Great ape skeletal collections: Making the most of scarce and irreplaceable resources in the digital age


  • Adam D. Gordon,

    Corresponding author
    1. Department of Anthropology, University at Albany, Albany, NY
    • Correspondence to: Adam Gordon, Department of Anthropology, CAS 237, University at Albany, 1400 Washington Avenue, Albany, NY 12222, USA. E-mail:

    Search for more papers by this author
  • Emily Marcus,

    1. Honors Program, George Washington University, Washington, DC
    2. Center for the Advanced Study of Hominid Paleobiology, Department of Anthropology, George Washington University, Washington, DC
    Search for more papers by this author
  • Bernard Wood

    1. Center for the Advanced Study of Hominid Paleobiology, Department of Anthropology, George Washington University, Washington, DC
    Search for more papers by this author


Information about primate genomes has re-emphasized the importance of the great apes (Pan, Gorilla, and Pongo) as, for most purposes, the appropriate comparators when generating hypotheses about the most recent common ancestor of the hominins and panins, or the most recent common ancestor of the hominin clade. Great ape skeletal collections are thus an important and irreplaceable resource for researchers conducting these types of comparative analyses, yet the integrity of these collections is threatened by unnecessary use and their availability is threatened by financial pressures on the institutions in which the collections reside. We discuss the general history of great ape skeletal collections, and in order to get a better sense of the utility and potential of these important sources of data we assemble the equivalent of a biography of the Powell-Cotton Collection. We explore the history of how this collection of chimpanzee and gorilla skeletons was accumulated, how it came to be recognized as a potentially important source of comparative information, who has made use of it, and what types of data have been collected. We present a protocol for collecting information about each individual animal (e.g., which bones are preserved, their condition, etc.) and have made that information about the Powell-Cotton Collection freely available in an online relational database (Human Origins Database, As an illustration of the practical application of these data, we developed a tabular summary of ontogenetic information about each individual (see Appendices A and B). Collections like the Powell-Cotton are irreplaceable sources of material regarding the hard-tissue evidence and recent history of the closest living relatives of modern humans. We end this contribution by suggesting ways that curators and the researchers who use and rely on these reference collections could work together to help preserve and protect them so that future generations can use and benefit from these priceless resources. Am J Phys Anthropol 57:2–32, 2013. © 2013 Wiley Periodicals, Inc.


New molecular (e.g., Arnold et al., 2010; Perelman et al., 2011; Prado-Martinez et al., 2013) and morphological (Diogo and Wood, 2011) evidence has confirmed the close relationship between modern humans and the great apes, and the particularly close relationship between modern humans and chimpanzees and bonobos. Great ape skeletal collections thus provide the core data required to study hominin fossils in a comparative evolutionary context. Unfortunately, these collections, which are uncommon to begin with, are, even with the best curatorial oversight, slowly but surely degrading over time. The collections are irreplaceable in at least two senses. First, the animals involved are endangered, and second many museum collections sample taxa from parts of their range that are no longer occupied by living animals. We suggest the time has come for a thorough evaluation of these critically important collections. This would include assembling a detailed inventory of each collection, devising and implementing ways of making use of information that has already been collected, and making recommendations about standard practices that will help preserve these precious collections for the future.

Not that long ago, when conducting comparative morphological research addressing questions about hominin evolution, it would have been deemed sufficient (even if not ideal) to collect hard-tissue data from the nearest skeletal collection of great apes. Most studies were content to use genus level samples (e.g., chimpanzees, lowland gorillas, or orangutans). However, the results of fine-grained morphological and genetic analyses (e.g., Groves, 2001; Pilbrow, 2006; Gonder et al., 2011; Prado-Martinez et al., 2013) make it clear that given the morphological variation that is distributed geographically among subspecies, when researchers assemble samples of great apes for comparative analyses care needs to be taken to assemble skeletal samples from regions that coincide with the known ranges of species or subspecies. Yet to our knowledge there is no single source of information about great ape skeletal collections that takes into account the important developments that have occurred since Groves' (2001) important review.

There is also a growing realization within the research community that we cannot take the continued availability of great ape skeletal collections for granted. We believe they are threatened by a combination of overuse, museums diverting resources from collections to what is perceived to be more important “basic” research, and the parlous financial state of host institutions that lack the prestige and funding of national museums. Thus, there is a pressing need for the community to maximize the utility of these collections by minimizing redundant data collection. No one should be turned away from examining the collections, but repeated and unnecessary measurement inevitably damages specimens and wastes the collective effort of the research community. That said, there is an argument that could be made for deliberately repeating some measurements in the interest of evaluating comparability across data sets collected by different researchers. We revisit this last point toward the end of this article.

