*For correspondence: Lee. Berger@wits.ac.za Competing interests: The authors declare that no competing interests exist. Funding: See page 57 Received: 12 December 2016 Accepted: 18 April 2017 Published: 09 May 2017 Reviewing editor: George H Perry, Pennsylvania State University, United States Copyright Hawks et al. This article is distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use and redistribution provided that the original author and source are credited. New fossil remains of Homo naledi from the Lesedi Chamber, South Africa John Hawks1,2, Marina Elliott1, Peter Schmid1,3, Steven E Churchill1,4, Darryl J de Ruiter1,5, Eric M Roberts6, Hannah Hilbert-Wolf6, Heather M Garvin1,7,8, Scott A Williams1,9,10, Lucas K Delezene1,11, Elen M Feuerriegel1,12, Patrick Randolph-Quinney1,13,14, Tracy L Kivell1,15,16, Myra F Laird1,17, Gaokgatlhe Tawane1, Jeremy M DeSilva1,18, Shara E Bailey9,10, Juliet K Brophy1,19, Marc R Meyer20, Matthew M Skinner1,15,16, Matthew W Tocheri21,22, Caroline VanSickle1,2,23, Christopher S Walker1,4,24, Timothy L Campbell5, Brian Kuhn25, Ashley Kruger1,26, Steven Tucker1, Alia Gurtov1,2, Nompumelelo Hlophe1, Rick Hunter1, Hannah Morris1,27, Becca Peixotto1,28, Maropeng Ramalepa1, Dirk van Rooyen1, Mathabela Tsikoane1, Pedro Boshoff1, Paul HGM Dirks6, Lee R Berger1* 1Evolutionary Studies Institute, University of the Witwatersrand, Wits, South Africa; 2Department of Anthropology, University of Wisconsin, Madison, United States; 3Anthropological Institute and Museum, University of Zürich, Winterthurerstr, Zürich, Switzerland; 4Department of Evolutionary Anthropology, Duke University, Durham, United States; 5Department of Anthropology, Texas A&M University, College Station, United States; 6Geosciences, College of Science and Engineering, James Cook University, Townsville, Australia; 7Department of Anthropology/ Archaeology, Mercyhurst University, Erie, United States; 8Department of Applied Forensic Sciences, Mercyhurst University, Erie, United States; 9Center for the Study of Human Origins, Department of Anthropology, New York University, New York, United States; 10New York Consortium in Evolutionary Primatology, New York, United States; 11Department of Anthropology, University of Arkansas, Fayetteville, United States; 12Department of Anthropology, University of Washington, Seattle, United States; 13School of Anatomical Sciences, University of the Witwatersrand Medical School, Johannesburg, South Africa; 14School of Forensic and Applied Sciences, University of Central Lancashire, Preston, United Kingdom; 15School of Anthropology and Conservation, University of Kent, Canterbury, United Kingdom; 16Department of Human Evolution, Max Planck Institute for Evolutionary Anthropology, Leipzig, Germany; 17Department of Organismal Biology and Anatomy, University of Chicago, Chicago, United States; 18Department of Anthropology, Dartmouth College, Hanover, United States; 19Department of Geography and Anthropology, Louisiana State University, Baton Rouge, United States; 20Department of Anthropology, Chaffey College, Rancho Cucamonga, United States; 21Department of Anthropology, Lakehead University, Thunder Bay, Canada; 22Human Origins Program, Department of Anthropology, National Museum of Natural History, Smithsonian Institution, Washington, United States; 23Department of Anthropology, Bryn Mawr College, Bryn Mawr, United States; 24Department of Molecular Biomedical Sciences, College of Veterinary Medicine, North Carolina State University, Raleigh, United States; 25Department of Geology, University of Johannesburg, Johannesburg, South Africa; 26School of Geosciences, University of the Witwatersrand, Johannesburg, South Africa; 27Department of Hawks et al. eLife 2017;6:e24232. DOI: 10.7554/eLife.24232 1 of 63 RESEARCH ARTICLE http://creativecommons.org/licenses/by/4.0/ http://creativecommons.org/licenses/by/4.0/ http://dx.doi.org/10.7554/eLife.24232 https://creativecommons.org/ https://creativecommons.org/ http://elife.elifesciences.org/ http://elife.elifesciences.org/ http://en.wikipedia.org/wiki/Open_access http://en.wikipedia.org/wiki/Open_access Forestry and Natural Resources, University of Georgia, Athens, United States; 28Department of Anthropology, American University, Washington, United States Abstract The Rising Star cave system has produced abundant fossil hominin remains within the Dinaledi Chamber, representing a minimum of 15 individuals attributed to Homo naledi. Further exploration led to the discovery of hominin material, now comprising 131 hominin specimens, within a second chamber, the Lesedi Chamber. The Lesedi Chamber is far separated from the Dinaledi Chamber within the Rising Star cave system, and represents a second depositional context for hominin remains. In each of three collection areas within the Lesedi Chamber, diagnostic skeletal material allows a clear attribution to H. naledi. Both adult and immature material is present. The hominin remains represent at least three individuals based upon duplication of elements, but more individuals are likely present based upon the spatial context. The most significant specimen is the near-complete cranium of a large individual, designated LES1, with an endocranial volume of approximately 610 ml and associated postcranial remains. The Lesedi Chamber skeletal sample extends our knowledge of the morphology and variation of H. naledi, and evidence of H. naledi from both recovery localities shows a consistent pattern of differentiation from other hominin species. DOI: 10.7554/eLife.24232.001 Introduction The Rising Star cave system (26˚10130 0 S; 27˚4204300 E, Figure 1) in the Cradle of Humankind World Heritage Site, Gauteng Province, South Africa, is known for the discovery in 2013 of more than 1,550 fossils representing a novel hominin species, Homo naledi (Berger et al., 2015; Dirks et al., 2015). These remains, representing at least 15 individuals of various ages at death, were recovered from a deep chamber (30 m below ground surface), named the Dinaledi Chamber. Additional fossil hominin material was subsequently discovered in the Lesedi Chamber of the cave system in November 2013 by Rick Hunter and Steven Tucker. The deposition of sediment and skeletal remains in the Lesedi Chamber has no direct geological connection to the Dinaledi Cham- ber. In the time following the first discovery of hominin material in the Lesedi Chamber, excavators have recovered 131 hominin specimens within three discrete collection areas. The sedimentary con- text of the three collection areas is broadly similar, but we have not yet established whether the fos- sil material resulted from a single depositional episode or from multiple distinct events. We approached the hominin skeletal remains from the Lesedi Chamber with the aim of identifying elements, assessing the number of individuals represented by the material, and determining the tax- onomic identity of the sample. Preliminary examination of the hominin remains suggested that they are morphologically consistent with H. naledi. To test this hypothesis, we carried out systematic com- parisons, employing the taxonomic diagnosis of this species (Berger et al., 2015) and focusing upon those characters that distinguish H. naledi from other hominin taxa. We also present essential con- textual information to place the specimens within the Lesedi Chamber and provide descriptions of the hominin specimens, focusing upon those features that contribute to the taxonomic diagnosis of the sample. All identifiable hominin fragments, including those that do not present information use- ful to taxonomic diagnosis, are listed in Table 1. Results Name of the chamber Following the University of the Witwatersrand’s fossil-numbering system (Zipfel and Berger, 2009), this second H. naledi locality has been designated U.W. 102. The chamber itself has been named the Lesedi Chamber, a word meaning ‘light’ in Setswana. By contrast, the Dinaledi Chamber was num- bered site U.W. 101. Excavations in the Lesedi Chamber have been carried out in three areas, desig- nated U.W. 102a, U.W. 102b, and U.W. 102c. Hawks et al. eLife 2017;6:e24232. DOI: 10.7554/eLife.24232 2 of 63 Research article Genomics and Evolutionary Biology http://dx.doi.org/10.7554/eLife.24232.001 http://dx.doi.org/10.7554/eLife.24232 Location of the Lesedi Chamber The Lesedi Chamber is in the central sector of the Rising Star system (Figure 2), at a depth of ~30 m from the surface directly above the chamber. All measurements reported here are approximate. The first fossil deposit to be recognized (U.W. 102a) is located just off the southwest corner of the North-South Fracture Passage, a northern arm of the Lesedi Chamber. This fossil deposit is approxi- mately 60 m NNE in a straight line from the Dinaledi Chamber. There is no straight-line route between the Dinaledi and Lesedi Chambers, and the shortest traversable route between the two areas is almost 145 m. There are currently four access routes from the surface to the Lesedi Cham- ber. The most accessible of these currently follows an 86 m downward-sloping path with several nar- row passages and short climbs, but only one squeeze and no significant crawls. This has been the main access route for excavators. The other three routes are each substantially more challenging. Location of skeletal material within the Lesedi Chamber In addition to the first fossil deposit to be recognized in the chamber, two additional concentrations of skeletal material have been identified to date (Figure 3), and we have designated these as areas 102a, 102b, and 102c. We began investigating each of these areas because team members noticed hominin fossil material exposed on sediment surfaces. The discovery of 102a by Rick Hunter and Ste- ven Tucker led to the initial scientific investigation of the chamber; discoveries of both 102b and eLife digest Species of ancient humans and the extinct relatives of our ancestors are typically described from a limited number of fossils. However, this was not the case with Homo naledi. More than 1500 fossils representing at least 15 individuals of this species were unearthed from the Rising Star cave system in South Africa between 2013 and 2014. Found deep underground in the Dinaledi Chamber, the H. naledi fossils are the largest collection of a single species of an ancient human- relative discovered in Africa. After the discovery was reported, a number of questions still remained. These questions included: why were so many fossils from a single species found at the one site, and how did they come to rest so far into the cave system? Possible explanations such as H. naledi living in the cave or being washed in by a flood were considered but ruled out. Instead, the evidence was largely consistent with intact bodies being deliberately disposed of in the cave and then decomposing. Now, Hawks et al. – who include many of the researchers who were involved in the discovery of H. naledi – report that yet more H. naledi fossils have been unearthed from a second chamber in the Rising Star cave system, the Lesedi Chamber. The chamber is 30 meters below the surface and there is no direct route between it and the Dinaledi Chamber. Again, the evidence is most consistent with the bodies arriving intact into the chamber, and there were no signs that the remains had been exposed to the surface environment. Also like the Dinaledi Chamber, no remains of other ancient humans or their relatives were found in the Lesedi Chamber. In total, 133 fossils of H. naledi have been found in this second chamber representing at least three individuals: two adults and a juvenile. However, and as Hawks et al. point out, only a small volume of the chamber has been excavated so far, and so there are likely more fossils still to be found. The fossils in the Lesedi Chamber are similar to those found before but include intact examples of bones, like the collarbone, that were previously known only from fragments. Perhaps the most impressive among the new fossils is a relatively complete skull that is part of a partial skeleton. The skull could have housed a brain that was 9% larger than the maximum estimate calculated from the previous H. naledi fossils. Though these new fossils provide us with yet more information about H. naledi, some questions