance between the population means will be tested with a two-tailed t-test. The complete set of non-metric data collected for the modern human sample is available in Appendix A. The complete set of non-metric data for the fossil sample is located in Appendix B. The raw measurements collected for the foramen magnum and occipital condyle can be found in

Appendix C.

Three-dimensional examination of the mandibular fossa

In addition to the non-metric observations taken on the original fossils, three-dimensional laser scans were taken of the mandibular fossae of several fossil specimens in the Applied

Morphometrics Laboratory at the National Museum of Natural History, Washington D.C.

The scans were taken on high-quality casts of Ngandong 7, 10, and 12; OH 9; Sangiran

17; and Sambungmacan 3 using an NVision H40 3D Surface Laser Scanner mounted to a

FARO Silver articulated positioning arm. The raw data points were captured into

ModelMaker v.6.3.0.0., the proprietary software required to run the laser scanner. This program takes the data from the scanner and edits it to remove extraneous or redundant points, then creates a polygonal model that can be exported for analysis.

The 3-D data were analyzed in the Biomedical Engineering computer lab at the

University of Tennessee. The scanned points were populated (“spray painted”) in

Raindrop GeoMagic Studio 6, and a surface wrap was done to create a model from the point cloud. Then an infinitely thin plane was created to use as a reference plane. The reference plane was oriented in a transverse plane roughly perpendicular to the coronal plane of the skull. Using the 3-D compare feature in Raindrop GeoMagic Qualify 6, the software compared the distance of the points on the test model (the surface created from the point cloud obtained from the scan) to the reference plane. Based on the distance measured between each point and the reference, a color-coded topographic map of the mandibular fossa could be constructed.

Notes about possible sources of error

Every effort was made to control for error during the collection of data for this project. In order to control for interobserver error, the author collected all measurements and observations personally. Any measurements or observations taken from the literature are noted as such. Nearly all of the original fossils included in this study were examined twice on different days, and at least 10% of the modern human samples were also re- examined in this way to check for the consistent scoring of characters. The same set of calipers was used throughout the project. The data itself was entered into an Excel spreadsheet by the author and was visually examined for any obvious errors.

Certain cranial samples contained characteristics worthy of note here. The majority of the crania examined from the Senckenberg collection were either sagittally or coronally sectioned, or both. While this usually did not affect the collection of non-metric traits it might have affected the measurements of the foramen magnum by ~1 mm.

Several modern crania studied had adherent soft tissue, matrix, and/or cultural modifications and decorations, and characters that were bilaterally obscured by these materials were not scored. If the majority of the base was affected the cranium was not used. This philosophy was applied to modern crania with damaged areas as well.

Chapter 3: Descriptions of Indonesian cranial bases

Because descriptions of the cranial base are rare in the literature, it would be useful to

provide a synopsis of the form and condition of each of the specimens under

consideration. Photographs of the cranial bases and relevant structures on these

specimens are included in Appendix E.

Sangiran 2

Sangiran 2 was obtained in August 1937 when von Koenigswald purchased it from local collectors. The specimen consists of a well-preserved skullcap that retains only the peripheral areas of the cranial base. However, the temporomandibular joint (TMJ) is preserved bilaterally and the area surrounding opisthion is present as well. Preservation of these areas is generally good, though some breakage and reconstruction does intrude onto the tympanic portions of the temporal.

The squamotympanic (S-Q) fissure of this specimen is well preserved on both sides, and is clearly posterior to the apex of the mandibular fossae. Small postglenoid tubercles are preserved bilaterally as well. The tympanic plates of Sangiran 2 are thick and heavily built, and do not show evidence of housing a styloid process. A styloid pit does appear on the left side, and styloid grooves can be seen on both plates. An area approximately 1 cm in breadth remains of the posterior foramen magnum surrounding opisthion. The rim of this structure appears to have been wide and rounded with no constriction, and no rugosity or tubercle development can be discerned on the remaining bone. Neither an occipitomastoid or juxtamastoid crest was noted in the mastoid region. 53

Due to the extremely small size of the mastoid processes it is difficult to diagnose the

presence or absence of the tympanomastoid fissure. The tympanic plate does not contact

the mastoid, but given the size of the mastoid this is not surprising. Thus, this feature may

not be taxonomically informative on this specimen.

Sangiran 4

Sangiran 4 was discovered in 1939 by von Koenigswald’s fossil hunters, and consists of three parts: a fairly complete cranial rear that is separated into an upper and lower half, and a palate whose association with the skull has been questioned (Krantz, 1975, 1994).

This specimen retains much of the cranial base, though parts have been reconstructed.

Many of the casts of this specimen do not differentiate between the preserved portions and the reconstructed areas, so some morphologies must be approached with caution when dealing with casts. This specimen is also clearly distorted, and has been somewhat twisted, so that the lateral ends of the sphenoid bone are not in the same plane. It is unclear which side of the specimen is more or less distorted than the other, and efforts to reconstruct the specimen with CT scans and various 3-D methods have proven to be unhelpful for this problem (Hertler, personal communication). This distortion has resulted in slightly different orientations for the tympanic plate, and has produced two differing morphologies in the tympanomastoid fissure. This feature has been noted by workers such as Andrews (1984) and Stringer (1984) as potentially autapomorphic for Asian H.

erectus, but the degree of its manifestation in Sangiran 4 is rendered debatable due to the

warping of this specimen in this area.

The S-Q fissure of this specimen is preserved bilaterally, though the left side is

more complete and retains the postglenoid tubercle. The S-Q fissure is posterior to the

apex of the fossa. The tympanic plates of Sangiran 4 are large and heavily built and

project inferiorly to a degree not seen in other Indonesian specimens. As noted above,

they are oriented somewhat differently due to warping in this specimen, but it is clear that

at least a small tympanomastoid fissure is probably present. Styloid pits are present

bilaterally in conjunction with styloid grooves on the posterior edges of the tympanic

plates, but no sign of a styloid process is present. The foramen lacerum is not preserved,

but it is likely that it would have been constricted or absent based on the surviving

morphology of the petrous temporals and the reconstructed basioccipital. This condition

has been described as typical for Asian H. erectus (Jacob, 1976). The foramen ovale is

preserved on the right side and is a single foramen with no evidence of doubling. The

occipital condyles are large both relatively and absolutely compared to the area of the

foramen magnum (Durband, 2002b), and their anterior ends are slightly angled toward

the midline. The outline of the foramen magnum is irregular due to damage, but it does

not appear to be constricted at opisthion or broader at basion. There is no PCT

development or rugosity around the posterior rim of the foramen. A large juxtamastoid

tubercle is preserved on the right side (and reconstructed in mirror image on the left), and

there is no indication of an occipitomastoid process.

Sangiran 12

Sangiran 12 was discovered in 1965. The fossil consists of the rear portion of the braincase and what is present is well preserved with no visible distortion. There is very

little of interest on this specimen for the present project. The TMJs are missing on both

sides, and none of the cranial base has survived. However, some peripheral areas near the

mastoid region and the occipital can be examined.

From what remains of the occipital bone it is clear that no significant PCT

development was present on this specimen. The broken edge of the occipital lies either at

the posterior edge of the foramen magnum, so if any significant rugosity or PCT

development were present it would be apparent. There is a small juxtamastoid tubercle

present on the left side, but because of some weathering on the specimen it is difficult to

tell how large the structure was.

Sangiran 14

Sangiran 14 was discovered by local collectors in 1966 and was briefly described by

Jacob (1972a, b). It consists of a basioccipital with the occipital condyles preserved and a number of other small fragments, including portions of a mastoid process and petrous pyramids.

The morphology of the basioccipital fragment is very similar to that seen in

Ngandong 7 and 12. The basioccipital is broad and there is strong development of the alar tubercles. No pharyngeal tubercle is present, though there is a low crest. The dimensions of the occipital condyles are very similar to the surviving condyle of Ngandong 7 and the reconstructed dimensions for Ngandong 12 taken from Weidenreich (1951). However, the condyles on Sangiran 14 are not parallel and have axes that diverge posteriorly, which is unlike the condition seen in Ngandong. However, other details of the condylar articular facets are very similar to the morphology of Ngandong 7 and Sangiran 4 (Jacob, 1972a,

b). The associated mastoid process is small with numerous air cells exposed, and Jacob

(1972a) notes that it is quite similar to the mastoid of Sangiran 10.

Sangiran 17

Sangiran 17 was discovered in 1969 by a local farmer as he plowed his field with a heavy metal rod. It is the most complete cranium yet recovered from Sangiran, and it has been recently reconstructed (Aziz et al., 1996; Baba et al., 1998, 2000). It retains the majority of the cranial base, missing only the sphenoid bone and the medial ends of the petrous temporals.

The S-Q fissure is preserved bilaterally on this specimen, and clearly lies posterior to the apex of the temporal fossa. Small postglenoid tubercles are present, though they are weathered. The tympanic plates are thick and heavily built and do not show evidence of a styloid groove or pit, and no styloid processes are present. Neither the foramen lacerum nor the foramen ovale is preserved. The basioccipital is fairly broad with well-defined alar tubercles. There is a weak but palpable pharyngeal tubercle present. The occipital condyles are large and their anterior ends angle towards the midline. The foramen magnum is well preserved except for slight damage to the anterior rim and the right edge near the posterior end of the occipital condyle. The foramen is roughly diamond shaped and is slightly constricted near opisthion. While there is slight rugosity around the posterior rim of the foramen magnum no PCT development is present. No juxtamastoid crests are present on this specimen.

Sambungmacan 1

Sambungmacan 1 was found on February 26, 1973 along the south bank of the Solo

River near the town of Ngadirojo in the Sambungmacan district of Central Java (Jacob,

1984). Most of the cranial base is missing on this specimen, though both TMJs and much of the left side of the base is present.

The left TMJ is the best preserved of the two, though in both fossae it is clear that the S-Q fissure runs entirely in the apex of the joint. Very small postglenoid processes are present bilaterally as well. The tympanic plates are thick, as in other Javan specimens, and no styloid processes are present. Unfortunately the foramen ovale is not preserved. A small tympanomastoid fissure is present, and its morphology is quite similar to Ngandong in being parallel-sided. A small juxtamastoid crest is present on the left side of the specimen.

Sambungmacan 3

This specimen has a rather colorful history, changing hands through a number of “antique dealers” before finally being returned to Indonesia in 2000 (Laitman and Tattersall, 2001;

Delson et al., 2001). Discovered during dredging operations in the Solo river in 1977, the specimen next appeared in a Jakarta antique shop in 1998 (Boedhihartono, 1998, 2001).

At that time the owner expressed interest in selling the fossil to foreign scientists, and indeed this was attempted in the summer of 1999. F. Aziz (personal communication and author’s observations) received faxed offers to sell the skull for $400,000. Fortunately, an arrangement was worked out that allowed Indonesian scientists to obtain the specimen

from Mr. Henry Galiano and the fossil was given to Teuku Jacob in 2000 (Laitman and

Tattersall, 2001).

The specimen consists of a calvarium missing the face and much of the cranial base. The fossil does retain many important basilar features, however, including the right

TMJ, the area surrounding opisthion, and some morphologies of the tympanic and mastoid regions.

Both S-Q fissures are preserved on this specimen, and they course in the apex of the TMJ similar to the morphology seen in the Ngandong specimens (Delson et al., 2001;

Márquez et al., 2001; Mowbray et al., 2000; Durband, 2002a). From what is preserved the postglenoid tubercle is expressed as a low ridge similar to that seen in Ngandong 6 or

12 and differing from the larger tubercles seen in the Sangiran specimens. The tympanic plate is broken, but what is present is indicative of a thick, robust structure similar to other Indonesian erectines. A styloid pit is present, though it does not appear to have housed a styloid process. There is a small tympanomastoid fissure preserved on the right side, though it is difficult to diagnose the width of this feature because of damage to both the mastoid and the tympanic plate. Well developed occipitomastoid crests are present bilaterally, but no juxtamastoid crests are visible. Much of the right side of the posterior foramen magnum is preserved, and the foramen shows narrowing to an opisthionic recess characteristic of the Ngandong fossils (Delson et al., 2001; Márquez et al., 2001; Baba et al., 2003). A well-developed postcondyloid tuberosity is also present on the left side, and it appears that a small portion of the right one may also be present. Structures that many correspond to Weidenreich’s postjugular septum and fossa are also present on this specimen, particularly on the left side.

Sambungmacan 4

This specimen was announced by Aziz (2002) and described more fully by Baba and colleagues (2003). The cranium was found along the Solo River during the collection of sand for construction, and the find site is allegedly only 100 meters from where the

Sambungmacan 3 specimen was recovered in 1977 (Aziz, 2002; Baba et al., 2003, 2004).

The calvarium is beautifully preserved, and retains a complete cranial base with no obvious distortion (Baba et al., 2003). While it was not possible to view the original fossil, Dr. Hisao Baba graciously supplied a detailed photograph of the cranial base of

Sambungmacan 4 for this analysis.

Both TMJs are well preserved on this specimen. The course of the S-Q fissure is said to be intermediate between Sangiran 17 and Ngandong 12, in that it has migrated superiorly relative to the Sangiran condition yet is not fully in the roof of the fossa as in the Ngandong hominids (Baba et al., 2003, 2004). The basioccipital is broad, and foramen lacerum is constricted to a thin slit. The tympanic plate is thick and heavily constructed with no visible styloid process, though it does appear to have styloid pits bilaterally as in the Ngandong skulls. Sm 4 also possesses a bifurcated foramen ovale that is located at the base of a deep pit, a feature not found outside the Ngandong population

(Baba et al., 2003, 2004; Durband 2002b, 2004). The occipital condyles are eroded, but much of their bases are present at the anterior end of the foramen magnum. The surviving morphology is broadly similar to the broken bases present on Ngandong 12, which

Weidenreich (1951) reconstructed as being similar to the surviving condyle of Ngandong

7. It is clear that the condyles are small relative to the area of the foramen magnum, as in the Ngandong crania (Durband 2002b, 2004). The foramen magnum is large and narrows

to an opisthionic recess (Baba et al., 2003, 2004). Postcondyloid tuberosities are present, and the nuchal area is strongly excavated (Baba et al., 2003).

Ngandong

The Ngandong assemblage was recovered between September 1931 and November 1933 by ter Haar, Oppenoorth (1932a, b, c) and Von Koenigswald (1951, 1956). The discoveries were made near the village of Ngandong, about 6 miles north of Ngawi, along the banks of the Solo River. Twelve cranial specimens and two tibiae were collected during these expeditions, several of which are exquisitely preserved. A later expedition by Jacob in 1976 recovered two additional specimens, including some postcranial material (Jacob, 1981). These later finds have not been described and are not currently available for study (Indriati, personal communication).

It is important to note that there are a couple of different numbering strategies for these hominids, and this makes it easy to confuse specimens between references. To ease translation with different sources I have detailed these schemes in the glossary of fossil nomenclature included at the front of this dissertation. For this project I will utilize what I call the “Modern” numbering sequence, which is the convention used in the majority of the recent literature on these fossils. However, be aware that this differs from the two primary sources on these specimens, Weidenreich (1951) and Santa Luca (1980), who both use the “Original” numbering system. Jacob uses yet another numbering system, and

Schwartz and Tattersall (2003) utilized this less-common convention in their recent survey of the Ngandong material. The specimens relevant to this study are Ngandong 1,

6, 7, 10, 11, and 12 (or Solo I, V, VI, IX, X, and XI).

Another point germane to the discussion of these specimens is the controversy surrounding a structure that Weidenreich (1951: 265-266) dubbed the postjugular fossa.

Several of the specimens exhibit fossae posterior to the jugular foramen, and are separated from that structure by a thin wall of bone that Weidenreich (1951) referred to as the postjugular septum. The floor of this fossa is typically perforated by a small foramen, though in the case of Ngandong 6 the foramen is located in the lateral wall

(Weidenreich, 1951). There is a bit of confusion in the attribution of this feature, however, in that Weidenreich (1951) notes in the text that Ngandong 7 (Solo VI) does not possess postjugular fossae, yet in the diagram of the skull base for that specimen (Plate

38 B) this structure is indicated. My own observations found that a foramen perforates the right side of the skull base posterior to the jugular foramen of Ngandong 7, and indentations are present on the right. Likewise, I noted foramina in the postjugular area on Ngandong 1. Jacob (1969, 1981) contends that the postjugular fossa is nothing more than damage to the skull base. “[B]ecause of its irregular shape, varied location, the thinness and diploetic appearance of its floor, its continuity with adjacent broken areas, its inconstant presence, and the presence of the real jugular fossa, I am convinced that the

[postjugular fossa] is artificial” (Jacob, 1981: 97). For the purposes of describing the

Ngandong fossils the presence of the postjugular fossa will be noted, and the possible significance, if any, of this structure will be discussed later in this dissertation.

Ngandong 1 is a calvarium that lacks most of the cranial base. The specimen is also somewhat twisted, with the right side depressed inward and the left side turned outward (Weidenreich, 1951; author’s observations). This warping appears to be plastic deformation that occurred during fossilization of the specimen. It is interesting to note

that the earliest photographs of this specimen (Oppenoorth, 1932a) appear to show a more complete cranial base prior to the cleaning of the fossil. This might have been an illusion created by the matrix embedded in the skull, but it could also mean that the base was subsequently destroyed during cleaning and lost. Presently, Ngandong 1 preserves the right TMJ and a portion of the petrous temporal on that side as well as the majority of the rim of the foramen magnum.

The S-Q fissure of Ngandong 1 runs wholly in the apex of the mandibular fossa.

The postglenoid tubercle is broken, however from the preserved morphology it is clear that a small tubercle was present. The tympanic plate is very heavily constructed, and a styloid pit and groove are located on the posterior side. No styloid process is present or indicated. A wide parallel-sided tympanomastoid fissure is preserved on the right side, as well as a wide digastric fossa bounded by a low occipitomastoid crest. No juxtamastoid crest is noted. The foramen magnum is constricted posteriorly into a well-developed opisthionic recess. The posterior border of the foramen magnum is also bounded by a large postcondyloid tuberosity on the right side. Posterior to the jugular fossa perforations are present, and these correspond to the location of Weidenreich’s (1951) postjugular fossa and foramina.

Ngandong 6 is striking in its overall size and robusticity. At 221 mm this specimen is one of the longest hominid skulls ever recorded. The majority of the base is missing on this calvarium, however the left TMJ and mastoid portion are preserved. The

S-Q fissure of Ngandong 6 courses in the highest point of the mandibular fossa for its entire length. A small postglenoid tubercle is present as well. The tympanic plate is thick and heavily constructed, and houses a styloid pit with a faint styloid groove. No bony

styloid process is present or indicated by the surviving morphology. A broad, parallel- sided tympanomastoid fissure is present. The mastoid area on this specimen is heavily damaged, so it is unclear whether occipitomastoid or juxtamastoid crests were present.

Finally, Ngandong 6 has a postjugular septum and the lateral portion of a postjugular fossa on what is preserved of the left side. Immediately adjacent to this structure, however, is a large air cell in the mastoid region that was opened by the damage to that area. This air cell bears a striking resemblance to the alleged postjugular fossa (e.g.

Jacob, 1969, 1981).

Ngandong 7 is a beautifully preserved calvarium that retains the entire cranial base. This specimen, along with Ngandong 12, provides the most detail regarding basilar morphology in this series. Both TMJs are preserved, though the right side is more complete. The morphology of the TMJ of Ngandong 7 is quite interesting and differs somewhat from that seen in the rest of the Ngandong series. In this specimen the lateralmost edges of the TMJ curve inferiorly, forming somewhat of a basin if the TMJ is viewed from below. Mowbray and colleagues (2002; Mowbray, personal communication) feel that this morphology is very similar to that seen on specimens like

OH 9 and Sangiran 17, and negates the potentially autapomorphic nature of the TMJ morphology in the Ngandong series. However, the position of the zygomatic root in

Ngandong 7 appears to be depressed inferiorly and suggests that some plastic deformation has taken place in this specimen. A similar depression of the zygomatic root can be seen on Ngandong 10 (Solo IX), and a similar alteration of the TMJ morphology in that specimen is apparent. More detailed examination of this morphology using 3-D digital images will be presented later in this dissertation.

Other aspects of Ngandong 7’s TMJ morphology are similar to the rest of the

Ngandong series. It possesses reduced postglenoid processes bilaterally, and the posterior wall of the fossa is formed solely by the tympanic plate. The tympanic plates themselves are thick and blunted, and styloid grooves are present bilaterally. No bony styloid processes are indicated. The tympanomastoid fissure is broad and parallel-sided. Well- developed occipitomastoid and juxtamastoid crests are present bilaterally. The foramen lacerum is present but quite constricted on Ngandong 7.

One aspect of the morphology of Ngandong 7 is particularly worthy of note. On this specimen the foramen ovale is doubled on the left side, as noted by both Weidenreich

(1951) and Jacob (1969). The entire f. ovale complex is located at the base of a deep pit, with an accessory foramen medial to the main foramen, and the two foramina are separated by a broad septum (Weidenreich, 1951). Jacob (1969) has noted that the septum has been broken on the right side of Ngandong 7, and the author’s observations support this assertion. Thus, both foramina ovale in Ngandong 7 exhibit this peculiar morphology, and this condition has not been noted outside the Ngandong and

Sambungmacan hominids (Weidenreich, 1951; Jacob, 1969, 1975, 1981; Durband,

2002b, 2004; Baba et al., 2003). It should be noted here that Weidenreich (1951: 281) misattributed this feature to appearing in “three of the four skulls” in the Ngandong series when it actually occurs it in three of four foramina that are preserved. And, as noted earlier, the fourth foramen (the right side of Ngandong 7) only fails to show this morphology because of damage to the ovale septum.

The remainder of the cranial base of Ngandong 7 shows a number of interesting morphologies. The entirety of the foramen magnum is preserved, as is one occipital

condyle, the left (though the right condyle has been reconstructed on the original). The occipital condyles of Ngandong 7 are remarkably small and are oriented parallel to the midline. Durband (2002b, 2004) has pointed out that the occipital condyles in the

Ngandong series are both absolutely small and also relatively small to the area of the foramen magnum when compared to other fossil and modern humans. The foramen magnum is distinctly teardrop shaped, with a broad and rounded anterior end and a markedly constricted posterior portion. This constriction at opisthion, dubbed the opisthionic recess, has been noted as a potential autapomorphy for the Ngandong and

Sambungmacan hominids (Delson et al., 2001; Márquez et al., 2001; Durband, 2002b,

2004; Baba et al., 2003). Well-developed postcondyloid tuberosities are present bilaterally on the posterior rim of the foramen magnum. The basioccipital is quite broad with moderate development of the alar tubercles. No pharyngeal tubercle was noted, though there is a slight ridge that is probably an analogous structure.

The cranial base of Ngandong 10 has been nearly completely destroyed, but the right TMJ and the rim of the foramen magnum around opisthion has been preserved. The specimen has an interesting history, in that it was presented to the Emperor of Japan as a birthday gift during World War II. It was recovered through the intervention of the

American forces stationed in Japan after the end of the war and returned to the collection in December 1946 (Koenigswald, 1951). In the preserved right TMJ of Ngandong 10 the

S-Q fissure runs solely in the roof of the structure. The morphology of the TMJ is somewhat altered by clear plastic deformation of the zygomatic root, and this has pushed the lateral edge of the fossa several mm inferiorly from its original position. Similar deformation has also been noted for Ngandong 7. A small postglenoid tubercle is

preserved on the right side. The right mastoid region is preserved, and an occipitomastoid crest is present. A small juxtamastoid crest is visible. The very edge of the foramen magnum is present at opisthion. From what is preserved it does not appear as though the posterior foramen was as constricted as Ngandong 1 or 7, though it might have resembled

Ngandong 12 in the degree of narrowing at opisthion. A postcondyloid tuberosity is present on the left side.

The base of Ngandong 11 is similar to the other specimens in the series. Both

TMJs are preserved, and the S-Q fissure is located in the roof of the fossa bilaterally.

Small postglenoid tubercles are also located in both fossae. The tympanic plates are extremely thick and heavily constructed. Styloid pits are present bilaterally, with no evidence of a bony styloid process present. The surviving rim of the foramen magnum is more rounded at opisthion than Ngandong 1 or 7, though there still appears to be some narrowing of the posterior foramen. A large postcondyloid tuberosity survives on the left side, and this structure is somewhat exceptional in its degree of inferior projection. Its location on the skull and orientation closely match the tuberosities present on the other

Ngandong and Sambungmacan crania. The posterior rim of a postjugular fossa may be present on the left side.

Ngandong 12, like Ngandong 7, retains a nearly complete cranial base. While it lacks the pterygoid plates and orbital roofs present on skull 7, Ngandong 12 preserves a wealth of information relevant to this project. Both TMJs are present and well-preserved, though the right side is more complete and retains much of the zygomatic root. The S-Q fissure courses in the very roof of the fossa bilaterally, and the posterior wall of the TMJ is formed solely by the tympanic plate. Small postglenoid tubercles/ridges are present.

As in the other crania in the series the tympanic plates are heavily constructed. No bony styloid is indicated, though styloid grooves and pits appear bilaterally. The tympanomastoid fissure is broad and parallel-sided. Well-developed occipitomastoid and juxtamastoid crests are present bilaterally. Foramen lacerum is present but constricted.

The foramina ovale both exhibit the morphology noted for Ngandong 7, with an accessory foramen located medially to the main foramen. These foramina are separated from one another by a broad septum, and the pair are further located at the base of a pit that is several mm deep. The occipital condyles are missing, but enough of their bases remain to allow Weidenreich (1951) to reconstruct them. Based on this work the condyles were found to be parallel to the midline, similar to Ngandong 7, and also relatively small in size (Weidenreich, 1951). The foramen magnum appears to be long and ovate in shape, but this is due to damage at the anterior end (Santa Luca, 1980). When this damage is corrected for the shape of the foramen magnum is similar to that of Ngandong 7, though skull 12 is not constricted at opisthion to the degree seen in specimen 7. Very well- developed postcondyloid tuberosities are present bilaterally just anterior to opisthion along the rim of the foramen magnum. The basioccipital is quite broad and large alar tubercles are present on the postero-lateral corners of this structure just anterior to the broken bases for the occipital condyles. No pharyngeal tubercle is present, though there is a foramen perforating the basioccipital where the tubercle would typically be found.

Finally, postjugular fossae are present bilaterally, and both are perforated by a foramen in their roof.

Ngawi 1

Ngawi is a virtually complete calvarium that was found in August 1987 by a student swimming in the Solo River (Sartono, 1990). As the skull was found when the individual kicked it on the riverbed, no stratigraphic information or even a reliable provenance could be discerned (Aziz, personal communication). Ngawi has been discussed in a number of publications (e.g. Sartono, 1990; Widianto and Grimaud-Hervé, 1993; Sartono et al.,

1996; Grimaud-Hervé et al., 1998), and has been more recently described by Widianto and Zeitoun (2003). The specimen lacks only the face and has a nearly complete cranial base, though much of the basicranium is still obscured by matrix.

Both TMJs of the specimen are well preserved, but they are still filled with matrix so details are not discernable. However, the general form of the TMJ is very similar to the

Ngandong and Sambungmacan fossils, making it probable that the S-Q fissure is located in the apex of the fossa. Widianto and Zeitoun (2003) describe the tympanic plate as forming the entire posterior wall of the mandibular fossae, which is consistent with the

Ngandong morphology. The tympanic plate is heavily built with no obvious indication that it housed a styloid process (though due to the adhering matrix the base of the tympanic plates are not visible). The foramen magnum is relatively large, however it is probably damaged at the anterior end. The foramen does narrow to an opisthionic recess posteriorly. Small postcondyloid tuberosities are also present bilaterally on this specimen, and the nuchal area is reminiscent of the condition seen on Sambungmacan 3.

Wajak 1

Wajak 1 (also spelled as Wadjak) is a virtually complete skull that was discovered by

B.D. van Rietschoten in October 1888 (Storm, 1995). This specimen has undergone several restorations, including work by Dubois, Jacob, and Stringer and Parsons (Storm,

1995). My observations were done on the most recent reconstruction, that of Stringer and

Parsons, which was done in the 1970’s (de Vos, personal communication). It should be noted that this reconstruction differs considerably from most of the casts of this skull that have been widely circulated. The most obvious difference between this most recent restoration and earlier attempts is the marked reduction in the occipital region. Prior reconstructions had a large bulge in the occipital that has been lessened considerably in the current version. It is unknown how much this might have affected the basal morphologies to be discussed. Faunal material recovered from the cave site has been dated to between 12,140 and 12,930 calibrated 14C years (Storm, 1995; Shutler et al., in

press). This recent date would remove Wajak from the ancestry of modern Australians

(Shutler et al., in press), but this specimen is still important as an example of late

Pleistocene modern morphology in Australasia.

Much of the morphology of interest in this study is obscured either by matrix,

plaster, or damage on this individual. However, there are many details worthy of note for

comparison with the Indonesian H. erectus sample. The squamotympanic fissure is not

visible on the right side due to the adherence of the mandibular condyle to the TMJ. On

the left side the S-Q fissure is not in the apex of the fossa, and it lies posterior to both the

roof of the joint surface as well as a small postglenoid tubercle. This tubercle is

weathered, so it is difficult to determine its original size. Wajak 1 has very large occipital

condyles both absolutely and relative to the area of the foramen magnum. Unfortunately the condyles were damaged so reliable measurements could not be taken from them. The foramen magnum is damaged at opisthion, so the presence or absence of an opisthionic recess cannot be reliably determined, but the anterior edge of the foramen does not have the more flattened “teardrop” shape distinctive of the Ngandong fossils. There are no

PCT nor is there any rugosity surrounding the posterior rim of the foramen magnum. The basioccipital is relatively narrow with no development of the alar tubercles visible. The pharyngeal tubercle is weathered but palpable.

Chapter 4: Descriptions of Australian cranial bases

Because descriptions of the cranial bases present in the Pleistocene Australian record are

exceedingly rare in the literature, it would be useful to provide a synopsis of the form and

condition of each of the specimens under consideration. Photographs of the cranial bases

and relevant structures on these specimens are included in Appendix F.

Keilor

Keilor is a well-preserved, virtually complete skull discovered in 1940 near the town of

Keilor, Victoria. The skull has been dated to approximately 15,000 BP by faunal correlation (Bowler, 1976) and a collagen date on a subsequently discovered femur provided an age of 12,000 ± 100 BP (Brown, 1987). The skull is large and robust, but lacks the sloping frontal profile and supraorbital development common in the similarly dated Kow Swamp series (Brown, 1992). This specimen retains a fairly complete cranial base, missing only the right anterior border of the foramen magnum, the right temporal, and the posterior half of the right sphenoid.