We are conscious that some paleoanthropologists may not be aware of the history of research about the relationships among the great apes and modern humans, so because this is the core of the case for the importance of these collections we begin by briefly reviewing this. We then set out recently accumulated morphological and genetic evidence about the taxonomy and biogeography of each of the great apes. To our knowledge there is no single source of information about great ape skeletal collections, so in the next section we briefly survey information published in print and available online regarding the major collections. We then suggest what types of information it would be useful to know about a collection before visiting it. In the main section of this review, we present a “biography” of one of the more comprehensive collections; for various reasons the exemplar we have chosen is the Powell-Cotton Collection. We set out its history, how it came to be recognized as a potentially important source of comparative information about the African apes, then we review what use has been made of it, and what types of data were collected. We introduce the Human Origins Database (, a freely-available online relational database we constructed as a pilot of the type of detailed information we suggest should be available for all great ape collections. Finally, we present suggestions about how the community might collectively move forward to help preserve and protect great ape skeletal collections for future generations.


In an essay entitled On the Relations of Man to the Lower Animals that formed the middle section of a small book entitled Evidence as to Man's Place in Nature, Huxley (1863) concluded that the phenetic differences between modern humans and the gorilla (and by inference the chimpanzee) were less marked than the differences between the gorilla and the orangutan and gibbons. Most recent attempts to use gross morphological evidence to generate hypotheses about higher primate relationships have confirmed the close relationship between modern humans and the African apes (e.g., Gibbs et al., 2002; Diogo and Wood, 2011), although some still claim to see gross morphological arguments in favor a modern human–orangutan sister relationship (Grehan and Schwartz, 2009).

During the first half of the 20th century the focus of the search for evidence about higher primate relationships shifted from gross morphology to the morphology of molecules (e.g., Grünbaum, 1902; Nuttall, 1904). In the early 1960s, Zuckerkandl et al. (1960) and Goodman (1963) used hemoglobin and albumin, respectively, to investigate the relationships among higher primates and they both concluded that chimpanzees were more closely related to modern humans than to gorillas. Others concurred (Sarich and Wilson, 1967) suggesting that 99% of the amino-acid sequences of chimpanzee and modern human proteins were identical (King and Wilson, 1975).

Initial attempts to compare the DNA of higher primates used a method called DNA hybridization (e.g., Caccone and Powell, 1989), but once sequencing methods became available they rapidly replaced hybridization as the preferred method for generating hypotheses about the relationships among living hominoids and the number of sequence-based studies increased each year (see Bradley, 2008; Arnold et al., 2010; Perelman et al., 2011; for reviews). When these DNA differences were calibrated using what was then the best paleontological evidence for the split between the apes and the Old World Monkeys, it was predicted that the hypothetical ancestor of modern humans and chimpanzees/bonobos lived between about 5 and 8 million years ago (Ma) (Bradley, 2008). New estimates based on empirical data about generation times (Langergraber et al., 2012) suggest that the date is probably closer to 8 than to 5 Ma, although a more recent analysis of a larger data set (Prado-Martinez et al., 2013) suggests it is closer to 5 Ma, and this date may be further affected by recalibrations of the molecular clock based on the newly discovered Oligocene catarrhine Rukwapithecus fleaglei, which has been argued to be a basal hominoid (Stevens et al., 2013).

Whole genomes can now be sequenced with acceptable levels of coverage, and in the last few years researchers have published good draft sequences of the genomes of the chimpanzee (Chimpanzee Sequencing and Analysis Consortium, 2005), the orangutan (Locke et al., 2011), the gorilla (Scally et al., 2012), and the bonobo (Prüfer et al., 2012). Scally et al. (2012) sampled two Western Lowland and one Eastern Lowland gorilla and showed that if you take all of the genome then the greatest number of similarities are between modern humans and chimpanzees. However, they also found that “in 30% of the genome, gorilla is closer to human and chimpanzee than the latter are to each other” (ibid, p. 169), a phenomenon referred to as incomplete lineage sorting (ILS). The Prüfer et al. (2012) study showed that bonobos and common chimpanzees are 99.7% alike, whereas 98.7% of the bonobo genome resembles that of modern humans. The latter authors also found evidence of ILS in their study in that ∼3% of the modern human genome is more closely related to bonobos or to common chimpanzees than either of these taxa is related to the other; they also suggest that 25% of all genes contain evidence of ILS. That said, a recent comparative study of 79 great ape genomes representing all six currently recognized species has also emphasized the presence of genetically distinct populations within each great ape species (Prado-Martinez et al., 2013).

It has been clear for several decades now that phylogenetic relationships must be taken into account in comparative studies (e.g., see Felsenstein, 1985; Harvey and Pagel, 1991; Garland et al., 1992; Martins and Hansen, 1997; Nunn, 2011), and thus the correct taxonomic identification of specimens in comparative collections is important in the context of correct phylogenetic placement. However, taxonomic information associated with individual museum specimens is often dated and may lag behind contemporary consensus taxonomy. Correct taxonomic identification is also inextricably bound up with accurate biogeographic provenance of specimens, information that regrettably is often poorly documented. In the next two section, we review the current consensus on the taxonomy and geographic distribution of Pan, Gorilla, and Pongo (Fig. 1).

Figure 1.