still remain unanswered – the material from the Lesedi Chamber is undated, for example. However, a related study by Dirks et al. does give an estimate for the age of the fossils from the Dinaledi Chamber, while Berger et al. provide an explanation for why this date might be much younger than was previously predicted. DOI: 10.7554/eLife.24232.002 Hawks et al. eLife 2017;6:e24232. DOI: 10.7554/eLife.24232 3 of 63 Research article Genomics and Evolutionary Biology http://dx.doi.org/10.7554/eLife.24232.002 http://dx.doi.org/10.7554/eLife.24232 102c were made by Hannah Hilbert-Wolf during the course of geological sampling of the chamber. These three areas do not represent a systematic sampling of the chamber’s contents and we have excavated only a very small sediment volume, less than 200 L (<0.2 m3) in total from all three areas. The chamber contains a much greater volume of sediment and we do not know what density of fossil bone it may contain beyond our samples. U.W. 102a is located at the entrance of a 20–50-cm-wide blind tunnel, which is 1.8 m long in total. The blind tunnel leads off of the southwest corner of the North-South Fracture Passage (Figure 3). Fossil material was exposed on the surface within this blind tunnel at the time of discovery. We have excavated the proximal 1.5 m of this blind tunnel, which has a tapering width of less than 50 cm in our excavation unit. The depth of excavation in this area is a maximum of 40 cm. The deposit in this area is a weakly stratified, unlithified mud-clast breccia. Most hominin material has been recovered from an approximately 10-cm-thick horizon of fine-grained mud-clast breccia, beneath a surface layer of ~2 cm of lighter brown-colored mudstone. This deposit is the source of at least some of the sediments that slope from the blind tunnel into the Antechamber. Fossil material attributed to 102a has also been recovered from the surface within the North-South Fracture Passage. U.W. 102b is a sediment deposit on a horizontal chert shelf 80 cm above the cave floor along the western wall of the Antechamber. It is also dominated by unlithified mud-clast breccia. The 102b deposit is located ~3.8 m to the south and 1.8 m below the 102a deposit. After the discovery of hominin fossil material on the surface here, we undertook limited excavations, with a total volume of ~20 L. U.W. 102c is a small unlithified sediment deposit within an irregular dissolution cavity on the north wall of the east–west-running Cake-Icing Fracture. This deposit is 1.3 m above the current cave floor. It is 11.6 m from U.W. 102a, and 0.3 m below the level of the 102a fossils. We have excavated this small sediment pocket in its entirely, with a total volume of approximately 2 L. Geological work to characterize the Lesedi Chamber depositional history is underway. The stratig- raphy is complex, with some hominin and faunal material concentrated in deposits of poorly consoli- dated mud-clast breccia, generally similar to the facies in the Dinaledi Chamber (Dirks et al., 2015). Notably, the fossil material in the Lesedi Chamber is concentrated in minor side fractures, dissolution cavities, or on chert shelves well above the current chamber floor. Our working hypothesis is that the chamber once held a greater volume of sediment than is present today, and when sediment eroded from the chamber, erosional remnants remained in protected fractures, wall cavities, and on chert shelves along the chamber walls. This and other indications of reworking of the deposits make it uncertain how much of the hominin assemblage may remain in its primary depositional context. Figure 1. Geographical location of the Rising Star cave in the Cradle of Humankind UNESCO World Heritage Site. DOI: 10.7554/eLife.24232.003 Hawks et al. eLife 2017;6:e24232. DOI: 10.7554/eLife.24232 4 of 63 Research article Genomics and Evolutionary Biology http://dx.doi.org/10.7554/eLife.24232.003 http://dx.doi.org/10.7554/eLife.24232 Table 1. Hominin fossil material from the Lesedi Chamber. All diagnostic hominin specimens are listed, with attribution to element. Specimens that have been refitted are not listed separately. Most Locality 102a cranial fragments are presumed to be part of LES1 and are not listed separately. Specimen number Element Notes LOCALITY 102a LES1 cranium constituted of 57 specimens, not listed separately U.W. 102a-001 proximal right femur U.W. 102a-002 proximal right humerus U.W. 102a-003 proximal left femur U.W. 102a-004 distal left femur U.W. 102a-010 right scapula fragment acromion U.W. 102a-013 humeral head fragments U.W. 102a-015 right proximal ulna U.W. 102a-018 long bone fragment immature U.W. 102a-019 partial rib U.W. 102a-020 right ulna fragment U.W. 102a-021 right clavicle U.W. 102a-025 right radius shaft fragment U.W. 102a-028 right fourth metacarpal U.W. 102a-036 T10 vertebra U.W. 102a-039 rib fragments U.W. 102a-040 long bone shaft fragment U.W. 102a-117 right scaphoid U.W. 102a-138 right ilium fragments immature U.W. 102a-139 L5 vertebra fragments U.W. 102a-148 sternum fragment U.W. 102a-151 T11 vertebra U.W. 102a-152 rib fragments U.W. 102a-154 T12 and L1 vertebrae found in articulation U.W. 102a-155 mid-thoracic vertebral body U.W. 102a-171 atlas fragment U.W. 102a-172 atlas fragment U.W. 102a-189 rib fragment U.W. 102a-195 rib fragment U.W. 102a-206 left clavicle fragment U.W. 102a-207 rib fragment U.W. 102a-210 sacral element immature, possibly S1 U.W. 102a-231 rib fragment U.W. 102a-232 rib fragment U.W. 102a-236 humerus head fragment U.W. 102a-239 left clavicle fragment U.W. 102a-247 right scapula fragment coracoid process U.W. 102a-250 right first rib U.W. 102a-252 rib fragment U.W. 102a-256 left scapula fragment portion of body, spine, and acromion U.W. 102a-257 left proximal humerus Table 1 continued on next page Hawks et al. eLife 2017;6:e24232. DOI: 10.7554/eLife.24232 5 of 63 Research article Genomics and Evolutionary Biology http://dx.doi.org/10.7554/eLife.24232 Hominin material from 102a Hominin material from the 102a area includes 118 identifiable specimens (Table 1; Figure 4). Fifty- seven of these are cranial and dental specimens that either refit directly or are morphologically com- patible with a nearly complete fossil cranium, designated LES1 (Figure 5). Hominin postcranial remains from locality 102a include 61 identified specimens that represent a minimum of 31 postcra- nial elements, not counting ribs. These include a minimum of two partial femora, two partial humeri, one complete clavicle and two clavicular fragments, two partial ulnae, several fragments of scapula and radius, many rib fragments, a near-complete first rib, a partial sternum, four hand and wrist elements, an immature ilium and sacrum fragment, and a partial thoracic and lumbar vertebral column. Every anatomical region of the skeleton is represented with the notable exceptions of tibia, fibula and pedal remains. Table 1 continued Specimen number Element Notes U.W. 102a-279 left scapula fragment partial glenoid fossa U.W. 102a-280 rib fragment U.W. 102a-300 vertebral fragment U.W. 102a-306 L4 vertebra body U.W. 102a-322 L2 vertebra body U.W. 102a-337 vertebral fragment neural arch U.W. 102a-348 right pubic ramus fragment U.W. 102a-349 vertebral fragment neural arch U.W. 102a-358 rib fragments U.W. 102a-360 vertebral fragment U.W. 102a-455 ulna shaft fragment U.W. 102a-456 ulna shaft fragment U.W. 102a-470 rib fragments U.W. 102a-471 right distal radius fragment U.W. 102a-474 long bone fragment immature U.W. 102a-476 right capitate U.W. 102a-477 partial right lunate U.W. 102a-479 rib fragment LOCALITY 102b U.W. 102b-178 LI2 U.W. 102b-437 rdm2 U.W. 102b-438 right mandibular corpus fragment immature, RP4 in crypt U.W. 102b-502 cranial fragments U.W. 102b-503 RP4 crown U.W. 102b-506 cranial fragment U.W. 102b-507 cranial fragment U.W. 102b-509 cranial fragment U.W. 102b-511 LC1 crown U.W. 102b-514 cranial fragment U.W. 102b-515 LI2 U.W. 102b-516 cranial fragment LOCALITY 102c U.W. 102 c-589 left mandibular fragment LM1 and LM2 in place DOI: 10.7554/eLife.24232.004 Hawks et al. eLife 2017;6:e24232. DOI: 10.7554/eLife.24232 6 of 63 Research article Genomics and Evolutionary Biology http://dx.doi.org/10.7554/eLife.24232.004Table%201.Hominin%20fossil%20material%20from%20the%20Lesedi%20Chamber.%20All%20diagnostic%20hominin%20specimens%20are%20listed,%20with%20attribution%20to%20element.%20Specimens%20that%20have%20been%20refitted%20are%20not%20listed%20separately.%20Most%20Locality&x00A0;102a%20cranial%20fragments%20are%20presumed%20to%20be%20part%20of%20LES1%20and%20are%20not%20listed%20separately.%2010.7554/eLife.24232.004Specimen%20numberElementNotesLOCALITY%20102a%20LES1craniumconstituted%20of%2057%20specimens,%20not%20listed%20separatelyU.W.%20102a-001proximal%20right%20femurU.W.%20102a-002proximal%20right%20humerusU.W.%20102a-003proximal%20left%20femurU.W.%20102a-004distal%20left%20femurU.W.%20102a-010right%20scapula%20fragmentacromionU.W.%20102a-013humeral%20head%20fragmentsU.W.%20102a-015right%20proximal%20ulnaU.W.%20102a-018long%20bone%20fragmentimmatureU.W.%20102a-019partial%20ribU.W.%20102a-020right%20ulna%20fragmentU.W.%20102a-021right%20clavicleU.W.%20102a-025right%20radius%20shaft%20fragmentU.W.%20102a-028right%20fourth%20metacarpalU.W.%20102a-036T10%20vertebraU.W.%20102a-039rib%20fragmentsU.W.%20102a-040long%20bone%20shaft%20fragmentU.W.%20102a-117right%20scaphoidU.W.%20102a-138right%20ilium%20fragmentsimmatureU.W.%20102a-139L5%20vertebra%20fragmentsU.W.%20102a-148sternum%20fragmentU.W.%20102a-151T11%20vertebraU.W.%20102a-152rib%20fragmentsU.W.%20102a-154T12%20and%20L1%20vertebraefound%20in%20articulationU.W.%20102a-155mid-thoracic%20vertebral%20bodyU.W.%20102a-171atlas%20fragmentU.W.%20102a-172atlas%20fragmentU.W.%20102a-189rib%20fragmentU.W.%20102a-195rib%20fragmentU.W.%20102a-206left%20clavicle%20fragmentU.W.%20102a-207rib%20fragmentU.W.%20102a-210sacral%20elementimmature,%20possibly%20S1U.W.%20102a-231rib%20fragmentU.W.%20102a-232rib%20fragmentU.W.%20102a-236humerus%20head%20fragmentU.W.%20102a-239left%20clavicle%20fragmentU.W.%20102a-247right%20scapula%20fragmentcoracoid%20processU.W.%20102a-250right%20first%20ribU.W.%20102a-252rib%20fragmentU.W.%20102a-256left%20scapula%20fragmentportion%20of%20body,%20spine,%20and%20acromionU.W.%20102a-257left%20proximal%20humerusU.W.%20102a-279left%20scapula%20fragmentpartial%20glenoid%20fossaU.W.%20102a-280rib%20fragmentU.W.%20102a-300vertebral%20fragmentU.W.%20102a-306L4%20vertebra%20bodyU.W.%20102a-322L2%20vertebra%20bodyU.W.%20102a-337vertebral%20fragmentneural%20archU.W.%20102a-348right%20pubic%20ramus%20fragmentU.W.%20102a-349vertebral%20fragmentneural%20archU.W.%20102a-358rib%20fragmentsU.W.%20102a-360vertebral%20fragmentU.W.%20102a-455ulna%20shaft%20fragmentU.W.%20102a-456ulna%20shaft%20fragmentU.W.%20102a-470rib%20fragmentsU.W.%20102a-471right%20distal%20radius%20fragmentU.W.%20102a-474long%20bone%20fragmentimmatureU.W.%20102a-476right%20capitateU.W.%20102a-477partial%20right%20lunateU.W.%20102a-479rib%20fragmentLOCALITY%20102b%20U.W.%20102b-178LI2U.W.%20102b-437rdm2U.W.%20102b-438right%20mandibular%20corpus%20fragmentimmature,%20RP4%20in%20cryptU.W.%20102b-502cranial%20fragmentsU.W.%20102b-503RP4%20crownU.W.%20102b-506cranial%20fragmentU.W.%20102b-507cranial%20fragmentU.W.%20102b-509cranial%20fragmentU.W.%20102b-511LC1%20crownU.W.%20102b-514cranial%20fragmentU.W.%20102b-515LI2U.W.%20102b-516cranial%20fragmentLOCALITY%20102c%20U.W.