The TMJs of Keilor are large and very wide at their anterior end with little development of the articular tubercle. The S-Q fissure is located posterior to the apex of the joint surface and posterior to a well-developed postglenoid tubercle. Basioccipital is fairly narrow anteriorly and the right side is damaged posteriorly. The alar tubercles are not developed, though there is a prominent pharyngeal tubercle on this specimen. The tympanic plates are thin and delicate, and while styloid processes were not noted they

appear to have been present. A tympanomastoid fissure is not present on Keilor, nor is the carotid foramen obscured at all by the tympanic plate. A large occipitomastoid crest is present, but no juxtamastoid crests are discernable. Foramen lacerum is present bilaterally. The anteromedial edges of both foramina ovale are present, with the right being better represented. From the portions preserved it is clear that these foramina are not doubled. The surviving right occipital condyle is large, and may have been slightly larger as there is some evidence of damage at the anterior end. The condyle is oriented predominantly parallel to the midline, though the anterior edge does flare medially. The condyle is also small considering the overall size of the cranium. The foramen magnum is ovate/round in shape and there is no evidence of narrowing or constriction at opisthion.

There is some rugosity around the posterior rim of the foramen, but there is no development of discrete PCT development.

Talgai

The Talgai cranium was discovered in 1884 in Dalyrymple Creek near Talgai Station southwest of Brisbane. The specimen is heavily crushed and distorted and has suffered through various reconstruction attempts that have done little but add to the destruction of the fossil (Macintosh, 1967a, b). The skull is clearly robust and has a retreating forehead, low profile, and prognathic face, but it is unknown how much the obvious deformation that has occurred might have contributed to these features. Very little of the cranial base of this specimen is clear. The only point of morphology relevant to this project that this researcher was able to clearly discern was the lack of a tympanomastoid fissure.

Kow Swamp

The site of Kow Swamp was first discovered in 1962 when a burial was disturbed during the deepening of a channel (Thorne, 1975; West, 1977). Alan Thorne examined those bone fragments in the National Museum of Victoria in 1967, and in February, 1968 he and Alan West rediscovered and excavated the site and recovered a skeleton later designated Kow Swamp 1 (West, 1977). Later in 1968 fragments of the Kow Swamp 2 burial were discovered after work to lay telephone cables (Thorne, 1975). Finally, an individual named Gordon Spark investigated some disturbed soil thrown up during the digging of an irrigation channel at Kow Swamp and discovered a number of bone fragments (Thorne, 1975). Subsequent work at the area uncovered a number of burials, including Kow Swamp 5, and these later excavations were designated as the Spark site

(Thorne, 1975). Excavations at Kow Swamp between 1969-1973 recovered a total of 22 specimens complete enough to be assigned numbers, plus hundreds of additional bone fragments and isolated teeth that could not be assigned to any of the numbered individuals (Thorne, 1975). Radiocarbon dates from several graves have been obtained, including dates of 13,000 ± 280 years on Kow Swamp 5 (shell), 10,930 ± 125 for Kow

Swamp 14 (shell), and 9,590 ± 130 for Kow Swamp 9 (charcoal) (Thorne, 1975). The site has recently been redated using OSL on sediments, and these have produced ages of between 22-19 kyrs (Stone and Cupper, 2003). However, these dates have been met with a great deal of skepticism because the sediments sampled lack context with the burials themselves (Thorne, personal communication; Bulbeck, personal communication). Thus, the radiocarbon dates obtained by Thorne (1975) remain the best estimates of antiquity for the Kow Swamp remains.

The majority of the Kow Swamp skeletons are quite fragmentary and the preservation in the basicranium is particularly poor. Thus, the only specimens examined by the author for this project are Kow Swamp 5 and 8. Additional observations on these and other specimens have been provided by Thorne (1975).

Kow Swamp 5 is a fairly complete cranium, but it lacks much of the left side of the cranial vault and the central cranial base. However, both TMJs are preserved, as are the petrous, tympanic, and mastoid region on the right side and the area surrounding opisthion, so a great deal of morphology remains that is of relevance to this study.

In both TMJs of Kow Swamp 5 there are large postglenoid tubercles, and the S-Q fissure runs posteriorly to the tubercles and quite posterior to the apex of the fossa. The

TMJs are deep and the articular processes are prominent. The tympanic plate is thin and forms a vaginal process, and, while broken, does not appear to have overhung the carotid foramen. A tympanomastoid fissure is not present on Kow Swamp 5. No juxtamastoid crest is present on the right, but a small portion of one is visible on the left at the edge of the broken mastoid. No occipitomastoid crest is present on the right, and the left mastoid region is too damaged to allow a diagnosis. Much of the left rim of the foramen magnum is present, though it has sustained a number of chips to its ectocranial surface. This rim is somewhat raised and rugose, but no PCT development is present.

Kow Swamp 8 preserves the right TMJ and a portion of the right tympanic. The rest of the cranial base on this specimen is missing.

The right TMJ of Kow Swamp 8 contains a large postglenoid tubercle, and the S-

Q fissure is located posterior to this structure and posterior to the apex of the fossa. The

surviving tympanic plate is thin and forms a vaginal process, which was also noted by

Thorne (1975).

Lake Mungo

The Lake Mungo I cremation was discovered by J.M. Bowler in July, 1968 (Bowler et al., 1970) and later excavated by H. Allen, R.M. Jones, and D.J. Mulvaney in March,

1969 (Thorne, 1975). The remains were extremely fragmentary and cemented together in a calcrete block that had been exposed after eroding out on the deflation surface (Bowler et al., 1970). The Lake Mungo III skeleton was discovered on February 26, 1974 eroding out of a lunette approximately 600 meters from the Mungo I site (Thorne et al., 1999). As noted earlier, the ages of the Lake Mungo burials have been the subject of considerable recent debate (e.g. Thorne et al., 1999; Bowler and Magee, 2000; Brown and Gillespie,

2000; Gillespie and Roberts, 2000; Grün et al., 2000; Bowler et al., 2003). Thorne and colleagues (1999) used a number of dating methods to obtain ages of 62 ± 6 kyr for the

Mungo III burial. These dates have been roundly criticized for seemingly ignoring the pre-existing chronology established for the Mungo sites (Bowler and Magee, 2000;

Brown and Gillespie, 2000; Gillespie and Roberts, 2000). The most recent dates for these specimens, provided by Bowler and colleagues (2003), place both Mungo burials at 40 ±

2 kyr.

Neither of the crania discovered at Lake Mungo are complete. Lake Mungo I does retain much of the cranial base, but, as mentioned earlier, access to the original specimen was not possible and the specimen was too fragile to allow any casts to be made (Thorne, personal communication). However, the detailed description provided by Thorne (1975)

provides much of the information that will be relevant to this study. The Lake Mungo III cranium lacks the entire base of the skull, but a small area surrounding opisthion survives that is germane to this work.

The Lake Mungo I cranial vault retains a portion of midline occipital bone that approaches opisthion. The present author did not get the impression that the posterior rim of the foramen magnum was preserved based on the morphology displayed by the cast, but Thorne (1975) does seem to contend that opisthion is present. The surviving area of occipital near (or at) opisthion exhibits a markedly projecting crest that is bilaterally flanked by excavations. The nuchal area is otherwise fairly smooth and without pronounced muscle markings. Thorne (1975) provides further descriptions of the Mungo

I cranial base:

“The basilar portion of the occipital bone is intact. It is triangular, narrowing

sharply at its anterior end. Although the atlas and axis remain in situ, enough is

visible at present to indicate that the occipital condyles are placed well forward on

the foramen magnum, their long axes being widely divergent posteriorly.” (pg.

204)

“The right glenoid fossa is broad and shallow with an almost vertical anterior

wall. The squamotympanic fissure lies at an angle of 35 degrees to the sagittal

axis of the base. It does not lie on the floor of the fossa, an important point in

relation to the condition in the Solo crania.” (pgs. 205-206)

“Neither styloid process is preserved. However, the greater parts of the styloid

sheaths are present, strongly grooved medially. Their form suggests that though

not necessarily long, well-developed styloid processes were present.” (pg. 206)

“In Lake Mungo [I] both foramina [ovale] are oval, the right foramen having a

long axis of 7 mm. In view of the Solo condition … it is important to record that

the foramina are not divided and do not lie in pits.” (pgs. 206-207)

The Lake Mungo III cranial vault clearly retains opisthion and a portion of the right posterior border of the foramen magnum. As with the Mungo I skull, there is a marked crest that begins at opisthion and runs along the midline of the nuchal plane for approximately 28-30 mm. Bilaterally adjacent to this crest are shallow excavations very similar in morphology to Mungo I. The surviving border of the foramen magnum is uniformly thickened, and this raised edge is connected to the cresting at opisthion. This thickening is commonly encountered in modern populations, however the connection of any such thickening to an external occipital crest at opisthion is unusual.

Nacurrie 1

The Nacurrie 1 cranium (also sometimes referred to as “Murrabit”) was excavated by

G.M. Black in August, 1949 from a sandhill north of Swan Hill and near the Nacurrie railway siding (Brown, website). It was dated to 11,440 ± 160 years BP through an AMS date on collagen (Brown, 1994b). The skull appears to be artificially deformed, with a low, sloping frontal bone and pronounced bossing on the parietal bones. Very little of the

cranial base is preserved on this specimen, though the TMJ, tympanic, petrous, and mastoid areas are preserved on the left side.

The Nacurrie 1 TMJ has a large postglenoid process, and the S-Q fissure is located posterior to this process and well posterior to the apex of the fossa. The tympanic plate is thin and forms a vaginal process. There is no expression of a tympanomastoid fissure. On the portion of the mastoid area that survives a small juxtamastoid crest and moderate occipitomastoid crest can be discerned.

Nacurrie 2

The Nacurrie 2 cranium was found by a Mrs. O’Donell, and while it has not been directly dated Peter Brown believes it to be of similar age to Nacurrie 1 (Westaway, personal communication). At the time of this writing the specimen has not yet been cast, and attempts to view the original were unsuccessful. However, some morphological details for Nacurrie 2 could be discerned from a detailed photograph of the cranial base

(Westaway, 2002b). The specimen is quite complete, though there is arthritic remodeling of the left TMJ and some abrasion to both occipital condyles and the tip of the right mastoid. In addition, some structures are obscured by adherent matrix. Despite these limitations, many points of interest on Nacurrie 2 can be seen.

Although the morphology of the left TMJ has been largely destroyed by arthritis, the right TMJ appears to be undistorted. It sports a large postglenoid tubercle, and the S-

Q fissure, though filled with matrix, can be seen posterior to the aforementioned tubercle.

The tympanic plates are both damaged, though a vaginal process can be seen on the more complete right side. A tympanomastoid fissure is not present. Small juxtamastoid crests

are present bilaterally, as are narrow occipitomastoid crests. The right foramen ovale is present and free of matrix, and it is undivided. Well-developed foramina lacerum are present bilaterally. The occipital condyles, though damaged, are large and angle toward the midline. The foramen magnum is complete, and is quite rounded with no narrowing at the posterior end. The posterior rim of the foramen magnum appears rugose, but does not have any development of discrete tubercles.

Cossack

The Cossack specimen was first discovered in 1972 eroding out of a sand dune near

Cossack, Western Australia, and subsequent recovery of additional skeletal material was performed at the site in 1974 (Freedman and Lofgren, 1979). Upon reconstruction, most parts of the cranium were represented, though the skull is mediolaterally compressed.

Absolute dating on the specimen was not possible, however Freedman and Lofgren

(1979) note that the geological history of the site limits the age to no greater than 6,500 years. Unfortunately for the purposes of this study, much of the cranial base is missing.

The cast available for study included only the parietals and some of the right temporal bone including the TMJ. From the published photographs (Freedman and Lofgren, 1979), it does not appear that any relevant portions of the occipital bone were preserved, and thus their absence from the cast does not materially affect this project.

The right TMJ of the Cossack specimen is complete and appears undistorted. It has a large postglenoid process, and the S-Q fissure is posterior to this process and well posterior to the apex of the fossa.

Mossgiel

The Mossgiel cranium was discovered as it eroded out of the ground at the Tumbridge

Station homestead, 24 km west of Mossgiel (Macintosh, 1965; Freedman, 1985). Dating of bone carbonate from the specimen gave a minimum age of 6010 ± 125 years BP

(Macintosh, 1967b). Much of the skeleton is present. While the cranial vault is relatively complete, the base is represented only by portions of the occipital, the left sphenoid, and the left TMJ.

The surviving TMJ appears to have been heavily modified by arthritis. However, a large postglenoid tubercle is evident and the overall shape of the fossa is similar to those seen in other Pleistocene Australians already mentioned. The left foramen ovale is roughly 60% complete, and it is undivided with no trace of a septum or an ovale pit.

Enough of the occipital squama is preserved near the area surrounding the left mastoid to suggest that there was no development of a juxtamastoid crest. Approximately 25 mm of the foramen magnum border is present, including much of the posterior left side and opisthion. There is some shallow excavation posterior to opisthion similar to that seen in the Mungo crania, though lacking the pronounced cresting seen in those skulls. The rim of the foramen magnum is somewhat rugose, but there is no formation of PCT. There is also no narrowing of the foramen near opisthion.

Cohuna

Exceptional reviews of the early history and controversies surrounding the Cohuna specimen are provided by Macintosh (1952, 1953). The specimen was discovered in

November, 1925 when it was struck by a plow during the excavation of an irrigation

canal (Macintosh, 1952). Though remains attributable to at least 11 other individuals were found at the site, chemical analysis of Cohuna found that that cranium possessed a different pattern of mineralization than the other remains (Macintosh, 1953). Thus,

Macintosh (1953) felt that Cohuna likely derived from different sediments and had probably been redeposited in that area, perhaps after an initial interment in the region of

Mt. Hope. Macumber and Thorne (1975) revisited the site and rejected the argument for secondary redeposition of the specimen. Instead, the cranium probably came from a local burial that was disturbed during digging, and the associated postcranial skeleton was simply not discovered (Macumber and Thorne, 1975). Based on the sedimentation at the site, Thorne and Macumber (1975) felt that the specimen was probably between 9-13 kyr.

Brown (1987, 1989) feels that a terminal Pleistocene date is appropriate for this specimen based on its morphological similarities to the Kow Swamp specimens. The skull is quite heavily built with well-developed cranial superstructures. Its cranial base is relatively complete, though it is missing much of the occipital bone.

TMJs are present bilaterally, though only the left one is visible as the right appears to be filled with matrix. The postglenoid tubercle is large, and the S-Q fissure is clearly posterior to the apex of the fossa. Both tympanic plates are present, and they are thin with vaginal processes. The mastoid processes are broken, but the surviving bases are pneumatized and exceptionally large. There is no tympanomastoid fissure. Neither a juxtamastoid or occipitomastoid crest is developed. F. lacerum is present on the more complete right side, and is manifested as a constricted slit. F. ovale is present on the left and is not divided nor does it reside in a pit. Roughly 40% of the basioccipital is preserved on the right, including the right occipital condyle. This condyle is slightly

damaged at the anterior end, but is clearly relatively large and angled toward the midline at its anterior end. Too little of the foramen magnum survives to allow any conjecture about its overall shape.

Lake Nitchie

In December of 1969 J.M. Bowler and J.M. Urquhart discovered evidence of human burials eroding from a dune on the northern edge of Lake Nitchie (Bowler, 1970).

Subsequent excavations by Macintosh in January, 1970, revealed one burial of a fairly complete skeleton wearing a necklace of approximately 178 canine teeth of Sarcophilus harrisi (the Tasmanian Devil) (Macintosh, 1971). Initial work on the site suggested a minimum age of at least 3,000 years (the date that the Tasmanian Devil disappeared on the mainland), though an age in the range of 14-17,000 years was suggested by Bowler

(1970). Radiocarbon dating on bone collagen from the right femur produced an age of

6,820 ± 200 years BP (Macintosh, 1971).

At 187.5 cm, the Lake Nitchie individual is exceptionally tall for an Australian aboriginal male (Macintosh, 1971; Macintosh and Larnach, 1976). The skeleton is very nearly complete and is well preserved (Macintosh, 1971). The skull is extraordinarily large and quite massive. Macintosh (1971) notes that its cranial contours are quite modern but it retains a number of more primitive non-metric characteristics, such as a heavy and undivided supraorbital torus, a wide mound-shaped occipital torus, and angular tori. The cranial base is relatively complete, with damage only to the body of the sphenoid and the area surrounding the foramen magnum. Both basion and opisthion are

present, as are both occipital condyles, so the outline of the foramen magnum can be reasonable estimated.

Both TMJs are complete and undamaged, and sport large postglenoid tubercles.

The S-Q fissure lies posterior to those tubercles and does not coincide with the apex of the fossa. The tympanic plates are thin and form vaginal processes. The right styloid process is present, and quite large. Tympanomastoid fissures are not present. Both mastoid processes are well-preserved and quite large. On the left side there is no occipitomastoid crest but a well-developed juxtamastoid crest. On the right this is reversed, and there is only a large occipitomastoid crest present. Foramen lacerum is present bilaterally. The occipital condyles are large both relatively and absolutely, and are angled toward the midline at their anterior end. The posterolateral borders of the foramen magnum are reconstructed, but based on the photographs published by Macintosh (1971) this restoration seems reasonable. The foramen is fairly round, particularly posteriorly, though it is somewhat more constricted at its anterior end. There is little rugosity of the surviving border at opisthion. The basioccipital is broad and rugose. Both the pharyngeal tubercle and the alar tubercles are well-developed.

It is interesting to note that the Lake Nitchie skull has what would appear to be a postjugular fossa present on the right side. There is a bony septum separating a round fossa from the more anteriorly located jugular foramen. From the cast the author was unable to tell if the floor of this structure is perforated, however in large original photographs of the specimen provided by Ann Macintosh it did not appear that perforations or foramina are present. Clearly, this feature is the result of damage to the skull base, echoing the explanation for the postjugular fossae seen in the Ngandong

fossils by Jacob (1969, 1981). Examination of the undamaged left side reveals a large, extensively pneumatized jugular process of the occipital bone. Obviously, damage to this structure on the left side simply opened a large air pocket in the bone. The broken air cell resembles other foramina and various “nooks and crannies” present on the skull base, and could be mistaken for a genuine feature if it were present bilaterally (as it is on Ngandong

12). Even though this “fossa” is not a legitimate point of anatomy, it is quite interesting in that it illustrates how similar “features” that appear in the Ngandong fossils are the result of taphonomic processes.

Chapter 5: Results of non-metric examinations

Proponents of the Multiregional hypothesis of modern human origins contend that a suite

of non-metric features provides evidence for the maintenance of genetic continuity

between the Pleistocene inhabitants of Java and later modern humans in Australia (e.g.

Thorne and Wolpoff, 1981; Wolpoff, 1989, 1992, 1999; Wolpoff et al., 1984, 1994,

2001; Hawks et al., 2000). Features such as “the marked ridge paralleling the

zygomaxillary suture, the eversion of the lower border of the zygomatic, the rounding of

the inferolateral orbital border, the lack of a distinct line dividing the nasal floor from the

subnasal face of the maxilla, the curvature of the posterior alveolar plane of the maxilla

that corresponds to the mandibular ‘curve of Spee’, the inferior border of the supraorbital

torus, and the marked and dramatic nature of the supraorbital or superciliary expression”

have been cited as showing regional continuity in Australasia (Wolpoff, 1992: 51).

However, many of the features noted above are located on the face, which make

them somewhat problematic for the Indonesian and Australian record. While many of the

Pleistocene Australians retain faces, only one specimen of Indonesian Homo erectus preserves a relatively complete face, Sangiran 17. Other studies have linked the WLH 50 cranial vault with the Ngandong fossils (Hawks et al, 2000; Wolpoff et al., 2001). In both of these cases, however, only one specimen is cited as providing the link between the

Indonesian fossils and later Australians. This “holy grail” approach to regional continuity in Australasia has drawn criticism from replacement advocates (e.g. Brown, 1987, 1995).

A study involving the cranial base provides a solution to this dilemma, as a

number of individuals in each of the samples purportedly involved in the transition from

Indonesian H. erectus to modern Australians retain at least a portion of the cranial base.

In addition, as noted earlier, cranial base characters have been shown to exhibit similar levels of heritability as neurocranial or facial characters (Cheverud, 1995), and have proven useful in phylogenetic analyses (Strait, 1998). Thus, a more comprehensive survey made possible by the greater preservation in the cranial base, in conjunction with the applicability of this data to the problem, makes the current study well suited for the task at hand.

Seventeen non-metric characters were chosen for this project. Many were culled from a series of features recognized by Weidenreich (1951) and later Larnach and

Macintosh (1974) as being relevant to the question of continuity in Australasia. Some of these features have been proposed to be unique to the Ngandong and Sambungmacan fossils (e.g. Weidenreich, 1951; Larnach and Macintosh, 1974; Delson et al., 2001;

Durband, 2002b). As a group, these characters provide coverage for each of the major bones and features of the cranial base. Each of the characters examined will be discussed in turn, with a summary table and discussion offered at the end of the section. Complete non-metric data for each modern specimen examined is provided in Appendix A, while the data for the fossils are located in Appendix B. Statistical significance for these traits was assessed with Manova tests, and the p value for significance is .05. The results of these tests can be found in Appendix D.

Individual character comparisons

Pharyngeal tubercle

This trait could be scored on two specimens from Sangiran, 14 and 17. Both specimens exhibit midline structures on the basioccipital. Sangiran 14 has a low crest, while on

Sangiran 17 a small but palpable tubercle is present. Ngandong 7 and 12 also preserve the basioccipital. Ngandong 7 lacks a tubercle, but does possess a low ridge on the midline.

Ngandong 12 possesses a small foramen in the midline at about the typical location for a pharyngeal tubercle, and a low crest originates here and courses posteriorly.

Sambungmacan 4 appears to possess a well-developed pharyngeal tubercle.

Three Late Pleistocene modern humans preserve the basioccipital, Keilor, Lake

Nitchie, and Wajak 1. All three specimens exhibit well-developed pharyngeal tubercles.

The basioccipital of Wajak 1 is weathered, but the tubercle is readily palpable.

The recent samples of modern humans are fairly uniform in their expression of this trait. Its expression in Australia is statistically significantly different from only 3/10 populations: Austria, Czech, and Greece. These latter three groups have relatively strong expression of the pharyngeal tubercle, and are significantly separated from a number of other modern groups for this trait.

Tympanic plate contact with mastoid

This trait could be scored on Sangiran 2, 4, and 17. It is somewhat problematic to score on Sangiran 2, as the mastoid processes are only a few mm in height. Thus, the tympanic plate projects well below the level of the mastoids and as a result it is essentially impossible for contact to be achieved. Sangiran 4, on the other hand, is problematic due

to the obvious warping that has affected at least one of the tympanic plates. Due to the nature of this deformation it is unclear which side is the most (or the least) affected. It is clear that Sangiran 4 exhibits a tympanomastoid fissure bilaterally, so this confusion is limited to the breadth of this structure. Sangiran 17 does not have tympanomastoid fissures. Each of the Ngandong and Sambungmacan crania possess tympanomastoid fissures that are quite broad and roughly parallel-sided.

Nine Late Pleistocene modern humans retained this area of the skull: Cohuna,

Cossack, Keilor, Kow Swamp 5, Kow Swamp 8, Lake Nitchie, Nacurrie 1, Nacurrie 2, and Wajak 1. Each of these specimens shows full contact between the tympanic plate and the mastoid process.

The recent samples of modern humans are quite similar to one another in this trait. Australia was statistically significantly different from only 1/10 comparative samples, the Czech group. Overall very few modern humans were scored as having a fissure present, and only Australia, Czech, and W. Africa contained any crania scored as

“no contact.”

Presence of F. ovale pit/F. ovale accessorium

While scored separately, these traits will be combined for this discussion. Unfortunately, among the Sangiran hominids only skull 4 could be scored for this trait. It possesses single foramina bilaterally. In the Ngandong and Sambungmacan sample, each cranium that preserves this foramen exhibit both the presence of the ovale pit as well as the accessory foramen ovale located medially to the main foramen. These occur on both

Ngandong 7 and 12, and Sambungmacan 4.

In the Late Pleistocene modern humans, Cohuna, Keilor, Lake Mungo I,

Mossgiel, and Nacurrie 2 all possess at least one F. ovale that is complete enough to score. None of these specimens have either an ovale pit or an accessory foramen ovale.

Likewise, none of the recent samples of modern humans had any examples of crania with ovale pits or accessory foramina ovale.

Styloid process/Vaginal process

Once again, though scored separately these characteristics are best treated as a unit for the purposes of this discussion. Each of the Sangiran hominids that preserve this area of the skull, crania 2, 4, and 17, lack any development of styloid processes. In addition, each of these fossils exhibit a thick, rugose inferior border of the tympanic plate that

Weidenreich (1943, 1951) dubbed the crista petrosa. The Ngandong and Sambungmacan hominids also exhibit this pattern in each instance where the tympanic plates are present.

Each of the Late Pleistocene modern humans that could be scored for these traits,

Cohuna, Cossack, Keilor, Kow Swamp 5, Kow Swamp 8, Lake Mungo I, Nacurrie 1,

Nacurrie 2, and Wajak 1, provide evidence for both the presence of the styloid process as well as the development of the thin vaginal process of the tympanic bone.

Once again, the recent samples of modern humans were invariant in their expression of this trait. Each cranium examined possessed styloid processes as well as vaginal processes.

Development of the alar tubercles

This trait could be scored on Sangiran 14 and 17. On both crania the alar tubercles are present and well defined. Ngandong 7 and 12 also exhibit moderate and strong development of the alar tubercles, respectively. Sambungmacan 4 also appears to show moderate development of these structures.

Only Cohuna, Lake Nitchie, and Wajak 1 could be scored for this trait. Cohuna has moderate development of tubercles, while Lake Nitchie shows strong development.

Wajak 1 does not exhibit alar tubercle development, though it is possible that erosion has affected the preservation of this trait.

In the recent modern human samples, expression of this trait is fairly uniform.

Australia is statistically significantly different from only 2/10 populations. The Austrian sample exhibited particularly weak development in this feature, and was significantly distant from 8/10 populations.

Orientation of the occipital condyles

Sangiran 4, 14, and 17 could each be scored for this trait. Each of these fossils has condyles that taper toward the midline at their anterior end. Ngandong 7 and 12 were also examined for this character. The left condyle of Ngandong 7 is preserved and the right has been restored (Weidenreich, 1951). The long axes of these condyles are parallel to the midline. The condyles of Ngandong 12 are missing, but their bases are preserved. From these, Weidenreich (1951) deduced that they are similar in orientation to those on

Ngandong 7. A similar dilemma exists for Sambungmacan 4. Based on the photo supplied by Dr. Baba, the broken condylar bases on that specimen suggest that the

occipital condyles could either have been fully parallel or slightly canted toward the midline. Until an examination of the original fossil can be made, or at the very least a published description of the condylar area is offered, it is best to refrain from further speculation.

Cohuna, Lake Nitchie, Keilor, Nacurrie 2, and Wajak 1 could be scored for this trait. Only Keilor has condyles that could be interpreted as parallel to the midline. In this specimen the body of the condyle and the majority of the articular surface is indeed parallel, however a small projection at the anterior end of the condyle does make the condyle appear to angle toward the midline. In the remaining specimens the condyles are clearly angled medially at their anterior end.

Parallel condyles are quite uncommon in the modern human sample examined, with only nine out of 305 individuals exhibiting this feature. Australia was significantly different from only the Czech population in this trait, and overall this feature is quite consistent in the crania studied.

Opisthionic recess

Sangiran 2, 4, 12, and 17 could be scored for this character. Both Sangiran 2 and 12 preserve only a very short section of the foramen magnum border at opisthion, but neither shows evidence of the posterior foramen being narrowed into a recess. Sangiran 4 and 17, which both preserve essentially complete foramina, show varying morphologies.

Sangiran 4 does not exhibit any narrowing in the posterior foramen magnum. In Sangiran

17, however, the posterior foramen does narrow appreciably. Of the Ngandong crania, specimens 1, 7, 10, 11, and 12 can be scored for this trait. Ngandong 10 is the least

complete, offering only a very short segment of the foramen rim at opisthion. From what is preserved it does appear that the foramen was constricted at opisthion. Likewise on crania 1 and 11, which both preserve much of the posterior rim of the foramen and exhibit constriction. The two most complete cranial bases in this sample, Ngandong 7 and

12, both provide excellent examples of this morphology. In each case the posterior foramen magnum is narrowed into a triangular shape with its apex at opisthion, with

Ngandong 7 being slightly more constricted. In the Sambungmacan crania, both skulls 3 and 4 preserve at least portions of the foramen magnum. Sambungmacan 3 is the more fragmentary of the two but nonetheless provides clear evidence of constriction in the posterior foramen magnum. Sambungmacan 4 retains a complete foramen, and shows marked narrowing near opisthion. As in Ngandong 7 (and Ngandong 12 when damage to the anterior foramen is corrected for), the foramen magnum of Sambungmacan 4 has a teardrop shape.

Of the Late Pleistocene modern humans sampled, Keilor, Kow Swamp 5, Lake

Mungo III, Lake Nitchie, Nacurrie 2, and Wajak 1 could be examined for this trait. None of these specimens expressed any narrowing of the posterior foramen magnum. Kow

Swamp 5 retains only a portion of the posterior foramen, and this surviving rim was damaged, but from what is preserved it does not appear that an opisthionic recess is present on this individual.

The recent modern humans sampled were remarkably consistent in their expression of this character. Only a small percentage of the crania sampled showed even a slight narrowing of the posterior foramen magnum, and a degree of narrowing similar to that seen in the Ngandong and Sambungmacan fossils was not encountered. No

modern human group was statistically significantly separated from any other modern sample in this trait.