Current consensus of phylogenetic relationships among modern humans and great ape species, and current subspecies designations in their presumed phylogenetic position (dotted lines indicate branches leading to subspecies). Branch lengths within genera are not necessarily proportional to time. Note that the division of the eastern chimpanzee into Pan troglodytes schweinfurthii and P. t. marungensis is based on a morphological distinction that is not supported by molecular evidence (Gonder et al., 2011). Phylogenetic relationships within each genus are drawn from a variety of sources (Pan: Groves, 2005; Gonder et al., 2011; Gorilla: Groves, 2001; Scally et al., 2012; Pongo: Brandon-Jones et al., 2004; Singleton et al., 2004; Locke et al., 2011); more recently, Prado-Martinez et al. (2013) suggested that P. t. verus is sister to P. t. ellioti rather than sister to a clade including all other P. troglodytes subspecies.



The present consensus (e.g., Groves, 2001; Prüfer et al., 2012) is that the genus Pan includes at least two species: Pan paniscus, the bonobo, and Pan troglodytes, the common chimpanzee (Fig. 1).

The range of P. paniscus is presently confined to the region in the Democratic Republic of the Congo that lies to the south of the Congo River. Prüfer et al. (2012) found “no indication of preferential gene flow between bonobos and any of the chimpanzee groups tested,” and they suggest that this is “consistent with the suggestion that the formation of the Congo River c.2.5–1.5 Ma created a barrier to gene flow that allowed bonobos and chimpanzees to evolve different phenotypes over a relatively short time” (ibid, p. 528). The time range of estimates for the separation of the two extant species matches the range of 2.6–1.5 Ma cited by Langergraber et al. (2012), and it is also consistent with the older end of the slightly later range of 1.7–0.75 Ma based on genome-level comparisons (Prado-Martinez et al., 2013).

The geographical range of P. troglodytes extends from Senegal and Guinea in the west to Uganda and Tanzania in the east, but it is not continuous between these two longitudinal extremes for there are no chimpanzees in the relatively arid Dahomey Gap. The range of the most western Pan isolate, the subspecies Pan troglodytes verus (also known as the Western chimpanzee) extends from Senegal in the west, through Liberia to the Ivory Coast in the east. Morin et al. (1994) used evidence from mitochondrial DNA (mtDNA) to suggest that this isolate should be recognized as a separate species, but few have adopted this suggestion.

Gonder et al. (1997) suggested that the mtDNA of chimpanzees from western Nigeria and from localities along the Nigeria-Cameroon border is distinct from that of P. t. verus, and this interpretation was confirmed by a more recent study of microsatellites (Gonder et al., 2011). Groves (2001) suggested that the latter populations should be recognized as a separate subspecies, P. troglodytes vellerosus, also called the Nigerian-Cameroonian chimpanzee. However, for reasons of priority these populations have since been renamed as P. troglodytes ellioti (Oates et al., 2009).

The range of the subspecies P. troglodytes troglodytes, or the Central chimpanzee, extends from its northern boundary with P. t. ellioti in Northern Cameroon (this approximates to the Sanaga River) to the Oubangui River in the east and the Congo River to the east and south. The Oubangui River separates the range of P. t. troglodytes from that of the Eastern chimpanzee, P. troglodytes schweinfurthii, and the Congo River separates it from the range of the bonobo (see above).

The Eastern chimpanzee is found from the Oubangui River eastwards into Western Uganda, Rwanda, and Tanzania; its southern range extends to the Congo River. Although often recognized as a single subspecies, P. t. schweinfurthii, Groves (2005) has argued that this group should be divided into two subspecies on the basis of craniometric variation, but Gonder et al. (2011) found no genetic support for such a division. In Groves' scheme, P. t. schweinfurthii occupies the northwestern part of the range of the Eastern chimpanzee, including the Ituri district in the Democratic Republic of the Congo as well as north and west of there. P. troglodytes marungensis is found in Uganda and in the Maniema district in the DRC, as well as in the southeast extent of the range.

Pilbrow (2006) provides a useful detailed review of the literature covering the molecular and morphological evidence for P. troglodytes. Her odontometric data are consistent with other morphological evidence (Braga, 1995; Uchida, 1996; Braga, 1998; Guy et al., 2003; Taylor and Groves, 2003; Lockwood et al., 2004) and with genetic evidence that all suggest that P. t. verus is the most distinctive subspecies of P. troglodytes. Her data also support the genetic and other morphological evidence for the distinctiveness of P. t. ellioti.


The present consensus (e.g., Groves, 2001; Pilbrow, 2010; Scally et al., 2012) is that the genus Gorilla includes two species: Gorilla gorilla, the Western gorilla, and Gorilla beringei, the Eastern gorilla (Fig. 1). When Scally et al. (2012) compared the genomes of the Western and Eastern lowland gorilla individuals, they estimated “an average sequence divergence time 1.75 Ma, but with evidence for more recent genetic exchange and a population bottleneck in the eastern species” (ibid, p. 169); this is also consistent with the results of Prado-Martinez et al. (2013) (i.e., divergence 1.7–0.75 Ma). The original dates proposed by Becquet and Przeworski (2007) and Thalmann et al. (2007) for the split, 1.29 Ma and 2.13–1.2 Ma, respectively, have recently been recalibrated by Langergraber et al. (2012) to 1.8 and 3.01–1.69 Ma, respectively.