%20102%20c-589%20left%20mandibular%20fragmentLM1%20and%20LM2%20in%20place http://dx.doi.org/10.7554/eLife.24232 LES1 The LES1 cranium is fragmented but is represented by most of the vault and part of the face (Fig- ure 5). To date, we have successfully refitted the near-complete mandible, the near-complete right maxilla, a partial palate and a partial left maxillary dental row, and a partial vault including the near- complete frontal, left and right nasal and left lacrimal bones, near-complete left parietal and tempo- ral, partial right parietal, and a portion of left occipital. LES1 has a complete adult dentition except for the crowns of the lower left central and lateral incisors. The face is reconstructed from the partial right maxillary bone, including the frontal process, which refits to the right nasal bone and frontal. Figure 2. Location of the Lesedi Chamber (U.W.102) in the Rising Star system (red circle). The Dinaledi Chamber (U.W. 101) is marked by a yellow circle, while three surface entrances into the system are marked by blue circles. DOI: 10.7554/eLife.24232.005 Hawks et al. eLife 2017;6:e24232. DOI: 10.7554/eLife.24232 7 of 63 Research article Genomics and Evolutionary Biology http://dx.doi.org/10.7554/eLife.24232.005 http://dx.doi.org/10.7554/eLife.24232 Figure 3. Schematic of the Lesedi Chamber, showing the three hominin-bearing collection areas: U.W.102a, 102b, and 102c. DOI: 10.7554/eLife.24232.006 Hawks et al. eLife 2017;6:e24232. DOI: 10.7554/eLife.24232 8 of 63 Research article Genomics and Evolutionary Biology http://dx.doi.org/10.7554/eLife.24232.006 http://dx.doi.org/10.7554/eLife.24232 Figure 4. Skeletal material from locality 102a provisionally assigned to the LES1 skeleton. The adult cranial material from 102a all belongs to a single cranium; most of the adult postcranial material probably belongs to the same individual. The adult cranial and postcranial material is shown here, except for the U.W. 102a-001 femur. The possibility that the femora represent two adult individuals makes it unclear which femur may be attributable to the skeleton; for the purposes of illustration, the U.W. 102a-003/U.W. 102a-004 femur is included in this photograph. DOI: 10.7554/eLife.24232.007 Hawks et al. eLife 2017;6:e24232. DOI: 10.7554/eLife.24232 9 of 63 Research article Genomics and Evolutionary Biology http://dx.doi.org/10.7554/eLife.24232.007 http://dx.doi.org/10.7554/eLife.24232 Figure 5. LES1 cranium. Clockwise from upper left: three-quarter, frontal, superior and left lateral views. Fragments of the right temporal, the parietal and the occipital have also been recovered (not pictured), but without conjoins to the reconstructed vault or face. Scale bar = 5 cm. DOI: 10.7554/eLife.24232.008 Hawks et al. eLife 2017;6:e24232. DOI: 10.7554/eLife.24232 10 of 63 Research article Genomics and Evolutionary Biology http://dx.doi.org/10.7554/eLife.24232.008 http://dx.doi.org/10.7554/eLife.24232 The left mandibular ramus is well-enough preserved to allow a rough estimation of the condyle posi- tion, enabling an approximation of the midsagittal contour of the face (Figure 5). All additional cranial fragments in the present 102a collection are non-duplicative with this refit- ted vault and face, and where they represent the opposite side of the vault, they match in morpho- logical detail. However, many of the fragments lack clear refits with the existing vault or maxillary portions. Further physical reconstruction of the cranium will await fragments that may emerge from excavation in the future. The refitted vault, with the application of virtual mirror reconstruction, is sufficient to allow an estimate of endocranial volume of approximately 610 ml (Figure 6). Most of the features of the LES1 vault are characteristic of H. naledi from the Dinaledi Chamber (Supplementary file 1; Figure 7). The LES1 vault is relatively short anteroposteriorly, without the elongation and sharp occipital angulation found in H. erectus. LES1 exhibits mild frontal and parietal bossing, similar to H. naledi DH3. Other features on the vault that are consistent with H. naledi include prelambdoidal flattening, limited postorbital constriction, widely spaced temporal lines, a continuous supraorbital torus with a supratoral sulcus, an occipital torus, and a marked angular torus. In the temporal region, LES1 has an anteroinferiorly oriented root of the zygomatic process of the temporal, a medially positioned mandibular fossa, a small and obliquely oriented external auditory meatus, a projecting Eustachian process, a small vaginal process, a weak crista petrosa, a triangular- shaped mastoid process, and a small suprameatal spine. Each of these traits characterizes the Dina- ledi H. naledi sample (Berger et al., 2015; Laird et al., 2017). Some of these traits occur individually in other species, including H. erectus, H. habilis, H. rudolfensis, and Australopithecus sediba, but they have never been found in combination except in H. naledi (Figure 7). The maxilla and mandible of LES1 are also consistent with the Dinaledi H. naledi sample (Supplementary file 1; Figures 8, 9, 10 and 11). The maxilla has a mediolaterally flattened subnasal region, a parabolic dental arcade, and an anteriorly shallow palate. The mandible of LES1 has a grac- ile mandibular corpus, a vertical mandibular symphysis with weak mentum osseum, a steeply inclined lingual alveolar plane, weak inferior and absent superior transverse tori, continuous and deeply exca- vated anterior and posterior subalveolar fossae, mental foramina positioned above mid-corpus height, well defined ectoangular and endoangular tuberosities, and a root of the ascending ramus that originates at the mesial border of the M2. Again, many of these traits can be found individually in other hominin species, but in combination, they are uniquely found in H. naledi. The teeth exhibit moderate occlusal wear on the second and third molars, trending toward near- complete dentine exposure on the occlusal surfaces of first molars and substantial removal of occlu- sal detail of the anterior dentition. The dental morphology of LES1 is entirely consistent with the Figure 6. Digital reconstruction of endocranial volume in LES1. The refitted calvaria was mirrored and filled, resulting in a volume estimate of 610 ml. Scale sphere = 10 mm. DOI: 10.7554/eLife.24232.009 Hawks et al. eLife 2017;6:e24232. DOI: 10.7554/eLife.24232 11 of 63 Research article Genomics and Evolutionary Biology http://dx.doi.org/10.7554/eLife.24232.009 http://dx.doi.org/10.7554/eLife.24232 Figure 7. Frontal and vault morphology in H. naledi compared to that in other hominin species. Several of the crania pictured here are similar to H. naledi in endocranial volume, including Sts 5, MH1, KNM-ER 1813, and D2282, representing four different species. However, these skulls contrast strongly in other features. H. erectus is highly variable in size, as illustrated here by D2282 from Dmanisi, Georgia, one of the smallest and earliest H. erectus crania, and the L2 cranium from Zhoukoudian, China, one of the largest and latest H. erectus specimens. Figure 7 continued on next page Hawks et al. eLife 2017;6:e24232. DOI: 10.7554/eLife.24232 12 of 63 Research article Genomics and Evolutionary Biology http://dx.doi.org/10.7554/eLife.24232 Dinaledi sample of H. naledi (Figures 10 and 12). The mesiodistal and buccolingual (or labiolingual) crown dimensions of all the LES1 teeth fall within the range of the Dinaledi dental sample, except Figure 7 continued The relatively early KNM-ER 3733 has a size and endocranial volume close to the mean for H. erectus. Cranial remains that are attributed to H. erectus share a combination of anatomical features despite their diversity in size. Many such features of H. erectus are also shared with H. naledi, H. habilis, or Au. sediba, and notably, the differences in the frontal and vault between KNM-ER 1813 (H. habilis) and KNM-ER 1470 (H. rudolfensis) are mostly features that the smaller KNM-ER 1813 shares with H. naledi, H. erectus, and Au. sediba. The H. naledi skulls share some aspects of frontal morphology with Au. sediba, H. habilis and H. erectus that are not found in Au. africanus or H. rudolfensis, including frontal bossing and a supratoral sulcus. Two additional traits of the H. naledi anterior vault are shared with Au. sediba and H. erectus:slight postorbital construction and a posterior position of the temporal crest on the supraorbital torus. More posteriorly on the vault, H. naledi further shares an angular torus with H. erectus, and some individuals also have sagittal keeling. Both of these traits are also present in some archaic humans. Some H. naledi crania, such as DH3, are substantially smaller than any H. erectus cranium, and the small size and thin vault bone of even the largest H. naledi skull, LES1, are outliers compared to H. erectus, matched only by some Dmanisi crania. The facial morphology of H. naledi is more distinct from those of H. erectus and H. habilis. The nasal bones of LES1 do not project markedly anteriorly, although like many specimens of H. erectus, LES1 has a projecting nasal spine. LES1 has a relatively flat lower face, with a transversely concave clivus and incisors that project only slightly past the canines. This morphology is similar but less extreme than that found in KNM-ER 1470 of H. rudolfensis, and is not shared with the other species pictured here. H. naledi has several distinctive features of the temporal bone that are absent from or found in only a few specimens of the other species pictured, including a laterally inflated mastoid process (comparable to some specimens of Au. afarensis), a weak or absent crista petrosa (comparable to Au. afarensis), and a small external auditory meatus (comparable to KNM-WT 40000 of Kenyanthropus platyops [Leakey et al., 2001]). In this illustration, KNM-ER 1813, KNM-ER 1470, KNM-ER 3733, and ZKD L2 are represented by casts. Because these images are in a nonstandard orientation, scale is approximate. DOI: 10.7554/eLife.24232.010 Figure 8. LES1 mandible compared to the DH1 holotype mandible of H. naledi. In each pair, LES1 is on the left and DH1 on the right. Top left: anterior view. Top right: occlusal view. Bottom left: left lateral view. Bottom right: posterior view. Scale bar = 2 cm. DOI: 10.7554/eLife.24232.011 Hawks et al. eLife 2017;6:e24232. DOI: 10.7554/eLife.24232 13 of 63 Research article Genomics and Evolutionary Biology http://dx.doi.org/10.7554/eLife.24232.010 http://dx.doi.org/10.7554/eLife.24232.011 http://dx.doi.org/10.7554/eLife.24232 for those teeth where interproximal wear has clearly reduced the mesiodistal dimension (Table 2; Figure 13). The P3 crowns are worn, but they are roughly symmetrical about their mesiodistal axis in occlusal view; they are fully bicuspid and multirooted, with a smaller circular mesiobuccal root and larger, more platelike, distal root. This configuration is repeated throughout the Dinaledi dental assemblage. The shared overall P3 morphology of LES1 and the Dinaledi sample is distinctive in H. naledi and not observed in other species of hominins (Figure 12; Berger et al., 2015). The P3 and P4 are both three-rooted, with two ovoid roots present buccally and a larger ovoid root present lin- gually. The roots are not widely splayed as in some other multi-rooted hominins, and especially for the P4, the buccal roots are closely packed in buccal view. This root configuration is seen in the H. naledi type specimen, U.W. 101–1277. The mandibular canine crowns have asymmetrically placed crown shoulders, with the mesial more apically placed than the distal. Further, the distal shoulder is formed by an accessory cuspule. These features are strongly distinctive in H. naledi (Berger et al., 2015), with only a few specimens of H. erectus approaching this canine configuration. None of the molars exhibit any evidence of supernumerary cusps, and cingular features, such as the protostylid and Carabelli’s feature, are either absent or weakly developed and are expressed independently of the grooves of the crown. The molar size gradient in the LES1 mandible is M1 < M2 < M3 as in the Dinaledi Chamber sample of H. naledi (Figure 10). The Dinaledi Chamber includes no maxillary den- tition with all three molars in place, but U.W. 101–1269 is a LM3 that exhibits a mesial interproximal facet that matches the distal facet of the LM2 present in the U.W. 101–1277 (DH1) maxilla. If these specimens do represent a single individual, then the maxillary molar gradient for this specimen would be M1 < M3 < M2, which is also seen in the LES1 maxilla. In total, these dental features are within the known range for H. naledi in every instance and distinguish LES1 clearly from all other hominin species. The LES1 cranium does exhibit some traits that differ from comparable examples in the Dinaledi Chamber. The cranium is slightly larger overall, with an estimated endocranial volume of approxi- mately 610 ml, and this larger size is reflected in the external vault measurements. Previously, the largest known H. naledi endocranium was DH1 at approximately 560 ml (Berger et al., 2015). LES1 contrasts