Postcondyloid tuberosity (PCT) development

None of the Sangiran hominids that could be scored for this character, crania 2, 4, 12, and

17, show either rugosity or tubercle development flanking the posterior rim of the foramen magnum. As mentioned previously, very little of the foramen magnum is preserved on Sangiran 2 and 12. The relevant portions of those crania which are preserved do not indicate the presence of PCT. Sangiran 4 and 17 both have good preservation of the posterior foramen magnum, and neither specimen exhibits either rugosity or PCT development in the area. The Ngandong sample preserves a number of crania which preserve at least a portion of the posterior foramen magnum, as noted above. Ngandong 1, 7, 10, 11, and 12 could be scored for this trait, and each of these specimens provides evidence of well developed PCT. In each of these specimens this rugosity is limited to the lateral borders of the foramen magnum immediately posterior to the occipital condyles, and the area immediately posterior to opisthion is smooth. Thus, there are two discrete tuberosities that are separated by a gap of several mm at the midline. PCT are also present on the only two Sambungmacan crania that retain this area of the skull, crania 3 and 4. Sambungmacan 3 shows a PCT on the surviving left border of the foramen magnum, while Sambungmacan 4 exhibits large PCT bilaterally.

In the Late Pleistocene sample of modern humans, Keilor, Kow Swamp 5,

Mossgiel, Nacurrie 2, and Wajak 1 could be scored for this trait. With the exception of

Wajak 1, each of these specimens exhibit some rugosity around the borders of the

foramen magnum but no discrete tubercle development. Wajak 1 has smooth borders with no rugosity or tubercle development.

The recent modern humans sampled show a wide range of variation in this trait.

The Australian sample has the most well developed PCT of all the modern groups studied, and it is statistically significantly different from 6/10 of the comparative populations. The remaining samples show considerably less variability amongst one another, with only one other sample (W. Africa) being significantly distant from more than two other groups in this feature.

It should be noted that the PCT exhibited in the modern populations have a different morphology than those seen in the Ngandong fossils. While the Ngandong PCT are large, discrete tuberosities, as described previously, the PCT on modern humans are considerably thinner, shorter, and often involve rugosity that extends over opisthion and completely around the posterior border of the foramen magnum. This characteristic is much more variable in its expression than the PCT seen in the Ngandong fossils.

Foramen lacerum development

None of the fossils from Sangiran retain the foramen lacerum, though Sangiran 4 and 17 are complete enough to suggest that it would have been constricted or absent. Ngandong

7 and 12 both retain this foramen bilaterally, and in each skull it is present but constricted. Sambungmacan 4 likewise exhibits constricted foramina lacerum.

Cohuna, Lake Nitchie, and Nacurrie 2 preserve this structure in the Late

Pleistocene sample of modern humans. In Cohuna the foramen is present but restricted, while in the remaining two specimens it is broad.

The recent modern human groups sampled again show a high degree of variability in the expression of this feature. The Australian sample is statistically significantly different than 6/10 of the comparative groups due to its propensity for more constricted foramina. The S. African group shows an even greater frequency of constricted foramina, and is statistically separated from 8/10 samples.

Juxtamastoid crest development

Sangiran 2, 4, and 17 were complete enough to be scored for this trait. Neither Sangiran 2 nor 17 show any development of a juxtamastoid crest. Sangiran 4 exhibits an exceptionally large swelling medial to the mastoid process on the preserved right side.

This swelling is not linear in shape, nor does it demarcate a digastric groove, and its form is unique amongst the fossil and modern humans examined for this study. It should also be noted that Sangiran 12 possesses what appears to be the edge of a juxtamastoid crest at the edge of the break medial to the broken mastoid process, but with so little of this structure remaining it is difficult to assess. In the Ngandong sample, crania 1, 7, 10, 11, and 12 preserved this area of the skull well enough for a diagnosis to be made. Ngandong

7 and 12 each exhibit very well developed juxtamastoid crests that run the full length of the occipitomastoid crests. Their degree of expression is quite exceptional. The remaining crania in this sample, with the exception of Ngandong 1, likewise exhibit juxtamastoid crest development, though to a much lesser extent. Juxtamastoid crests are not present on either Sambungmacan 3 or 4, but a small crest is present on the left side of

Sambungmacan 1.

In the Late Pleistocene modern human sample, Cohuna, Keilor, Kow Swamp 5,

Lake Nitchie, Mossgiel, Nacurrie 1, and Nacurrie 2 could be scored for this trait. Of these, only Lake Nitchie and Nacurrie 2 have small juxtamastoid crests. It should be noted that these examples are considerably less extensive than those seen in the

Ngandong crania.

The recent modern human crania show considerable variation in the expression of the juxtamastoid crest. The Australian group is statistically significant from 5/10 of the comparative groups, and in each of those cases the Australians show significantly less development of this feature. The Austrian and Czech samples showed the greatest mean development in this trait.

Occipitomastoid crest development

Sangiran 2, 4, and 17 were complete enough to be scored for this trait. Sangiran 2 and 4 lack expression of the occipitomastoid crest, while Sangiran 17 has a well-developed occipitomastoid crest forming the medial border of a deep digastric groove. Ngandong 1,

7, 10, and 12 each exhibit occipitomastoid crest development. This structure is particularly well developed in Ngandong 7 and 12. Both Sambungmacan 3 and 4 exhibit strong occipitomastoid crests.

In the Late Pleistocene modern sample, Cohuna, Keilor, Kow Swamp 5, Lake

Nitchie, Mossgiel, Nacurrie 1, Nacurrie 2 and Wajak 1 could be scored for this feature.

Only Cohuna and Lake Nitchie lacked development of an occipitomastoid crest, while

Keilor and Wajak 1 show strong development.

The recent modern human samples are fairly uniform in their expression of this trait, with the exception of Australia and S. Africa. Australia is significantly separated from 4/10 comparative groups, while S. Africa is significantly distant from 6/10, and this can be attributed to the exceptionally strong occipitomastoid crest development in these two samples.

Relative position of the carotid foramen to the squamotympanic fissure

Sangiran 4 and 17 could be scored for this trait, and in both crania the carotid foramen is well posterior to the S-Q fissure. Ngandong 6, 7, 10, 11, and 12 could be scored for this character, and in all cases the foramen is either fully posterior to the fissure (Ngandong 6,

7, 10, and 12) or the leading edge of the foramen is even with the fissure (Ngandong 11).

On Sambungmacan 4 the foramen is fully posterior to the S-Q fissure.

In the Late Pleistocene modern sample, Cohuna, Keilor, Kow Swamp 5, Nacurrie

2, and Lake Nitchie could be scored for this trait. Of these, Kow Swamp 5 had carotid foramina posterior to the S-Q fissure, Lake Nitchie had foramina whose leading edges were even with the fissure, and the remainder had foramina bisected by the S-Q fissure.

The modern human crania sampled tend to show little variation in this trait, generally exhibiting foramina that are bisected by the S-Q fissure. The exceptions to this general pattern are Australia and W. Africa, whose foramina lie more posterior (and thus farther from the S-Q fissure) than is typical for the overall sample, and the Utah sample, whose foramina lie more anterior than the norm. Australia possesses the most posterior position of the carotid foramen on average and is significantly different from 8/10 samples in this feature, while W. Africa differs from 5/10 and Utah from 6/10.

Size of postglenoid tubercle

Postglenoid tubercle height could be scored on Sangiran 2, 4, and 17. Each of these specimens exhibit clear postglenoid tubercles that exceed 3 mm in height. Ngandong 1, 6,

7, 10, 11, and 12 all retain this portion of the skull, and none of these specimens exhibit a postglenoid tubercle. Likewise, Sambungmacan 3 does not have postglenoid tubercles.

Based on the photo examined, Sambungmacan 4 may have postglenoid tubercles, but examination of the fossil will be required to determine this.

In the late Pleistocene modern human sample, every specimen examined retained at least one mandibular fossa for scoring purposes. Each of these specimens exhibits clear postglenoid tubercles that exceed 3 mm in height.

Prominent postglenoid tubercles were common among the recent modern humans included in this study. All modern humans studied have postglenoid tubercles, and an inferiorly projecting posterior wall of the mandibular fossa anterior to the S-Q fissure, but a small percentage have relatively small postglenoid tubercles. The Australian sample was significantly different from 3/10 comparative samples, and in all cases these samples

(Czech, Utah, and Greece) were significantly smaller than the Australians.

Orientation of Squamotympanic (S-Q) fissure

Of the Sangiran hominids, crania 2, 4, and 17 could be scored for this feature. Sangiran 2 and 17 have S-Q fissures that are roughly perpendicular to the midline. On Sangiran 4 the right S-Q fissure is perpendicular and the left is at an oblique angle posteriorly. This discrepancy is likely due to the deformation that has taken place on this specimen. As mentioned previously, it is unknown which side of the skull is least affected by this

deformation. Ngandong 1, 6, 7, 10, 11, and 12 could be used to examine this feature, and each of these specimens show S-Q fissures that are perpendicular to the midline.

Sambungmacan 3 and 4 likewise show S-Q fissures roughly perpendicular to the midline.

Each of the Late Pleistocene modern humans examined were able to be scored for this trait. Cohuna and Kow Swamp 5 exhibit fissures that are perpendicular to the midline, while the remainder of this sample shows oblique angles posteriorly.

The recent modern humans sampled typically show S-Q fissures that angle obliquely posteriorly, but it is not that uncommon to find perpendicular S-Q fissures amongst these crania. Perpendicular fissures occurred most frequently in the S. African and W. African groups, at rates of 23.33% and 27.59% respectively, while the Australian sample had a rate of 22.47%. The Australians were statistically significantly separated from 4/10 samples (China; Czech, Utah, and Java), each of whom had lower frequencies of perpendicular S-Q fissures.

Relative area of occipital condyle to foramen magnum

This feature could be scored on Sangiran 4 and 17, with these specimens scoring 39.40% and 31.19% respectively. In the Ngandong fossils this measure could be taken on two specimens, crania 7 and 12. Ngandong 7 has a ratio of 17.96, while Ngandong 12 (based on condylar measurements reconstructed by Weidenreich [1951]) has a ratio of 20.75.

None of the Sambungmacan specimens can be used to score this feature, though it should be noted that the broken bases present on Sambungmacan 4 are quite small relative to the area of the foramen magnum.

100

In the Late Pleistocene modern sample, this feature can be scored on Keilor and

Lake Nitchie and estimated for Nacurrie 2 and Wajak 1. Keilor shows a ratio of 23.60, while Lake Nitchie has a ratio of 32.39. Nacurrie 2 has damage to the occipital condyles that affect breadth, while Wajak 1 has damage to both the condyles as well as the rim of the foramen magnum. In both cases, however, the condyles are quite large relative to the area of the foramen magnum. Based on repeated observations of this feature, this researcher feels that a conservative estimate of the ratio for both specimens is between

25.00 and 30.00.

There is considerable variability amongst the recent modern human samples in this feature. Of the 262 total modern crania scored, only 3 (1.15%) have condyles that are less than 21% of the foramen magnum area. Of these, two are S. African and one is

Austrian. None of the crania that scored below 21% are Australian (or even Asian). Table

5.1 provides the mean and range for the modern human groups sampled. Table 5.2 provides the results of independent samples t-tests for equality of means. The Native

American sample from Utah has the relatively largest condyles, while the Czech sample has the relatively smallest condyles. The Australian sample falls near the middle of these groups in terms of its average relative size, and is statistically significantly different from

4/10 samples. In three of the four statistically significant comparisons, the Australian crania are significantly larger than the comparative population. Only Utah is significantly larger than Australia in this trait. The scores obtained for the Ngandong crania do not fit within the range sampled for Australia, and the Australian condyles are clearly dissimilar to the Ngandong crania in both their absolute and relative size.

101

Table 5.1: Summary statistics for the relative area of occipital condyle to foramen magnum in the modern sample.

Population N Mean Range Australia 73 33.38 21.94-51.06 Austria 15 30.17 19.34-38.54 China 10 31.77 25.24-43.41 Czech 21 29.81 22.46-38.37 Egypt 13 33.24 24.68-53.27 Greece 17 31.10 22.23-37.93 Utah 17 39.92 31.01-54.18 India 10 33.05 25.88-44.44 Java 23 32.72 24.44-41.12 S. Africa 42 29.84 17.44-41.36 W. Africa 21 36.32 25.74-51.27

Table 5.2: Results of t-test for equality of means for the relative area of occipital condyle to foramen magnum area in the modern sample. Equal variances are not assumed in this test. Listed score reflects the p-value for that pairwise comparison. Statistically significant scores (p=.05) are listed in bold.

Australia 0.000 Austria 0.040 0.000 China 0.448 0.508 0.000 Czech 0.004 0.828 0.386 0.000 Egypt 0.955 0.251 0.624 0.178 0.000 GG, Utah 0.001 0.000 0.004 0.000 0.022 0.000 Greece 0.068 0.583 0.767 0.370 0.396 0.000 0.000 India 0.880 0.254 0.655 0.176 0.949 0.015 0.409 0.000 Java 0.571 0.131 0.669 0.041 0.832 0.000 0.259 0.887 0.000 S. Africa 0.001 0.833 0.380 0.982 0.173 0.000 0.343 0.170 0.027 0.000 W. Africa 0.084 0.004 0.080 0.001 0.267 0.104 0.007 0.213 0.053 0.001 0.000

Australia Austria China Czech Egypt GG, Utah Greece India Java S. Africa W. Africa

102

Discussion

These non-metric characters provide an opportunity to test the assumptions of the

Multiregional hypothesis of modern human origins. According to this model, modern populations that inhabit a particular region should share a suite of characters with the archaic populations that preceded them. The features examined here provide broad coverage of the cranial base, and provide evidence from each fossil and modern population that has been proposed to form a part of the regional continuity sequence in

Australasia.

Table 5.3 shows the character states assigned to the non-metric features studied here. Based on these data, a hypothesis of continuity between the Sangiran fossils and

Ngandong/Sambungmacan cannot be rejected. None of the features examined is inconsistent with the development of the Ngandong pattern of morphologies from an ancestral group that includes the known Sangiran hominids. These same data, however, are not consistent with a hypothesis of continuity between the hominids from

Ngandong/Sambungmacan and later humans in Australia. No single feature, let alone a suite of features, can be cited that supports a link between these samples. Even the PCT, which is both characteristic of the Ngandong fossils and relatively well developed in modern Australian crania, cannot be used to support an argument for continuity because of the different manifestations of this trait in each group. There is also no evidence for

PCT development in any of the Late Pleistocene or early Holocene samples from

Australasia that could serve as intermediaries morphologically.

Even more damning is the presence of a number of potential autapomorphies in

103

Table 5.3: Non-metric traits used in this study and their frequencies in each segment of the proposed regional continuity sequence for Australasia. Potential autapomorphies in Ngandong are noted in bold type.

Feature Sangiran Ngandong/Sam. L. Pleist. Aust. Modern Australian Pharyngeal tub. Trace Variable Present Present Tympanomastoid fis. Variable Strong Absent Absent Ovale pit/accessory Absent Present Absent Absent Styloid/vaginal proc. Absent Absent Present Present Alar tubercles Present Present Present Present Orientation of condyles Angled Parallel Angled Angled Opisthionic recess Trace? Present Absent Absent PCT development Absent Present Absent Variable Foramen lacerum ? Constricted Variable Present Juxtamastoid Variable Present Variable Variable Occipitomastoid Present Present Present Present Carotid relative to S-Q Posterior Posterior/even Posterior/even Variable Postglenoid size >3mm Absent >3mm >3mm Orientation of S-Q Perpen. Perpen. Oblique Oblique Relative size of OC/FM >30 <21 >23 and + Large

the Ngandong/Sambungmacan cranial base. These are shown in bold in Table 5.3 and have not been found outside the Ngandong and Sambungmacan fossils (Baba et al., 2003,

2004; Durband 2002a, b, 2004). These characters provide evidence of a clear morphological break between the Ngandong/Sambungmacan people and all later populations in Australasia and elsewhere. Tables 5.4-5.9 provide the results of a series of

Fisher’s Exact tests that were run on these data. Every possible pairwise comparison was tested within the proposed Australasian continuity sequence of Sangiran-Ngandong-

Pleistocene Australia-Modern Australia. These tests provide further support for discontinuity within this alleged sequence in a number of characteristics. Clearly, the

Ngandong hominids show the greatest dissimilarity with each of the other samples tested.

104

Table 5.4: Results of Fisher’s Exact tests comparing data for Sangiran and Ngandong/Sambungmacan. Statistically significant values (p=.05) are shown in bold. Categories that did not differ between the groups are scored “No Diff.”

Character p value Pharyneal tubercle development 1.000 Tympanic plate contact w mastoid 0.250 Ovale pit/accessorium 0.250 Styloid/vaginal No Diff. Alar tubercle development No Diff. Condyle orientation 0.100 Opisthionic recess 0.024 PCT development 0.003 Foramen lacerum development NA Juxtamastoid process 1.000 Occipitomastoid process 0.067 Carotid relative to S-Q fissure 0.222 PGT size 0.018 Orientation of S-Q fissure 0.250

Table 5.5: Results of Fisher’s Exact tests comparing data for Sangiran and the Pleistocene Australians. Statistically significant values (p=.05) are shown in bold. Categories that did not differ between the groups are scored “No Diff.”

Character p value Pharyneal tubercle development No Diff. Tympanic plate contact w mastoid 0.038 Ovale pit/accessorium No Diff. Styloid/vaginal 0.006 Alar tubercle development 1.000 Condyle orientation 1.000 Opisthionic recess 1.000 PCT development No Diff. Foramen lacerum development NA Juxtamastoid process 0.576 Occipitomastoid process 0.500 Carotid relative to S-Q fissure 1.000 PGT size No Diff. Orientation of S-Q fissure 0.236

105

Table 5.6: Results of Fisher’s Exact tests comparing data for Sangiran and modern Australians. Statistically significant values (p=.05) are shown in bold. Categories that did not differ between the groups are scored “No Diff.”

Character p value Pharyneal tubercle development 1.000 Tympanic plate contact w mastoid 0.004 Ovale pit/accessorium No Diff. Styloid/vaginal 0.000 Alar tubercle development 1.000 Condyle orientation 1.000 Opisthionic recess 0.430 PCT development 1.000 Foramen lacerum development NA Juxtamastoid process 0.606 Occipitomastoid process 0.086 Carotid relative to S-Q fissure 0.490 PGT size 1.000 Orientation of S-Q fissure 0.153

Table 5.7: Results of Fisher’s Exact tests comparing data for Ngandong/Sambungmacan and the Pleistocene Australians. Statistically significant values (p=.05) are shown in bold. Categories that did not differ between the groups are scored “No Diff.”

Character p value Pharyneal tubercle development 0.400 Tympanic plate contact w mastoid 0.000 Ovale pit/accessorium 0.018 Styloid/vaginal 0.000 Alar tubercle development 1.000 Condyle orientation 0.048 Opisthionic recess 0.003 PCT development 0.001 Foramen lacerum development No Diff. Juxtamastoid process 0.315 Occipitomastoid process 0.462 Carotid relative to S-Q fissure 0.364 PGT size 0.000 Orientation of S-Q fissure 0.002

106

Table 5.8: Results of Fisher’s Exact tests comparing data for Ngandong/Sambungmacan and the modern Australians. Statistically significant values (p=.05) are shown in bold. Categories that did not differ between the groups are scored “No Diff.”

Character p value Pharyneal tubercle development 0.019 Tympanic plate contact w mastoid 0.000 Ovale pit/accessorium 0.000 Styloid/vaginal 0.000 Alar tubercle development 1.000 Condyle orientation 0.000 Opisthionic recess 0.000 PCT development 0.000 Foramen lacerum development 1.000 Juxtamastoid process 1.000 Occipitomastoid process 1.000 Carotid relative to S-Q fissure 0.183 PGT size 0.000 Orientation of S-Q fissure 0.000

Table 5.9: Results of Fisher’s Exact tests comparing data for the Pleistocene Australians and the modern Australians. Statistically significant values (p=.05) are shown in bold. Categories that did not differ between the groups are scored “No Diff.”

Character p value Pharyneal tubercle development 1.000 Tympanic plate contact w mastoid 1.000 Ovale pit/accessorium No Diff. Styloid/vaginal No Diff. Alar tubercle development 0.295 Condyle orientation 0.123 Opisthionic recess 1.000 PCT development 1.000 Foramen lacerum development 1.000 Juxtamastoid process 0.095 Occipitomastoid process 0.041 Carotid relative to S-Q fissure 1.000 PGT size 1.000 Orientation of S-Q fissure 1.000

107

The Sangiran-Ngandong sequence shows significant differences in 3/14 characteristics,

while Ngandong is significantly different from the Pleistocene Australians in 8/14

characters and the modern Australians in 9/14. Sangiran, conversely, is significantly

different from the Pleistocene and modern Australians in only 2/14 characteristics. It

should be noted that both features that are significantly different in the Sangiran/Australia

comparisons involve the tympanic plate, and both features have been named as potential

autapomorphies for Asian H. erectus (Andrews, 1984). These tests lend additional weight to the conclusion that a pattern of morphologies clearly separates the Ngandong group from the Australian populations, a finding that runs counter to the expectations of any of the theories espousing continuity between the archaic and modern populations of

Australasia.

However, since each of these features is located on the cranial base, with some found on the same structure (i.e. the size and orientation of the occipital condyles), it could be argued that these changes are somehow related as part of a complex and therefore should not be considered individually. This would reduce the significance of these autapomorphies, and suggest that these changes could be relatively minor in scope.

To test this possibility, polychoric correlations were run on a sample of 201 modern humans from the AMNH used in this project. These tests examined variation in four characters: condyle orientation, relative condyle size, opisthionic recess, and PCT development, in an effort to determine the degree of interrelatedness. The morphology of the foramen ovale could not be tested in this way because it did not vary in any of the modern populations. The fossil groups surveyed were likewise not included in the polychoric correlation because the features under scrutiny were relatively invariant in

108

these specimens. Their group structure could have considerably influenced the correlation measures and potentially invalidated the results. For these tests the modern samples were pooled.

The results of the polychoric correlation, presented below in Table 5.10, provide correlation values that can be examined for statistical significance. These tests confirm the very low, often statistically insignificant levels of interdependence between the features studied. The majority of the Pearson and Spearman correlation scores do not exceed the .05 or .10 p-values for statistical significance, especially when the asymptotic error values (ASE) are considered. These error values reflect the disproportionate numbers of individuals in the various cells created during the computation of the chi-square matrix.

The polychoric correlation value also allows an estimation of the relative effect that each feature has on the others. The square of the polychoric correlation score provides a percentage score reflecting how change in one feature is influencing the other trait of interest. For example, the highest polychoric correlation score obtained during these examinations is 0.2868 for the combination of condyle orientation and relative condyle size. The square of this number, or 8.22%, is the influence that relative condyle size has on the manifestation of condyle orientation (or vice versa) in this sample of modern humans. These scores indicate that there is a very low level of correlation between the basicranial features studied in this analysis, and is consistent with an interpretation of independent development of these features.

From these results it would appear that the Ngandong and Sambungmacan

109

Table 5.10: Polychoric correlation results for the potential Ngandong and Sambungmacan autapomorphies.

______

Statistics for Condyle orientation by Opisthionic recess Statistic Value ASE Pearson correlation 0.0048 0.0670 Spearman correlation 0.0027 0.0700 Polychoric correlation 0.0158 0.2589

Statistics for Condyle orientation by PCT Statistic Value ASE Pearson correlation 0.0923 0.0672 Spearman correlation 0.0924 0.0715 Polychoric correlation 0.0158 0.1796

Statistics for Condyle orientation by Relative condyle size Statistic Value ASE Pearson correlation 0.1042 0.0683 Spearman correlation 0.1034 0.0645 Polychoric correlation 0.2868 0.1971

Statistics for Opisthionic recess by PCT Statistic Value ASE Pearson correlation 0.0306 0.0716 Spearman correlation 0.0362 0.0758 Polychoric correlation 0.0473 0.1177

Statistics for Opisthionic recess by Relative condyle size Statistic Value ASE Pearson correlation -0.0332 0.0739 Spearman correlation -0.0380 0.0790 Polychoric correlation -0.0425 0.1300

Statistics for PCT by Relative condyle size Statistic Value ASE Pearson correlation 0.1780 0.0782 Spearman correlation 0.1789 0.0762 Polychoric correlation 0.2081 0.0871

______

110

hominids had undergone both genetic and morphological changes from the pattern observed in the earlier Sangiran hominids. To date, this Ngandong pattern has not been observed elsewhere, and indeed none of the individual characters has been observed elsewhere to my knowledge. An additional feature thought to be autapomorphic for the

Ngandong and Sambungmacan fossils, the location of the S-Q fissure in the apex of the mandibular fossa, will be discussed in detail in the next chapter.

The evidence considered here also suggests that the modern human groups under consideration are not characterized by regional patterns of features in the cranial base.

None of the individual samples were consistently statistically significantly different from any other such sample for any groups of features. Thus, at least amongst this set of characters, it would be difficult to find support for any regional suites of features that differentiated these modern human groups from one another. Instead, these results are more consistent with the hypothesis of a single “modern” morphological pattern in the cranial base. While there is certainly variability in this pattern, shown by the statistically significant differences that were encountered for some individual features, overall these samples are much more similar to one another than they are different. The modern crania also differ from the Sangiran and Ngandong crania in a fairly consistent manner, with no individual samples evincing a greater degree of similarity to those Indonesian fossils.

Indeed, features such as the lack of a styloid process and the presence of a tympanomastoid fissure served to separate those ancient Indonesians from the entirety of the modern human sample. Based upon this evidence, an argument of regional continuity would not appear to be tenable. Instead, these findings, which provide evidence for a morphological break in the proposed continuity sequence of Australasia as well as a

111

single “modern” morphological pattern, are more consistent with a replacement hypothesis.

112

Chapter 6: Temporomandibular joint (TMJ) morphology

Introduction

The potentially unique nature of the Ngandong TMJ was first noted by Weidenreich

(1951) in his superlative monograph on those fossils. His description is as follows (note that he is writing from the point of view of looking down upon the base of the skull while examining it, rather than referring to proper anatomical position):

“The deepest point of the [mandibular] fossa in modern man and in Sinanthropus

is situated farther forward than in Solo man, and the floor of the fossa ascends

towards the posterior wall before it reaches the Glaserian [squamotympanic]

fissure. However, in Solo man this fissure runs along the very floor of the fossa;

the wall ascends in front of and behind it. The posterior wall is formed for its full

extent solely by the tympanic plate” (Weidenreich, 1951: 273-274).

Weidenreich (1951: 273) further states that the Zhoukoudian fossils retain only “the vestigial equivalent of a postglenoid process” and that in “Solo man this structure is reduced to a short, thin line. It cannot form the posterior wall of the fossa, not only because it is much too small, but, above all, because it is a structure of the anterior and not of the posterior wall of the fossa.” In other words, the thin ridge that represents the postglenoid tubercle in the Ngandong fossils is anterior to the apex of the mandibular fossa. Based on his comparison to the fossil material available at the time (namely the

Zhoukoudian crania, Sangiran 4, Kabwe 1, and a few Neandertal skulls), as well as some

113

representatives of modern chimpanzee, gorilla, and one Tasmanian skull (AMNH VL

275), Weidenreich (1951) felt that the TMJ in the Ngandong skulls was “peculiar.”

Larnach and Macintosh (1974) studied these morphologies further. Their work

involved a study of 18 character traits from casts of Ngandong 7 and 12 (Solo VI and XI)

in a series of modern human crania in an effort to determine the persistence of these traits

in modern populations. From this project they concluded “that there is a great

morphological difference between the Solo crania we know and the crania of any modern

group” (Larnach and Macintosh, 1974: 101). Regarding the specific morphologies of the

mandibular fossa, Larnach and Macintosh (1974) found that postglenoid tubercles, while

present in other H. erectus fossils, were absent in the Ngandong fossils. Likewise,

Ngandong was unique in the location of the location of the squamotympanic (S-Q) fissure in the very floor of the fossa. These authors hypothesized that “[t]hese traits must have appeared in Solo man after he evolved from Homo erectus, for they distinguish Solo

man from Homo erectus no less than they do from modern man” (Larnach and

Macintosh, 1974: 101).

Debate concerning this feature came to the fore in 2001 with the publication of

the Sambungmacan 3 specimen. Márquez et al (2001: 363) state that “the

squamotympanic fissure, sometimes incorrectly referred to as the “Glaserian” fissure, …

runs the entire length of the floor in the deepest portion of the fossa.” Delson and

colleagues (2001), in their comparative analysis of the specimen, note that

Sambungmacan 3 shares this morphology with the Ngandong fossils and Sambungmacan

1. However, those authors further claim that this morphology is shared with other

114

specimens from Java and even Africa, namely Sangiran 4 and 17 and the OH 9 and

KNM-WT 15000 crania (Delson et al., 2001).

This assertion was met with skepticism by Durband (2002a), who noted the presence of postglenoid tubercles in each of the fossils from Sangiran and Africa that were claimed to exhibit the Ngandong/Sambungmacan TMJ morphology. These tubercles form a portion of the posterior wall of the fossa, and preclude the S-Q fissure from running along the apex of the fossa for its entire length. While in many crania the S-Q fissure does indeed run in the apex of the fossa at its most medial aspect, as the fissure nears the lateral edge of the fossa it moves posterior to the postglenoid tubercle and out of the deepest portion of the fossa (Durband, 2002a). Mowbray and colleagues (2002) responded to this criticism and reasserted their claims that the form of the

Ngandong/Sambungmacan TMJ was not limited to that group and could be found in the

Sangiran and African fossils previously mentioned. Further personal communication with two of the responding authors (K. Mowbray and S. Márquez) revealed that the morphology of Ngandong 7 had been a particular focus of their assessment. They felt that the right TMJ of that specimen preserved clear evidence of a well-developed postglenoid tubercle, and its morphology was very similar to that of OH 9 in particular.

Antón (2002, 2003) has scored this characteristic following Delson and colleagues (2001), and used this perceived similarity (amongst many others) to group the

Indonesian fossils within a lineage that excludes the Chinese H. erectus specimens. The location of the S-Q fissure in the deepest point of the glenoid fossa was thus seen to unite a lineage of “small brained” Indonesian fossils (Sangiran 4), a “large brained, early”

Indonesian category (represented solely by Sangiran 17), and the “large brained, late”

115

Indonesian sample (which includes Sambungmacan 1 and the Ngandong sample) (Antón,

2002). While Antón (2002) could not use Sambungmacan 3 in her study because it had not yet been published, its inclusion would have provided additional support for her groupings.