The Western gorilla, or G. gorilla, is divided into two subspecies (Groves, 2001). The bigger of the two in terms of geographical range and population size is G. gorilla gorilla, whose range extends from Cameroon (south of the Sanaga River) southwards through Equatorial Guinea, the Gabon and Congo down to the mouth of the Congo River. Eastward, the range extends into the Congo and the Central African Republic close to the Oubangui River. In the southern part of this range gorillas are not known beyond the Congo River in the Democratic Republic of the Congo. The second subspecies of the Western gorilla, G. gorilla diehli, or the Cross River gorilla, is a small isolated population that straddles the Nigeria-Cameroon border in the upper Cross River highlands.

The Eastern gorilla, or G. beringei, is also divided into two subspecies (Groves, 2001). The larger of the two in terms of geographical range and population size is G. beringei graueri, the Eastern Lowland or Grauer gorilla, whose range extends east from the Lualaba River into the Mitumba Mountains and as far south as Lake Tanganyika. The second eastern gorilla subspecies, G. beringei beringei, or the Mountain gorilla, is represented in two small, isolated populations. One, in the Virunga Volcanoes, straddles the border between Rwanda, Uganda, and the Democratic Republic of the Congo; the other is located in the Bwindi Impenetrable Forest in southwestern Uganda.

Pilbrow (2010) provides a useful detailed review of the literature covering the molecular and morphological evidence for Gorilla. Her odontometric data are consistent with recognizing the distinctiveness of G. g. diehli (see above) and they also support the distinctiveness of the regional subpopulations of G. b. graueri in Tshiaberimu and Kahuzi-Biega. Long ago these two populations had been flagged as a separate subspecies, G. beringei rex-pygmaeorum (Schwarz, 1927).


The present consensus (e.g., Groves, 2001; Locke et al., 2011) is that the genus Pongo includes two species: Pongo pygmaeus, the Bornean orangutan, and Pongo abelii, the Sumatran orangutan (Fig. 1).

The ancestors of orangutans most likely migrated from the mainland to Sumatra and from there to Borneo. Steiper (2006) is in favor of a deep (c.2 Ma) divergence and suggests that the “Bornean and Sumatran orangutans are independent lineages” that are “in an early stage of speciation” (ibid, p. 520). However, when Locke et al. (2011) compared the genomes of the two species they estimated the split time to be more recent, c. 1 Ma (ibid, Fig. 1, p. 530); this is also consistent with the results of Prado-Martinez et al. (2013).

Warren et al. (2001), on the basis of variation in the control region of the mtDNA, identified four distinct subpopulations with particular regional diversity and geographic clustering: (1) Southwest and Central Kalimantan; (2) Northwest Kalimantan and Sarawak; (3) Sabah; and (4) East Kalimantan. Brandon-Jones et al. (2004) recognized two subspecies of the Bornean orangutan: P. p. pygmaeus and P. p. wurmbii. An additional subspecies, P. p. morio, was recognized by Singleton et al. (2004).

The names and distributions of the three proposed subspecies of P. pygmaeus are as follows: P. p. pygmaeus: Northwest Bornean Orangutan, known from Sarawak (Malaysia) and Northwest Kalimantan (Indonesia); P. p. wurmbii: Central Bornean orangutan, known from Southwest and Central Kalimantan (Indonesia), and P. p. morio: Northeast Bornean orangutan, known from East Kalimantan (Indonesia) and Sabah (Malaysia).

The review presented above uses the contemporary names for the various species and subspecies, but many museum collections antedate contemporary usage. The junior synonyms of the current taxa may have been used in some of the older collections, so we list these in Table 1. In Table 2 we list the names that have previously been used for the countries we refer to above. Maps delineating the boundaries between taxa can be found on the website of the IUCN Red List of Threatened Species (

Table 1. Junior synonyms for the genus, species, and subspecies of Gorilla, Pan, and Pongo adapted (with the author's permission) from Groves (2001, 2005)
Genus Gorilla (I. Geoffroy, 1853)
  1. Current designations are presented with taxon name first; junior synonyms are preceded by the date of publication. Depending on the date when specimens were accessioned in collections, these taxon names may have been used in place of the currently valid taxa. Note the prevailing use of Simia, Troglodytes, and Anthropopithecus as the genus name for chimpanzees, and Pithecus as the genus name of orangutans. Readers wanting further information about these taxa and their geographical distribution should consult Groves (2001).