in morphological features with the small DH3 cranium in ways that have been observed Figure 9. Comparison of LES1 maxilla to the DH1 holotype maxilla of H. naledi. In each pair, LES1 is on the left and DH1 on the right. Top left: anterior view. Top right: right (LES1) and left (DH1) lateral view. Bottom: occlusal view. Scale bar = 2 cm. DOI: 10.7554/eLife.24232.012 Hawks et al. eLife 2017;6:e24232. DOI: 10.7554/eLife.24232 14 of 63 Research article Genomics and Evolutionary Biology http://dx.doi.org/10.7554/eLife.24232.012 http://dx.doi.org/10.7554/eLife.24232 when comparing male individuals with female individuals in other hominin species. The supraorbital torus of LES1 is more pronounced than that in the small DH3 individual. LES1 has a stronger supra- mastoid/suprameatal crest, a larger mastoid process, and a more marked pterygoid insertion when compared to the U.W. 101–361 mandible. Although LES1 is outside of the endocranial volume range of specimens presently attributed to H. naledi, the larger size and more robust features of LES1 are consistent with the combination of cranial and mandibular characters in H. naledi. Figure 10. Mandibular and dental anatomy in H. naledi compared to other species of Homo. Right demi-mandibles attributed to H. rudolfensis, H. habilis, H. naledi, H. erectus, and H. sapiens are pictured. All mandibles are aligned using the line marking the distal edge of the first molar. Each of the six horizontal lines corresponds to the edges of teeth in the DH1 mandible, the holotype specimen of H. naledi, with corresponding teeth labeled to the left. Using these lines, it is apparent which specimens have longer premolars and first molars, and which have longer second and third molars compared to DH1. The dentition of the LES1 mandible has been affected by interproximal wear, resulting in shorter mesiodistal measurements. Mandibular morphology and dental proportions vary slightly among most species of Homo, particularly in comparison with the large differences in dental proportions among species of Australopithecus and Paranthropus. Still, H. naledi is clearly distinguishable from other species of Homo (Berger et al., 2015; Laird et al., 2017). Fossils of H. rudolfensis, H. habilis, and H. erectus differ from H. naledi in the proportions of different parts of the postcanine tooth row and in features of the mandibular corpus. H. erectus. While fossils attributed to H. erectus vary in dental proportions, the early African and Georgian fossil specimens (here represented by KNM-ER 992, D211 and D2600) have larger first molars than H. naledi, comparable premolar sizes, and highly variable second and third molar sizes. The mandibles attributed to H. erectus mostly have greater corpus height than H. naledi mandibles and are highly diverse in corpus breadth, symphyseal thickness, and robusticity. Many have a strong post-incisive planum, most obvious in D2600 (shown). All three also differ from H. naledi in the crown complexity of their molars and premolar morphology, as illustrated in more detail in Figure 12. Some specialists would attribute these three mandibles of H. erectus to three different species. H. habilis. The two Olduvai mandibles of H. habilis are themselves quite different from each other in size. Both have similar dental proportions to H. naledi with bigger teeth across the postcanine dentition. Tobias (1967) viewed O.H. 13 as being similar to H. erectus and described it as an ‘evolved H. habilis’. Its occlusal morphology and dental proportions do resemble KNM-ER 992 (Wood, 1991), although the mandibular corpus is thinner and shallower, with a curved base in lateral profile. A strong post-incisive planum is evident in both mandibles. H. rudolfensis. The KNM-ER 1802 and Uraha (UR 501) mandibles have often been attributed to H. rudolfensis, although both attributions may be doubtful (Leakey et al., 2012). However, both lack any special similarities with contemporary australopiths and represent a megadont early Homo morphology with corpus size and robusticity much greater than those of H. naledi. Au. sediba. Molar sizes in the MH2 mandible are around 1 mm larger than the average for H. naledi, but the proportions are very similar to those of H. naledi, and like H. naledi, MH2 has a weak post-incisive planum and a small symphysis area. H. sapiens. The modern human mandible shown here, from a recent South African individual, has similar first molar size to the H. naledi mandibles, but much smaller premolars and second and third molars. The crown complexity in this individual, which is not unusual for African population samples, is substantially greater than evidenced in H. naledi. The mandibular corpus is smaller and much less robust than H. naledi. KNM-ER 1802, UR 501, O.H. 13, O.H. 7, and KNM-ER 992 are illustrated here with casts; the remainder are original specimens. The left side of O.H. 7 is shown here mirrored. DOI: 10.7554/eLife.24232.013 Hawks et al. eLife 2017;6:e24232. DOI: 10.7554/eLife.24232 15 of 63 Research article Genomics and Evolutionary Biology http://dx.doi.org/10.7554/eLife.24232.013 http://dx.doi.org/10.7554/eLife.24232 Figure 11. Comparison of H. naledi mandibles to other hominin species, from lateral view. The DH1 holotype mandible and the LES1 mandible of H. naledi have a moderately deep mandibular corpus compared to other species of Homo; the LES1 mandible has a slightly greater corpus height anteriorly (at P3) than posteriorly (at M2). LES1 has rather a high coronoid process; the height of the condyle was probably lower than this. The mental foramen is at the midpoint or slightly higher in both H. naledi mandibles, and in both, the symphysis is nearly Figure 11 continued on next page Hawks et al. eLife 2017;6:e24232. DOI: 10.7554/eLife.24232 16 of 63 Research article Genomics and Evolutionary Biology http://dx.doi.org/10.7554/eLife.24232 Comparative cranial anatomy The comparative anatomy of the H. naledi cranial remains from the Dinaledi Chamber was presented in detail by Laird et al. (2017), and morphometric comparative analyses of that collection and of other hominin samples were presented by Schroeder et al., 2017. Additionally, Rightmire et al., 2017 addressed the morphological features of H. naledi in comparison with the Dmanisi fossil crania, in particular the robust D4500 cranium. The anatomy of the LES1 cranium reinforces the conclusions of those studies in most respects (Supplementary file 1, 2; Figure 7). Crania of H. naledi are most similar in cranial vault shape to other Homo or Homo-like australo- pith crania with small endocranial volume, including D2700, D2280, MH1, KNM-ER 1470, and KNM- ER 1813 (Schroeder et al., 2017). These shape similarities do not reflect small size alone: for exam- ple, H. naledi cranial material is quite distinct in shape from LB1, and the small DH3 calvaria of H. naledi is also notably different from KNM-ER 42700 in shape analyses (Schroeder et al., 2017). Additional differences between H. naledi and other small specimens of Homo are evident among the morphological characters of the skull (Laird et al., 2017). Compared to specimens attributed to H. erectus, H. habilis or H. rudolfensis, the temporal bone of H. naledi exhibits a small external audi- tory meatus, a shallow and relatively narrow mandibular fossa, a small postglenoid process, and a laterally inflated mastoid process. The features of the occiput that distinguish H. naledi from H. erec- tus (Laird et al., 2017) are not part of the preserved sections of LES1. However, the maxilla of LES1 is better preserved than DH1, and like the latter specimen, presents a transversely flat nasoalveolar clivus, similar to H. rudolfensis but not H. erectus or H. habilis, a shallow anterior palate, unlike H. erectus or H. habilis, and an anteriorly projecting anterior nasal spine, comparable to H. erectus but not present in H. habilis or H. rudolfensis. The LES1 cranium is similar to the Dinaledi H. naledi sam- ple in its morphological differences from the H. floresiensis LB1 cranium (Berger et al., 2015). All H. naledi crania are estimated to have been larger than LB1 in endocranial volume, while LES1 and the other H. naledi cranial remains lack the reduced cranial height, marked frontal keel, canine juga and anterior pillars of the LB1 cranium. The cranial anatomiesof H. naledi and LES1 share a unique set of traits that otherwise distinguish Homo species from each other. The LES1 mandible shares very similar overall dimensions, shape, and morphological features with DH1 from the Dinaledi Chamber (Supplementary file 1; Figures 8, 10 and 11). Comparative analysis of overall mandibular shape places H. naledi closer to australopith mandibles such as Sts 36 and Sts 52b than to any specimens of H. habilis, H. rudolfensis, or H. erectus, despite the large range of shape variation among H. erectus mandibular specimens (Schroeder et al., 2017). Dmanisi man- dibular specimens, including D2735 and D2600, are different from the DH1 mandible despite the similarity in vault shape between D2700 and D2280 and H. naledi DH3, and these mandibular differ- ences are likewise reflected in the LES1 mandible. The morphological features of the LES1 mandible align it clearly with DH1 and other partial H. naledi mandibles from the Dinaledi Chamber, setting it apart from other species of Homo, including those with similar-sized molars (Figures 10 and 12). Unlike the H. floresiensis mandibles LB1 and LB2, the mandibular symphyses of LES1 and DH1 have steeply inclined posterior faces that lack any post-incisive planum or superior transverse torus. The mandibular molar gradient of H. naledi and the morphology of the mandibular premolars also Figure 11 continued vertical. These features vary substantially within Homo. Modern humans (bottom) typically have a chin, but otherwise vary substantially in corpus height, whether the base of the corpus is parallel with the alveolar portion or with the occlusal surfaces of the teeth. Here that variability is illustrated with two modern human mandibles of male individuals, one from island Melanesia (left), and one from southern Africa (right). H. erectus exhibits very extensive variation in corpus height and thickness. D2600 (shown) is extremely thick and robust, but is not an outlier; other H. erectus mandibles approach or equal its corpus dimensions. The position of the mental foramen also varies, as does the relative anterior versus posterior corpus height and the symphyseal profile, from more sloping to near vertical (as illustrated by KNM-ER 992, although this specimen is damaged at the symphysis). MH2 (Au. sediba) has comparable corpus height and robusticity to the H. naledi mandibles, with a more sloping symphysis. O.H. 13 is a more gracile mandible than the H. naledi specimens in many respects. It has a curved base and a sloping symphysis. The more complete left side of LES1 is shown here and mirrored for comparison to other specimens. KNM-ER 992 and O.H. 13 are represented here by casts. DOI: 10.7554/eLife.24232.014 Hawks et al. eLife 2017;6:e24232. DOI: 10.7554/eLife.24232 17 of 63 Research article Genomics and Evolutionary Biology http://dx.doi.org/10.7554/eLife.24232.014 http://dx.doi.org/10.7554/eLife.24232 distinguish it from H. floresiensis (Brown and Maeda, 2009; Kaifu et al., 2011, 2015). The symphy- seal morphology of H. naledi likewise distinguishes LES1 and DH1 from mandibular remains attrib- uted to H. habilis and H. rudolfensis, such as OH 7, OH 13, KNM-ER 60000 and KNM-ER 1802. Claviculae U.W. 102a-021 is a nearly complete right clavicle, missing only the articular surface of the sternal end, where trabecular bone is exposed over the entire articular area, including a small bit of the anterior surface (Figure 14). The shaft is broken into two pieces near the midshaft but the two pieces conjoin cleanly. There is also a small bit of damage to the acromial end. On the posterior sur- face, the medial part of the crest for the conoid tubercle is broken off. The specimen exhibits a dark Figure 12. Occlusal view of H. naledi mandibular teeth compared to those of other hominins. Teeth from the canine to the third molar are shown, if present, in the orientation in which they are found within the mandible. All individuals are aligned vertically by the distal margin