While seemingly a minor and relatively esoteric feature, the shape of the TMJ is important for reconstructing the evolution of hominids in Australasia. Whether the course of the S-Q fissure is unique to Ngandong and Sambungmacan, and therefore indicative of morphological change between these specimens and the Sangiran hominids (e.g. Larnach and Macintosh, 1974; Durband, 2002a, b, 2004) or instead simply highlights regional variation between geographic variants of Homo erectus (e.g. Delson et al., 2001; Antón,

2002, 2003), is still open to question. Therefore, it is necessary to revisit the morphology of this feature. In this section the TMJs from several fossils will be analyzed in an effort to determine the specific manifestation of this feature in each of the hominids mentioned by Durband (2002a) and Delson and colleagues (2001; Mowbray et al., 2002).

Results

The Ngandong fossils

Casts of three Ngandong specimens were available to be scanned for this project, crania

7, 10, and 12. As mentioned earlier, Ngandong 7 was noted by Mowbray and colleagues

(2002; Mowbray, personal communication) as exhibiting TMJ morphology similar to other specimens of both African and Indonesian H. erectus. Both the right and left fossae of Ngandong 7 were scanned for this project. For Ngandong 10 and 12 only the right

116

TMJ was scanned either because it was the only one remaining (Ngandong 10) or the more complete of the two (Ngandong 12).

Ngandong 12 provides one of the best preserved examples of the TMJ in the series, and as such provides an excellent type specimen to represent the typical Ngandong morphologies that will be discussed. As noted by Weidenreich (1951) and Larnach and

Macintosh (1974) previously, the Ngandong TMJ is characterized by the extreme reduction (or even absence) of the postglenoid tubercle, and the location of the S-Q fissure in the apex of the fossa along its entire length. Ngandong 12 provides an excellent example of these morphologies, and Figures 6.1-6.2 clearly show this distinct Ngandong pattern. The fossa is undamaged in this specimen, and examination of the original fossil did not suggest that weathering or any other agent had obscured or distorted the morphology of this structure.

Ngandong 7, on the other hand, does exhibit a somewhat different morphology in the right TMJ (shown in Figure 6.3), as noted by Mowbray and colleagues (2002).

However, it appears likely that the area in question has undergone some plastic deformation during fossilization. The zygomatic root appears to have been bent inferiorly at its lateral edge, and as a result the squamous portion of the mandibular fossa has been similarly affected. This has resulted in the displacement of the entire lateral portion of the fossa, with the fossa forming a basin rather than being open laterally as in the other crania in the series. Further examination reveals that the most posterior portion of the fossa does not curve inferiorly to form a postglenoid tubercle, but instead is essentially uniformly flat. Hence, while the most posterior and lateral corner of the fossa is indeed inferior to

117

Figure 6.1: 3-D scans of the right TMJ of Ngandong 12, inferior view. Top of page is anterior, left side is lateral. Note the location of the S-Q fissure on the top view, it is in the apex of the fossa. It is also important to note that no postglenoid tubercle is located anterior to the fissure.

118

Figure 6.2: Right TMJ of Ngandong 12 (cast). Note the location of the S-Q fissure in the apex of the mandibular fossa, and the lack of a postglenoid tubercle anterior to the fissure. The fissure runs along the very highest extent of the fossa.

119

Figure 6.3: 3-D scans of the right TMJ of Ngandong 7, inferior view. Left side of picture is anterior, top of page is medial. Plastic deformation has resulted in the inferior displacement of the lateral edge of the TMJ, with the result that the lateral edge of the fossa curves inferiorly in a uniform fashion. Note the lack of any discrete swellings at the posterior edge of the fossa that would be indicative of a postglenoid tubercle. Instead, the lateral edge of the fossa is uniformly depressed, indicating that no tubercle is present.

120

Figure 6.4: 3-D scan of the left TMJ of Ngandong 7, lateral view. External auditory meatus is visible at left (the hole in the scan represents an area inside the tympanic tube that could not be reached with the scanning laser).

the S-Q fissure, and thus superficially similar to specimens such as OH 9 or Sangiran 17, this examination indicates that the morphology of Ngandong 7 is not analogous to other non-Ngandong hominids. Study of the less-complete left TMJ supports this diagnosis.

Figure 6.4 provides a lateral view of this structure, and demonstrates that this fossa exhibits a morphology broadly similar to that seen in Ngandong 12. Though the most lateral parts of the fossa are missing, the surviving roof of the structure does slope superiorly to the S-Q fissure. Additional support for the notion that the right TMJ of

Ngandong 7 suffers from plastic deformation is provided by Ngandong 10. In this specimen, the zygomatic root is twisted inferiorly, as in skull 7, and the surviving

121

Figure 6.5: 3-D scan of the right TMJ of Ngandong 10, lateral view. The zygomatic root curves inferiorly and is twisted laterally, leading to the inferior displacement of the lateral mandibular fossa.

lateral portions of the fossa have been displaced inferiorly. Figure 6.5 shows a lateral view of the depressed zygomatic root, and the subsequent displacement of the lateral mandibular fossa in Ngandong 10. It is not unreasonable that the plate of bone at the zygomatic root, a relatively thin structure even in the heavily built Ngandong crania, could be deformed by taphonomic processes. Undeformed examples of the zygomatic process of the temporal bone can be seen on two Ngandong crania besides the aforementioned skull 12, crania 6 and 11. In both specimens the morphology of the TMJ is virtually identical to that seen in Ngandong 12.

122

Figure 6.6: 3-D scan of the left TMJ of Sambungmacan 3, inferior view. Top of page is anterior, the right side is lateral. Note the absence of a postglenoid tubercle on the posterior edge of the fossa.

Sambungmacan 3

Sambungmacan 3 (Figure 6.6) likewise displays the Ngandong morphology in the TMJ

(Delson et al., 2001; Márquez et al., 2001; Durband, 2002a, b, 2004). Only the left TMJ of this specimen is preserved well enough to be included in this study. As in the

Ngandong crania, the S-Q fissure is clearly in the apex of the fossa. Interestingly,

Sambungmacan 3 preserves a ridge on the anterior wall of the TMJ that is similar to structures found in some Ngandong skulls. This low ridge is located in the same area

123

referred to by Weidenreich (1951: 273) as the “short, thin line” equivalent to a greatly reduced postglenoid tubercle in Ngandong. Similar ridges appear on the anterior slope of the Ngandong 12 TMJ.

Sangiran 4

Sangiran 4 was the first specimen mentioned by Delson and colleagues (2001) as being similar to the Ngandong fossils in the morphology of the TMJ. According to those authors, Weidenreich “observed that in the Ngandong specimens and in Sangiran 4 the squamotympanic fissure courses mediolaterally and coincides with the deepest portion of the mandibular fossa” (Delson et al., 2001: 390). This statement, however, does not agree with the present author’s reading of the source to which they refer. Weidenreich (1951:

274), does, in fact, express the opinion “that the mandibular fossa of Pithecanthropus robustus [Sangiran 4] generally has exactly the same form as that of Solo man.” As he continues, however, it is clear that he does not consider the course of the S-Q fissure in

Sangiran 4 to be identical to that of Ngandong; stating that “[t]here are, however, some differences. The tubercle is shorter and higher in Pithecanthropus. A postglenoidal process separates the most lateral part of the fossa from the vestibulum of the ear entrance” (Weidenreich, 1951: 274). As noted previously, Weidenreich (1951) specifically stated that no postglenoid structures appear in the Ngandong fossils. Sagittal sections of Ngandong 12 and Sangiran 4 provided by Weidenreich (1951: 273, Figure 26) likewise reflect his opinion that both the shape of the fossa and the location of the S-Q fissure differ between those two specimens. Figure 6.7 clearly shows the presence of a

124

Figure 6.7: Sangiran 4 left mandibular fossa (cast). Note the inferior curvature of the posterior mandibular fossa, and the large postglenoid tubercle at the posterior edge of the fossa.

large postglenoid tubercle forming a portion of the posterior wall of the Sangiran 4 mandibular fossa. Anterior to this tubercle the roof of the TMJ curves inferiorly, and the

S-Q fissure is posterior to the apex of the fossa.

Sangiran 17

Sangiran 17 was likewise mentioned by Delson et al (2001: 390) as “show[ing] a morphology similar to that preserved in the Ngandong group.” Mowbray and colleagues

(2002) elaborate, claiming that Sangiran 17 is similar to the condition seen in Ngandong

125

7. As shown earlier, in Figure 6.4, the entire lateral edge of the Ngandong 7 TMJ curves inferiorly in the transverse plane and does so uniformly, suggesting deformation of the surface. A scan of the Sangiran 17 TMJ shows a different morphology. Figure 6.8 shows a view of the right TMJ of this specimen, and indicates the presence of a postglenoid tubercle at the posterior edge of the fossa. The topography of the posterior fossa supports this diagnosis, and clearly shows that the roof of the fossa slopes inferiorly in the sagittal plane towards the tympanic plate anterior to the S-Q fissure. This does not conform to the morphology seen in Ngandong 7, where the lateral edge of the fossa does not curve inferiorly in the sagittal plane near the tympanic plate but instead stays essentially flat in the transverse plane.

OH 9

Delson and others (2001: 392) also contend that “OH 9 presents a mandibular fossa that is similar both in its shape and in the position of the squamotympanic fissure to the

Ngandong and other Indonesian specimens.” Mowbray et al (2002) elaborate on this point. “Of perhaps greater importance,” they note, “is our observation that several

African members of the H. erectus (s.l.) group, such as NMT OH 9 and KNM-WT

15000, exhibit an SQF aligned similarly to those of the Indonesian fossils” (Mowbray et al., 2002: 144). Those authors explain that “[c]lose examination of a cast of NMT OH 9 reveals that the SQF does run in the highest portion of the mandibular fossa roof.

Tracking it from its medial border, the SQF runs along the highest part of the roof and continues laterally demarcating the posterior edge of the temporally-derived portion of

126

Figure 6.8: 3-D scan of the right TMJ of Sangiran 17. Top of page is anterior, left side is lateral. Note the large postglenoid tubercle, which can be seen on the topographic scan as well.

127

the glenoid fossa, which has become slightly lower at its most lateral roof edge”

(Mowbray et al., 2002: 144). I would agree that this is a perfectly accurate description of the OH 9 mandibular fossa and the role of the S-Q fissure. However, the location of that fissure and the topography of the OH 9 TMJ differs significantly from that seen in any of the Ngandong specimens, including skull 7.

Figures 6.9-6.10 provide views of the OH 9 TMJ. This specimen possesses a large postglenoid tubercle, a structure that does not appear on any of the Ngandong specimens.

In addition, the lateral side of the fossa is open and does not curve inferiorly to any significant degree to form a basin, which is different from the morphology previously discussed for Ngandong 7 (contra Mowbray et al., 2002). The roof of the OH 9 TMJ curves inferiorly in the sagittal plane from its apex to the postglenoid tubercle, which is located anterior to the S-Q fissure. This morphology is unlike that seen in any of the

Ngandong fossils or Sambungmacan 3.

WT 15000

As mentioned previously, both Delson et al (2001) and Mowbray et al (2002) claim that the juvenile KNM-WT 15000 (Figures 6.11-6.12) exhibits TMJ morphology similar to that seen in the Ngandong specimens. Delson and colleagues (2001) are careful to note that this specimen does represent a subadult, and thus the morphologies displayed in this fossil are potentially unrepresentative of the adult condition. With this in mind, WT

15000 does exhibit a slightly different TMJ morphology than those mentioned previously. While the specimen does possess a clear postglenoid tubercle, it is short and bounds a very shallow mandibular fossa. There is also a 2-3 mm gap between the

128

Figure 6.9: Right TMJ of OH 9 (cast). Note the large postglenoid tubercle, and the inferior curvature of the mandibular fossa anterior to the tubercle. The S-Q fissure is well posterior to the apex of the fossa, and is unlike the condition seen in Ngandong.

129

Figure 6.10: 3-D scans of the right TMJ of OH 9. Top of page is anterior, left side is lateral. Note the presence of a discrete postglenoid tubercle at the posterior edge of the fossa. The lateral edge of the fossa does not curve inferiorly in a uniform fashion as seen in the right TMJ of Ngandong 7.

130

Figure 6.11: 3-D scans of the left TMJ of WT 15000. Note the location of the S-Q fissure posterior to the postglenoid tubercle. 131

Figure 6.12: Right TMJ of WT 15000 (cast). Though these features are not as well expressed as in the adult specimens studied, a postglenoid tubercle is present anterior to the S-Q fissure. The roof of the fossa curves inferiorly to the tubercle.

132

postglenoid tubercle and the anterior edge of the tympanic bone, which is unusual. From the preserved morphologies, however, it is readily apparent that WT 15000 bears a much closer resemblance to OH 9 than it does to the Ngandong hominids. Figures 6.11-6.12 illustrate both TMJs for this specimen, and illuminate all the relevant points of interest to this study. In this specimen there is a clear inferior curvature of the roof of the mandibular fossa in the sagittal plane culminating in the formation of a postglenoid tubercle, and the S-Q fissure is posterior to this structure. The roof of the fossa does not continue to slope superiorly until its apex coincides with the fissure.

Discussion

These examinations support the conclusion that the morphology seen in the Ngandong and Sambungmacan TMJs is autapomorphic for those populations. Studies by

Weidenreich (1951), Larnach and Macintosh (1974), and Durband (2002a, b, 2004) have pointed out the unique nature of the Ngandong and Sambungmacan TMJ in comparisons with both fossil and modern hominid samples. While there is some variability within the

Ngandong sample in the mandibular fossa, as pointed out by Mowbray and colleagues

(2002), these differences can be attributed to plastic deformation suffered during fossilization. The characteristic combination of the lack of a postglenoid tubercle and the location of the S-Q fissure lying solely in the apex of the mandibular fossa is currently limited to only the Ngandong and Sambungmacan hominids.

While no casts of Late Pleistocene Australian fossils were available to be scanned for this particular facet of the study, this trait could easily be examined in every specimen examined for my dissertation. As shown in the previous chapter, each of the fossil

133

Australians studied possess a well-developed postglenoid tubercle, indicating the inferior projection of the TMJ roof anterior to the S-Q fissure. None of the Pleistocene

Australians studied resembled the Ngandong condition in this feature.

In their recent publication on the newly discovered Sambungmacan 4 cranium,

Baba and colleagues (2003) described the location of the S-Q fissure in that specimen as intermediate between the Sangiran and Ngandong conditions. Comparisons of various cranial characteristics place Sambungmacan 4 with the other Sambungmacan and

Ngandong crania (Baba et al., 2003, 2004), and the specimen exhibits many of the derived basicranial characters that are potentially unique to those populations (Baba et al.,

2003, 2004; Durband, 2004). The TMJ in Sambungmacan 4 is very important in that it provides evidence for the evolution of this suite of features, and potentially represents a link between the more primitive Sangiran hominids and the later, more derived

Ngandong and Sambungmacan people. This is consistent with the other non-metric data presented in the previous chapter.

134

Chapter 7: Discussion and Conclusions

The cranial base is considered by many to be one of the more evolutionarily conservative

portions of the cranium, left relatively free to develop without influence from

surrounding structures (Van Limborgh, 1970; Burdi, 1976; Šmahel and Škvařilova,

1988). While further data is needed to confirm this phylogenetic conservatism in the basicranium (Lieberman et al., 2000), previous studies have shown that basal characters are as useful as facial or neurocranial characters for phylogenetic reconstruction (e.g.

Strait, 1998; Lieberman et al., 2000). As such, this region of the skull provides an excellent opportunity to test a number of possible hypotheses regarding the possible evolutionary fate of the Indonesian hominids. As noted earlier in this dissertation, there are a number of models that have been constructed by various workers in an attempt to explain the origins of modern humans. Of course, the most widely publicized (and polarized) theories are the Multiregional hypothesis (e.g. Wolpoff, 1989, 1992, 1999;

Wolpoff et al., 1984; Hawks et al., 2000) and the Replacement hypothesis, which is often referred to as the “Eve” or “Out of Africa” model (e.g. Stringer, 1989, 1992, 1994;

Stringer and Andrews, 1988). More centrist models have also been proposed that envision varying levels of gene flow between resident archaic populations and subsequent invasions of more modern humans. Of these, the Afro-European Sapiens hypothesis of Brauer (1984a, b, 1989, 1992) posits relatively minor interbreeding between these groups, while the Assimilation hypothesis of Smith and colleagues (1989;

Smith, 1992) proposes a more significant contribution by these indigenous archaics. This

135

study provides an opportunity to test the assumptions of not only the Multiregional hypothesis, the stated goal of this work, but also the rest of these theories in turn to see which model (if any) provides the best fit with the current body of evidence.

To achieve this goal, it is first necessary to apply the general assumptions provided for each model to the specific scenario of the Australasian evidence. Thus, for the Multiregional hypothesis to be supported one would expect to see evidence for a transition between the earlier hominids from Sangiran, through the specimens from

Ngandong, to fossil and modern Australians. This transition will be characterized by a group of features peculiar to this region that will unify the lineage relative to other regional lineages (e.g. Thorne and Wolpoff, 1981; Hawks et al., 2000). Conversely, if the

Replacement hypothesis is supported one would anticipate evidence for a morphological break between the archaic hominids (e.g. Sangiran and Ngandong) and the later modern

Australians. In addition, the modern humans sampled from Australia would share a single

“modern” morphological pattern with other groups sampled from throughout the world that would be unlike that seen in the preceding archaic hominids from the region. The

Afro-European Sapiens and/or Assimilation hypotheses would be supported by varying levels of similarity between the archaic and modern populations of Australasia. More trivial levels of continuity would be consistent with the Afro-European Sapiens model, while a more striking level of morphological similitude could be argued as evidence for the Assimilation model (though it might be difficult in practice to differentiate this model from a strict Multiregional interpretation in many instances).

Clearly, the results of this study indicate that a rejection of the strict Multiregional hypothesis is required in this case. None of the features examined support a specific

136

regional link between the hominids from Java and Australia, and several of these characters appear to be autapomorphic for the hominids from Ngandong and

Sambungmacan (and probably Ngawi as well, but until the matrix obscuring the cranial base of that specimen is cleaned the form of many of these characters remains unknown).

These findings run counter to the expectations of a continuity model between the early

Indonesians and later Australians (e.g. Weidenreich, 1943, 1951; Thorne and Wolpoff,

1981; Hawks et al., 2000). Likewise, those models that posit more modest genetic contributions from archaic populations to waves of incoming modern humans are not consistent with the present findings. Even the Afro-European Sapiens hypothesis, which would require a minimum of morphological resemblance, would be difficult to argue based on the evidence examined in this project.

“Genetic swamping” has been argued for the European evidence by many authors, most notably Frayer (1992). Under this scenario, character traits that are allegedly diagnostic of the Neandertals (the suprainiac fossa, H-O foramen, and the axillary border of the scapula) can be shown to appear in relatively higher rates in Early Upper

Paleolithic modern humans before diminishing over time (Frayer, 1992; though the utility of these traits has been questioned by workers like Trinkaus [2004]). This gradual decrease in frequency, rather than an abrupt disappearance of those traits, has been explained as “swamping” of the Neandertal genes in the modern human gene pool

(Frayer, 1992). One could attempt to extend this argument to the Australasian data, and simply theorize that any genetic contribution made by the Ngandong hominids to later modern human groups was relatively small and has subsequently been reduced to the point of being invisible. This is, of course, not outside the realm of possibility. Based on

137

the current evidence, however, this interpretation would seem to be incompatible with the data. I would argue that, unlike the European fossil record, there is no evidence of any transitional morphology in the late Pleistocene/early Holocene record of Java or

Australia. None of the features shown here to be diagnostic of the Ngandong group of fossils have been found outside of that sample, and the morphological pattern that emerges on Australia (and at Wajak) is fully modern in its cranial base morphology. No evidence of any transitional morphologies or diminishing frequencies of certain characters has been identified. While it is certainly possible that evidence to the contrary could be discovered at some point in the future, the current reading of the existing fossil record does not support the type of arguments put forth by Frayer (1992) as being applicable in Australasia.

In the absence of any indications of morphological continuity between the

Ngandong folk and later Australians, a replacement model provides the best fit for these data. As mentioned earlier, a morphological break or gap between the archaic and modern populations in a particular region is a central tenet of any replacement model.

The data presented in this dissertation show clear evidence for such a gap in the

Australasian fossil record. While there is some support for a hypothesis of continuity between the Sangiran fossils and the Ngandong/Sambungmacan specimens, the high number of seemingly autapomorphic features in this latter population would preclude them from any role in the ancestry of the Pleistocene Australians. The subsequent disappearance of these features in every other human group, fossil or modern, implies extinction of the Ngandong hominids and replacement by modern humans (Durband

2002b, 2004).

138

A replacement model is consistent with the genetic data compiled by various workers for the Australian record (e.g. Adcock et al., 2001; Kayser et al., 2001; Redd et al., 2002; Ingman and Gyllensten, 2003). Adcock and colleagues (2001) demonstrated that the so-called “robust” and “gracile” specimens from Australia are accommodated within a single mtDNA lineage. “A heterogenous source population” is one of the possibilities forwarded by Ingman and Gyllensten (2003: 1605) after their analysis of

Australian mtDNA, and even multiple colonization events would be compatible with my results if one assumes waves of fully modern humans. Further, the Y-chromosome work reported earlier (e.g. Kayser et al., 2001; Redd et al., 2002) suggests a population expansion that may have started with only a few hundred individuals. This scenario would clearly support a single replacement event as being responsible for the colonization of Australia.

The timing and nature of this replacement event is unknown at this time. As mentioned earlier, Swisher and colleagues (1996, 1997; Rink et al., 1997) contend that the Ngandong hominids survived until as recently as 27 kyr. This late date would indicate certain contemporaneity with anatomically modern humans, as the colonization of

Australia was underway by at least 38-42 kyr (Bowler et al., 2003). There has been considerable opposition to this timeline, however, with many authorities dismissing the

Ngandong U-series and ESR/EPR dates as potentially technically accurate but having no bearing on the age of the hominids themselves. Grün and Thorne (1997) raise a number of different issues with the Swisher et al. (1996) study. They note differences in fossilization between the hominids and faunal material, with the former being “dark brown and black, dense, and ceramic-like in texture” while the faunal remains are

139

“generally grey with bluish manganese staining, with a crumbly texture,” and contend that this is evidence that these remains probably fossilized in differing environments

(Grün and Thorne, 1997: 1575). Preliminary work by Westaway (2002a: 191) on this problem led him to a similar conclusion, suggesting “that the faunal remains at Ngandong have not had the same length of burial time when compared with the H. erectus fossils as they are not at the same stage of mineral replacement as the human remains.” Westaway and others (2003) have also alleged that the bovid teeth used to obtain the 27 kyr date at

Ngandong probably came from disturbed sediments, largely backfill from the original

Dutch excavations at the site in the 1930’s. These objections cast serious doubt on the conclusions offered by Swisher and colleagues (1996, 1997; Rink et al., 1997).

Alternative timelines for the Ngandong assemblage, their extinction, and the subsequent arrival of modern humans have been offered by other workers. Santa Luca

(1980) felt that the hominids probably dated from the late Middle Pleistocene/early Upper

Pleistocene, and were redeposited in Upper Pleistocene sediments after weathering out of their original locations. Since the hominid assemblage consists predominantly of easily transportable crania that show signs of damage and/or weathering, and the specimens were scattered irregularly throughout the site, an argument could be made that transport has occurred, and, therefore, that the hominid crania are older than the faunal remains discovered near them (Santa Luca, 1980). As mentioned earlier, this argument had been raised by Koenigswald (1951) and reiterated by Grün and Thorne (1997) and Westaway

(2002a) based on the different levels of fossilization observed between the fauna and the hominid fossils. De Vos (1983, 1985; personal communication) and Storm (2000, 2001a, b) also support an earlier date for the Ngandong assemblage based on the biostratigraphic

140

data. The Ngandong fauna, as well as the associated palynological data, is consistent with an open woodland environment present during warm and dry conditions on Java (de Vos,

1983; Storm, 2000). This faunal period ended at approximately 126 kyr, and was followed by a ~40 kyr period of warm and humid conditions characterized by a tropical rainforest fauna, the Punung (Storm, 2000). The Ngandong-Punung transition is characterized by a major faunal turnover event, likely due to a southern migration of tropically adapted species in conjunction with this environmental change (Storm, 2000,

2001a, b). During this time period most of the species characteristic of the Ngandong fauna (including the hominids) disappear and are replaced by modern species, one of which might have been H. sapiens (de Vos, 1985; Storm, 2000, 2001a, b). This scenario is consistent with previous estimates for the age of the Ngandong hominids as late Middle or early Upper Pleistocene (e.g. Santa Luca, 1980; Grün and Thorne, 1997), as well as the extinction and eventual replacement of said hominids. Unlike the scenario envisioned by

Swisher and colleagues (1996), however, this ecological model does not favor an overlap between the Ngandong folk and modern humans. Instead, the last archaic residents of

Java would have been driven extinct by changes in the environment prior to the arrival of modern humans on Java (Storm, 2000, 2001a, b).

Equivocations aside, these models are both consistent with the results obtained in this study indicating a morphological break in Australasia near the end of the Pleistocene.

As such, they provide a starting point for our understanding of hominid evolution on Java and the nature of modern human origins in the area. Presently, it would appear that Java and Australia have little to offer us as far as elucidating the patterns of modernization between archaic and modern forms, though this issue will be revisited again toward the

141

end of this section. Despite this apparent lack of transitional morphologies leading to modern humans, however, the Javan record is still important for our understanding of late

Pleistocene hominid evolution. A number of interesting questions remain pertinent for this fascinating fossil sample.

One of these questions surrounds the taxonomic attribution of the Ngandong and

Sambungmacan hominids themselves. While most workers would place these specimens in H. erectus (e.g. Santa Luca, 1977, 1980; Rightmire, 1981, 1990, 1994, 2004; Antón,

2001, 2002, 2003), others prefer to place them in the pseudo-taxon of archaic H. sapiens

(e.g. Bräuer, 1992; Bräuer and Mbua, 1992). The latter assignment is generally defended by claims that the Ngandong fossils exhibit traits that are derived in the direction of modern humans and are therefore transitional between earlier Indonesian erectines and later modern Australians. This assertion has been called into question by various authors

(e.g. Rightmire, 1990, 1994, 2004; Stringer, 1992) as well as the present work. Based on the evidence gleaned from the current project, however, it can be argued that neither species would be appropriate for the Ngandong and Sambungmacan hominids.

To begin, cranial capacities in this group of fossils do not appear to be markedly higher than in other middle and late Pleistocene hominid populations. Brain size ranges from 1,013 cm2 (Ngandong 7) to 1,251 cm2 (Ngandong 6) (Holloway, 1980). Meanwhile,

Zhoukoudian X, a specimen that is widely accepted as an example of Homo erectus

(Santa Luca, 1980; Kramer, 1993; Antón, 2002, 2003; Rightmire, 2004) has a cranial capacity of 1,225 cm2 (Weidenreich, 1943). While the average cranial capacity for the

Ngandong sample may be slightly higher than that found at Zhoukoudian, this difference

142

is certainly not marked enough to imply that the Ngandong and Sambungmacan hominids have become significantly advanced in terms of their level of encephalization.

Non-metric examinations of various cranial superstructures have also failed to provide evidence of any incipient modernity in the Ngandong and Sambungmacan fossils. Comparisons of brow ridge form between the Ngandong series and the Willandra

Lakes hominids (including WLH 50) by Webb (1989) indicate a very different pattern of supraorbital development between those two samples. The typical pattern of brow ridge development in early Australians, including not only the Willandra Lakes sample but also

Kow Swamp and Coobool Creek, is typified by the heaviest development occurring at the medial portion of the brow around the superciliary ridges, with the lateral section being the most weakly developed (Webb, 1989). This arrangement is the opposite of that seen in Ngandong, where the knob-like lateral trigone is the thickest portion of the supraorbital torus and the medial portion is the thinnest (Webb, 1989). At the rear of the cranium, the

Ngandong hominids are characterized by a pronounced excavation of the nuchal area, with the inferior edge of the occipital torus undercut by the large nuchal muscles (Santa

Luca, 1980; Rightmire, 1990). This exceptional hollowing under the occipital torus is considerably more pronounced than that seen in the earlier hominids from Sangiran

(Rightmire, 1990), and is likewise not seen on any Pleistocene Australian crania, each of which exhibit a typically modern pattern of muscle markings on the nuchal plane of the occipital. The Ngandong and Sambungmacan hominids express angular tori (except perhaps Sambungmacan 3), another archaic trait typical in Pleistocene hominids

(Rightmire, 1990; Baba et al., 2003). The squamosal suture in the Ngandong and

Sambungmacan series is low and straight, giving the temporal squama a triangular shape

143

(Santa Luca, 1980; Rightmire, 1990), and is unlike the high and rounded temporal squama seen in later hominids. The Ngandong and Sambungmacan hominids are much like earlier specimens from Sangiran in their sagittal contour (Rightmire, 1990, 1994), and, as in all earlier hominids, their maximum cranial breadth is located quite inferiorly on the skull near the supramastoid crests (Baba et al., 2003). Thus, it would be difficult to point to any significant indications of modernization in the Ngandong and

Sambungmacan crania that would signal a transition between the Sangiran hominids and modern humans.

Other aspects of cranial vault form, however, do hint at some level of differentiation within the late Pleistocene record of Java. Several metric studies have suggested that the hominids from Ngandong had evolved beyond the level of other specimens widely regarded as Homo erectus. Santa Luca (1977: 67) noted that cranial vault shape in Ngandong is “at the advanced end of the within group H. erectus trends.”

These distinctions are manifested primarily in an occipital plane that is generally longer than the nuchal plane (in most, but not all Ngandong crania), and a maximum cranial breadth that is slightly above the level of the supramastoid crests (Santa Luca, 1980).

However, Santa Luca (1980: 120) is careful to note that this more derived morphological pattern is “developed over a structure that is typically H. erectus” and he did not feel that it was in any way intermediate between earlier H. erectus and H. sapiens. Grimaud-Hervé

(1986) reached a similar conclusion after an analysis of the parietal bones of Indonesian erectines. She found that Ngandong and Sambungmacan 1 exhibit temporal lines that are less accentuated and located more inferiorly on the parietals in conjunction with a maximum cranial breadth that is slightly higher and more anterior than in the Sangiran

144

hominids (Grimaud-Hervé, 1986). Kidder (1996: 151) reported similar results, finding

that the Ngandong fossils “[demonstrate] metric patterns more advanced than Homo

erectus in the shape of the vault.” Here again, however, this more progressive metric

pattern is not similar to that found in modern humans (Kidder, 1996). Each of these

previous studies indicate that the cranial vaults in the Ngandong sample have become

derived relative to the hominids from Sangiran, yet are not evolving in the direction of

modern humans.