Gorilla gorilla gorilla (Savage, 1847) 
1847Troglodytes gorilla Savage.
1848Troglodytes savagei Owen.
1855Gorilla gina I. Geoffroy.
1856Satyrus adrotes Mayer. Replacement for Troglodytes gorilla.
1856Sat[yrus] africanus Mayer. Replacement for Troglodytes gorilla.
1862Gorilla castaneiceps Slack.
1877Gorilla mayêma Alix and Bouvier.
1903Gorilla gigas Haeckel.
1905Gorilla gorilla matschiei Rothschild.
1905Gorilla jacobi Matschie.
1912Gorilla gorilla schwarzi Fritze.
1914Gorilla hansmeyeri Matschie.
1914Gorilla zenkeri Matschie.
1927G[orilla] uellensis Schouteden.
1927Gorilla gorilla halli Rothschild.
1943Gorilla (Pseudogorilla) ellioti Frechkop.
Gorilla gorilla diehli (Matschie, 1904) 
1904Gorilla diehli Matschie.
Gorilla beringei beringei (Matschie, 1903) 
1903Gorilla beringeri Matschie.
1905Gorilla beringei Matschie.
1917Gorilla beringei mikenensis Lönnberg.
Gorilla beringei graueri (Matschie, 1914) 
1908Gorilla manyema Rothschild. Lapsus for mayêma Alix and Bouvier, 1877.
1914Gorilla graueri Matschie.
1927Gorilla gorilla rex-pygmaeorum Schwarz.
Genus Pan (Oken, 1816) 
1812Troglodytes E. Geoffroy. Troglodytes niger E. Geoffroy, 1812.
1816Pan Oken. Pan africanus Oken, 1816.
1828Theranthropus Brookes. Troglodytes niger E. Geoffroy.
1838Anthropopithecus de Blainville. Simia troglodytes Blumenbach, 1799.
1841Hylanthropus Gloger. Simia troglodytes Blumenbach, 1799.
1860Pseudanthropus Reichenbach. Replacement for Troglodytes.
1866Pongo Haeckel. Replacement for Troglodytes. Not of Lacépède, 1799 (Ponginae).
1866Engeco Haeckel. Simia troglodytes Blumenbach, 1799.
1895Anthropithecus Haeckel. Emendation of Anthropopithecus.
1905Fsihego de Pauw. Fsihego ituriensis de Pauw, 1905.
1954Bonobo Tratz and Heck. Pan satyrus paniscus Schwarz, 1929.
Pan troglodytes (Blumenbach, 1799) 
Pan troglodytes troglodytes (Blumenbach, 1799) 
1758Simia satyrus Linnaeus (in part). Name suppressed by the International Commission on Zoological Nomenclature (1929), Opinion 114.
1792Simia satyrus pongo Kerr.
1792Simia satyrus jocko Kerr.
1799Simia troglodytes Blumenbach.
1812Troglodytes niger E. Geoffroy.
1816Pan africanus Oken.
1831Troglodytes leucoprymnus Lesson.
1840Anthropopithecus pan Lesson.
1855Troglodytes tschego Duvernoy.
1856Satyrus lagaros Mayer.
1856Satyrus chimpanse Mayer.
1860Troglodytes calvus du Chaillu.
1860Troglodytes kooloo-kamba du Chaillu.
1862Troglodytes vellerosus Gray.
1866Troglodytes aubryi Gratiolet and Alix.
1870Pseudanthropus fuliginosus Schaufuss.
1876Anthropopithecus angustimanus Brehm. Nomen nudum.
1895Anthropopithecus fuscus Meyer.
1899Troglodytes livingstonii Selenka. Nomen nudum.
1903Anthropithecus mafuca Haeckel. Nomen nudum.
1905Simia pygmaeus raripilosus Rothschild.
1914Anthropopithecus reuteri Matschie.
1914Anthropopithecus ochroleucus Matschie.
1919Anthropopithecus schneideri Matschie.
1919Anthropopithecus pusillus Matschie.
1932Anthropopithecus heckii Koch.
Pan troglodytes verus (Schwarz, 1934) 
1904Simia chimpanse Matschie. Not of Mayer, 1856.
1934Pan satyrus verus Schwarz.
Pan troglodytes ellioti (Gray, 1862) 
1914Anthropopithecus ellioti Matschie.
1914Anthropopithecus oertzeni Matschie.
1919Anthropopithecus papio Matschie.
Pan troglodytes schweinfurthii (Giglioli, 1872) 
1872Troglodytes schweinfurthii Giglioli.
1905Fsihego ituriensis de Pauw.
1912Simia (Anthropopithecus) nahani Matschie.
1912Simia (Anthropopithecus) ituricus Matschie.
1912Simia (Anthropopithecus) kooloo-kamba yambuyae Matschie.
1914Anthropopithecus schubotzi Matschie.
1914Anthropopithecus steindachneri Lorenz.
Pan troglodytes marungensis (Noack, 1887) 
1887Troglodytes niger var. marungensis Noack.
1899Troglodytes livingstonii Selenka. Nomen nudum.
1912Simia (Anthropopithecus) cottoni Matschie.
1912Simia (Anthropopithecus) adolfi-friederici Matschie.