of the first molar. Mandibles from the Lesedi Chamber, U.W. 102 c-589 and LES1 are shown next to DH1 and U.W. 101–377 (Berger et al., 2015). The mandibles illustrated from H. erectus have relatively little occlusal wear, so their morphology can be seen more clearly than that of worn mandibles. The immature U.W. 101–377 (H. naledi) is comparable in developmental age and wear to O.H. 7 (H. habilis), as well as to D2735 and KNM-WT 15000 (H. erectus). When compared to H. habilis, H. erectus, and australopiths, H. naledi is notable for its relatively small first molars, its relatively small canines, and its lack of supernumerary cusps and crenulation on the molars. The complexity of molar cusp and groove patterns is especially evident in the chronologically early H. erectus specimens from Africa and Georgia shown here. For example, the unworn M2 of the immature U.W. 101–377 mandible of H. naledi has a relatively simple crown anatomy with very little wrinkling or crenulation. By comparison, the M2 of D2735, D211, and KNM-WT 15000, all with minimal occlusal wear, show extensive crenulation and supernumerary cusps. Canine size and molar crown complexity vary substantially among modern human populations, but the southern African individual illustrated on the right is not atypical for its population, and has greater molar crown complexity and larger canine dimensions than any of the H. naledi mandibular dentitions. The morphology of the third premolar varies extensively among these hominin species and within H. erectus. The H. naledi P3 anatomy can be seen clearly in the immature U.W. 101–377 individual. It is characterized by roughly equally prominent lingual and buccal cusps and an expanded talonid. In H. naledi, this tooth is broadly similar in morphology and size to the P4. This configuration of the P3 is not present in the other species, with only KNM-WT 15000 exhibiting some expansion of the lingual cusp in what remains an asymmetrical and rounded P3. A.L. 400–1, O.H. 7 and KNM-WT 15000 are represented by casts; The left dentition of U.W. 102 c-589 and O.H. 7 have been mirrored to compare to right mandibles. Images have been scaled by measured first molar dimensions. DOI: 10.7554/eLife.24232.015 Hawks et al. eLife 2017;6:e24232. DOI: 10.7554/eLife.24232 18 of 63 Research article Genomics and Evolutionary Biology http://dx.doi.org/10.7554/eLife.24232.015 http://dx.doi.org/10.7554/eLife.24232 Table 2. Dental measurements for Lesedi Chamber specimens. Specimen Mesiodistal diameter Buccolingual (or labiolingual) diameter U.W. 102b-437 ldm2 10.7 8.7 U.W. 102b-503 RP4 8.4 10.9 U.W. 102b-515 LI2 6.8 6.5† U.W. 102b-178 LI2 5.6 5.9 U.W. 102b-511 LC1 6.8 6.8† U.W. 102 c-589 LM1 11.4 10.6 U.W. 102 c-589 LM2 13.1 11.3 LES1 maxillary RI1 7.6* 6.9 RI2 6.8* 7.0 RC1 7.5 8.7 RP3 8.1 10.8 RP4 8.1 11.3 RM1 10.6* 11.8 RM2 11.7 12.7 RM3 11.4 12.7 LI1 7.4* 6.9 LI2 6.1* 6.8 LC1 7.4 8.7 LP3 8.0 10.9 LP4 8.1 11.3 LM1 10.7* 11.9 LM2 12.1 12.8 LM3 11.4 13.6 LES1 mandibular RI1 5.8* 6.3 RI2 5.4* 6.1 RC1 7.1 7.7 RP3 8.4 9.3 RP4 8.2 9.1 RM1 10.8* 10.6 RM2 12.3 11.5 RM3 13.3 11.7 LC1 7.8 7.5 † LP3 8.4 9.3 LP4 8.2 9.1 LM1 11.2* 10.6 LM2 12.3 11.5 LM3 13.3 11.7 *Denotes measurements where the tooth is extremely worn, and mesiodistal diameter reported here has not been corrected for the degree of wear. †Denotes instances where we report a minimum value for labiolingual measurements because the crown is not complete or is broken. DOI: 10.7554/eLife.24232.017 Hawks et al. eLife 2017;6:e24232. DOI: 10.7554/eLife.24232 19 of 63 Research article Genomics and Evolutionary Biology http://dx.doi.org/10.7554/eLife.24232.017Table%202.Dental%20measurements%20for%20Lesedi%20Chamber%20specimens.%2010.7554/eLife.24232.017SpecimenMesiodistal%20diameterBuccolingual%20(or%20labiolingual)%20diameterU.W.%20102b-437%20ldm210.78.7U.W.%20102b-503%20RP48.410.9U.W.%20102b-515%20LI26.86.5&x2020;U.W.%20102b-178%20LI25.65.9U.W.%20102b-511%20LC16.86.8&x2020;U.W.%20102%20c-589%20LM111.410.6U.W.%20102%20c-589%20LM213.111.3LES1%20maxillary%20RI17.6&x002A;6.9RI26.8&x002A;7.0RC17.58.7RP38.110.8RP48.111.3RM110.6&x002A;11.8RM211.712.7RM311.412.7LI17.4&x002A;6.9LI26.1&x002A;6.8LC17.48.7LP38.010.9LP48.111.3LM110.7&x002A;11.9LM212.112.8LM311.413.6LES1%20mandibular%20RI15.8&x002A;6.3RI25.4&x002A;6.1RC17.17.7RP38.49.3RP48.29.1RM110.8&x002A;10.6RM212.311.5RM313.311.7LC17.87.5%20&x2020;LP38.49.3LP48.29.1LM111.2&x002A;10.6LM212.311.5LM313.311.7&x002A;Denotes%20measurements%20where%20the%20tooth%20is%20extremely%20worn,%20and%20mesiodistal%20diameter%20reported%20here%20has%20not%20been%20corrected%20for%20the%20degree%20of%20wear.&x2020;Denotes%20instances%20where%20we%20report%20a%20minimum%20value%20for%20labiolingual%20measurements%20because%20the%20crown%20is%20not%20complete%20or%20is%20broken. http://dx.doi.org/10.7554/eLife.24232 Figure 13. Metric comparisons of the Lesedi Chamber dental material. H. naledi is clearly differentiated in first molar and canine dimensions from other species with broadly similar cranial and dental morphology, including Au. sediba, H. habilis, H. rudolfensis, and early H. erectus samples from Africa and Georgia. The material from the Lesedi Chamber is within the range of or similar to H. naledi in these dimensions and well differentiated from the other samples. Top left: mandibular first molar dimensions. Top right: maxillary first molar dimensions. Bottom left: mandibular canine dimensions. Bottom right: maxillary canine dimensions. The LES1 first molars and maxillary canines have a substantial degree of interproximal wear, and the values plotted here are not corrected for this wear, which shortened the mesiodistal dimension by as much as a millimeter. The values plotted here should thus be regarded as minimum values. The H. erectus sample here includes specimens from the Lake Turkana area, Konso, Tighenif (Ternifine), Thomas Quarry, and Dmanisi; Asian H. erectus specimens are omitted. Attributions of H. habilis and H. rudolfensis specimens are indicated in the Materials and methods. DOI: 10.7554/eLife.24232.016 Hawks et al. eLife 2017;6:e24232. DOI: 10.7554/eLife.24232 20 of 63 Research article Genomics and Evolutionary Biology http://dx.doi.org/10.7554/eLife.24232.016 http://dx.doi.org/10.7554/eLife.24232 surface coating on the anterior aspect of the sternal half and patchy areas of black staining on its acromial half. There are fine hairline longitudinal cracks on much of the acromial half of the bone. U.W. 102a-206 is a ca. 41-mm-long fragment of left clavicular shaft, preserving the midshaft region (based on anatomical comparisons with U.W. 102a-021). The shaft anteroposterior (AP) and superoinferior (SI) dimensions are slightly smaller than those of the right side clavicle at this position. The fragment compares favorably to U.W. 102a-021 in overall size, curvature, and shaft morphology. U.W. 102a-239 is the acromial end of a left clavicle, including the lateral 51.5 mm, preserving the conoid tubercle (but not its medial crest) and the articular surface for the acromion of the scapula. This is slightly larger than the acromial end of U.W. 102a-021, but otherwise fairly similar in morphology. Comparative clavicular anatomy The clavicular anatomy of H. naledi is comparable to that present in Au. sediba (Churchill et al., 2013) and H. habilis (Oxnard, 1969; Ohman, 1986; Voisin, 2001), suggesting a superiorly posi- tioned shoulder (Feuerriegel et al., 2017). The U.W. 102a-021 and U.W. 102a-206 claviculae are Figure 14. U.W. 102a-021 right clavicle from the Lesedi Chamber. Left, from top: superior, anterior, inferior, posterior views. Right, from top: medial and lateral views. Scale bar = 2 cm. DOI: 10.7554/eLife.24232.018 Hawks et al. eLife 2017;6:e24232. DOI: 10.7554/eLife.24232 21 of 63 Research article Genomics and Evolutionary Biology http://dx.doi.org/10.7554/eLife.24232.018 http://dx.doi.org/10.7554/eLife.24232 consistent with the morphology noted in the clavicular material from the Dinaledi Chamber. As with clavicular specimens from U.W. 101, the overall appearance of the clavicle is smooth (non-rugose), with only a few weakly developed entheses. The deltoid crest is present as a mildly rugose line on the anterior surface of the lateral curvature. The conoid tubercle appears well-developed and forms a posteriorly projecting flange, producing a pronounced border to a deep subclavian sulcus. How- ever, unlike some specimens from the Dinaledi Chamber (such as U.W. 101–258), the conoid is not centrally positioned on the shaft, but rather occurs on the posterior margin. Humerus U.W. 102a-002 is a proximal shaft fragment of a right humerus (Figure 15). The head and greater tubercle are missing, as is all but the very distal base of the lesser tubercle. From this metaphyseal region, the fragment preserves approximately 50–60% of the shaft, with a total fragment length of 85 mm. U.W. 102a-013 includes two fragments identified as humeral head, each with some articular subchondral bone, which may derive from the same element as U.W. 102a-002. They appear to be consistent in curvatures of the articular surface with U.W. 102a-257. The specimen is mostly coated with a brown to dark-brown mineral patina, the surface is unweathered with only slight surface removal on the distal end of the anterior surface. The breaks, both proximal and distal, are sharp. U.W. 102a-257 is a fragment of left humerus, including the head and proximal shaft, and is largely complete from head to just around midshaft (Figure 16). There is corrosion to the superior aspect of the proximal articular surface (which precludes an accurate measurement of the superoin- ferior diameter of the head), and to the articular margin of the superolateral head. The surface of the specimen is otherwise very well preserved with no signs of weathering. A dark-brown to black patina covers much of the posterior surface of the shaft, wrapping around to the anterior surface on the most distal part. The proximal 40 mm or so of the lateral crest of the deltoid tuberosity is pres- ent. U.W. 102a-257 is consistent with the morphology and size of U.W. 102a-002, and they may rep- resent left and right humeri of the same individual. Figure 15. U.W. 102a-002 right humerus fragment. From left: posterior, medial, anterior and lateral views. Right, from top: Scale bar = 5 cm. DOI: 10.7554/eLife.24232.019 Hawks et al. eLife 2017;6:e24232. DOI: 10.7554/eLife.24232 22 of 63 Research article Genomics and Evolutionary Biology http://dx.doi.org/10.7554/eLife.24232.019 http://dx.doi.org/10.7554/eLife.24232 The proximal humerus fragments from the Lesedi Chamber have morphology consistent with the Dinaledi Chamber collection of H. naledi, both in the size of the head and in the shaft diameter. In both assemblages, the bicipital groove appears deep and narrow, and the lesser tubercle is projec- ting. The most distinctive aspect of the humerus material of H. naledi in comparison with other homi- nin species is the very low humeral torsion angle in the adult U.W. 101–283 humerus, in which the head faces nearly directly posteriorly (Feuerriegel et al., 2017). This aspect cannot be assessed directly in the fragments from the Lesedi Chamber. Ulna U.W. 102a-015 is a right proximal ulna, on which much of the trochlear notch and olecranon process are preserved in addition to the proximal half of the diaphysis (Figure 17). There is erosion to the anterior tips and margins of the coronoid and olecranon process. The surface of the shaft is well-preserved and exhibits very slight hairline longitudinal cracks. The break to the distal end is sharp and cleanly transverse. The olecranon process is mediolaterally narrow and the trochlear notch appears to have opened anterosuperiorly, as in modern humans but not Neandertals. While there is only one fragmentary (and probably immature) proximal ulna from the Dinaledi Chamber, U.W. 102a-015 generally compares well with U.W. 101–560 in terms of morphology and gracility. Hand and wrist Four adult hand and wrist elements have been recovered from the 102a locality (Figure 18). U.W. 102a-028 is a right fourth metacarpal (RMc4), with a small base and a relatively radioulnarly broad Figure 16. U.W. 102a-257 left proximal humerus fragment. From left: posterior, medial, anterior and lateral views. Top: proximal