The limited postcranial evidence available from Ngandong also exhibits traits that

serve to differentiate this sample. Nelson (1995: 140) notes that “[t]he two Ngandong

tibiae are remarkable in terms of their absolutely thick proximal cortical diameters and

elevated cortical indices.” He further states that “[this] pattern … appears to distinguish

the Ngandong tibiae from all other tibiae for which comparable data currently exist”

(Nelson, 1995: 140).

Thus, from the evidence examined both in this project as well as others (e.g.

Macintosh and Larnach, 1972; Larnach and Macintosh, 1974; Santa Luca, 1980;

Grimaud-Hervé, 1986; Rightmire, 1994; Nelson, 1995; Kidder, 1996) it is clear that

while in many ways the hominids from Ngandong and Sambungmacan are still quite

primitive relative to modern H. sapiens, they also display a number of significant dissimilarities from earlier Javan forms. Zeitoun (2002; Widianto and Zeitoun, 2003) has suggested the resurrection of the species Homo soloensis to encompass the hominids from Ngandong, Sambungmacan, and Ngawi. This solution is based on the premise that this particular group of specimens represents “a particular geographically and chronologically restricted human group defined by unique morphological and biometrical

145

data” (Widianto and Zeitoun, 2003: 349). The resurrection of Homo soloensis is supported by the present study, which has highlighted a number of autapomorphic features on the cranial base that are both distinctive for and diagnostic of this group. This combination of unique traits, in conjunction with the previously cited studies expounding the progressive metric features of the Ngandong, Sambungmacan, and Ngawi cranial vaults (Santa Luca, 1980; Grimaud-Hervé, 1986; Kidder, 1996; Widianto and Zeitoun,

2003), provides a strong case for separating these crania from Homo erectus.

The following scenario for the course of hominid evolution in Southeast Asia and

Australia is suggested by these data. Java was originally populated by early hominids at approximately 1.5 mya (Larick et al., 2001; Morwood et al., 2003). Current evidence suggests that these hominids were quite robust, and they are represented by the Sangiran

4 cranium and perhaps by Sangiran 27 and 31 as well (Kaifu et al., in press). Later hominids, those recovered from within or above the Grenzbank at Sangiran (e.g. Sangiran

2, 12, 14, 17), date to greater than 1.02 mya (Larick et al., 2001) and are typically placed in Homo erectus (e.g. Rightmire, 1990; Antón, 2002, 2003). During this time period Java is intermittently connected with the Asian mainland (Aziz et al., 1995), and the island is characterized by an open woodland environment (de Vos, 1983). In later Ngandong times the fauna shows signs of endemism (van den Bergh, 1999; van den Bergh et al., 2001), indicating a long period of isolation from the mainland. A long period of endemism would be consistent with the appearance of autapomorphic features in the Ngandong and

Sambungmacan crania. Assuming a conservative Middle Pleistocene date for Ngandong

(e.g. Santa Luca, 1980; Grün and Thorne, 1997), humans had occupied Java for approximately 1.25 million years. Much of this time the island was separated from the

146

mainland (Aziz et al., 1995), and even when it was connected to Asia it remained one of the most peripheral areas of the known range of H. erectus. During this length of time, in virtual (if not actual) allopatry, genetic drift acted upon the early inhabitants of Java and culminated in the evolution of Homo soloensis. This species was extant until approximately 126 kyr, when a faunal turnover commenced following a climatic shift to more humid conditions (Storm 2000, 2001a, b). Modern humans arrived on Java possibly as early as Punung times (de Vos, 1985) but certainly by the early Holocene at Wajak

(Storm, 1995; Shutler et al., in press). Meanwhile, Australia was colonized by at least 38-

42 kyr (Bowler et al., 2003), probably via the Indonesian archipelago.

This, then, would seem to suggest extinction of all Javan hominid lineages that were present prior to about 126 kyr. This may not be the case, however. Though the present study provides strong evidence for the differentiation and extinction of the

Ngandong and Sambungmacan hominids, this does not necessarily condemn the Sangiran assemblage to a similar fate. Thorne and Wolpoff (1981) advocated continuity between the Sangiran hominids and later Australians, namely the Kow Swamp sample, and emphasized the differences between Sangiran 17 and the Ngandong fossils. Kramer

(1989, 1991) has also shown evidence for morphological continuity between the Sangiran mandibular sample and modern Australians. Thus, it is still possible that Ngandong could represent only one of at least two fossil morphs on Java during the Pleistocene, and that one (or more) of these other types of hominids contributed to the gene pool of modern

Australians.

Baba and colleagues (1998, 2000, 2004; Aziz et al., 1996) dispute evidence of continuity between Sangiran and Kow Swamp. As mentioned earlier in this dissertation,

147

their reconstruction of Sangiran 17 has cast doubt upon many of the facial features

thought to indicate continuity in Australasia by Thorne and Wolpoff (1981). In addition,

another of these characters, flatness of the frontal bone, has been influenced by artificial

cranial deformation in several Pleistocene samples from Australia, including Kow

Swamp, Cohuna, and Coobool Creek (Brown, 1981, 1989; Antón and Weinstein, 1999).

The studies by Kramer (1989, 1991) have not been similarly assailed. This is not to say

that aspects of these studies are not questionable, however. While Kramer (1989, 1991)

provides sound results for the samples utilized, namely the Sangiran H. erectus mandibles and modern samples from Kenya and Australia, to date no studies have incorporated the

Pleistocene mandibles from Australia. Therefore, it is unknown whether these earliest inhabitants of Australia would fit the pattern predicted by the Multiregional model.

The weight of the evidence presented strongly suggests that the Middle/Upper

Pleistocene populations of Java, represented by the crania from Ngandong,

Sambungmacan, and Ngawi, went extinct after undergoing a speciation event. These specimens are differentiated from earlier specimens at Sangiran by differences in craniometric pattern, a number of autapomorphies in the cranial base, and the distribution of cortical bone thickness in the tibia. In short, every skeletal element for which we have evidence shows evolution away from the patterns present at Sangiran and in later modern humans. These groups of Homo soloensis did not make any contributions to the gene pools of later Australians, and quite possibly did not even encounter any other species of hominid. Despite the eloquence of Weidenreich’s (1943: 276) original statement, it does not accommodate our knowledge of the fossil evidence. While we may state with

148

reasonable certainty that “[t]here is an almost continuous line leading from

Pithecanthropus through Homo soloensis,” that is where the story must come to an end.

149

References Cited

150

Abbie, AA (1951) The Australian Aboriginal. Oceania 22: 91-100.

Abbie, AA (1963) Physical characteristics of Australian Aborigines. In (Shiels H, Ed.) Australian Aboriginal Studies. Oxford: Oxford University Press. Pgs. 89-107.

Abbie, AA (1966) Physical characteristics. In (Cotton BC, Ed.) Aboriginal Man in south and central Australia, Part 1. Adelaide: Government Printers. Pgs. 9-45.

Adcock GJ, Dennis ES, Easteal S, Huttley GA, Jermiin LS, Peacock WJ, and Thorne A (2001) Mitochondrial DNA sequences in ancient Australians: Implications for modern human origins. Proceedings of the National Academy of Sciences 98: 537-542.

Andrews, P (1984) An alternative interpretation of the characters used to define Homo erectus. Courier Forsch. Inst. Senckenberg 69: 167-175.

Antón, SC (2001) Cranial evolution in Asian Homo erectus: The Ngandong hominids. Geological Research Development Centre Special Publication No. 27: 39-46.

Antón, SC (2002) Evolutionary significance of cranial variation in Asian Homo erectus. Am. J. Phys. Anthropol. 118: 301-323.

Antón, SC (2003) Natural history of Homo erectus. Yearbook of Physical Anthropology 37: 126-170.

Antón SC, Marquez S and Mowbray K (2002) Sambungmacan 3 and cranial variation in Asian Homo erectus. Journal of Human Evolution 43: 555-562.

Antón, SC and Weinstein, KJ (1999) Artificial cranial deformation and fossil Australians revisited. Journal of Human Evolution 36: 195-209.

Aziz, F (1990) Pleistocene mammal faunas of Sulawesi and their bearings on paleozoogeography. Unpublished Ph.D dissertation. Kyoto University, Japan.

Aziz, F (2002) New discovery of a hominid skull from Cemeng, Sambungmacan, Central Java: an announcement. Jurnal Geologi dan Sumberdaya Mineral 12: 2-7.

Aziz F, Sondaar PY, de Vos J, Van Den Bergh GD, and Sudijono (1995) Early dispersal of man on islands of the Indonesian archipelago: facts and controls. Anthropological Science 103: 349-368.

Aziz F, Baba H, and Watanabe N (1996) Morphological study on the Javanese Homo erectus Sangiran 17 based upon the new reconstruction. GRDC Paleontology Series No. 8: 11-25.

151

Baba H, Aziz F, and Narasaki S (1998) Restoration of head and face in Javanese Homo erectus Sangiran 17. Bull. Natn. Sci. Mus., Tokyo, Series D 24: 1-8.

Baba H, Aziz F, and Narasaki S (2000) Restoration of the face of Javanese Homo erectus Sangiran 17 and re-evaluation of regional continuity in Australasia. Acta Anthropologica Sinica Volume 19 Supplement: 57-63.

Baba H, Aziz F, Kaifu Y, Suwa G, Kono RT, and Jacob T (2003) Homo erectus calvarium from the Pleistocene of Java. Science 299: 1384-1388.

Baba H, Aziz F, Kaifu Y, Kono T, and Jacob T (2004) Morphology of Sambungmacan 4 skull and the evidence of discontinuity in Australasia. American Journal of Physical Anthropology Supplement 38: 57.

Bartstra, GJ (1983) Comments and reply on “The fauna from Trinil, type locality of Homo erectus: a reinterpretation.” Comment I: The vertebrate-bearing deposits of Kedungbrubus and Trinil, Java, Indonesia. Geologie en Mijnbouw 62: 329-336.

Bartstra, GJ (1987) Late Homo erectus or Ngandong Man of Java. Paleohistoria 29: 1-7.

Bartstra GJ, Soegondho S, and van der Wijk A (1988) Ngandong man: Age and artifacts. Journal of Human Evolution: 17: 325-37.

Birdsell, JB (1949) The racial origin of the extinct Tasmanians. Records of the Queen Victoria Museum 2: 105-122.

Birdsell, JB (1950) Some Implications of the Genetical Concept of Race in Terms of Spatial Analysis. Cold Spring Harbor Symposia on Biology 15: 259-314.

Birdsell, JB (1967) Preliminary data on the trihybrid origin of the Australian Aborigines. Archaeology and Physical Anthropology in Oceania 2: 100-155.

Boedhihartono (1998) A new Homo erectus finding. Antropologi 54: 121-125.

Boedhihartono (2001) A new Homo erectus finding: a preliminary observation did Homo erectus undergo a sapientization before giving rise to Homo erectus soloensis? . In (Simanjuntak T, Prasetyo B, and Handini R, Eds.) Sangiran: Man, Culture and Environment in Pleistocene Times. Jakarta: Yayasan Obor Indonesia. Pgs. 130- 139.

Bowdler, S (1993) Sunda and Sahul: A 30 Kyr BP culture area? In (Smith MA, Spriggs M, and Fankhauser B, eds) Sahul in Review: Pleistocene archaeology in Australia, New Guinea and Island Melanesia. Occasional Papers in Prehistory 24. Australian National University: Department of Prehistory. Pgs. 60-70.

152

Bowler, JM (1970) Lake Nitchie skeleton – Stratigraphy of the burial site. Archaeology and Physical Anthropology in Oceania 5: 102-113.

Bowler, JM (1976) Recent developments in reconstructing late Quaternary environments in Australia. In (Kirk, RL and Thorne, AG, Eds) The Origin of the Australians. Canberra: Australian Inst. For Aboriginal Studies. Pgs. 55-77.

Bowler, JM (1998) Willandra Lakes revisited: environmental framework for human occupation. Archaeology in Oceania 33: 120-155.

Bowler JM, Johnston H, Olley JM, Prescott JR, Roberts RG, Shawcross W, and Spooner NA (2003) New ages for human occupation and climatic change at Lake Mungo, Australia. Nature 421: 837-840.

Bowler JM, Jones R, Allen H and Thorne AG (1970) Pleistocene human remains from Australia: a living site and human cremation from Lake Mungo, western New South Wales. World Archaeology 2: 39-60.

Bowler, JM and Magee, JW (2000) Redating Australia’s oldest human remains: a skeptic’s view. Journal of Human Evolution 38: 719-726.

Bowler, JM and Price, DM (1998) Luminescence dates and stratigraphic analyses at Lake Mungo: review and new perspectives. Archaeology in Oceania 33: 156-168.

Bowler, JM and Thorne, AG (1976) Human remains from Lake Mungo: discovery and excavation of Lake Mungo III. In (Kirk, RL and Thorne, AG, Eds.) The Origin of the Australians. New Jersey: Humanities Press. Pgs. 127-138.

Bräuer, G (1984a) A craniological approach to the origin of anatomically modern Homo sapiens in Africa and implications for the appearance of modern Europeans. In Smith and Spencer (eds.) The Origins of Modern Humans: A Survey of the Fossil Evidence. New York: AR Liss. Pgs. 327-410.

Bräuer, G (1984b) The “Afro-European sapiens-hypothesis,” and hominid evolution in East Asia during the late Middle and Upper Pleistocene. Courier Forsch. Inst. Senckenberg 69: 145- 165.

Bräuer, G (1989) The evolution of modern humans: A comparison of the African and non-African evidence. In Mellars and Stringer (eds.) The Human Revolution. Edinburgh: Edinburgh Univ. Press. Pgs. 123-54.

Bräuer, G (1992) Africa’s place in the evolution of Homo sapiens. In Bräuer and Smith (eds.) Continuity or Replacement: Controversies in Homo sapiens Evolution. Rotterdam: AA Balkema. Pgs. 83-98.

153

Bräuer, G and Mbua, E (1992) Homo erectus features used in cladistics and their variability in Asian and African hominids. Journal of Human Evolution 22: 79- 108.

Brown, P (1981) Artificial cranial deformation: A component in the variation in Pleistocene Australian Aboriginal crania. Archaeology in Oceania 16:156-167.

Brown, P (1987) Pleistocene homogeneity and Holocene size reduction: the Australian human skeletal evidence. Archaeology in Oceania 22: 41-71.

Brown, P (1989) Coobool Creek. A morphological and metrical analysis of the crania, mandibles and dentitions of a prehistoric Australian human population. Terra Australis, 13. Department of Prehistory, Australian National University, Canberra.

Brown, P (1992) Recent Human Evolution in East Asia and Australasia. Philosophical Transactions of the Royal Society of London, Series B 337: 235-242.

Brown, P (1994a) A flawed vision: sex and robusticity on King Island. Australian Archaeology 38: 1-7.

Brown, P (1994b) Human skeletons. In (Horton D, Ed.) The Encyclopedia of Aboriginal Australia. Canberra: Australian Aboriginal Studies.

Brown, P (1995) Still flawed: a reply to Pardoe (1994) and Sim and Thorne (1994). Australian Archaeology 41: 26-29.

Brown, P (2000) Australian Pleistocene variation and the sex of Lake Mungo 3. Journal of Human Evolution 38: 743-749.

Brown, P and Gillespie, R (2000) Mungo claims unsubstantiated. Australasian Science 21: 26-27.

Bulbeck, FD (2001) Robust and gracile Australian Pleistocene crania: The tale of the Willandra Lakes. In (Simanjuntak T, Prasetyo B, and Handini R, Eds.) Sangiran: Man, Culture and Environment in Pleistocene Times. Jakarta: Yayasan Obor Indonesia. Pgs. 60-106.

Burdi, AR (1976) Early development of the human basicranium: Its morphogenic controls, growth patterns, and relations. In Bosma (ed.) Development of the Basicranium. DHEW Publ. No. (NIH) 76-989. Bethesda: NIH. Pgs. 81-90.

Cheverud, JM (1995) Morphological integration in the saddle back tamarin. Am. Nat. 145: 63-89.

154

Corner, EM (1898) The processes of the occipital and mastoid regions of the skull. Journal of Anatomy and Physiology 30: 386-389.

Corruccini, RS (1974) An examination of the meaning of cranial discrete traits for human skeletal biological studies. American Journal of Physical Anthropology 40: 425- 446.

Curtis, GH (1981) Man’s immediate forerunners: Establishing a relevant time scale in anthropological and archaeological research. Philosophical Transactions of the Royal Society of London, Series B 292: 7-20. de Vos, J (1983) The Pongo faunas from Java and Sumatra and their significance for biostratigraphical and paleoecological interpretation. Proc. Kon. Ned. Akad. Wetensch. B 86: 417-425.

de Vos, J (1985) Faunal stratigraphy and correlation of the Indonesian hominid sites. In (Delson E, ed) Ancestors: The Hard Evidence. New York: AR Liss. Pgs. 215-220.

de Vos, J (1987) The environment of Homo erectus from Trinil H.K. Proceedings of the 2nd International Congress of Human Paleontology, Turin, Italy. Pgs. 225-229.

de Vos, J (1995) The migration of Homo erectus and Homo sapiens in south-east Asia and the Indonesian archipelago. In (J.R.F. Bower & S. Sartono, Eds) Human Evolution in its Ecological Context, Vol. 1, Evolution and Ecology of Homo erectus, pp. 239-260. Leiden: Pithecanthropus Centennial Foundation.

de Vos J, Sartono S, Hardja-Sasmita S, and Sondaar PY (1982) The fauna from Trinil, type locality of Homo erectus; a reinterpretation. Geologie en Mijnbouw 61: 207- 211. de Vos, J and Sondaar, PY (1994) Dating hominid sites in Indonesia. Science 266: 1726- 1727.

de Vos J, Sondaar PY, Van Den Bergh GD, and Aziz F (1994) The Homo bearing deposits of Java and its ecological context. Cour. Forsch. Inst. Senckenberg 171: 129-140.

Delson E, Harvati K, Reddy D, Marcus LF, Mowbray K, Sawyer GJ, Jacob T, Márquez S (2001) The Sambungmacan 3 Homo erectus calvaria: a comparative morphometric and morphological analysis. The Anatomical Record 262: 380-397.

Dubois, E (1923) Phylogenetic and ontogenetic increase of the volume of the brain in vertebrata. Proceedings of the Koninklijke Akademie van Wetenschappen 25: 230-255.

155

Dubois, E (1924) On the brain quantity of specialized genera of mammals. Proceedings of the Koninklijke Akademie van Wetenschappen 27: 430-437.

Dubois, E (1928) The law of the necessary phylogenetic perfection of the psychoencephalon. Proceedings of the Koninklijke Akademie van Wetenschappen te Amsterdam 31: 304-314.

Dubois, E (1935) On the gibbon-like appearance of Pithecanthropus erectus. Proceedings of the Koninklijke Akademie van Wetenschappen te Amsterdam 38: 578-585.

Dubois, E (1936) Racial identity of Homo soloensis Oppenoorth (including Homo modjokertensis von Koenigswald) and Sinanthropus pekinensis Davidson Black. Proceedings of the Koninklijke Akademie van Wetenschappen te Amsterdam 39: 1180-1185.

Dubois, E (1937a) Early Man In Java And Pithecanthropus Erectus. In MacCurdy, GG (ed.) Early Man. New York: JB Lippincott Co. Pgs. 315-322.

Dubois, E (1937b) On the fossil human skulls recently discovered in Java and Pithecanthropus erectus. Man 37: 1-7.

Dubois, E (1938a) The mandible recently described and attributed to the Pithecanthropus by G.H.R von Koenigswald, compared with the mandible of Pithecanthropus erectus described in 1924 by Eug. Dubois. Proceedings of the Koninklijke Akademie van Wetenschappen te Amsterdam 41: 139-147.

Dubois, E (1938b) On the fossil human skull recently described and attributed to Pithecanthropus erectus by G.H.R. von Koenigswald. Proceedings of the Koninklijke Akademie van Wetenschappen 41: 380-386.

Dubois, E (1940a) The fossil human remains discovered in Java by Dr. G.H.R. von Koenigswald and attributed by him to Pithecanthropus erectus, in reality remains of Homo wadjakensis (syn. Homo soloensis). Proceedings of the Koninklijke Akademie van Wetenschappen 43: 494-496.

Dubois, E (1940b) The fossil human remains discovered in Java by Dr. G.H.R. von Koenigswald and attributed by him to Pithecanthropus erectus, in reality remains of Homo wadjakensis (syn. Homo soloensis) (Continuation). Proceedings of the Koninklijke Akademie van Wetenschappen 43: 842-851.

Dubois, E (1940c) The fossil human remains discovered in Java by Dr. G.H.R. von Koenigswald and attributed by him to Pithecanthropus erectus, in reality remains of Homo wadjakensis (syn. Homo soloensis) (Conclusion). Proceedings of the Koninklijke Akademie van Wetenschappen 43: 1268-1275.

156

Durband, AC (1998) The Cranial Base of the Ngandong Hominids: Implications for Modern Human Origins. American Journal of Physical Anthropology Supplement 26: 79.

Durband, AC (2002a) The Squamotympanic Fissure in the Ngandong and Sambungmacan Hominids: A Reply to Delson et al. The Anatomical Record 266: 138-141.

Durband, AC (2002b) The View from Down Under: A Test of the Multiregional Hypothesis of Modern Human Origins using the Basicranial Evidence from Australasia. Invited paper presented at the 17th Congress of the Indo-Pacific Prehistory Association, Taipei, Taiwan, September 9-15.

Durband, AC (2004) A test of the multiregional hypothesis of modern human origins using the basicranial evidence from Southeast Asia and Australia. American Journal of Physical Anthropology Supplement 38: 90-91.

Frayer, DW (1992) The persistence of Neanderthal features in post-Neanderthal Europeans. In (Bräuer G, Smith FH, eds) Continuity or Replacement: Controversies in Homo sapiens evolution. Rotterdam: AA Balkema Press. Pgs. 179-188.

Freedman L (1985) Human skeletal remains from Mossgiel, N.S.W. Archaeology in Oceania 20: 21-31.

Freedman, L and Lofgren, M (1979) The Cossack skull and a dihybrid origin of the Australian Aborigines. Nature 282: 298-300.

Fullagar RLK, Price DM, and Head LM (1996) Early human occupation of northern Australia: Archaeology and thermoluminescence dating of Jinmium rock-shelter, Northern Territory. Antiquity 70: 751-773.

Gillespie, R (1997) Burnt and unburnt carbon: dating charcoal and burnt bone from the Willandra Lakes, Australia. Radiocarbon 39: 225-236.

Gillespie, R (1998) Alternative timescales: a critical review of Willandra Lakes dating. Archaeology in Oceania 33: 169-182.

Gillespie, R and Roberts, RG (2000) On the reliability of age estimates for human remains at Lake Mungo. Journal of Human Evolution 38: 727-732.

Gray, H (1985) Gray’s Anatomy. Thirtieth American Edition (Clemente, C, Ed.) Philadelphia: Lea & Febiger.

157

Grimaud-Hervé, D (1986) The parietal bone of Indonesian Homo erectus. Human Evolution 1: 167-182.

Grimaud-Hervé D, Sartono S, Widianto H, and Djubiantono T (1998) The fossil hominid from Ngawi. Bulletin of the Indo-Pacific Prehistory Association 17: 42.

Grün R, Spooner NA, Thorne A, Mortimer G, Simpson JJ, McCulloch MT, Taylor L, and Curnoe D (2000) Age of the Lake Mungo 3 skeleton, reply to Bowler & Magee and to Gillespie & Roberts. Journal of Human Evolution 38: 733-741.

Grün, R and Thorne, A (1997) Dating the Ngandong humans. Science 276: 1575.

Habgood, PJ (1986) The origin of the Australians: a multivariate approach. Archaeology in Oceania 21: 130-137.

Hawks J, Oh S, Hunley K, Dobson S, Cabana G, Dayalu P, and Wolpoff MH (2000) An Australasian test of the recent African origin theory using the WLH-50 calvarium. Journal of Human Evolution 39: 1-22.

Hauser, G and De Stefano, GF (1989) Epigenetic Variants of the Human Skull. Stuttgart: E. Schweizerbart’sche Verlagsbuchhandlung.

Holloway, RL (1980) Indonesian “Solo” (Ngandong) endocranial reconstructions: preliminary observations and comparisons with Neandertal and Homo erectus groups. American Journal of Physical Anthropology 53: 285-295.

Hooijer, DA (1983) Comment II: Remarks upon the Dubois collection of fossil mammals from Trinil and Kedungbrubus in Java. Geologie en Mijnbouw 62: 337-338.

Howells, WW (1969) Criteria for selection of osteometric dimensions. American Journal of Physical Anthropology 30: 451-458.

Howells, WW (1973a) Cranial Variation in Man. Papers of the Peabody Museum of Archaeology and Ethnology, Vol. 67, Harvard Univ., Cambridge.

Howells, WW (1973b) The Pacific Islanders. New York: Charles Scribner’s Sons

Howells, WW (1977) The Sources of Human Variation in Melanesia and Australia. In: Allen J, Golson J, and Jones R, (Eds.) Sunda and Sahul: Prehistoric Studies in Southeast Asia, Melanesia, and Australia. New York: Academic Press. Pgs 169- 186.

Hyodo M, Watanabe N, Sunata W, Susanto EE, and Wahyono H (1993) Magnetostratigraphy of hominid fossil bearing formations in Sangiran and Mojokerto, Java. Anthropological Science 101: 157-186.

158

Ingman, M and Gyllensten, U (2003) Mitochondrial genome variation and evolutionary history of Australian and New Guinean aborigines. Genome Research 13: 1600- 1606.

Ishida, H and Dodo, Y (1993) Nonmetric cranial variation and the populational affinities of the Pacific peoples. American Journal of Physical Anthropology 90: 49-57.

Jacob, T (1969) Some problems pertaining to the racial history of the Indonesian Region. Utrecht: Drukkerij Neerlandia.

Jacob, T (1972a) New hominid finds in Indonesia and their affinities. Mankind 8: 176- 181.

Jacob, T (1972b) The problem of head-hunting and brain-eating among Pleistocene men in Indonesia. Archaeology and Physical Anthropology in Oceania 7: 81-91.

Jacob, T (1975) Morphology and paleoecology of early man in Java. In (Tuttle R, ed) Paleoanthropology, Morphology, and Paleoecology. Mouton: The Hague. Pgs. 311-25.

Jacob, T (1976) Early populations in the Indonesian region. In (Kirk RL and Thorne AG, Eds) The origin of the Australians. Human Biology Series No. 6, Australian Institute for Aboriginal Studies. New Jersey: Humanities Press.

Jacob, T (1978a) The puzzle of Solo Man. Modern Quaternary Research in Southeast Asia 4: 31-40.

Jacob, T (1978b) New finds of lower and middle Pleistocene hominines from Indonesia and an examination of their antiquity. In (Ikawa-Smith F, Ed.) Early Paleolithic in South and East Asia. Paris: Mouton Publ. Pgs. 13-22.

Jacob, T (1979) Hominine evolution in south East Asia. Archaeology and Physical Anthropology in Oceania 14: 1-10.

Jacob, T (1981) Solo Man and Peking Man. In (Sigmon and Cybulski, eds.) Homo erectus: Papers in Honor of Davidson Black. Toronto: Univ. of Toronto Press. Pgs. 87-104.

Jacob, T (1984) The fossil skull cap from Sambungmacan and its implication for human evolution. Berkala Bioanthropologi Indonesia 5: 19-27.

Jacob, T (2001) Biological aspects of Homo erectus through the time-space continuum. In (Simanjuntak T, Prasetyo B, and Handini R, Eds.) Sangiran: Man, Culture and Environment in Pleistocene Times. Jakarta: Yayasan Obor Indonesia. Pgs. 19-23.

159

Jones, FW (1934) Tasmania’s Vanished Race. Australian Broadcasting Commission. Pgs. 1-32.

Kaifu Y, Baba H, Aziz F, Indriati E, Schrenk F and Jacob T (in press) Taxonomic affinities and evolutionary history of the early Pleistocene hominids of Java: dentognathic evidence. American Journal of Physical Anthropology.

Kayser M, Brauer S, Weiss G, Schiefenhovel W, Underhill PA, and Stoneking M (2001) Independent histories of human Y chromosomes from Melanesia and Australia. American Journal of Human Genetics 68: 173-190.

Keith, A (1925) The Antiquity of Man. 2nd edition. London: Williams and Norgate, Ltd.

Kennedy, GE (1991) On the autapomorphic traits of Homo erectus. Journal of Human Evolution 20: 375-412.

Kidder, JH and Durband, AC (2000) The question of speciation in Homo erectus revisited I: the metric evidence. American Journal of Physical Anthropology Supplement 30: 195.

Kidder, JH and Durband, AC (2004) A re-evaluation of the metric diversity within Homo erectus. Journal of Human Evolution 46: 297-313.

Koenigswald, GHR von (1937) A review of the stratigraphy of Java and its relations to early man. In (MacCurdy GG, ed.) Early Man. Pgs. 23-33.

Koenigswald, GHR von (1951) Introduction. In: Weidenreich, F. Morphology of Solo Man. Anthropological Papers of the American Museum of Natural History 43: 205-90.

Koenigswald, GHR von (1956) Meeting Prehistoric Man. New York: Thames and Hudson.

Koenigswald, GHR von and Weidenreich, F (1939) The relationship between Pithecanthropus and Sinanthropus. Nature 144: 926-929.

Kondo M, Matsu’ura S, Fujimori E, Sawatari H, and Haraguchi H (1995) A multielement approach to the relative dating of Java man fossils from Sangiran. Anthropological Science 103: 161.

Konigsberg LW, Kohn LAP, and Cheverud JM (1993) Cranial deformation and nonmetric trait variation. American Journal of Physical Anthropology 90: 35-48.

Kramer, A (1989) The Evolutionary and Taxonomic Affinities of the Sangiran Mandibles of Central Java, Indonesia. Unpubl. Ph.D. Dissertation, University of Michigan.

160

Kramer, A (1991) Modern human origins in Australasia: Replacement or evolution? American Journal of Physical Anthropology 86:455-473.

Kramer, A (1993) Human taxonomic diversity in the Pleistocene: Does Homo erectus represent multiple hominid species? American Journal of Physical Anthropology 91:161-171.