1914Anthropopithecus purschei Matschie.
1914Anthropopithecus pfeifferi Matschie.
1914Anthropopithecus graueri Matschie.
1914Anthropopithecus calvescens Matschie.
1914Anthropopithecus castanomale Matschie.
Pan paniscus (Schwarz, 1929) 
1929Pan satyrus paniscus Schwarz.
Genus Pongo (Lacépède, 1799) 
Pongo pygmaeus pygmaeus (Linnaeus, 1760) 
1760Simia pygmaeus Linnaeus.
1896Pithecus satyrus landakkensis Selenka.
1896P[ithecus] satyrus batangtuensis Selenka.
1896P[ithecus] satyrus dadappensis Selenka.
1896P[ithecus] satyrus genepaiensis Selenka.
1896P[ithecus] satyrus skalauensis Selenka.
1896P[ithecus] satyrus rantaiensis Selenka.
1896P[ithecus] tuakensis Selenka.
 Pongo pygmaeus morio (Owen, 1837) 
1837Simia morio Owen. Borneo.
1853Pithecus brookei Blyth. Sarawak.
1853Pithecus owenii Blyth. Sarawak.
1855Pithecus curtus Blyth. Sarawak.
Pongo abelii (Lesson, 1827) 
1827Pongo abelii Lesson.
1841Simia gigantica Pearson.
1896P[ithecus] satyrus deliensis Selenka.
1896P[ithecus] satyrus langkatensis Selenka.
1896P[ithecus] satyrus abongensis Selenka.
Table 2. History of names of nations in which great apes are found
Contemporary country namesNames used during the accumulation of Western primate collections
CameroonGerman colony of Kamerun: 1884–1919
French Cameroun and British Cameroons: 1919–1961
Republic of Cameroon: 1961 to present
Central African RepublicOubangi-Chari: 1903–1908
Part of Oubangi-Chari-Tchad in French Equatorial Africa: 1908–1920
Oubangi-Chari (part of French Equatorial Africa): 1920–1958
Central African Republic: 1968–1958
Central African Empire: 1976–1979
Central African Republic: 1979 to present
CongoFrench Congo: 1882–1903
Middle Congo: 1903–1910
Congo (part of French Equatorial Africa): 1910–1960
Republic of the Congo: 1960 to present
Côte d'Ivoire (Ivory Coast)Côte d'Ivoire: 15th century to 1986 (independence from France in 1960)
République de Côte d'Ivoire: 1986 to present
Democratic Republic of the CongoCongo Free State: 1885–1908
Belgian Congo: 1908–1960
Republic of the Congo: 1960–1964
Democratic Republic of the Congo: 1964–1971
Zaire: 1971–1997
Democratic Republic of the Congo: 1997 to present
Equatorial GuineaSpanish Guinea: 1926–1968
Republic of Equatorial Guinea: 1968 to present
GabonGabon (French colony): 1885–1910
Gabon (part of French Equatorial Africa): 1910–1960
Gabonese Republic: 1960 to present
GuineaFrench Guinea: 1895–1958
Republic of Guinea: 1958 to present
LiberiaReferred to as the Grain or Pepper Coast prior to colonization in 1822
Liberia: 1822–1847
Republic of Liberia: 1847 to present
NigeriaUnification of Southern and Northern Nigeria: 1914
Colony and Protectorate of Nigeria: 1914–1960
Republic of Nigeria: 1960 to present
RwandaRwanda (part of German East Africa): 1885–1919
part of Ruanda-Urundi (Belgian colony): 1919–1962
Republic of Rwanda: 1962 to present
SenegalPart of French West Africa: 1895–1958
Part of Mali Federation: 1959–1960
Part of Senegambia: 1982–1989
Republic of Senegal: 1960 to present
TanzaniaPart of German East Africa: 1885–1919
Tanganyika (British East Africa): 1920–1961
Republic of Tanganyika: 1961–1964
Zanzibar and Pemba merged with Tanganyika to form the United Republic of Tanzania: 1964
United Republic of Tanzania: 1964 to present
UgandaBritish East Africa or East Africa Protectorate: 1890–1962
Republic of Uganda: 1962 to present
Indonesia (includes Sumatra and part of Borneo)Republic of Indonesia: 1949 to present
Malaysia (includes part of Borneo)Straits Settlements: 1826–1946
British North Borneo: 1882–1963
Malay States: 1895–1946
Federation of Malaya: 1948–1963
Sabah and Sarawak merged with Malaya: 1963
Malaysia: 1963 to present
Brunei (part of Borneo)Sultanate of Brunei: 15th century to 1959
Nation of Brunei, the Abode of Peace: 1959 to present