view. Bottom: distal view. Scale bar = 5 cm. DOI: 10.7554/eLife.24232.020 Hawks et al. eLife 2017;6:e24232. DOI: 10.7554/eLife.24232 23 of 63 Research article Genomics and Evolutionary Biology http://dx.doi.org/10.7554/eLife.24232.020 http://dx.doi.org/10.7554/eLife.24232 head (Figure 15). The metacarpal shaft shows substantial curvature and is relatively robust for its length, although it still falls within the upper range of variation found in modern humans (Figure 16). U.W. 102a-117 is a complete right scaphoid; U.W. 102a-476 is a complete right capitate; and U.W. 102a-477 is a partial right lunate. The scaphoid, lunate, and capitate are consistent in size and appear to match each other when placed in anatomical articulation; the RMc4 is likewise a good match in size, with a lateral base matching in dorsopalmar contour the base of the capitate. Thus, these four bones may represent the right hand of one individual. These four bones are qualitatively similar in overall shape to that described for H. naledi, but they are absolutely larger in most of their overall dimensions (Kivell et al., 2015). The lunate is missing a large portion of its articular surface for the radius and adjacent areas, precluding quantitative com- parisons of its morphology. A canonical variates analysis of scaphoid and capitate comparative mor- phology in African apes and hominins demonstrates that the Dinaledi and Lesedi scaphoids and capitates fall together within a distinct space relative to other fossil and extant hominids. Along the first canonical axis, Dinaledi and Lesedi wrist remains cluster with modern humans and Neandertals because they all share derived features relative to those of African apes (Figure 19; Supplementary file 3). For instance, the scaphoid’s trapezium facet extends further onto the Figure 17. U.W. 102a-015 ulna fragment. From left: anterior, medial, posterior and lateral views. Right from top: proximal and distal views. Scale bar = 2 cm. DOI: 10.7554/eLife.24232.021 Hawks et al. eLife 2017;6:e24232. DOI: 10.7554/eLife.24232 24 of 63 Research article Genomics and Evolutionary Biology http://dx.doi.org/10.7554/eLife.24232.021 http://dx.doi.org/10.7554/eLife.24232 tubercle, and together, the trapezium and trapezoid facets are relatively large, as in modern humans and Neandertals (Supplementary file 3). The Dinaledi and Lesedi scaphoid and capitate morphol- ogy are distinguished from those of modern humans and Neandertals on the second canonical axis because the Mc2 facet orientation in H. naledi is roughly intermediate between that of modern humans and Neandertals on the one hand and that of African apes on the other. In this respect, the H. naledi capitates are more similar to those of H. floresiensis and several australopiths. Vertebrae Seven vertebrae have been recovered in the 102a assemblage, all from the lower thoracic and lum- bar region of the spine. These vertebrae are roughly equivalent in preservation. The thin cortical bone of the vertebral bodies is eroded in large patches on these elements with exposure of underly- ing trabeculae. They have minimal surface staining or patination, and where the vertebral arches are present, the cortical surface is well-preserved. U.W. 102a-036 is a largely complete antepenultimate thoracic vertebra, inferred as T10, with lim- ited erosion to the anterior surface of the vertebral body and some damage distally on the trans- verse processes, particularly on the right side, and to the posterior end of the spinous process (Figure 20). The vertebral body is ovoid to kidney-shaped and the ring apophyses are relatively thick, covering approximately three-quarters of the vertebral body surface. The spinal canal is ovoid in shape and about one-third the size of the vertebral body. Facets for the tenth rib are posteriorly positioned, almost entirely on the pedicles. The pedicles themselves are transversely thick, as are the transverse processes, which are strongly posteriorly oriented. Together, the transverse processes and pedicles form nearly continuous, robust lateral structures for anchoring epaxial muscles and for transmitting forces to and from the vertebral body, respectively. U.W. 102a-151 is a nearly complete penultimate thoracic vertebra, inferred as T11, with some erosion and loss to the left side of the vertebral body and missing the left-side transverse process (Figure 21). Portions of the superior vertebral body surface are eroded away, revealing trabeculae. Figure 18. U.W. 102a-028 right fourth metacarpal. From left: dorsal, ulnar, palmar and radial views. Right from top: distal and proximal views. Scale bar = 1 cm. DOI: 10.7554/eLife.24232.022 Hawks et al. eLife 2017;6:e24232. DOI: 10.7554/eLife.24232 25 of 63 Research article Genomics and Evolutionary Biology http://dx.doi.org/10.7554/eLife.24232.022 http://dx.doi.org/10.7554/eLife.24232 The superior articular facets are planiform and posteriorly oriented. The inferior articular facets are asymmetrical – the right side is curved and anterolaterally oriented, whereas the left side is planiform and oriented anteriorly on the coronal plane, as in the transitional vertebra. Costal facets are large, extending from the posterior aspect of the body inferiorly and posteriorly onto the pedicle. The ver- tebral body is kidney- to heart-shaped and the spinal canal is ovoid, with a slightly triangular shape. The spinous process is relatively long and relatively horizontal in its orientation, with its major axis deflecting inferiorly at an angle of approximately 20˚ from the surface of the superior vertebral body. U.W. 102a-154a is a nearly complete last thoracic vertebra, inferred as T12. The right inferior articular facet, distal spinous process, and anterior aspect of the inferior vertebral body are broken away. The anterior portion of the body is eroded on the right side, as are the lateral aspects of the superior vertebral body. The superior articular facets are asymmetrical, matching the inferior articular Figure 19. Quantitative comparisons of hand and wrist material from the Lesedi Chamber. Left: ratios of fourth metacarpal dimensions in H. naledi compared to those in other hominin and great ape samples. Right: canonical variates analysis of capitate and scaphoid morphology in humans, chimpanzees, gorillas, and fossil hominins. H. naledi from the Dinaledi Chamber occupies a unique position in scaphoid and capitate joint configurations, which is closely matched by the capitate and scaphoid from the Lesedi Chamber. In this analysis, no a priori groups are assumed; we also examined the scenario in which Homo naledi and other fossil specimens are included as a priori groups and the results are essentially identical. DOI: 10.7554/eLife.24232.023 Hawks et al. eLife 2017;6:e24232. DOI: 10.7554/eLife.24232 26 of 63 Research article Genomics and Evolutionary Biology http://dx.doi.org/10.7554/eLife.24232.023 http://dx.doi.org/10.7554/eLife.24232 facets of the superjacent vertebra (U.W. 102a-151): the left superior articular facet is planiform and posteriorly oriented on the coronal plane, whereas the right superior articular facet is curved and posterolaterally oriented. The right superior articular facet is comparatively diminutive in size, partic- ularly in transverse dimension. The vertebral body is kidney- to heart-shaped and transversely wide. The costal facets are positioned at the body-pedicle border but are eroded on both sides; thus, their morphology cannot be fully appreciated. The pedicles themselves are anteroposteriorly short and contribute to a wide, ovoid spinal canal. U.W. 102a-154b, U.W. 102a-322, and U.W. 102a-306 are vertebral bodies associated with little or no vertebral arch structures. U.W. 102a-139 is a lumbar vertebra preserving most aspects of the vertebral body and neural arch, but it is broken into five pieces that refit reasonably well, although the spinous process is missing. None of the bodies or preserved aspects of pedicles bear costal fac- ets. U.W. 102a-154b nicely articulates with U.W. 102a-154a superiorly and U.W. 102a-322 inferiorly, and U.W. 102a-306 and U.W. 102a-139 articulate with each other; however, U.W. 102a-322 and U. W. 102a-306 do not articulate. The lumbar transverse processes of U.W. 102a-139 are anteroposter- iorly wide, emerging anteriorly from the posterior aspect of the vertebral body along the pedicles and posteriorly to the bases of the superior articular processes. Its body is clearly posteriorly wedged in lateral view. Together, these features indicate that U.W. 102a-139 is the last lumbar ver- tebra. Therefore, the likely seriation is as follows: U.W. 102a-154b is L1, U.W. 102a-322 is L2, U.W. 102a-306 is L4, U.W. 102a-139 is L5, and L3 is missing. Comparative vertebral anatomy The U.W. 102a-036 T10 and U.W. 102a-151 T11 vertebrae are directly comparable to the near-com- plete U.W. 101–855 T10 and U.W. 101–1733 T11 vertebrae from the Dinaledi Chamber. The Dinaledi Figure 20. U.W. 102a-036 vertebra, T10. Clockwise from top left: posterior, superior, inferior, left, right, and anterior views. Scale bar = 2 cm. DOI: 10.7554/eLife.24232.024 Hawks et al. eLife 2017;6:e24232. DOI: 10.7554/eLife.24232 27 of 63 Research article Genomics and Evolutionary Biology http://dx.doi.org/10.7554/eLife.24232.024 http://dx.doi.org/10.7554/eLife.24232 and Lesedi Chamber pairs are comparable in size, but the Lesedi vertebrae clearly belong to a larger, more robust (presumed male) individual. Although the Dinaledi transverse processes are bro- ken at their bases, the preserved aspects are strongly posteriorly oriented, albeit to a lesser degree than those from the Lesedi Chamber (Figure 22). The lower thoracic transverse processes of Au. afarensis, Au. africanus, and Au. sediba possess more laterally oriented transverse processes. Only SKX-41692, a presumed P. robustus T10, possesses similarly posteriorly oriented transverse pro- cesses among australopiths. However, its relatively large vertebral body, small spinal canal, and over- all shape contrast with the lower thoracic transverse processes of H. naledi (Williams et al., 2017). The combination of a relatively large vertebral body and spinal canal is present in both the Dinaledi and the Lesedi Chamber T10 vertebrae, but not inAustralopithecus and Paranthropus specimens. The Dinaledi T11 bears planiform articular facets superiorly and inferiorly and is therefore not the transitional vertebra. In the Lesedi Chamber vertebral column, the change in articular facet orienta- tion occurs asymmetrically across the T11 and T12 vertebrae, as occurs in <4% of modern humans (Williams et al., 2017). In all known Australopithecus and H. erectus specimens, the transitional ver- tebra is T11 (Haeusler et al., 2002, 2011; Williams et al., 2013; Meyer et al., 2015). Costae U.W. 102a-250 is a nearly complete right first rib, with erosion and breakage to the head, tubercle, lateral border and distal end. The neck is flattened in its superior-inferior dimension and descends in the vertebro-inferior direction. The tubercle and the posterior angle coincide. The facet of the articu- lar tubercle was damaged post-mortem. Figure 21. U.W. 102a-151 vertebra, T11. Clockwise from top left: posterior, superior, inferior, left, right, and anterior views. Scale bar = 2 cm. DOI: 10.7554/eLife.24232.025 Hawks et al. eLife 2017;6:e24232. DOI: 10.7554/eLife.24232 28 of 63 Research article Genomics and Evolutionary Biology http://dx.doi.org/10.7554/eLife.24232.025 http://dx.doi.org/10.7554/eLife.24232 Two partial first ribs (U.W. 101–083 and U.W. 101–621) of H. naledi are preserved in the Dinaledi Chamber hominin sample, but neither rib has its head nor enough of the shaft preserved to allow accurate estimation of curvature (Williams et al., 2017). The angulation and shape of these frag- ments appears comparable to those of MH2 Au. sediba and A.L. 288–1 Au. afarensis. U.W. 102a- 250 is more complete than the Dinaledi first rib fragments, and is similar in morphology in the over- lapping regions. This rib is slightly more curved than the Sterkfontein first rib, StW 670 (Tawan et al., 2016). The anatomy of U.W. 102a-250 is entirely compatible with attribution to H. naledi, although the bone is also similar in morphology and size to known australopith first ribs. Thirteen additional