Kramer, A (1994) A critical analysis of claims for the existence of Southeast Asian australopithecines. Journal of Human Evolution 26:3-21.

Kramer, A and Konigsberg, LW (1994) The phyletic position of Sangiran 6 as determined by multivariate analyses. Courier Forschungsinstitut Senckenberg 171:105-114.

Krantz, GS (1975) An explanation for the diastema of Javan erectus skull IV. In (Tuttle, RH, ed.) Paleoanthropology, Morphology, and Paleoecology. The Hague: Mouton. Pgs. 361-372.

Krantz, GS (1994) The palate of skull Sangiran 4 from Java. Courier Forsch. Inst. Senckenberg 171: 69-74.

Laitman, JT and Tattersall, I (2001) Homo erectus newyorkensis: An Indonesian fossil rediscovered in Manhattan sheds light on the middle phase of human evolution. The Anatomical Record 262: 341-434.

Larick R, Ciochon RL, and Zaim Y (1999) Fossil farming in Java. Natural History 108: 54-57.

Larick R, Ciochon RL, Zaim Y, Sudijono, Suminto, Rizal Y, Aziz F, Reagan M, and Heizler M (2001) Early Pleistocene 40Ar/39Ar ages for Bapang formation hominids, Central Jawa, Indonesia. Proceedings of the National Academy of Science 98: 4866-4871.

Larnach SL (1974) The origin of the Australian Aboriginal. Archaeology and Physical Anthropology in Oceania 9: 206-213.

Larnach SL and Macintosh NWG (1974) A comparative study of Solo and Australian Aboriginal crania. In (Elkin, AP and Macintosh, NWG, eds.) Grafton Elliot Smith: The man and his work. Adelaide: Griffin Press. Pgs. 95-102.

Le Gros Clark, WE (1937) The status of Pithecanthropus. Man 37: 60-62.

Lieberman DE, Ross CF, and Ravosa MJ (2000) The primate cranial base: ontogeny, function, and integration. Yearbook of Physical Anthropology 43: 117-169.

161

Long VT, de Vos J, and Ciochon RL (1996) The fossil mammal fauna of the Lang Trang Caves, Vietnam, compared with southeast Asian fossil and recent mammal faunas: the geographical implications. Bulletin of the Indo-Pacific Prehistory Association 14(1): 101-109.

Macintosh, NWG (1952) The Cohuna Cranium. Mankind 4: 307-329.

Macintosh, NWG (1953) The Cohuna cranium: physiography and chemical analysis. Oceania 23: 277-296.

Macintosh, NWG (1963) Origin and physical differentiation of the Australian Aborigines. Australian Natural History 14: 248-252.

Macintosh, NWG (1965) The physical aspect of man in Australia. In (Berndt, RM and Berndt, CH, Eds.) Aboriginal Man in Australia. Sydney: Angus and Robertson. Pgs. 29-70.

Macintosh, NWG (1967a) Fossil Man in Australia: with particular reference to the 1965 discovery at Green Gully near Keilor, Victoria. Australian Journal of Science 30: 86-98.

Macintosh, NWG (1967b) Recent discoveries of early Australian man. Annals of the Australian College of Dental Surgeons 1: 104-126.

Macintosh, NWG (1971) Analysis of an aboriginal skeleton and a pierced tooth necklace from Lake Nitchie, Australia. Anthropologie 9: 49-62.

Macintosh NWG and Larnach SL (1972) The persistence of Homo erectus traits in Australian Aboriginal crania. Oceania 7: 1-7.

Macintosh NWG and Larnach SL (1976) Aboriginal affinities looked at in world context. In (Kirk, RL and Thorne, AG, Eds.) The Origin of the Australians. New Jersey: Humanities Press. Pgs. 114-126.

Macumber, PG and Thorne, R (1975) The Cohuna Cranium Site – a re-appraisal. Archaeology and Physical Anthropology in Oceania 10: 67-72.

Márquez S, Mowbray K, Sawyer GJ, Jacob T, and Silvers A (2001) New fossil hominid calvaria from Indonesia – Sambungmacan 3. The Anatomical Record 262: 344- 368.

Matsu’ura, S (1982) A chronological framing for the Sangiran hominids: Fundamental study by the fluorine dating method. Bulletin of the National Science Museum, Tokyo, Series D (Anthropology) 8: 1-53

162

Matsu’ura, S (1986) Fluorine and phosphate analysis of fossil bones from the Kabuh formation of Trinil. Bulletin of the National Science Museum, Tokyo, Series D (Anthropology) 12: 1-9.

Matsu’ura S, Watanabe N, Aziz F, Shibasaki T, and Kondo M (1990) Preliminary dating of a hominid fossil tibia from Sambungmacan by fluorine analysis. Bulletin of the National Science Museum, Tokyo, Series D (Anthropology): 18: 19-29.

Matsu’ura S, Aziz F, and Watanabe N (1992) Probable chronological positions of two additional mandible specimens of the Sangiran hominids. J. Anthrop. Soc. Nippon 100: 262.

Matsu’ura S, Kondo M, and Aziz F (1995) A preliminary report on stratigraphic allocation of Sangiran hominid remains by multielement analysis of bone. Anthropological Science 103: 161

Matsu’ura S, Kondo M, Aziz F, Sudijono, Narasaki S, and Watanabe N (2000) First known tibia of an early Javanese hominid. Current Anthropology 41: 297-300

Miller GH, Magee JW, Johnson BJ, Fogel ML, Spooner NA, McCulloch MT, and Ayliffe LK (2003) Pleistocene extinction of Genyornis newtoni: Human impact on Australian megafauna. Science 283: 205-208.

Morrison, J (1967) The Biracial Origin of the Australian Aborigines. The Medical Journal of Australia 2(23): 1055-1056.

Morwood MJ, O’Sullivan P, Susanto EE and Aziz F (2003) Revised age for Mojokerto 1, and early Homo erectus cranium from East Java, Indonesia. Australian Archaeology 57: 1-4.

Mowbray KM, Márquez S, Antón SC, Swisher III CC, Jacob T, Sawyer GJ, Broadfield D, Laitman JT, Holloway RL, Delson E, and Tattersall I (2000) The newly recovered Poloyo hominin (PL-1) from Java. American Journal of Physical Anthropology Supplement 30: 233.

Mowbray K, Márquez S, and Delson E (2002) Mandibular fossa of Homo erectus: A response to Durband. The Anatomical Record 266: 142-145.

Nelson, AJ (1995) Cortical bone thickness in the primate and hominid postcranium: taxonomy and allometry. Unpublished Ph.D. Dissertation, University of California, Los Angeles.

Neves WA, Powell JF, and Ozolins EG (1999) Modern human origins as seen from the peripheries. Journal of Human Evolution 37: 129-133.

163

O’Connell, JF and Allen, J (2004) Dating the colonization of Sahul (Pleistocene Australia – New Guinea): a review of recent research. Journal of Archaeological Science 31: 835-853.

Oakley KP, Campbell BG and Molleson TI (1975) Catalogue of Fossil Hominids. Part III: Americas, Asia, Australasia. London: British Museum of Natural History.

Oppenoorth, WFF (1932a) Homo (Javanthropus) Soloensis. Een Plistoceene mensch van Java. Wetensch. Meded. Van den Dienst van den Mijnbouw Nederlandsch-Indië, No. 20: 49-63.

Oppenoorth, WFF (1932b) De vondst van paleolithische menschelijke schedels op Java. De Mijningenieur, Bandoeng, No. 5: 106-115.

Oppenoorth, WFF (1932c) Ein Neuer diluvialer Urmensch von Java. Nat. u. Mus., Frankfort 62 (9): 269-279.

Oppenoorth, WFF (1937) The Place of Homo Soloensis Among Fossil Men. In MacCurdy, GG (ed.) Early Man. New York: JB Lippincott Co. Pgs. 349-360.

Oyston, B (1996) Thermoluminescence age determinations for the Mungo III human burial, Lake Mungo, southeastern Australia. Quatern. Sci. Rev. (Quatern. Geochronol.) 15: 739-749.

Pardoe, C (1991a) Review article: Competing paradigms and ancient human remains: the state of the discipline. Archaeology in Oceania 26: 79-85.

Pardoe, C (1991b) Isolation and evolution in Tasmania. Current Anthropology 31: 1-21.

Pardoe, C (1993) The Pleistocene is still with us: analytical constraints and possibilities for the study of ancient human remains in archaeology. In (Smith MA, Spriggs M, and Fankhauser B, eds) Sahul in Review: Pleistocene archaeology in Australia, New Guinea and Island Melanesia. Occasional Papers in Prehistory 24. Australian National University: Department of Prehistory. Pgs. 81-94.

Pietrusewsky, M (1979) Craniometric variation in Pleistocene Australian and more recent Australian and New Guinea populations studied by multivariate procedures. Occasional Papers in Human Biology 2: 83-123.

Pope, G (1985) Taxonomy, dating, and paleoenvironment: The paleoecology of the early far eastern hominids. Modern Quaternary Research in Southeast Asia 9: 65-80.

164

Pope, G (1995) The influence of climate and geography on the biocultural evolution of the far eastern hominids. In (Vrba ES, Denton GH, Partridge TC, and Burckle LH, eds.) Paleoclimate and Evolution, with Emphasis on Human Origins. New Haven: Yale University Press. Pgs. 493-506.

Redd AJ, Roberts-Thompson J, Karafet T, Bamshad M, Jorde LB, Naidu JM, Walsh B, and Hammer MF (2002) Gene flow from the Indian subcontinent to Australia: Evidence from the Y chromosome. Curr. Biol. 12: 673-677.

Relethford, JH (2001a) Genetics and the Search for Modern Human Origins. Hoboken: Wiley-Liss.

Relethford, JH (2001b) Ancient DNA and the origin of modern humans. Proceedings of the National Academy of Science 98: 390-391.

Richtsmeier JT, Cheverud JM, and Buikstra JE (1984) The relationship between cranial metric and nonmetric traits in the rhesus macaques from Cayo Santiago. American Journal of Physical Anthropology 64: 213-222.

Rightmire, GP (1981) Patterns in the evolution of Homo erectus. Paleobiology 7: 241- 246.

Rightmire, GP (1984) Comparisons of Homo erectus from Africa and Southeast Asia. Courier Forsch. Inst. Senckenberg 69: 83-98.

Rightmire, GP (1990) The Evolution of Homo erectus: Comparative anatomical studies of an extinct human species. Cambridge: Cambridge Univ. Press.

Rightmire, GP (1992) Homo erectus: Ancestor or evolutionary side branch? Evolutionary Anthropology 1: 43-49.

Rightmire, GP (1994) The relationship of Homo erectus to later middle Pleistocene hominids. Courier Forsch. Inst. Senckenberg 171: 319-326.

Rightmire, GP (2004) Brain size and encephalization in early to mid-Pleistocene Homo. American Journal of Physical Anthropology 124: 109-123.

Rink WJ, Schwarcz HP, Swisher III CC, Anton SC, Jacob T (1997) Late Pleistocene ESR dates for Homo erectus at Ngandong, Java; Contemporaneity with H. sapiens in Southeast Asia. American Journal of Physical Anthropology Supplement 24: 198.

Roberts RG, Jones R, and Smith MA (1990) Thermoluminescence dating of a 50,000- year-old human occupation site in northern Australia. Nature 345: 153-156.

165

Roberts R, Bird M, Olley J, Galbraith R, Lawson E, Laslett G, Yoshida H, Jones R, Fullagar R, Jacobsen G, and Hua Q (1998) Optical and radiocarbon dating at Jinmium rock shelter in northern Australia. Nature 393: 358-362.

Rogers, AR (1995) How much can fossils tell us about regional continuity? Current Anthropology 36: 674-676.

Santa Luca, AP (1977) A Comparative Study of the Ngandong Fossil Hominids. Unpublished Ph.D. Dissertation, Harvard University.

Santa Luca, AP (1980) The Ngandong fossil hominids: A comparative study of a Far Eastern Homo erectus group. Yale Univ. Publ. in Anthropol. 78. 175 Pgs.

Sartono, S (1964) On a new find of another Pithecanthropus skull: an announcement. Bulletin of the Geological Society of Indonesia 1: 2-5.

Sartono, S (1967) An additional skull cap of a Pithecanthropus. Journal of the Anthropological Society of Nippon 75: 754.

Sartono, S (1968) Early man in Java: Pithecanthropus VII, a male specimen of Pithecanthropus erectus (I). Proceedings of the Koninklijke Akademie van Wetenschappen, Amsterdam Series B 71: 5.

Sartono, S (1971) Observations on a new skull of Pithecanthropus erectus (Pithecanthropus VIII) from Sangiran, Central Java. Proceedings of the Koninklijke Akademie van Wetenschappen Series B 74: 2.

Sartono, S (1972) Discovery of another hominid skull at Sangiran, Central Java. Current Anthropology 13: 124-126.

Sartono, S (1975) Implications arising from Pithecanthropus VIII. In (Tuttle R, ed.) Paleoanthropology, Morphology, and Paleoecology. Mouton: The Hague. Pgs. 327-360.

Sartono, S (1976) Genesis of the Solo Terraces. Modern Quaternary Research in Southeast Asia 2: 1-21.

Sartono, S (1990) A new Homo erectus skull from Ngawi, east Java. Bulletin of the Indo- Pacific Prehistory Association 11: 14-22.

Sartono S, Grimaud-Hervé D, Sémah F, Sémah AM, Widianto H, and Djubiantono T (1996) L’homme fossile de Ngawi. XIII International Congress of Prehistoric and Protohistoric Sciences, September 8-14, 1996, Forli, Italy: 121.

166

Schmidt, E (1887): Die anthropologischen Privat-Sammlungen Vol. I: Catalog der im anatomischen Institut der Universität Leipzig aufgestellten craniologischen Sammlung. In (Schaafhausen H., ed.) Die anthropologischen Sammlungen Deutschlands. Ein Verzeichniss des in Deutschland vorhandenen anthropologischen Materials. Vieweg, Braunschweig. 181 Pgs.

Schwartz, JH and Tattersall, I (2003) The Human Fossil Record, Craniodental Morphology of Genus Homo (Africa and Asia). Hoboken: Wiley-Liss.

Shipman, P (2001) The Man Who Found The Missing Link: Eugene Dubois and his lifelong quest to prove Darwin right. New York: Simon & Schuster.

Shreeve, J (1995) The Neandertal Enigma: Solving the Mystery of Modern Human Origins. New York: William Morrow & Co.

Shutler Jr. R, Head JM, Donahue DJ, Jull AJT, Barbetti MF, Matsu’ura S, de Vos J, and Storm P (in press) AMS radiocarbon dates on bone from cave sites in southeast Java, Indonesia, including Wajak. Modern Quaternary Research in Southeast Asia 18.

Sighinolfi GP, Sartono S, and Artioli G (1993) Chemical and mineralogical studies on hominid remains from Sangiran, Central Java (Indonesia). Journal of Human Evolution 24: 57-68.

Sim, R and Thorne, AG (1990) Pleistocene human remains from King Island, Southeastern Australia. Australian Archaeology 31: 44-51.

Sim, R and Thorne, AG (1995) Reply to Brown: Australian Archaeology, Number 38, 1994. Australian Archaeology 41: 29.

Simpson, JJ and Grün, R (1998) Non-destructive gamma spectrometric U-series dating. Quaternary Geochronology 17: 1009-1022.

Singh G, Kershaw AP, and Clark R (1981) Quaternary vegetation and fire history in Australia. In (Gill AM, Groves RH, and Noble IR, Eds) Fire and the Australian Biota. Canberra: Australian Academy of Science. Pgs. 23-54.

Sjøvold, T (1984) A report on the heritability of some cranial measurements and nonmetrical traits. In (Van Vark, GN and Howells, WW, Eds.) Multivariate Statistical Methods in Physical Anthropology. Dordrecht: Reidel Publishing. Pgs. 223-246.

167

Smahel, Z and Skvarilova, B (1988) Multiple correlations between craniofacial characteristics: An x-ray study. American Journal of Physical Anthropology 77: 221-229.

Smith CI, Chamberlain AT, Riley MS, Stringer C, and Collins MJ (2003) The thermal history of human fossils and the likelihood of successful DNA amplification. Journal of Human Evolution 45: 203-217.

Smith, FH (1992) The role of continuity in modern human origins. In (Bräuer G, Smith FH, Eds) Continuity or Replacement: Controversies in Homo sapiens evolution. Rotterdam: AA Balkema Press. Pgs. 145-156.

Smith FH, Falsetti AB, and Donnelly SM (1989) Modern human origins. Yearbook of Phys. Anthropol. 32: 35-68.

Sondaar PY, de Vos J, and Leinders JJM (1983) Reply: Facts and fiction around the fossil mammals of Java. Geologie en Mijnbouw 62: 339-343.

Sperber, GH (1989) The cranial base. In (Sperber, GH, ed.) Craniofacial Embryology. Oxford: Wright. Pgs. 101-118.

Stone, T and Cupper, ML (2003) Last glacial maximum ages for robust humans at Kow Swamp, southern Australia. Journal of Human Evolution 45: 99-111.

Storm, P (1995) The evolutionary significance of the Wajak skulls. Scripta Geologica 110: 1-247.

Storm, P (2000) The evolutionary history of humans in Australasia from an environmental perspective. Anthropological Science 108: 225-244.

Storm, P (2001a) Life and death of Homo erectus in Australasia: an environmental approach to the fate of a paleospecies. In (Sémah F, Falguères C, Grimaud-Hervé D, and Sémah A-M, eds) Origine Des Peuplements Et Chronologie Des Cultures Paléolithiques Dans Le Sud-Est Asiatique. Paris: Semenanjung. Pg. 279-298.

Storm, P (2001b) The evolution of humans in Australasia from an environmental perspective. Palaeogeography, Palaeoclimatology, Palaeoecology 171: 363-383.

Strait, DS (1998) Evolutionary integration in the hominid cranial base. Unpubl. Ph.D. Dissertation, State University of New York, Stony Brook.

Stringer, CB (1984) The definition of Homo erectus and the existence of the species in Africa and Europe. Courier Forsch. Inst. Senckenberg 69: 131-143.

168

Stringer, CB (1989) The origin of early modern humans: A comparison of the European and non-European evidence. In (Mellars, P and Stringer, CB, eds.) The Human Revolution. Edinburgh: Edinburgh Univ. Press.

Stringer, CB (1992) Replacement, continuity, and the origin of Homo sapiens. In (Bräuer, G and Smith, FH, eds.) Continuity or Replacement: Controversies in Homo sapiens Evolution. Rotterdam: AA Balkema. Pgs. 9-24.

Stringer, CB (1994) Out of Africa - a personal history. In (Nitecki and Nitecki, eds.) Origins of Anatomically Modern Humans. New York: Plenum Press. Pgs. 150- 170.

Stringer, CB (1998) A metrical study of the WLH-50 calvaria. Journal of Human Evolution 34: 327-332.

Stringer, CB and Andrews, P (1988) Genetic and fossil evidence for the origin of modern humans. Science 239: 1263-1268.

Swisher III CC, Curtis GH, Jacob T, Getty AG, Suprijo A, and Widiasmoro (1994) Age of the earliest known hominids in Java, Indonesia. Science 263: 1118-1121.

Swisher III CC, Rink WJ, Antón SC, Schwarcz HP, Curtis GH, Suprijo A, and Widiasmoro. (1996) Latest Homo erectus of Java: Potential contemporaneity with Homo sapiens in Southeast Asia. Science 274: 1870-1874.

Swisher III CC, Rink WJ, Schwarcz HP, and Antón SC (1997) Response: Dating the Ngandong humans. Science 276: 1575-1576.

Taxman, RM (1963) Incidence and size of the juxtamastoid eminence in modern crania. American Journal of Physical Anthropology 21: 153-157.

Templeton, AR (2002) Out of Africa again and again. Nature 416: 45-51.

Theunissen, B (1989) Eugene Dubois and the Ape-Man from Java: The history of the first “missing link” and its discoverer. Dordrecht: Kluwer Academic Publishers.

Theunissen B, de Vos J, Sondaar PY, and Aziz F (1990) The establishment of a chronological framework for the hominid-bearing deposits of Java; A historical survey. In (Laporte LF, ed.) Establishment of a geologic framework for paleoanthropology. Boulder, Colorado, Geological Society of America Special Paper 242.

Thorne, AG (1971) Mungo and Kow Swamp: morphological variation in Pleistocene Australians. Mankind 8: 85-89.

169

Thorne, AG (1975) Kow Swamp and Lake Mungo: Towards an osteology of early man in Australia. Unpubl. Ph.D. Dissertation, University of Sydney.

Thorne, AG (1976) Morphological contrasts in Pleistocene Australians. In (Kirk RL and Thorne AG, eds) The Origin of the Australians. New Jersey: Humanities Press. Pgs. 95-112.

Thorne, AG (1977) Separation or reconciliation? Biological clues to the development of Australian society. In (Allen J, Golson J, and Jones R, eds) Sunda and Sahul: Prehistoric studies in Southeast Asia, Melanesia, and Australia. New York: Academic Press. Pgs. 187-204.

Thorne, AG (1980) The longest link: human evolution in Southeast Asia and the settlement of Australia. In (Fox JJ, Garnaut R, McCawley P, and Mackie JAC, eds) Indonesia: Australian Perspectives. Canberra: Australian National University. Pgs. 35-43.

Thorne, AG (1981) The center and the edge: The significance of Australian hominids to African palaeoanthropology. Proceedings of the Pan-African Congress of Prehistory (Nairobi). Nairobi: National Museums of Kenya. Pgs. 180-181.

Thorne, AG (1984) Australia’s Human Origins – How Many Sources? American Journal of Physical Anthropology Supplement 63: 227.

Thorne, AG (1989) The emergence of the Pacific peoples. Proceedings of the 1988 Conference of the Australian Society for Human Biology. University of Western Australia. Pgs. 103-111.

Thorne, AG (2002) Not out of Africa. Discover 23(8): 51-57.

Thorne AG, Grün R, Mortimer G, Spooner NA, Simpson JJ, McCulloch M, Taylor L, and Curnoe D (1999) Australia’s oldest human remains: age of the Lake Mungo 3 skeleton. Journal of Human Evolution 36: 591-612.

Thorne, AG and Sim, R (1994) The gracile male skeleton from late Pleistocene King Island, Australia. Australian Archaeology 38: 8-10.

Thorne, AG and Wilson SR (1977) Pleistocene and recent Australians: A multivariate comparison. Journal of Human Evolution 6: 393-402.

Thorne, AG and Wolpoff, MH (1981) Regional continuity in Australasian Pleistocene hominid evolution. American Journal of Physical Anthropology 55: 337-349.

Thorne, AG and Wolpoff, MH (1992) The multiregional evolution of humans. Scientific American 266: 76-83.

170

Trinkaus, E (2004) Eyasi 1 and the suprainiac fossa. American Journal of Physical Anthropology 124: 28-32.

Tyler, DE (1991) A taxonomy of Javan hominid mandibles. Human Evolution 6: 401- 420.

Tyler, DE (1994) The taxonomic status of "Meganthropus." Courier Forschungsinstitut Senckenberg 171: 115-121.

van den Bergh, GD (1999) The late Neogene elephantoid-bearing faunas of Indonesia and their palaeozoogeographic. A study of the terrestrial faunal succession of Sulawesi, Flores, and Java, including evidence for early hominid dispersal east of Wallace’s Line implications. Scripta Geol. 117: 1-419.

van den Bergh GD, de Vos J, and Sondaar PY (2001) The late Quaternary palaeogeography of mammal evolution in the Indonesian archipelago. Palaeogeography, Palaeoclimatology, Palaeoecology 171: 385-408.

Van Limborgh, J (1970) The role of genetic and local environmental factors in the control of postnatal craniofacial morphogenesis. Acta Morph. 8: 143-60.

Van Vark GE, Steerneman AGM, and Czarnetzki A (1989) Some notes on the definition and application of nonmetrical skeletal variables. Homo 40: 133-136.

Webb, SG (1989) The Willandra Lakes Hominids. Occasional Papers in Prehistory 16. Canberra: Australian National University.

Webb, SG (1990) Cranial thickening in an Australian hominid as a possible palaeoepidemiological indicator. American Journal of Physical Anthropology 82: 403-412.

Weeks RA, Bogard JS, Elam JM, Weinand DC, and Kramer A (2003) Effects of thermal annealing on the radiation produced electron paramagnetic resonance spectra of bovine and equine tooth enamel: fossil and modern. Journal of Applied Physics 93: 9880-9889.

Weidenreich, F (1943) The skull of Sinanthropus pekinensis: a comparative study on a primitive hominid skull. Palaeontologia Sinica, New Series D 10.

Weidenreich, F (1945a) Giant early man from Java and south China. Anthropological Papers of the American Museum of Natural History 40: 1-134.

Weidenreich, F (1945b) The puzzle of Pithecanthropus. Science and Scientists in the Netherlands Indies 1: 380-390.

171

Weidenreich, F (1945c) The Keilor Skull: A Wadjak Type from Southeast Australia. American Journal of Physical Anthropology 3: 21-33.

Weidenreich, F (1946) Apes, Giants, and Man. Chicago: University of Chicago Press.

Weidenreich, F (1951) Morphology of Solo Man. Anthropological Papers of the American Museum of Natural History 43: 205-90.

West, AL (1977) Aboriginal man at Kow Swamp, Northern Victoria: the problem of locating the burial site of the KS1 skeleton. The Artefact 2: 19-30.

Westaway, MC (2002a) Preliminary observations on the taphonomic processes at Ngandong and some implications for a late Homo erectus survivor model. Tempus 7: 189-193.

Westaway, MC (2002b) The Nacurrie 2 Cranium. A Plain Language Community Report to the Wemba Wemba Aboriginal Land Council. Unpublished report from the National Museum of Australia.

Westaway MC, Jacob T, Aziz F, Otsuka H and Baba H (2003) Faunal taphonomy and biostratigraphy at Ngandong, Java, Indonesia and its implications for the late survival of Homo erectus. American Journal of Physical Anthropology Supplement 36: 222-223.

White, TD (1991) Human Osteology. New York: Academic Press.

Widianto, H (2001) The perspective on the evolution of Javanese Homo erectus based on morphological and stratigraphic characteristics. In (Simanjuntak T, Prasetyo B, and Handini R, Eds.) Sangiran: Man, Culture and Environment in Pleistocene Times. Jakarta: Yayasan Obor Indonesia. Pgs. 24-45.

Widianto, H and Grimaud-Hervé, D (1993) Le crane de Ngawi. Dossiers d’Archeologie 184: 36.

Widianto, H and Zeitoun, V (2003) Morphological description, biometry and phylogenetic position of the skull of Ngawi 1 (East Java, Indonesia). International Journal of Osteoarchaeology 13: 339-351.

Wolpoff, MH (1989) Multiregional evolution: the fossil alternative to Eden. In (Mellars P, Stringer CB, eds) The Human Revolution. Edinburgh: Edinburgh University Press. Pgs. 62-108.

Wolpoff, MH (1992) Theories of modern human origins. In (Bräuer G, Smith FH, eds) Continuity or Replacement: Controversies in Homo sapiens evolution. Rotterdam: AA Balkema Press. Pgs. 25-64.

172

Wolpoff, MH (1999) Paleoanthropology, 2nd edition. Boston: McGraw Hill.

Wolpoff MH, Hawks JD, Frayer DW, and Hunley K (2001) Modern Human Ancestry at the Peripheries: A Test of the Replacement Theory. Science 291: 293-297.

Wolpoff MH, Thorne AG, Jelinek A and Zhang Yinyun (1994) The case for sinking Homo erectus: 100 years of Pithecanthropus is enough! Courier Forsch. Inst. Senckenberg 171: 341-61.

Wolpoff MH, Zhi WX and Thorne AG (1984) Modern Homo sapiens origins: a general theory of hominid evolution involving the fossil evidence from East Asia. In (Smith FH, Spencer F, eds) The Origins of Modern Humans: A Survey of the Fossil Evidence. New York: AR Liss. Pgs 411-484.

Wood, BA (1984) The origin of Homo erectus. Courier Forsch. Inst. Senckenberg 69: 99- 111.

Wood, BA (1991) Koobi Fora Research Project Vol. 4: Hominid Cranial Remains. Oxford: Clarendon Press.

Wright, RVS (1986) How old is zone Fat Lake George? Archaeology in Oceania 21: 138- 139.

Zeitoun, V (2002) Asian palaeoanthropology: Homo soloensis reconsidered? Paper presented at the 17th Congress of the Indo-Pacific Prehistory Association, Taipei, Taiwan, September 9-15.