Unfortunately, many collections do not provide either current or obsolete taxonomic information that corresponds to current subspecific or even specific designations, but rather lump all gorillas as G. gorilla, all common chimpanzees (as opposed to bonobos) as P. troglodytes, and all orangutans as P. pygmaeus. As a consequence, the information derived from field notes associated with each specimen are of critical importance. In the next section, we consider the acquisition and disposition of these collections in general, and the information about them that is available online.


The primary destinations of “natural history” material collected throughout the 19th and 20th centuries were Europe and the United States, thus it is not surprising that this is where most of the large collections of great ape skeletal material are found. Some of these collections were acquired through field trips designed for that express purpose by scientists with the backing of academic institutions and/or museums. A case in point is the Asiatic Primate Expedition of the late 1930s, which included the physical anthropologists Adolph Schultz, Clarence Ray Carpenter, and a young Sherwood Washburn, and which was backed by the Carnegie Institution, Columbia University, and Harvard University ([Anonymous], 1937). The specimens in this collection (which, in addition to orangutans, also include hylobatids, monkeys, lorisiforms, a tarsier, and many other mammals and birds; Coolidge et al., 1940) are associated with detailed field notes regarding collection locality, date collected, and in some cases, accurate drawings of skeletal elements and measurements made by Adolph Schultz. However, many other collections were acquired through purchases made from individual collectors or through companies that specialized in the purchase and distribution of skeletal specimens. The field data associated with such collections vary in quality: some individual collectors took careful notes (including geographic location, collection date, body mass, linear measurements of the body, etc.) that were sent along with specimens, while other specimens were purchased in markets in source countries from local hunters who provided no other information at all. Unfortunately, it is not that uncommon for collections to list the “country” of origin for some gorilla and chimpanzee specimens as “Africa.”

The field data that were recorded have been preserved in a variety of ways, some having to do with the collectors and methods of field preparation, others having to do with past and present curatorial practices of the institutions in which they reside. Field data may be recorded on tags attached to skins or skeletal elements, written on skulls, recorded on box labels or in museum log books, or preserved as separate field notes in an archive. This information may be variably recorded even within the same collection, and it is probably the case that many researchers have visited a museum, collected their data, and left, without ever knowing of the existence of some important field data relevant to their research.

Field data are also occasionally associated with the wrong specimen or cannot be attributed to a particular specimen. When field-prepared materials arrived at museums, in some cases they were incorrectly matched with their field notes, and in others cases elements from the same specimen were incorrectly labeled; one reviewer noted that he or she has had to match skeletal specimens and skins by their bullet holes. It is also not unheard of for sex to be recorded incorrectly. This is obvious in the case of adult gorillas and orangutans, but is much less so for chimpanzees or for juveniles of any species. In some cases curators recognized attribution errors at the time of preparation, resulting in cataloging practices that at least alert researchers to the problem (e.g., at the Royal Museum of Central Africa in Tervuren, Belgium, the long bones from several G. b. graueri individuals all received the same catalog number and are stored in the same box). Situations like this are frustrating, particularly when they involve rare taxa among the already rare great ape collections (as in the case above), but individual attribution of elements may be resolvable in time through genetic sampling if adequate financial resources can be found to devote to the problem.

Access to resources surely plays a role in determining whether specimen-level data from the field have been migrated to digital formats and made available online. Some collections are in publicly-funded national museums (e.g., the Smithsonian, the Natural History Museum in London) or major national or state-level research institutes (e.g., the Zoologische Staatssammlung München). Others are in municipal museums that are public–private partnerships (e.g., the American Museum of Natural History, the Cleveland Museum of Natural History), and some are in museums that are still in private hands (e.g., the Powell-Cotton Museum). A consequence of these different circumstances is that there is wide variation in the amount and type of information available online for each of these collections (Table 3).

Table 3. Summary of the information available online for major and commonly used great ape skeletal collections in the US and Europe
CollectionWeb address for collection informationCounts of specimens by taxon?Geographical information?Counts of specimens by skeletal element?Information on maturity of individual specimens?Information on the condition of specimens/ elements?Other specimen information (e.g., wild vs. captive, reproductive condition)?Information regarding previously collected data?Pubs citing the collection?Notes
  1. Collection abbreviations are as follows: AMNH: American Museum of Natural History; BMNH: British Museum, Natural History; CMNH: Cleveland Museum of Natural History; FMNH: Field Museum of Natural History, Chicago; MCZ: Museum of Comparative Zoology, Harvard University; MNHN: Muséum National d'Histoire Naturelle, Paris; MNK: Museum für Naturkunde, Berlin; MRAC: Musée Royal de l'Afrique Centrale; NHMV: Naturhistoriches Museum, Vienna; NMNH: National Museum of Natural History, Smithsonian Institution; NSF: Naturmuseum Senckenberg, Frankfurt; PCM: Powell-Cotton Museum, Birchington; RBINS: Royal Belgian Institute of Natural Sciences, Brussels; RMNH: Naturalis (formerly the Rijksmuseum van Natuurlijke History), Leiden; UZH: University of Zurich; YPM: Yale Peabody Museum; ZSM: Zoologische Staatssammlung München.