specimens from 102a are partial ribs or rib fragments, none are identifiable to element and none present anatomical information that is useful for testing the taxonomic affiliation of the sample. Figure 22. Vertebral transverse process orientation. H. naledi is distinctive when compared to many other hominin species in having T10 and T11 vertebral transverse processes oriented with a relatively low angle. Left: U.W. 102a-036 compared to U.W. 101–855 from the Dinaledi Chamber (top), and U.W. 102a-151 compared to U.W. 101–1733 (bottom). All of these vertebrae have transverse processes oriented more posteriorly than those of most other hominins, U.W. 102a-036 is the most extreme. Right: charts showing the comparative orientation of transverse processes in humans, living great apes, and fossil hominins. For the T10 (top), the U.W. 102a-036 value (labeled ‘Lesedi’) is lower than that for any other hominins, while the Dinaledi T10 is similar to the Neandertal value and extremely low compared to that for modern humans. The T11 (bottom) shows a similar but less extreme pattern. DOI: 10.7554/eLife.24232.026 Hawks et al. eLife 2017;6:e24232. DOI: 10.7554/eLife.24232 29 of 63 Research article Genomics and Evolutionary Biology http://dx.doi.org/10.7554/eLife.24232.026 http://dx.doi.org/10.7554/eLife.24232 Ossa coxae U.W. 102a-138 (Figure 23) is a fragmentary right ilium of an immature individual (as evident by the presence of triradiate cartilage, by an unfused apophysis at the anterior inferior iliac spine, and by very small overall size). The fragment is very light, with thin cortical bone, and is eroded around mar- gins of the acetabular portion. The iliac blade is mostly missing, but the auricular surface, greater sci- atic notch, acetabulosacral buttress, and anterior margin of the iliac blade are present. Despite the thin and fragile nature of this element, the surface is well-preserved. The adult pelvic material of H. naledi from the Dinaledi Chamber is notable in combining an Au. afarensis-like degree of iliac flare, a weak and anteriorly placed iliac pillar, and a narrow tuberoace- tabular sulcus on the ischium (Berger et al., 2015; VanSickle et al., personal communication). U.W. 102a-138 represents the most complete immature ilium fragment of H. naledi found to date, and its morphology is comparable to that of the juvenile U.W. 101–486 ilium fragment, and thus consistent with the morphology seen in H. naledi. It lacks the diagnostic characters that could differentiate it clearly from ilium fragments from other hominin species, as the iliac blade and iliac pillar are both poorly preserved. The lack of an accompanying ischial fragment precludes an evaluation of tuberoa- cetabular sulcus morphology in the 102a material. Femora U.W. 102a-001 is a proximal right femur, in which much of the head and neck, and the proximal sub- trochanteric shaft are preserved (Figure 24). The head is badly eroded, especially anteriorly, and Figure 23. U.W. 102a-138 immature right os coxa fragment. The medial view is at the center. Clockwise from top: superior, lateral, inferior and anterior views. The unfused triradiate suture is notable. DOI: 10.7554/eLife.24232.027 Hawks et al. eLife 2017;6:e24232. DOI: 10.7554/eLife.24232 30 of 63 Research article Genomics and Evolutionary Biology http://dx.doi.org/10.7554/eLife.24232.027 http://dx.doi.org/10.7554/eLife.24232 only a few small patches of subchondral articular bone are preserved on the posterior aspect. The posterior side of the neck is fairly well preserved from the head all the way to the lesser trochanter, which is planed off, with only the base remaining. The anterior side of the neck is missing. Trabecular bone is exposed from the anterior head all the way to the lateral surface at the base of the greater trochanter. The greater trochanter is missing entirely, save for a small bit of its distal lateral surface. The surface overall is marred by areas of post-depositional damage, including a number of trans- verse scratches on the shaft. U.W. 102a-003 is a left femoral shaft fragment, from the lesser trochanter proximally to about midshaft (Figure 25). Only the base of the lesser trochanter remains. The head and neck are not present. U.W. 102a-004 is a fragment of left distal femur, preserved from roughly midshaft to the distal subchondral bone surface of the intercondylar notch (Figure 26). Both condyles are missing. The shaft has surficial markings similar to those present on U.W. 102a-001. This fragment is morphologi- cally compatible with U.W. 102a-003 in shaft diameter and cross-section, and the two specimens exhibit no morphological overlap, suggesting that they may represent the same femur. U.W. 102a-003 and U.W. 102a-004 may conjoin with each other. Both fragments are morphologi- cally compatible in shaft diameter and cross-section, and at the broken distal end of U.W. 102a-003 and at the proximal end of U.W. 102a-004, a small part of the circumference of the shaft (approxi- mately 10 mm in total) on the posterolateral side appears to provide a refit. However, the edges of this apparent break are abraded, reducing the certainty of the association. Joining the bones at this point, U.W. 102a-003 and U.W. 102a-004 preserve 321 mm of a femoral shaft. Using the similarly sized U.W. 101–002 to represent the missing proximal end and KNM-ER 1481 to represent the distal end (Figure 27), we preliminarily estimate the femoral length of this individual to be ~375 mm. U.W. 102a-001 (right proximal) and U.W. 102a-003 (left proximal) are similar in size. They preserve an overlapping area of anatomy from just above the lesser trochanter down to around the midshaft area. However, despite their similarity in size, the two contrast in several Figure 24. U.W. 102a-001 proximal femoral fragment. From left: posterior, medial, anterior and lateral views. Right from top: proximal and distal views. Scale bar = 5 cm. DOI: 10.7554/eLife.24232.028 Hawks et al. eLife 2017;6:e24232. DOI: 10.7554/eLife.24232 31 of 63 Research article Genomics and Evolutionary Biology http://dx.doi.org/10.7554/eLife.24232.028 http://dx.doi.org/10.7554/eLife.24232 anatomical details. U.W. 102a-001 is more platymeric in the subtrochanteric area, with a greater mediolateral (ML) breadth than U.W. 102a-003. The lesser trochanter is abraded in both specimens, but the morphology of the inferior and medial aspects of it appear different in the two bones. U.W. 102a-001 has a shallow sloping border to the lesser trochanter medially, and the inferior aspect tails off into a broad, less marked line leading to a very slight linea aspera by midshaft. By contrast, U.W. 102a-003 has a steep medial aspect to the lesser trochanter, and it tails into a sharply defined crest that broadens around 15 mm down the shaft into a rugose, double crest, which narrows by midshaft into a strong linea aspera. The insertion for m. gluteus maximus is prominent and rugose in both femora but in U.W. 102a-001, the rugosity extends further down the shaft. Overall, the asymmetry of these two bones would be very unusual in the left and right femora of a single individual. We accept this provisionally as evidence for a second adult individual in the 102a assemblage. Comparative femoral anatomy The femoral morphology of H. naledi is a mosaic of features seen in Australopithecus species such as Au. afarensis, Au. africanus and Au. sediba, and features otherwise known in Homo, including Plio- Pleistocene H. erectus and fossils attributed to Homo sp. indet., such as KNM-ER 1472 and KNM-ER 1481 (Figure 28; Marchi et al., 2017). The femoral remains from the 102a locality share this mosaic of features, including: a marked linea aspera, with a weak pilaster in U.W. 102a-003; a strong muscle insertion for m. gluteus maximus; a platymeric shaft just inferior to the lesser trochanter; a long and anteroposteriorly flattened femoral neck; and an AP expanded femoral midshaft. The Dinaledi remains of H. naledi have markedly anteverted femoral necks. Damage to the femoral neck of the U. W. 102a-001 specimen limits the accuracy of an estimate of anteversion, but at 115 degrees, this Figure 25. U.W. 102a-003 left proximal femur fragment. From left: posterior, medial, anterior and lateral views. Right from top: proximal and distal views. Scale bar = 5 cm. DOI: 10.7554/eLife.24232.029 Hawks et al. eLife 2017;6:e24232. DOI: 10.7554/eLife.24232 32 of 63 Research article Genomics and Evolutionary Biology http://dx.doi.org/10.7554/eLife.24232.029 http://dx.doi.org/10.7554/eLife.24232 estimate is within the range for the Dinaledi specimens. The most distinctive feature of the H. naledi proximal femur is the presence of two mediolaterally oriented pillars on the superior femur neck, separated by a medially positioned shallow and vascularized groove where m. obturator internus and gemelli insert (Marchi et al., 2017). This configuration is not seen in other hominin species. The U.W. 102a-001 femoral neck is damaged superiorly, precluding a clear assessment of whether two distinct pillars were present. What is preserved posteriorly demonstrates the presence of a medially positioned vascularized groove, but not of the inferior pillar present in the Dinaledi femora. With this exception, the U.W. 102a-001, U.W. 102a-003 and U.W. 102a-004 femoral fragments are entirely consistent with the morphology known for H. naledi, and the combination of features in the U.W. 102a-001 proximal femur is not consistent with known fossil examples attributed to other species of Homo or Australopithecus. Individuals represented in 102a The hominin material from 102a appears to represent a minimum of two adult individuals and one immature individual. The inference of two adults is based upon the morphological incongruence of the left (U.W. 102a-003) and right (U.W. 102a-001) femoral elements (discussed above). Still, no adult element is clearly duplicated in the collection. The U.W. 102a-138 ilium, along with an immature sacrum fragment and two immature long bone fragments not described here, demonstrates the presence of at least one immature individual. The lack of duplication of elements suggests that much (but not all) of the adult material may rep- resent a single individual skeleton, which parsimoniously would also include the LES1 cranium. We Figure 26. U.W. 102a-004 left distal femur fragment. From left: posterior, medial, anterior and lateral views. Right from top: proximal and distal views. Scale bar = 5 cm. DOI: 10.7554/eLife.24232.030 Hawks et al. eLife 2017;6:e24232. DOI: 10.7554/eLife.24232 33 of 63 Research article Genomics and Evolutionary Biology http://dx.doi.org/10.7554/eLife.24232.030 http://dx.doi.org/10.7554/eLife.24232 accept this hypothesis provisionally. All elements in the current 102a collection were recovered from within an excavation area less than 50 cm x 70 cm, and 40 cm deep. Two vertebrae (U.W. 102a-154a and U.W. 102a-154b) were in articulation in situ. The articular morphology and sizes of seven verte- brae suggest strongly that they represent a single individual; the remainder of these elements were recovered in close physical proximity but not in articulation. All fragments attributed to the LES1 cra- nium were likewise recovered from a small area. The hand and wrist material is consistent with a sin- gle right hand on the basis of articular morphology. None of the other elements lend themselves to an evaluation of articular compatibility, but they are consistent in size. We consider it unlikely that the number of elements in the current collection would be recovered from a commingled assem- blage consisting of substantial parts of multiple skeletons without also introducing duplicate elements. This raises the problem of what seems to be a mismatch of the U.W. 102a-001 and U.W. 102a- 003 femora. These two are similar in size, and the difference in their shaft dimensions is not greater than that found in 95% of a large sample of paired left and right human femora. However, they are different in muscle attachments and diaphyseal morphology, to the extent that would represent unusual asymmetry in a single individual. The data do not allow us to discard the hypothesis that one of the femora represents a second adult individual in 102a, albeit an individual of similar body size. The descriptive