173

Appendix A

Raw non-metric data for the modern human sample

174

Number Sex 1. Phar tub 2. Tymp plate mast 3. Ovale pit 4. Ovale access 5. Styloid 6. Vaginal 7. Alar tub 8. Condyle orient 9. Opis recess 10. PCT 11. F. lacerum 12. Juxtamastoid 13. Occipitomastoid 14. Carotid rel to S-Q 15. PGT size 16. Orient of S-Q 17. Rel condyle sz Australia 99-8163M11001101021121103 99-8164M11001111031010113 99-8166F11001121121220113 99-8167M21001121012110101 99-8168F20001121011011112 99-8169F20001121010110111 99-8170F11001111021020113 99-8171F21001111122110111 99-8172F11001111021120103 99-8173M21001111021110111 99-8174F21001111002000112 99-8175F11001121021020112 99-8176M21001111011111112 99-8177M11001111021110112 99-8178M11001121021211101 99-8179M11001111031120111 99-8180M11001111012120112 99-8181M11001111011110102 VL 981M11001121012211001 99-8217M21001121012221112 99-8153M11001111022120113 99-8154F11001111022021113 99-8155F11001101012110103 99-8156M11001111022220102 99-8157F11001101032011112 99-8158F11001121002120102 VL 1207F11001121NA12010102 VL 1604M21001121011220111 VL 1412M11001111012220112 VL 1413F11001111012020112 99-8165M21001121022022111 VL 243F11001121021120112 VL 244M2100112102111011NA VL 245M21001111022222111 VL 246M21001111011122112 VL 247M21001111022021112 VL 625M11001111010221101 VL 626M2100112101122011NA VL 627F11001111021010102 99-68F21001111022111112 99-69F01001111022022112 VL 1578M21001111121120102 175

VL 1579M21001121012221110 226086F01001111000021112 226088F11001101012022112 226089M21001111002021101 226090F21001111010122101 258483F01001120011111113 329778F2100110101102011NA 329779M21001121022121103 331242M11NANA113111202110NA 331243M2100111101201111NA 344711M21001111012121113 344712F11001111012021112 344713F11001111010221113 344714F11001111012020112 344715F1100111100202111NA 350096M2100111101222211NA 350097M21001121012210113 350098M11001111011011103 380429M21001111012221113 380430F2100112101212211NA 380431M21001121022111113 380432F11001111011122111 380433F1100110112022211NA 380435F010011010NA110211NA 380437M11001111012NANA111NA 380447M11001111232211112 380448M21001121012121112 380449M21001111022011112 380450M2100111112112111NA 380451M21001111011011101 380452M11001101011021113 380453M11001111111011113 380464F2100111101211111NA 380465M2100111111211211NA 380475M11001131112111113 SCK 1214M21001111012111113 FM 272961M2100111102111111NA FM 272962M21001111012111113

Tasmania VL 269M11001121012111112 VL 270M01001111012221110 VL 271F11001111022210103 VL 272M11001121122221112 VL 273M11001121012021112 VL 274F11001111022121113 VL 275M01001121032220103 FM 272964M11001111032011112 FM 272965F11001101011001112 176

W. Africa VL 365M11001121011011102 VL 366F11001120102111111 VL 362F11001111002111112 VL 363F11001111001221103 VL 367M1000111101211110NA VL 368M1100111111211110NA VL 446F1100110101200011NA VL 447M2100111111221211NA VL 448F11001101001111112 VL 449F21001121001000112 VL 574F11001111011220002 VL 575F1000111100212111NA VL 452M21001111102112111 VL 576F20001111012120113 VL 577F21001111001220113 VL 578M20001121021121103 VL 579F1100111101210011NA VL 5264M21001111012001113 VL 5274F11001111102011113 VL 5284F21001101022121113 VL 400M21001111012211113 VL 593M11001111012122113 VL 470M21001121012021101 VL 471M11001111021011102 VL 1756F11001111012111112 VL 1757F21001121001210112 VL 4924bM21001111010101111 VL 2472M11001111002011111 SUL 761M11001111012212113

S. Africa VL 2793F11001101012022102 99-8428M21001121111021101 99-8429M21001121000220112 99-8431F11001111112121112 99-8433M21001101112212111 99-8436M21001121011020101 99-8442F21001111012021111 99-8446F11001111002021110 99-8449F11001111012020111 99-8451M2100111101102101NA 99-8452F21001111022221113 99-8453M11001111012211111 99-8454F2100110101211111NA 99-8456M11NANA111110202111NA 99-8458M11001101011211112 99-76M11001131011021111 99-77F11001121012120110 99-78M11001121001022110 177

VL 4313F21001101001022110 VL 2462F11001111011021113 VL 2463M11001111000021102 VL 2464M11001111010021112 VL 2469M1100111100112111NA VL 2470F21001111001022110 99-6913M11001121010021102 VL 3579M11001121002121110 VL 3280M11001111021121101 VL 3281F11001101021021112 SUL 2477M1100110111202211NA SUL 2478F2100111NA00211210NA

Greece VL 2217M11001101032221112 VL 2218M21001121002221112 VL 2219F21001101022212012 VL 2220F1100111101211111NA VL 2221M21001101002202NA1NA VL 2222M21001111012202112 VL 2223M2100112100201111NA VL 2225M21001111112121112 VL 2226M11001111012012111 VL 2227M21001111012211010 VL 2228F21001121012022010 VL 2229F21001121012021111 VL 2230F2100111101201111NA VL 2231M2100112101211111NA VL 2233M2100112101202110NA VL 2234M2100111100220111NA VL 2236M21001121012200102 VL 2237M21001111012NANA1111 VL 2238F21001111012211111 VL 2239F21001121002211112 VL 2241M21001111002111112 VL 2242M21001111012111111 VL 2243F2100112101201111NA VL 2247M21001111002201111 VL 2248M11001121122221111

Czech VL 3495M1100111102NA212111 VL 3496M1100111101201111NA VL 3497F21001110012201110 VL 3498F21001111002201111 VL 3499M21001111012111111 VL 3500M11001111012211011 VL 3501M2100111111220111NA VL 3502F11001111002221111 VL 3503M21001121012212110 178

VL 3504M1100111112221101NA VL 3505M21001121112201112 VL 3506M20001111002021112 VL 3507F11001111122021111 VL 3508F21001121002022112 VL 3509M11001131012221101 VL 3510M21001111012NANA2111 VL 3511M21001121102212113 VL 3513M21001111032221110 VL 3514M2100111102222111NA VL 3515M21001121022221111 VL 3516F20001111012222112 VL 3517M11001101012121010 VL 3518M21001111022221012 VL 3521F200011110222111NA2 VL 3522M21001110012111112

Austria VL 4F2100110100221211NA VL 874M21001101102212101 VL 875F1100110102NA00201NA VL 876M21001111212211111 VL 877M21001101012222112 VL 878F1100111101212211NA VL 879M2100111101221111NA VL 880M21001111012212112 VL 1100F2100110NA01212211NA VL 1101M2100110111222211NA VL 1102M2100111002220110NA VL 1103F1100112NA00201211NA VL 1109F21001111102221112 VL 1110M21001101012211110 VL 11M21001121012221003 VL 883F010011110NA210111NA VL 884F21001111002022111 VL 886F21001111022001112 VL 887F11001111022022112 VL 888F21001101012221112 VL 890F11001101012211112 VL 893F11001100002212111 VL 895M21001121122221110 VL 109M11001101002201011

Utah 99-7446M21001121002022111 99-7447M11001121011222112 99-7448M11001121111021111 99-7449M2100111101202111NA 99-7450M110011NA1011021113 99-7451M11001121122022113 179

99-7453F2100112111212111NA 99-7454F21001111011122013 99-7455F11001121012122112 99-7458F0100112001221111NA 99-7460F11001121011022113 99-7463F11001111011111112 99-7464M11001111011022013 99-7466F11001111012212013 99-7467M21001111011021113 99-7469F11001121011111113 99-7385F11001121131212113 99-7364M21001121012002113 99-7384F11001131002112113 99-7382M21001110002111112

Egypt SUL 267M11001101022210113 SUL 274M21001121002210102 SUL 280M11001111002222113 SUL 282M21001110012211113 SUL 284M2100112112221111NA SUL 301M110011110122NA111NA SUL 307M21001111012212113 SUL 308M1100110102221211NA SUL 311M11001121012211113 SUL 318F11001111022221111 SUL 319F11001111110022113 SUL 321M11001101022121111 SUL 324F11001121022111213 SUL 326F11001121011021112 SUL 330F1100111NA21102111NA SUL 336F11001121211120113 SUL 343F11001101121221112

Java SCK 1163M11001111012222113 SCK 1164M1100112112222211NA SCK 1165F21001111022122113 SCK 1166M11001111012201113 SCK 1167F11001111012122112 SCK 1184M21001121011222112 SCK 1187F11001101022012113 SCK 1230M21001121232201113 SCK 1291M21001111012101113 SUL 818M21001121002112113 SUL 820M11001121012122113 SUL 823M11001111022101113 SUL 825M11001111012212112 SUL 827M11001111NA1120111NA SUL 828M11001121012111112 180

SUL 832M11001121112202103 SUL 836M11001111022122113 SUL 838M11001111012212113 SUL 840M21001111012201112 SUL 849M11001111012121113 SUL 852M11001121022211113 SUL 853M0100111101222111NA SUL 856F11001111012221113 SUL 858F01001111021121112 SUL 868F21001121011122111 SUL 873M110011NA1022021113

India SCK 178M21001111022021112 SCK 185M11001111111121113 SCK 187M21001111012122112 SCK 192M2100110102211211NA SCK 193M11001111012020113 SCK 195M11001121012122113 SCK 1274M11001111012212113 SCK 1275F1100110101211111NA SCK 1276M11001101012112102 SCK 1277M21001121012222112 SCK 1280M21001101112111113 SCK 1284M11001131112101113

China SCK 1196M1100111101202111NA SCK 1201M21001111012122112 SCK 1203M21001121012211113 SCK 1206M21001121012222113 SCK 1208M21001111011221113 SCK 1209M21001121111022112 SCK 1210M21001121031012113 SCK 1211M21001121102221112 SCK 1200M2100113100221101NA SCK 1295M11001101022100113 SCK 5343M11001111012122NA13 SCK 5344M01001121002222112

181

Appendix B

Raw non-metric data for the fossil sample

182

Specimen 1. Phar tub 2. Tymp plate mast 3. Ovale pit 4. Ovale access 5. Styloid 6. Vaginal 7. Alar tub 8. Condyle orient 9. Opis recess 10. PCT 11. F. lacerum 12. Juxtamastoid 13. Occipitomastoid 14. Carotid rel to S-Q 15. PGT size 16. Orient of S-Q 17. Rel condyle sz Sangiran 2 NA 0 NA NA 0 0 NA NA 0 0 NA 0 0 NA 1 0 NA Sangiran 4NA00000NA100NA200113 Sangiran 12 NA NA NA NA NA NA NA NA 0 0 NA 1 NA NA NA NA NA Sangiran 14 NA NA NA NA NA NA 2 1 NA NA NA NA NA NA NA NA NA Sangiran 1711NANA002110NA022103 Sambungm 1 NA 0 NA NA 0 0 NA NA NA NA NA 1 NA NA 0 0 NA Sambungm 3NA0NANA00NANA23NA02200NA Sambungm 420110020?2310221?0NA Ngandong 1 NA 0 NA NA 0 0 NA NA 2 3 NA 01200NA Ngandong 6 NA 0 NA NA 0 0 NA NA NA NA NA NA NA 2 0 0 NA Ngandong 7 00110020231222000 Ngandong 10 NA 0 NA NA 0 0 NA NA 2 3 NA 12200NA Ngandong 11 NA 0 NA NA 0 0 NA NA 2 3 NA 12200NA Ngandong 12 00110020231222000

Wajak 1 11NANA1101NA0NANA2NANA13 Keilor 210011NA001NA022111 Talgai NA1 NANANANANANANANANANANANANANANA K. Swamp 5NA1NANA11NANANA1NA01010NA K. Swamp 8 NA1 NANANA1 NANANANANANANANA1 1 NA Lake Mungo I NA NA 0011NA1NANANANANANA1NANA Lake Mungo III NA NA NA NA NA NA NA NA 0 1 NA NA NA NA NA NA NA Nacurrie 1 NA 1 NA NA NA 1 NA NA NA NA NA 0 1 NA 1 1 NA Nacurrie 2NA10011NA101211NA112 Cossack NA1 NANA1 1 NANANANANANANANA1 1 NA Mossgiel NANA0 0 NANANANANA1 NA0 NANA1 NANA Cohuna NA1NANA1111NANA100210NA Lake Nitchie210011310NA2201113

183

Appendix C

Raw metric data for the comparative samples

184

Location Specimen FM L FM W FM area R Cond L R Cond W Cond area Ratio Australia 99-8163 35.0 26.2 720.21 26.0 11.0 286.00 39.71 Australia 99-8164 35.7 27.0 757.04 28.0 10.5 294.00 38.84 Australia 99-8166 35.5 30.0 836.45 25.5 12.6 321.30 38.41 Australia 99-8167 41.0 35.0 1127.05 26.2 12.0 314.40 27.90 Australia 99-8168 34.0 29.0 774.40 22.0 12.0 264.00 34.09 Australia 99-8169 35.0 28.7 788.93 21.0 10.5 220.50 27.95 Australia 99-8170 34.2 26.0 698.38 24.8 12.0 297.60 42.61 Australia 99-8171 40.0 30.2 948.76 21.8 12.0 261.60 27.57 Australia 99-8172 34.2 25.0 671.51 21.5 12.1 260.15 38.74 Australia 99-8173 40.5 31.0 986.07 24.0 11.4 273.60 27.75 Australia 99-9174 34.8 26.5 724.29 23.0 11.5 264.50 36.52 Australia 99-8175 31.8 25.7 641.87 20.0 12.1 242.00 37.70 Australia 99-8176 37.0 30.0 871.79 25.0 11.7 292.50 33.55 Australia 99-8177 32.7 26.1 670.31 23.8 10.0 238.00 35.51 Australia 99-8178 38.8 31.0 944.68 21.5 13.8 296.70 31.41 Australia 99-8179 34.0 29.5 787.75 22.9 10.2 233.58 29.65 Australia 99-8180 35.5 28.2 786.26 24.1 10.5 253.05 32.18 Australia 99-8181 32.1 29.8 751.30 23.4 10.5 245.70 32.70 Australia VL 981 35.5 33.0 920.09 23.0 11.2 257.60 28.00 Australia 99-8217 30.3 24.5 583.04 21.5 10.0 215.00 36.88 Australia 99-8153 35.7 26.5 743.03 27.1 14.0 379.40 51.06 Australia 99-8154 32.0 28.5 716.28 22.0 12.8 281.60 39.31 Australia 99-8155 30.0 25.5 600.83 24.5 10.0 245.00 40.78 Australia 99-8156 35.0 29.5 810.92 24.5 12.0 294.00 36.25 Australia 99-8157 35.0 28.5 783.43 23.0 11.5 264.50 33.76 Australia 99-8158 33.8 29.2 775.16 22.1 12.8 282.88 36.49 Australia VL 1207 33.5 30.0 789.32 24.0 11.0 264.00 33.45 Australia VL 1604 35.5 30.0 836.45 23.5 11.0 258.50 30.90 Australia VL 1412 32.8 26.8 690.40 21.8 11.7 255.06 36.94 Australia VL 1413 33.7 26.0 688.17 23.0 10.5 241.50 35.09 Australia 99-8165 35.7 29.7 832.75 21.0 11.3 237.30 28.50 Australia VL 243 32.2 28.5 720.76 19.2 12.8 245.76 34.10 Australia VL 245 38.5 29.6 895.04 23.7 11.5 272.55 30.45 Australia VL 246 33.0 29.0 751.63 23.0 11.5 264.50 35.19 Australia VL 247 35.5 30.1 839.24 23.5 12.4 291.40 34.72 Australia VL 625 35.2 28.0 774.09 20.7 11.2 231.84 29.95 Australia VL 627 32.5 28.2 719.82 19.5 12.2 237.90 33.05 Australia 99-68 32.0 28.4 713.77 20.3 13.2 267.96 37.54 Australia 99-69 36.5 27.7 794.08 26.0 10.0 260.00 32.74 Australia VL 1578 36.7 29.5 850.31 21.0 13.5 283.50 33.34 Australia VL 1579 38.0 33.0 984.89 23.0 10.1 232.30 23.59 Australia 99-8424 34.0 29.0 774.40 24.3 11.0 267.30 34.52 Australia 272962 36.8 31.0 895.98 25.0 14.9 372.50 41.57 Australia 226086 33.0 27.8 720.52 22.0 8.9 195.80 27.17 Australia 226088 33.1 27.8 722.71 22.8 9.1 207.48 28.71 Australia 226089 40.1 33.0 1039.32 25.5 10.0 255.00 24.54 Australia 226090 34.2 32.0 859.54 20.5 9.2 188.60 21.94 Australia 258483 32.5 26.9 686.63 20.0 12.0 240.00 34.95 Australia 329779 36.0 32.1 907.61 25.2 12.8 322.56 35.54 185

Australia 344711 33.9 30.2 804.07 22.0 13.1 288.20 35.84 Australia 344712 37.0 30.0 871.79 21.6 11.0 237.60 27.25 Australia 344713 34.8 30.2 825.42 24.0 11.8 283.20 34.31 Australia 344714 38.0 32.7 975.93 22.4 13.0 291.20 29.84 Australia 350097 36.9 31.0 898.42 23.5 12.5 293.75 32.70 Australia 350098 34.4 30.3 818.64 26.5 10.5 278.25 33.99 Australia 380429 34.7 31.0 844.85 24.5 11.0 269.50 31.90 Australia 380431 36.8 33.0 953.79 25.2 13.0 327.60 34.35 Australia 380432 38.5 32.1 970.63 22.2 10.8 239.76 24.70 Australia 380447 41.6 30.6 999.78 25.0 10.1 252.50 25.26 Australia 380448 34.5 30.1 815.60 21.0 11.5 241.50 29.61 Australia 380449 31.2 30.8 754.74 20.5 10.2 209.10 27.70 Australia 380451 33.5 28.5 794.63 19.8 10.0 198.00 24.92 Australia 380452 37.5 28.8 848.23 25.5 11.2 285.60 33.67 Australia 380453 41.0 30.0 966.04 23.1 13.5 311.85 32.28 Australia 380475 38.8 30.1 917.25 23.7 12.5 296.25 32.30 Australia SCK 1214 34.5 28.7 777.66 22.5 12.0 270.00 34.72 Tasmania VL 275 33.0 27.0 699.79 30.0 11.0 330.00 47.16 Tasmania VL 269 35.0 28.5 783.43 25.5 10.2 260.10 33.20 Tasmania VL 270 38.6 30.0 909.49 21.0 11.0 231.00 25.40 Tasmania VL 271 32.0 28.1 706.23 21.0 13.7 287.70 40.74 Tasmania VL 272 34.5 29.3 793.92 29.4 10.0 294.00 37.03 Tasmania VL 273 32.5 26.8 684.08 23.0 9.9 227.70 33.29 Tasmania VL 274 32.5 31.1 793.84 23.0 14.0 322.00 40.56 32.61

Austria VL 874 35.8 31.0 871.63 22.2 10.0 222.00 25.47 Austria VL 876 36.7 30.4 876.25 23.3 10.0 233.00 26.59 Austria VL 877 34.8 31.0 847.29 26.7 11.0 293.70 34.66 Austria VL 880 39.0 29.8 912.79 25.9 11.5 297.85 32.63 Austria VL 1109 32.5 26.0 663.66 20.0 10.8 216.00 32.55 Austria VL 1110 41.3 32.0 1037.98 24.5 10.0 245.00 23.60 Austria VL 11 35.5 30.0 836.45 24.8 13.0 322.40 38.54 Austria VL 884 39.1 32.1 985.76 21.5 13.6 292.40 29.66 Austria VL 886 38.0 31.2 931.17 27.7 11.0 304.70 32.72 Austria VL 887 37.0 33.0 958.97 26.1 13.0 339.30 35.38 Austria VL 888 34.4 30.8 832.14 25.0 10.6 265.00 31.85 Austria VL 890 37.2 32.7 955.39 23.8 13.5 321.30 33.63 Austria VL 893 35.7 28.7 804.71 22.7 10.8 245.16 30.47 Austria VL 895 39.2 33.5 1031.38 21.0 9.5 199.50 19.34 Austria VL 109 35.1 32.8 904.21 23.0 10.0 230.00 25.44 30.17

Czech VL 3495 34.0 29.3 782.41 20.0 11.6 232.00 29.65 Czech VL 3497 36.6 29.2 839.37 21.0 9.0 189.00 22.52 Czech VL 3498 32.1 28.6 721.04 21.2 9.2 195.04 27.05 Czech VL 3499 33.7 31.1 823.15 24.0 10.0 240.00 29.16 Czech VL 3500 36.0 29.8 842.57 26.5 10.0 265.00 31.45 Czech VL 3502 37.0 35.0 1017.09 24.2 13.2 319.44 31.41 186

Czech VL 3503 37.8 31.3 929.24 21.0 11.2 235.20 25.31 Czech VL 3505 34.1 27.8 744.54 24.0 10.0 240.00 32.23 Czech VL 3506 36.8 28.8 832.40 24.2 11.2 271.04 32.56 Czech VL 3507 37.0 32.4 941.53 26.6 9.9 263.34 27.97 Czech VL 3508 35.9 28.7 809.22 26.0 11.1 288.60 35.66 Czech VL 3509 34.9 28.7 786.68 21.0 10.0 210.00 26.69 Czech VL 3510 38.0 31.0 925.20 24.8 10.6 262.88 28.41 Czech VL 3511 34.8 25.1 686.03 23.5 11.2 263.20 38.37 Czech VL 3513 35.0 31.0 852.16 22.0 8.7 191.40 22.46 Czech VL 3515 36.2 31.8 904.12 23.2 10.3 238.96 26.43 Czech VL 3516 31.8 27.8 694.32 24.5 10.7 262.15 37.76 Czech VL 3517 34.3 29.0 781.23 19.0 10.1 191.90 24.56 Czech VL 3518 32.7 25.8 662.61 23.1 9.3 214.83 32.42 Czech VL 3521 32.0 29.0 728.85 22.8 11.0 250.80 34.41 Czech VL 3522 34.6 29.0 788.07 21.7 10.7 232.19 29.46 30.31

Greece VL 2217 40.0 32.7 1027.30 29.7 11.3 335.61 32.67 Greece VL 2218 36.8 29.0 838.18 25.4 10.9 276.86 33.03 Greece VL 2219 35.7 28.0 785.08 23.8 11.3 268.94 34.26 Greece VL 2222 35.9 28.0 789.48 25.7 10.0 257.00 32.55 Greece VL 2225 35.2 28.5 787.91 24.0 11.8 283.20 35.94 Greece VL 2226 33.4 29.3 768.61 21.8 9.9 215.82 28.08 Greece VL 2227 37.5 33.6 989.60 22.0 10.0 220.00 22.23 Greece VL 2228 40.6 33.3 1061.84 22.9 11.0 251.90 23.72 Greece VL 2229 34.0 27.4 731.68 19.3 11.0 212.30 29.02 Greece VL 2236 35.2 29.5 815.56 23.7 12.0 284.40 34.87 Greece VL 2237 37.3 32.2 943.31 27.2 11.0 299.20 31.72 Greece VL 2238 36.0 30.0 848.23 24.2 10.2 246.84 29.10 Greece VL 2239 34.1 26.0 696.33 24.0 10.0 240.00 34.47 Greece VL 2241 37.0 30.7 892.13 24.7 13.7 338.39 37.93 Greece VL 2242 38.5 30.5 922.25 22.2 11.0 244.20 26.48 Greece VL 2247 38.0 32.0 955.04 25.0 11.8 295.00 30.89 Greece VL 2248 33.4 29.2 765.98 24.3 10.0 243.00 31.72 31.10

S. Afr 99-8442 35.2 30.8 851.50 21.0 11.2 235.20 27.62 S. Afr 99-8446 39.0 30.2 925.04 17.5 12.0 210.00 22.70 S. Afr 99-8449 37.5 29.7 874.74 19.0 12.6 239.40 27.37 S. Afr 99-8452 35.0 25.3 695.47 21.8 13.0 283.40 40.75 S. Afr 99-8453 40.1 32.0 1007.82 22.3 13.0 289.90 28.76 S. Afr 99-8458 37.0 27.8 807.86 22.8 12.0 273.60 33.87 S. Afr 99-6913 35.6 28.2 788.48 24.0 11.0 264.00 33.48 S. Afr 99-8868 35.4 26.3 731.22 23.1 8.5 196.35 26.85 S. Afr 99-8882 37.0 29.1 845.64 19.1 11.6 221.56 26.20 S. Afr VL 898 34.0 23.6 630.20 22.0 11.2 246.40 39.10 S. Afr VL 5263 35.3 29.5 817.87 24.2 10.0 242.00 29.59 S. Afr VL 5281 37.8 27.0 801.58 20.2 11.5 232.30 28.98 S. Afr 42877 39.5 29.0 899.67 23.2 11.3 262.16 29.14

187

S. Afr 99-78 39.2 29.6 911.31 18.5 11.7 216.45 23.75 S. Afr VL 4313 36.6 30.0 862.37 16.9 8.9 150.41 17.44 S. Afr 99-6913 36.0 29.0 819.95 24.6 11.2 275.52 33.60 S. Afr VL 3280 34.1 30.6 819.53 21.2 10.5 222.60 27.16 S. Afr VL 3281 34.0 28.6 763.72 20.6 13.0 267.80 35.07 S. Afr 99-8428 35.7 29.8 835.55 21.0 11.1 233.10 27.90 S. Afr 99-8429 38.6 27.3 827.64 24.5 11.5 281.75 34.04 S. Afr 99-8430 41.6 33.4 1091.26 22.5 9.0 202.50 18.56 S. Afr 99-8431 36.0 28.2 797.34 23.6 11.0 259.60 32.56 S. Afr 99-8253 40.8 32.2 1031.82 23.0 12.5 287.50 27.86 S. Afr 99-76 38.2 30.2 906.07 22.8 10.2 232.56 25.67 S. Afr VL 2462 32.2 25.6 647.42 20.6 13.0 267.80 41.36 S. Afr VL 2463 35.0 30.5 838.41 22.8 12.0 273.60 32.63 S. Afr VL 2464 37.6 27.0 797.34 22.0 10.2 224.40 28.14 S. Afr VL 2469 37.0 29.1 845.64 24.0 11.2 268.80 31.79 S. Afr VL 2470 35.5 26.2 730.50 17.5 10.2 178.50 24.44 S. Afr VL 2793 35.4 27.0 750.68 22.5 10.8 243.00 32.37 S. Afr 99-8433 33.7 29.3 775.51 19.8 10.8 213.84 27.57 S. Afr 99-8436 36.0 28.4 802.99 22.0 10.1 222.20 27.67 S. Afr 99-77 37.2 31.7 926.17 20.4 10.9 222.36 24.01 S. Afr VL 3579 39.0 33.7 1032.25 24.6 10.0 246.00 23.83 S. Afr VL 3461 37.9 33.8 1006.11 26.8 10.3 276.04 27.44 S. Afr VL 3462 42.5 33.8 1128.22 26.2 13.7 358.94 31.81 S. Afr VL 3573 36.8 29.1 841.07 24.2 10.7 258.94 30.79 S. Afr VL 3574 34.0 30.0 801.11 19.0 13.2 250.80 31.31 S. Afr VL 3576 35.1 27.2 749.83 22.0 12.5 275.00 36.67 S. Afr VL 3578 36.0 31.1 879.33 19.2 13.9 266.88 30.35 S. Afr VL 3580 36.2 30.3 861.47 21.2 13.8 292.56 33.96 S. Afr 42876 37.9 31.5 937.65 25.2 14.5 365.40 38.97 29.61

Utah 99-7446 36.4 30.6 874.81 21.7 12.5 271.25 31.01 Utah 99-7447 37.0 26.9 781.71 23.9 11.8 282.02 36.08 Utah 99-7448 39.0 31.0 949.55 27.0 11.0 297.00 31.28 Utah 99-7450 32.1 26.2 660.54 24.0 13.9 333.60 50.50 Utah 99-7451 32.8 30.0 772.83 24.0 12.4 297.60 38.51 Utah 99-7454 30.5 25.9 620.42 19.0 13.0 247.00 39.81 Utah 99-7455 35.0 29.5 810.92 22.7 12.5 283.75 34.99 Utah 99-7460 36.7 27.7 798.43 23.6 13.0 306.80 38.43 Utah 99-7463 33.2 26.2 683.17 20.6 12.2 251.32 36.79 Utah 99-7464 30.8 25.1 607.18 20.0 12.8 256.00 42.16 Utah 99-7466 33.0 27.0 699.79 21.0 14.6 306.60 43.81 Utah 99-7467 34.2 26.9 722.55 24.8 12.8 317.44 43.93 Utah 99-7469 33.6 26.3 694.04 21.0 13.0 273.00 39.33 Utah 99-7385 35.3 25.7 712.52 24.6 12.8 314.88 44.19 Utah 99-7364 33.6 28.5 752.10 26.1 12.0 313.20 41.64 Utah 99-7384 29.7 23.5 548.17 22.5 13.2 297.00 54.18 Utah 99-7382 37.2 28.0 818.07 23.8 11.0 261.80 32.00 39.92

188

W. Afr VL 365 36.1 32.4 918.63 24.6 13.5 332.10 36.15 W. Afr VL 366 40.0 30.0 942.48 21.5 12.0 258.00 27.37 W. Afr VL 362 34.8 26.9 735.23 20.0 13.0 260.00 35.36 W. Afr VL 363 31.2 24.6 602.81 21.7 11.0 238.70 39.60 W. Afr VL 448 37.4 29.5 866.53 24.0 12.2 292.80 33.79 W. Afr VL 449 37.3 27.6 808.55 22.7 13.0 295.10 36.50 W. Afr VL 574 33.2 26.0 677.96 18.8 13.6 255.68 37.71 W. Afr VL 452 40.4 31.0 983.63 21.1 12.0 253.20 25.74 W. Afr VL 576 35.1 26.0 716.75 25.0 14.7 367.50 51.27 W. Afr VL 577 34.9 29.0 794.90 22.9 13.3 304.57 38.32 W. Afr VL 578 37.5 30.0 883.57 22.2 16.0 355.20 40.20 W. Afr VL 5264 35.2 27.4 757.50 22.0 14.8 325.60 42.98 W. Afr VL 5274 34.6 28.5 774.48 25.0 14.0 350.00 45.19 W. Afr VL 400 36.0 27.0 763.41 29.2 13.2 385.44 50.49 W. Afr VL 470 39.4 30.5 943.81 21.7 12.2 264.74 28.05 W. Afr VL 471 37.2 30.6 894.03 25.7 12.0 308.40 34.50 W. Afr VL 1756 38.0 29.9 892.37 22.7 13.0 295.10 33.07 W. Afr VL 1757 35.6 29.2 816.44 22.0 12.0 264.00 32.34 W. Afr VL 4924b 39.7 29.8 929.17 21.7 13.0 282.10 30.36 W. Afr VL 2472 38.9 28.7 876.84 21.0 12.4 260.40 29.70 W. Afr SUL 761 31.0 30.0 730.42 20.2 12.3 248.46 34.02 36.32

Egypt SUL 267 32.0 28.0 703.72 24.5 15.3 374.85 53.27 Egypt SUL 274 33.0 30.0 777.54 19.5 11.5 224.25 28.84 Egypt SUL 280 36.0 28.0 791.68 21.9 13.7 300.03 37.90 Egypt SUL 282 35.0 28.0 772.81 22.0 10.5 231.00 29.89 Egypt SUL 307 32.0 28.0 703.76 23.9 10.5 250.95 35.66 Egypt SUL 311 35.0 30.0 824.67 22.2 14.7 326.34 39.57 Egypt SUL 318 33.0 27.0 958.97 24.8 9.6 238.08 24.83 Egypt SUL 319 32.0 29.0 728.85 22.2 11.0 244.20 33.50 Egypt SUL 321 34.0 27.0 988.03 21.2 11.5 243.80 24.68 Egypt SUL 324 32.0 25.0 628.32 23.0 10.5 241.50 38.44 Egypt SUL 326 35.0 28.0 769.69 21.5 9.0 193.50 25.14 Egypt SUL 336 36.0 23.0 650.31 21.4 10.3 220.42 33.89 Egypt SUL 343 32.0 30.0 753.98 18.2 11.0 200.20 26.55 33.24