and measurement data do not indicate which (if either) of these femora may belong to the individual represented by LES1, nor which (if any) of the other postcranial elements are associated with either femur. On the basis of the non-duplication of elements, it seems likely that if there is a second individual, this second individual is represented by only a small number of elements, possibly just the femur. The two humerus specimens, U.W. 102a-002 and U.W. 102a-257, differ slightly in shaft diameter where it can be compared, but no other morphological differences are apparent on the preserved fragments, and a slight degree of upper limb asymmetry is not Figure 27. Length estimation of femur based on U.W. 102a-003 and U.W. 102a-004. Two specimens were used to estimate the missing proximal and distal ends of the femur. Top: U.W. 101–215 is a distal femur fragment that presents a similar morphology to the U.W. 102a-004 distal femur, while preserving the distal articular surface. Middle: U.W. 102a-004 and U.W. 102a-003 conjoined, in posterior view. Bottom: U.W. 102a-001 is comparable in size with U.W. 102a-003, and while the morphology of the muscle markings is different, the alignment of the lesser trochanters gives a good basis for estimating the proximal extent of the bone. The length estimate is 375 mm. DOI: 10.7554/eLife.24232.031 Hawks et al. eLife 2017;6:e24232. DOI: 10.7554/eLife.24232 34 of 63 Research article Genomics and Evolutionary Biology http://dx.doi.org/10.7554/eLife.24232.031 http://dx.doi.org/10.7554/eLife.24232 Figure 28. Comparison of H. naledi femora to those attributed to early Homo. Top: roximal femora attributed to Au. afarensis, Homo sp., H. naledi, and H. erectus, all shown in posterior view. The femora have been aligned by matching the inferior point on the lesser trochanter to the horizontal line on the figure, and all are shown at approximately the same shaft angle. Many of the H. naledi femora have notably thin shafts, although the largest shown here, U.W. 101–1475, is greater in shaft thickness than the complete KNM-ER 1480, KNM-ER 1472, or D4167 femora. The specimens here attributed to ‘Homo sp.’ were surface finds without associated cranial material. Their anatomy has been considered consistent with Homo, although Figure 28 continued on next page Hawks et al. eLife 2017;6:e24232. DOI: 10.7554/eLife.24232 35 of 63 Research article Genomics and Evolutionary Biology http://dx.doi.org/10.7554/eLife.24232 unusual in humans or in fossil Homo. We conclude that most of the 102a adult material probably represents a single skeleton, but we cannot assume that the ratios of either femur with other ele- ments in the sample would reflect the proportions of a single individual. To be conservative, any consideration of proportions should allow for the possibility that multiple individuals are present. Hominin material from 102b The material from area 102b includes 12 specimens identified as hominin. Most of this collection consists of small fragments of cranium, many of which are identifiable to element, but which pre- serve no diagnostic morphology that would assist in taxonomic assessment. One partial mandible and five possibly associated teeth preserve morphological characters that are useful for taxonomic attribution. Figure 28 continued some have suggested that KNM-ER 738 may instead be Paranthropus. U.W. 102a-001 is a right femoral fragment, and left femora here have been mirrored for comparison, including KNM-ER 738, KNM-ER 1481, U.W. 101–398, U.W. 101–1475, KNM-WT 15000, and A.L. 288–1. Bottom: H. naledi femur length compared to those of other fossil femora. U.W. 102a-003 is shown conjoined with U.W. 102a-004. The top and bottom horizontal lines correspond to the proximal and distal limits of the maximum length estimate for this femur as illustrated in Figure 24. Femur maximum lengths and length estimates are as reported by McHenry (1991), D4167 length is from Lordkipanidze et al. (2007). As above, all femora are aligned by the lesser trochanter, which is preserved in all of these shaft fragments. The H. naledi U.W. 102a-003 femur has a shaft diameter comparable to that of A.L. 288–1, or even slightly shorter, but at an estimated 375 mm, it is nearly the same length as the D4167 femur of H. erectus at 387 mm. The U.W. 101–484 tibia specimen from the Dinaledi Chamber is also long and relatively narrow, and has an estimated length greater than that of the D3901 tibia from Dmanisi (Marchi et al., 2017). Several very thick, large femora have been attributed to H. erectus from the Early Pleistocene of Africa; these are very different from U.W. 102a-003 in their size, robustness, and more platymeric proximal shafts. U.W. 101–938 is an immature femur with no fusion of the head or distal epiphysis; it represents a younger developmental age than KNM-WT 15000, the immature H. erectus skeleton. Even at its young age, this H. naledi specimen is nonetheless longer than the A.L. 288–1 femur of Au. afarensis. A.L. 288–1 is shown here as reconstructed by P. Schmid. For comparison to the right U.W. 102a-003 femur, left femora are shown mirrored here, these include KNM-ER 1472, U.W. 101–938, and D4167. In this figure, KNM-ER 739, KNM-ER 1472, KNM-ER 1481, KNM-WT 15000, O.H. 28, and A.L. 288–1 are represented by casts. DOI: 10.7554/eLife.24232.032 Figure 29. U.W. 102b-438 immature mandibular fragment. From left: basal, lingual, occlusal, buccal and anterior views. The RP4 is within its crypt. Scale bar = 2 cm. DOI: 10.7554/eLife.24232.033 Hawks et al. eLife 2017;6:e24232. DOI: 10.7554/eLife.24232 36 of 63 Research article Genomics and Evolutionary Biology http://dx.doi.org/10.7554/eLife.24232.032 http://dx.doi.org/10.7554/eLife.24232.033 http://dx.doi.org/10.7554/eLife.24232 U.W. 102b-438 is a fragment of right mandibular corpus from an immature individual (Figure 29). The alveoli for the RM1 are present; the alveoli for the rdm2 are also present and reveal the crypt with RP4 crown intact within. The fragment is broken anteriorly at the crypt for RP3 and posteriorly at the crypt for RM2, neither permanent tooth is present. The preserved corpus height at the mid- point of rdm2 is 14.9 mm; this is perhaps 1–2 mm less than the true value because of the erosion of the alveolar bone surface. Corpus breadth at the anterior edge of M1 is 15.2 mm; total length of the fragment as preserved is 43.5 mm. Several teeth excavated from 102b are compatible with the same approximate developmental stage as the U.W. 102b-438 mandibular fragment, and were recovered in close proximity to each other and to the mandible; no elements are duplicated. We hypothesize that these fragments represent the same individual, at least until the recovery of additional material makes us reassess this possible association. U.W. 102b-437 is the complete crown and two nearly complete roots of a ldm2 with moderate occlusal wear, including dentine exposure at the centra of cusps, and roots that appear to have begun to resorb at the tips. U.W. 102b-503 is a RP4 crown, nearly com- plete but not erupted. U.W. 102b-511 is a LC1 crown, nearly complete with no occlusal wear. U.W. 102b-515 is a LI2 that is nearly crown complete and unerupted. U.W. 102b-178 is a broken but apparently unworn probable RI2 crown that was recovered separately from these other teeth, which may also represent this individual. The teeth all fall within the size range of equivalent elements from the Dinaledi Chamber sample. The U.W. 102b-438 RP4 crown is partially obscured within its crypt and not complete, to the minimal extent it is visible at this time, it is consistent with H. naledi mandibular P4 morphology. The U.W. 102b-503 maxillary P4 is likewise consistent with comparable Dinaledi examples, although this tooth differs little among several species of Homo, including modern humans. Like the incisors of LES1, U. W. 102b-515 lacks prominent cervico-incisal crown curvature, has no prominent crests on its lingual surface, and is not shoveled. The U.W. 102b-437 ldm2 is buccolingually narrow and mesiodistally long, with an ovo-rectangular shape. Five well-developed cusps are present in a Y-fissure pattern, and the talonid is wider than the trigonid. Compared to the two lower dm2 specimens from the Dinaledi chamber, the metaconid and protoconid are relatively small, although it is not clear whether or not this is an artifact of mesial wear. Although the mesial aspect is worn, it is clear that the mesial marginal ridge was thick; it forms the distal border of a fissure-like anterior fovea. A thick distal marginal ridge is also present and it borders a fissure-like posterior fovea. Both the mesio- and disto-buccal grooves are deep and are associated with a wide V-shaped furrow. No protostylid is present. The mesial and distal roots each have a buccal and lingual canal connected by a dentin plate. The crown morphology is nearly identi- cal to that of the two analogous teeth from the Dinaledi chamber (U.W. 101–655 and U.W. 101– 1686), and supports their taxonomic designation to H. naledi. Although few have been found, the lower dm2s of H. erectus s.l. are relatively longer and more rectangular than those of H. naledi. In addition, the substantially wider talonid compared to the trigonid of both the 102 tooth and the H. naledi lower dm2s differentiates these dm2s from those of most other Homo. The most important of the 102b teeth for morphological assessment is the LC1. The U.W. 102b- 511 crown preserves its occlusal morphology and replicates with better detail the same morphology observed in the LES1 mandibular canines (Figure 30). It has asymmetrically placed crown shoulders, with the mesial shoulder more apically placed than the distal. Further, the distal shoulder is formed by an accessory cuspule and the mesial crest is shorter and more convex than the vertically disposed distal crest. These features are identical to those found in the Dinaledi permanent mandibular can- ines and are among the defining features of H. naledi. The asymmetrical shoulders of the canine crown and distal accessory cuspule are found to some degree in a number of specimens attributed to H. erectus, H. rudolfensis, H. habilis and Australopithecus, but the small canine and incisor sizes are inconsistent with attribution to any of these other taxa. On the basis of these observations, the immature dentition from 102b is identifiable as H. naledi. All of the cranial fragments recovered from 102b are compatible with an immature developmental stage, and may come from the same individual as U.W. 102b-438. The minimum number of individuals (MNI) in 102b is therefore one individual at present. Without more information about the depositional history of the 102b locality in comparison with that of 102a, we cannot say for sure whether any of the specimens collected at this locality may represent the same immature individual that is represented by the pelvic fragments in 102a. The evidence does not currently exclude the Hawks et al. eLife 2017;6:e24232. DOI: 10.7554/eLife.24232 37 of 63 Research article Genomics and Evolutionary Biology http://dx.doi.org/10.7554/eLife.24232 hypothesis that the U.W. 102a-138 ilium and other immature pelvic fragments may represent the same individual as the U.W. 102b-438 mandible. The 102a and 102b areas are separated by approxi- mately 3 m, and there is a possibility that slumping of material might have brought the remains of a single individual into both areas. Further work to establish the life-history stage of the immature remains may help to resolve this question. Hominin material from 102c The 102c deposit is a very small (~2 L) volume of sediment enclosed within a dissolution cavity in the cave wall. Only one morphologically identifiable hominin specimen has been recovered from this area. U.W. 102c-589 is a fragment of left mandibular corpus with worn LM1 and LM2 in situ (Fig- ure 31). These teeth have less occlusal wear than those of the LES1 mandible, but are morphologi- cally very similar. The mandibular corpus is broken irregularly at the alveoli and the base of the mandible is not present. Anteriorly, it is broken at the mesial M1 root; posteriorly, it is broken at the mesial alveolus for the M3. The root of the ascending ramus is preserved and becomes independent Figure 30. U.W. 10