Java SCK 1163 36.5 29.5 845.68 21.7 14.5 314.65 37.21 Java SCK 1165 33.2 29.2 761.40 21.5 11.0 236.50 31.06 Java SCK 1166 35.0 26.0 714.71 24.7 11.0 271.70 38.02 Java SCK 1167 35.5 28.0 780.69 24.0 9.5 228.00 29.20 Java SCK 1184 35.0 30.0 824.67 24.5 10.0 245.00 29.71 Java SCK 1187 32.7 27.0 693.43 23.3 10.5 244.65 35.28 Java SCK 1230 36.0 28.5 805.82 23.5 11.7 274.95 34.12 Java SCK 1291 33.0 29.0 751.63 21.0 12.0 252.00 33.53 Java SUL 818 35.0 28.0 769.69 25.0 12.0 300.00 38.98 Java SUL 820 32.0 29.0 728.85 22.2 13.5 299.70 41.12 Java SUL 823 35.0 27.0 742.20 21.3 10.5 223.65 30.13

189

Java SUL 825 36.0 34.0 961.33 24.7 11.2 276.64 28.78 Java SUL 828 35.0 29.0 797.18 23.4 9.3 217.62 27.30 Java SUL 832 34.0 30.0 801.11 23.6 11.1 261.96 32.70 Java SUL 836 36.0 32.3 913.26 21.6 13.3 287.28 31.46 Java SUL 838 32.0 28.0 703.72 22.8 12.2 278.16 39.53 Java SUL 840 37.0 33.0 958.97 22.0 11.4 250.80 26.15 Java SUL 849 35.0 30.0 824.67 25.5 11.5 293.25 35.56 Java SUL 852 30.0 26.0 848.23 22.7 12.0 272.40 32.11 Java SUL 856 35.4 30.7 853.55 24.6 11.2 275.52 32.28 Java SUL 858 33.0 30.0 777.54 18.5 11.1 205.35 26.41 Java SUL 868 33.0 30.0 777.54 19.0 10.0 190.00 24.44 Java SUL 873 38.0 32.0 955.04 25.0 14.3 357.50 37.43 31.40

India SCK 178 37.1 30.8 897.46 20.2 11.5 232.30 25.88 India SCK 185 38.7 32.8 996.95 22.2 13.8 306.36 30.73 India SCK 187 39.5 31.5 977.23 20.5 12.5 256.25 26.22 India SCK 193 35.7 31.2 874.81 24.0 14.1 338.40 38.68 India SCK 195 37.0 30.1 874.70 23.0 16.9 388.70 44.44 India SCK 1274 35.2 29.0 801.73 24.0 12.0 288.00 35.92 India SCK 1276 41.0 32.2 1036.88 25.1 12.2 306.22 29.53 India SCK 1277 37.8 31.0 920.33 22.8 11.0 250.80 27.25 India SCK 1280 36.0 28.0 791.68 24.2 10.3 249.26 31.48 India SCK 1284 33.0 27.0 699.79 22.4 12.6 282.24 40.33 33.05

China SCK 1201 35.5 31.7 883.85 21.2 10.7 226.84 25.66 China SCK 1203 36.5 31.5 903.01 23.5 12.0 282.00 31.23 China SCK 1206 34.3 29.5 794.70 23.0 15.0 345.00 43.41 China SCK 1208 32.0 28.0 703.72 22.8 12.2 278.16 39.53 China SCK 1209 35.0 30.5 838.41 20.5 11.5 235.75 28.12 China SCK 1210 35.0 30.9 849.41 23.1 11.0 254.10 29.91 China SCK 1211 43.0 34.5 1165.14 24.0 12.5 300.00 25.75 China SCK 1295 34.5 27.5 745.15 22.0 11.7 257.40 34.54 China SCK 5343 29.4 26.0 600.36 17.0 12.1 205.70 34.26 China SCK 5344 34.2 29.5 792.39 20.0 10.0 200.00 25.24 31.77

Ngandong 7 42.0 31.0 1022.59 16.7 11.0 183.70 17.96 Ngandong 12 40.0 29.0 911.06 21.0 9.0 189.00 20.75

Sangiran 4 38.8 29.0 883.73 25.6 13.6 348.16 39.40 Sangiran 17 36.8 29.9 864.19 24.5 11.0 269.50 31.19

Wajak 1 25.0 10.0 Cohuna 24.0 11.0 Keilor 41.0 35.0 1127.05 19.0 14.0 266.00 23.60 Lake Nitchie 35.8 38.5 801.34 22.5 11.5 258.75 32.29

190

Appendix D

Results of Manova tests on the modern comparative samples

191

Table D1: Results of Manova test for expression of the pharyngeal tubercle in the modern samples. Score denotes pairwise difference in mean expression of trait (bottom-left). Statistically significant comparisons (p=.05) are in bold text.

Australia 0.000 Austria -0.418 0.000 China -0.352 0.067 0.000 Czech -0.407 0.011 -0.056 0.000 Egypt 0.084 0.503 0.436 0.491 0.000 GG, Utah 0.003 0.421 0.354 0.410 -0.082 0.000 Greece -0.497 -0.079 -0.146 -0.090 -0.582 -0.500 0.000 India -0.085 0.333 0.267 0.322 -0.169 -0.087 0.413 0.000 Java 0.042 0.461 0.394 0.449 -0.042 0.040 0.540 0.127 0.000 S. Africa -0.018 0.400 0.333 0.389 -0.103 -0.021 0.479 0.067 -0.061 0.000 W. Africa -0.163 0.255 0.188 0.244 -0.247 -0.166 0.334 -0.078 -0.206 -0.145 0.000

Australia Austria China Czech Egypt GG, Utah Greece India Java S. Africa W. Africa

Table D2: Results of Manova test for expression of tympanic plate contact with mastoid in the modern samples. Score denotes pairwise difference in mean expression of trait (bottom-left). Statistically significant comparisons (p=.05) are in bold text.

Australia 0.000 Austria -0.027 0.000 China -0.027 0.000 0.000 Czech 0.084 0.111 0.111 0.000 Egypt -0.027 0.000 0.000 -0.111 0.000 GG, Utah -0.027 0.000 0.000 -0.111 0.000 0.000 Greece -0.027 0.000 0.000 -0.111 0.000 0.000 0.000 India -0.027 0.000 0.000 -0.111 0.000 0.000 0.000 0.000 Java -0.027 0.000 0.000 -0.111 0.000 0.000 0.000 0.000 0.000 S. Africa -0.027 0.000 0.000 -0.111 0.000 0.000 0.000 0.000 0.000 0.000 W. Africa 0.060 0.087 0.087 -0.024 0.087 0.087 0.087 0.087 0.087 0.087 0.000

Australia Austria China Czech Egypt GG, Utah Greece India Java S. Africa W. Africa

192

Table D3: Results of Manova test for expression of the alar tubercles in the modern samples. Score denotes pairwise difference in mean expression of trait (bottom-left). Statistically significant comparisons (p=.05) are in bold text.

Australia 0.000 Austria 0.580 0.000 China -0.309 -0.889 0.000 Czech -0.087 -0.667 0.222 0.000 Egypt 0.093 -0.487 0.402 0.179 0.000 GG, Ut ah -0.441 -1.021 -0.132 -0.354 -0.534 0.000 Greece -0.003 -0.583 0.306 0.083 -0.096 0.437 0.000 India 0.047 -0.533 0.356 0.133 -0.046 0.488 0.050 0.000 Java -0.072 -0.652 0.237 0.015 -0.164 0.369 -0.068 -0.118 0.000 S. Africa 0.080 -0.500 0.389 0.167 -0.013 0.521 0.083 0.033 0.152 0.000 W. Africa 0.073 -0.507 0.382 0.159 -0.020 0.514 0.076 0.026 0.144 -0.007 0.000

Australia Austria China Czech Egypt GG, Utah Greece India Java S. Africa W. Africa

Table D4: Results of Manova test for expression of the orientation of the occipital condyles in the modern samples. Score denotes pairwise difference in mean expression of trait (bottom-left). Statistically significant comparisons (p=.05) are in bold text.

Australia 0.000 Austria 0.530 0.000 China -0.014 -0.067 0.000 Czech 0.097 0.044 0.111 0.000 Egypt 0.063 0.010 0.077 -0.034 0.000 GG, Utah 0.049 -0.004 0.063 -0.049 -0.014 0.000 Greece -0.014 -0.067 0.000 -0.111 -0.077 -0.062 0.000 India -0.014 -0.067 0.000 -0.111 -0.077 -0.062 0.000 0.000 Java -0.014 -0.067 0.000 -0.111 -0.077 -0.063 0.000 0.000 0.000 S. Africa -0.014 -0.067 0.000 -0.111 -0.077 -0.063 0.000 0.000 0.000 0.000 W. Africa 0.030 -0.023 0.043 -0.068 -0.033 -0.019 0.043 0.043 0.043 0.043 0.000

Australia Austria China Czech Egypt GG, Utah Greece India Java S. Africa W. Africa

193

Table D5: Results of Manova test for expression of the the opisthionic recess in the modern samples. Score denotes pairwise difference in mean expression of trait (bottom- left). Statistically significant comparisons (p=.05) are in bold text.

Australia 0.000 Austria -0.224 0.000 China -0.113 0.111 0.000 Czech -0.057 0.167 0.056 0.000 Egypt -0.198 0.026 -0.085 -0.141 0.000 GG, Utah -0.078 0.146 0.035 -0.021 0.120 0.000 Greece -0.015 0.208 0.097 0.042 0.183 0.062 0.000 India -0.190 0.033 -0.078 -0.133 0.008 -0.112 -0.175 0.000 Java -0.027 0.197 0.086 0.030 0.171 0.051 -0.011 0.164 0.000 S. Africa -0.015 0.208 0.097 0.042 0.183 0.062 0.000 0.175 0.011 0.000 W. Africa -0.021 0.203 0.092 0.036 0.177 0.057 -0.005 0.170 0.006 -0.005 0.000

Australia Austria China Czech Egypt GG, Utah Greece India Java S. Africa W. Africa

Table D6: Results of Manova test for expression of the postcondyloid tuberosities in the modern samples. Score denotes pairwise difference in mean expression of trait (bottom- left). Statistically significant comparisons (p=.05) are in bold text.

Australia 0.000 Austria 0.599 0.000 China 0.355 -0.244 0.000 Czech 0.466 -0.133 0.111 0.000 Egypt 0.235 -0.364 -0.120 -0.231 0.000 GG, Ut ah 0.466 -0.133 0.111 0.000 0.231 0.000 Greece 0.466 -0.133 0.111 0.000 0.231 0.000 0.000 India 0.366 -0.233 0.011 -0.100 0.131 -0.100 -0.100 0.000 Java 0.148 -0.452 -0.207 -0.318 -0.087 -0.318 -0.318 -0.218 0.000 S. Africa 0.632 0.033 0.278 0.167 0.397 0.167 0.167 0.267 0.485 0.000 W. Africa 0.770 0.171 0.415 0.304 0.535 0.304 0.304 0.404 0.623 0.138 0.000

Australia Austria China Czech Egypt GG, Utah Greece India Java S. Africa W. Africa

194

Table D7: Results of Manova test for expression of the foramen lacerum in the modern samples. Score denotes pairwise difference in mean expression of trait (bottom-left). Statistically significant comparisons (p=.05) are in bold text.

Australia 0.000 Austria -0.479 0.000 China -0.146 0.333 0.000 Czech -0.479 0.000 -0.333 0.000 Egypt -0.095 0.385 0.051 0.385 0.000 GG, Utah 0.083 0.562 0.229 0.563 0.178 0.000 Greece -0.479 0.000 -0.333 0.000 -0.385 -0.563 0.000 India -0.379 0.100 -0.233 0.100 -0.285 -0.462 0.100 0.000 Java -0.343 0.136 -0.197 0.136 -0.248 -0.426 0.136 0.036 0.000 S. Africa 0.271 0.750 0.417 0.750 0.365 0.188 0.750 0.650 0.614 0.000 W. Africa -0.001 0.478 0.145 0.478 0.094 -0.084 0.478 0.378 0.342 -0.272 0.000

Australia Austria China Czech Egypt GG, Utah Greece India Java S. Africa W. Africa

Table D8: Results of Manova test for expression of the juxtamastoid process in the modern samples. Score denotes pairwise difference in mean expression of trait (bottom- left). Statistically significant comparisons (p=.05) are in bold text.

Australia 0.000 Austria -0.696 0.000 China -0.429 0.267 0.000 Czech -0.596 0.100 -0.167 0.000 Egypt -0.557 0.138 -0.128 0.038 0.000 GG, Utah 0.154 0.850 0.583 0.750 0.712 0.000 Greece -0.533 0.163 -0.104 0.063 0.024 -0.688 0.000 India -0.096 0.600 0.333 0.500 0.462 -0.250 0.437 0.000 Java -0.505 0.191 -0.076 0.091 0.052 -0.659 0.028 -0.409 0.000 S. Africa 0.321 1.017 0.750 0.917 0.878 0.167 0.854 0.417 0.826 0.000 W. Africa -0.052 0.643 0.377 0.543 0.505 -0.207 0.481 0.043 0.453 -0.373 0.000

Australia Austria China Czech Egypt GG, Utah Greece India Java S. Africa W. Africa

195

Table D9: Results of Manova test for expression of the occipitomastoid crest in the modern samples. Score denotes pairwise difference in mean expression of trait (bottom- left). Statistically significant comparisons (p=.05) are in bold text.

Australia 0.000 Austria 0.215 0.000 China -0.008 -0.222 0.000 Czech 0.159 -0.056 0.167 0.000 Egypt 0.009 -0.205 0.017 -0.150 0.000 GG, Utah 0.048 -0.167 0.056 -0.111 0.038 0.000 Greece 0.360 0.146 0.368 0.201 0.351 0.313 0.000 India 0.048 -0.167 0.056 -0.111 0.038 0.000 -0.312 0.000 Java 0.366 0.152 0.374 0.207 0.357 0.318 0.006 0.318 0.000 S. Africa -0.327 -0.542 -0.319 -0.468 -0.337 -0.375 -0.688 -0.375 -0.693 0.000 W. Africa 0.331 0.116 0.338 0.171 0.321 0.283 -0.030 0.283 -0.036 0.658 0.000

Australia Austria China Czech Egypt GG, Utah Greece India Java S. Africa W. Africa

Table D10: Results of Manova test for expression of the location of the carotid foramen relative to the squamotympanic fissure in the modern samples. Score denotes pairwise difference in mean expression of trait (bottom-left). Statistically significant comparisons (p=.05) are in bold text.

Australia 0.000 Austria -0.688 0.000 China -0.732 -0.044 0.000 Czech -0.510 0.178 0.222 0.000 Egypt -0.288 0.400 0.444 0.222 0.000 GG, Ut ah -0.975 -0.287 -0.243 -0.465 -0.688 0.000 Greece -0.475 0.213 0.257 0.035 -0.188 0.500 0.000 India -0.688 0.000 0.044 -0.178 0.400 0.288 -0.212 0.000 Java -0.833 -0.145 -0.101 -0.323 -0.545 0.142 -0.358 -0.145 0.000 S. Africa -0.329 0.358 0.403 0.181 -0.042 0.646 0.146 0.358 0.504 0.000 W. Africa -0.201 0.487 0.531 0.309 0.087 0.774 0.274 0.487 0.632 0.129 0.000

Australia Austria China Czech Egypt GG, Utah Greece India Java S. Africa W. Africa

196

Table D11: Results of Manova test for the size of the postglenoid tubercle in the modern samples. Score denotes pairwise difference in mean expression of trait (bottom-left). Statistically significant comparisons (p=.05) are in bold text.

Australia 0.000 Austria 0.120 0.000 China -0.014 -0.133 0.000 Czech 0.153 0.033 0.167 0.000 Egypt -0.091 -0.210 -0.077 -0.244 0.000 GG, Ut ah 0.174 0.054 0.188 0.021 0.264 0.000 Greece 0.174 0.054 0.188 0.021 0.264 0.000 0.000 India -0.014 -0.133 0.000 -0.167 0.077 -0.187 -0.188 0.000 Java -0.014 -0.133 0.000 -0.167 0.077 -0.188 -0.188 0.000 0.000 S. Africa -0.014 -0.133 0.000 -0.167 0.077 -0.188 -0.188 0.000 0.000 0.000 W. Africa 0.030 -0.090 0.043 -0.123 0.120 -0.144 -0.144 0.043 0.043 0.043 0.000

Australia Austria China Czech Egypt GG, Utah Greece India Java S. Africa W. Africa

Table D12: Results of Manova test for expression of the orientation of the squamotympanic fissure in the modern samples. Score denotes pairwise difference in mean expression of trait (bottom-left). Statistically significant comparisons (p=.05) are in bold text.

Australia 0.000 Austria -0.254 0.000 China -0.521 -0.267 0.000 Czech -0.409 -0.156 0.111 0.000 Egypt -0.367 -0.113 0.154 0.043 0.000 GG, Ut ah -0.521 -0.267 0.000 -0.111 -0.154 0.000 Greece -0.333 -0.079 0.188 0.076 0.034 0.188 0.000 India -0.321 -0.067 0.200 0.089 0.046 0.200 0.013 0.000 Java -0.430 -0.176 0.091 -0.020 -0.063 0.091 -0.097 -0.109 0.000 S. Africa -0.021 0.233 0.500 0.389 0.346 0.500 0.312 0.300 0.409 0.000 W. Africa 0.045 0.299 0.565 0.454 0.411 0.565 0.378 0.365 0.474 0.065 0.000

Australia Austria China Czech Egypt GG, Utah Greece India Java S. Africa W. Africa

197

Appendix E

Cranial base photographs of the Indonesian fossil sample

198

Figure E.1: Sangiran 2 cranial base. Approximately 1 cm of the posterior border of the foramen magnum survives at opisthion. Both TMJs are present, with the left fossa better represented. Both mastoid processes survive, as do their surrounding areas.

199

Figure E.2: Sangiran 2 posterior foramen magnum (cast). Approximately 1 cm of the posterior rim of the foramen magnum at opisthion survives. Though little is here, the surviving contour suggests a rounded border with no narrowing.

200

Figure E.3: Sangiran 2 left TMJ (cast). This specimen exhibits a clear postglenoid tubercle that is anterior to the S-Q fissure. The S-Q fissure is not coincident with the apex of the fossa. The mastoid process is extremely small, and it does not contact the tympanic plate.

201

Figure E.4: Sangiran 4 cranial base. Reconstructed areas are in dark brown. Both TMJs are present, and both tympanic plates are likewise preserved though these have undergone some deformation. The occipital condyles are present, though these have sustained some damage, and the foramen magnum is largely intact and relatively undamaged. Both f. ovale are present and exhibit the typical undivided morphology.

202

Figure E.5: Sangiran 4 left TMJ (cast). The red line denotes the course of the S-Q fissure. The large postglenoid tubercle diverts the course of the fissure posterior to the apex of the fossa, and separates the fossa from the anterior wall of the tympanic plate.

203

Figure E.6: Sangiran 4 foramina ovale. These foramina show the typical hominid morphology with no indication of doubling.

Figure E.7: Sangiran 4 foramen magnum and occipital condyles. The occipital condyles are large both absolutely as well as relative to the size of the foramen magnum, and are within the modern human range for relative size. The condyles are also angled toward the midline at their anterior end. The foramen magnum is fairly rounded with no constriction near opisthion. There is also no development of PCT.

204

Figure E.8: Sangiran 12 occipital. Little of the cranial base survives on this specimen, though the posterior border of the foramen magnum is preserved for ~2 cm (denoted by the white lines). From what is present it is clear that the foramen is not constricted into an opisthionic recess. PCT are also not present on this specimen.

205

Figure E.9: Sangiran 14 basioccipital. The occipital condyles are damaged, but are similar in size to those preserved in the Ngandong series. Unlike Ngandong, however, the long axes of the condyles are not parallel to the midline and angle towards the midline at their anterior end. The alar tubercles are strongly developed, and there is development of a midline ridge on the basioccipital instead of a single discrete pharyngeal tubercle. This morphology is similar to that found on the basioccipitals of Ngandong 7 and 12.

206

Figure E.10: Sangiran 17 cranial base. Note the large size of the occipital condyles both absolutely as well as relative to the size of the foramen magnum. The occipital condyles also angle towards the midline at the anterior end and are clearly not parallel to one another. The foramen magnum is roughly diamond shaped, with some narrowing near opisthion, however PCT are not present. A postglenoid tubercle is present on the more well-preserved right TMJ. 207

Figure E.11: Sangiran 17 right TMJ. This specimen possesses a postglenoid tubercle that lies anterior to the S-Q fissure. This fissure is posterior to the apex of the fossa along its entire length.

Figure E.12: Sangiran 17 foramen magnum. Note the lack of PCT development and the large, medially angled occipital condyles. The narrowing in the posterior foramen magnum does bear some similarities to the Ngandong fossils.

208

Figure E.13: Ngandong 7 cranial base. The number “6” on the left orbital roof refers to its number in the “Original” system. Note the small size of the occipital condyles both absolutely as well as relative to the area of the foramen magnum. The posterior foramen magnum exhibits an opisthionic recess, and PCT are present bilaterally as well. Juxtamastoid and occipitomastoid crests are present bilaterally.

209

Figure E.14: Ngandong 11 cranial base. The number “10” on the inside of the frontal bone refers to its number in the “Original” system. Much of the base is missing, however the specimen provides additional evidence for the consistency of the Ngandong morphological suite. Both TMJs are at least partially represented, and both show the S-Q fissure wholly in the apex of the fossa. Much of the left rim of the foramen magnum survives, and it exhibits an opisthionic recess. A well-developed PCT is also present.

210

Figure E.15: Ngandong 12 cranial base. Note the asymmetrical anterior border of the foramen magnum (due to breakage), the constricted posterior border of the foramen magnum (though a bit less constricted than in Ngandong 1 or 7), the large PCT, and the morphology of the TMJ. The septum of the doubled foramen ovale can also be seen on the left, visible at the medial edge of the foramen. Well-developed juxtamastoid and occipitomastoid crests are present bilaterally. 211

Figure E.16: Ngandong 12 right posterior cranial base. This diagram highlights many of the features discussed in the text.

212

Figure E.17: Ngandong 12 foramen magnum. Anterior border is damaged (Santa Luca, 1980), and photo at top shows a reconstructed view with the stippled area replacing damaged bone. The black outlines depict the probable location and orientation of the occipital condyles gleaned from the surviving condylar bases. 213

Figure E.18: Ngandong 7 foramen magnum. At the anterior end of the foramen note the inferiorly projecting rim, which is approximately 5-7 mm inferior to the rest of the basioccipital. The occipital condyles are set parallel to one another and are small both absolutely as well as relative to the area of the foramen magnum. The posterior end of the foramen is constricted into the opisthionic recess, which is flanked by well-developed PCT.

214

Figure E.19: Ngandong 11 posterior cranial base. Though damaged, this specimen exhibits several of the features noted in the more complete Ngandong specimens. The surviving border of the foramen magnum clearly shows the characteristic narrowing of an opisthionic recess. A pronounced PCT is also present. Much of the tympanic bone is missing on the right side, but in both mandibular fossae the S-Q fissure is located in the apex of the fossa.

215

Figure E.20: Ngandong 12 left foramen ovale. Two views of the f. ovale complex, including the broad septum (indicated by arrows) and the accessory foramen lying medially to the main foramen. The top of the page is medial in the upper photo, while the upper right corner is medial in the lower photo.

216

Figure E.21: Ngandong 7 left foramen ovale. Scale bar in cm. Main foramen is visible just above the center of the photo. Lying medial (left in photo) to that is the ovale septum, and medial to that and barely visible is the accessory foramen.

Figure E.22: Ngandong 12 left tympanomastoid fissure in lateral view. Also note the lack of a postglenoid tubercle separating the TMJ from the tympanic plate.

217

Figure E.23: Sambungmacan 1 left TMJ. Note the lack of a large postglenoid tubercle, and that the S-Q fissure (shown in red on the right) courses in the apex of the mandibular fossa.

218

Figure E.24: Sambungmacan 3 posterior foramen magnum (cast). Opisthion is preserved, as is much of the posterior right border of the foramen magnum. The specimen exhibits narrowing in the posterior foramen magnum, and the characteristic bulge of a PCT.

Figure E.25: Sambungmacan 3 left TMJ (cast). Note the lack of a well-defined postglenoid tubercle and the location of the S-Q fissure in the apex of the mandibular fossa. Also note the ridge anterior to the S-Q fissure on the anterior wall of the fossa.

219

Figure E.26: Sambungmacan 4 cranial base. Many pertinent characters are shown here, including the opisthionic recess, well-developed PCT with no rugosity at opisthion itself, the doubled foramen ovale, and the relatively small occipital condyles that are set relatively anteriorly on the borders of the foramen magnum. Also note the lack of a delicate vaginal process of the tympanic bone (which is instead a thickened, rugose structure), and the lack of styloid processes. Styloid pits are present bilaterally. Baba and colleagues (2003, 2004) contend that the location of the S-Q fissure in this specimen is transitional between the condition seen in the earlier Sangiran material and Ngandong.

220

Figure E.27: Wajak 1 cranial base. This specimen has been extensively reconstructed, and the detail in many areas is poor. Despite these limitations, some relevant information can still be discerned. The TMJ clearly possesses a postglenoid tubercle, though it is eroded. The shape of the foramen magnum is undistorted and does not exhibit any narrowing posteriorly. PCT are likewise not present. The occipital condyles are large both absolutely as well as relative to the area of the foramen magnum. The condyles are also clearly angled toward the midline anteriorly.

221

Appendix F

Cranial base photographs of the Australian fossil sample

222

Figure F.1: Lake Mungo 1 occipital bone near (or at) opisthion (cast). Note the pronounced occipital crest and the excavated areas flanking it bilaterally.

Figure F.2: Lake Mungo III occipital bone (cast). Specimen preserves opisthion and the posterior right border of the foramen magnum. Note the prominent occipital crest and uniform thickening of the foramen magnum border.

223

Figure F.3: Kow Swamp 5 cranial base (cast).

224

Figure F.4: Kow Swamp 5 left TMJ, mastoid, and occipital (cast). Note the prominent postglenoid tubercle, vaginal process of the tympanic bone, and lack of juxtamastoid and occipitomastoid cresting.

Figure F.5: Kow Swamp 5 midline occipital bone (cast). Note the undulating occipital crest, and the lack of thickening or tubercle development at the border of the foramen magnum.

225

Figure F.6: Keilor cranial base (cast).

226

Figure F.7: Keilor left TMJ and tympanic bone (cast). Note the presence of a postglenoid tubercle and the course of the S-Q fissure well posterior to the apex of the fossa. Also present is a vaginal process of the tympanic bone housing the root of a styloid process.

Figure F.8: Keilor foramen magnum (cast). Note the fairly small occipital condyle, and the well rounded borders of the foramen magnum with no development of postcondyloid tuberosities.

227

Figure F.9: Nacurrie 1 left TMJ (cast). This area represents the only preserved cranial base morphology on this specimen. Note the prominent postglenoid tubercle and the location of the S-Q fissure posterior to the apex of the fossa. A vaginal process of the tympanic bone is present.

228

Figure F.10: Nacurrie 2 cranial base from Westaway (2002b). Note the rounded borders of the foramen magnum with some thickening and tubercle development, single foramen ovale with no bifurcation or ovale pit, large postglenoid tubercle on undeformed right side, and large occipital condyles.

229

Figure F.11: Mossgiel cranial base (cast).

230

Figure F.12: Mossgiel left TMJ (cast). Though it is arthritically deformed, a large postglenoid tubercle is visible.

Figure F.13: Mossgiel foramen magnum (cast). Note the smooth border of the foramen with no thickening or tubercle development. The surviving contour also indicates that the posterior foramen was not narrowed into an opisthionic recess.

231

Figure F.14: Cossack right TMJ (cast). Note the large postglenoid tubercle, with the S-Q fissure well posterior to the apex of the fossa.

232

Figure F.15: Cohuna cranial base (cast).

233

Figure F.16: Cohuna left TMJ and tympanic bone (cast). Note the large postglenoid tubercle. Although the tympanic bone is broken, the vaginal process and styloid sheath can be seen just to the left of center.

234

Figure F.17: Lake Nitchie cranial base (cast). Note the ritually evulsed central incisors.

235

Figure F.18: Lake Nitchie posterior cranial base (cast). The borders of the foramen magnum are quite rounded and broad posteriorly, and these contours are confirmed by surviving internal anatomy. Occipital condyles are large and angled medially at their anterior end. Both TMJs have large postglenoid tubercles and the S-Q fissure is posterior to the apex of the fossa. Also note the damage to the right jugular process of the occipital bone and the heavily pneumatized, undamaged left jugular process.

Figure F.19: Lake Nitchie right posterior cranial base (cast). Detail of postjugular fossa damage.

236

Vita

Arthur C. Durband was born on February 4, 1972 in Evanston, Il. He was raised in

Buffalo Grove, Il and attended Louisa May Alcott Elementary School (grades K-3),

Joyce Kilmer Elementary School (grades 4-6), and James Fenimore Cooper Junior High

School (grades 7-9), before finishing his secondary education at Buffalo Grove High

School, graduating in May, 1990. From there he attended Northern Illinois University in

DeKalb, receiving a Bachelor of Arts degree in English and Anthropology in December,

1994. While he had originally planned on pursuing a graduate degree in English with a focus on the American short story, that course was abandoned after he took his first upper division course in human paleontology. Arthur stayed on at Northern Illinois University for his early graduate work in Anthropology, completing his Master of Arts degree in

December, 1997. His thesis, oddly enough, involved a test of the Multiregional hypothesis of modern human origins using the cranial base evidence from Australasia.

After spending seven long years in DeKalb, it was time for a change of scenery.

In August, 1997, Arthur began his doctorate at the University of Tennessee. After another seven long (but generally very pleasant) years in Knoxville he earned his Ph.D. in

Anthropology in December, 2004.

Arthur is currently employed as a Visiting Assistant Professor of Anthropology at his old Alma Mater, Northern Illinois University.

237