AMNH (some to species, some to subspecies)Variable: from no information to specific locality within a provinceSkin/Skull/ SkeletonNoOnly “disarticulated” or “mounted”Occasionally data on captive/wildOccasionally in log book scansNoAttached to each specimen page there is a “multimedia” category. Some photographs of the specimens are available and a scanned image of their entry in the catalogue of specimens.
BMNH (species only)RarelySkin/Skull/ Skeleton in some casesAge categories (undefined)NoNoNoneNoAttached to many specimens there is a low resolution image of the collections register, but any associated information is not legible.
CMNH (species only)Most identify nearest townPostcranium/Skull (present or absent)Age categories (undefined)NoOccasionally data on captive/wild, no reproductive status dataNoneNoAvailable as spreadsheet, not online database
FMNH (some to species, some to subspecies)Variable: from no information to specific locality with latitude and longitudeSkin/Skull/ Skeleton and a few specific elementsNoRarely in “collection notes” for elementsOften data on captive/wild, no reproductive status dataNoneNoCollection notes field with useful but non-standardized information
MCZ (some to species, some to subspecies)Variable: from no information to specific locality within a provinceSkin/Skull/ Skeleton and a few specific elementsAge categories (undefined) but room for remarksRarely in log book remarks (attached as media file to record), general condition sometimes noted in databaseUsually data on captive/wildMeasurements occasionally providedNoAttached to many specimens there is an image of the collections register as well as any associated data such as standard measurements, information about specific elements, etc.
MNHN mammal collection is currently being entered into a database
MNK museum web site includes a link to the GBIF portal (see text), but appears to only include animal audio recording and fossil data.
MRAC (subspecies)CountrySkin/Skull/ SkeletonAge categories (undefined), notes latest erupted tooth for each specimenNoWild status can usually be inferred from locality dataNoneNoAvailable as spreadsheet, not online database
NHMV (species only)CountryNoNoNoNoNoneNoDatabase provides tallies by prepartion (Skin, Skull, Skeleton, Dermoplast, Fluid) and sex, not individual specimen data
NMNH (subspecies)Country, Relation to town, RegionSkin/Skull/ Skeleton and most elements in many cases“Stage” and “Remarks” category for different ages usually blankCondition category for specimens usually blankUsually data on captive/wild, “Reproductive Condition” field present in database but typically emptyMeasurements often providedRarelySome specimens have attachments such as cranial X-rays, pathology reports
NSF (most to species, some to subspecies)CountrySkin/Skull/ SkeletonAge categories occasionally provided (undefined)NoWild status can be inferred from locality dataMeasurements occasionally providedNo 
PCM*No*No*No*No*No*None*No**No institutional database, but see text
RBINS (subspecies)CountryNoAge (undefined)NoWild status can be inferred from locality dataNoneNo 
RMNH specific information about collection available online
UZH (subspecies)NoTallies for Skeleton/Crania/Postcrania/CadaversTallies for adults (undefined)NoNoNoneNoNot a database, tallies of totals not individual specimen data
YPM (some to species, some to subspecies)Country or no informationSkin/Skull/ SkeletonRarely age descriptor in “Other attributes” fieldNoOccasional data on captive/wild in “Collected” fieldNoneNo 
ZSM information about type specimens is available online; no information about the extensive orangutan collection is available

Historically, obtaining detailed information regarding great ape skeletal collections without visiting them has been difficult. Researchers and students usually learned about the existence, scope, and content of the various collections from acknowledgment sections in journal articles, from their academic supervisor, or from other researchers in the field. In our experience such information is often vague. A few journal articles from the late 1960s to the early 1980s provided summaries of counts of specimens by species at some museums (Tappen, 1969; Almquist, 1973; Albrecht, 1982), but they do not go beyond basic “skin, skull, skeleton” distinctions, nor do they provide geographical information. The most detailed published information about primate collections began as a catalog of primates in the Natural History Museum in London that was later expanded to include the primates in all British collections. The first volume focused on platyrrhines (Napier, 1976) with the volume on hominoids coming out fourteen years later (Jenkins, 1990). Individual specimens located in British museums are listed in the latter volume by taxon, plus, where available, information about locality and age (adult, juvenile, or infant), but there is no detailed information about the elements present nor about the condition of the specimens. In any event, the hominoid volume is limited to collections housed in the UK, and because it is out of print it is difficult to get access to a copy.

With the advent of the internet, which became available to the general public in the 1990s, the better-resourced museums began building websites, and now even small local museums have some web presence. However, most of these websites tend to be geared primarily to the public visitor rather than the research community. Even in those cases where some collection information is available online, the types of data and the manner in which they are presented (e.g., specimen-level data versus species-level aggregates) vary from institution to institution, and information that is key for researchers intending to use a skeletal collection (see below) is often lacking.

We suggest that there are several categories of information that it would be useful to have before making plans to visit one of the institutions housing a collection of great apes. They include the following:

  1. Sex-specific counts of specimens identified to the lowest taxonomic level possible. In the past, researchers were often content to collect data at the genus level, but as noted above, over the past few decades taxonomic designations of Gorilla, Pan, and Pongo have been revised in favor of finer-scale divisions, and these divisions are thought to represent distinct lineages with differences that may be relevant to evolutionary questions (e.g., Tocheri et al., 2011). Although museums usually identify specimens to at least the species level, in many cases these designations were made when all gorillas and all orangutans were considered to belong to a single species in their respective genera.
  2. Geographical information for specimens identified to the smallest geographical region possible, and preferably by latitude and longitude. Biogeographic provenance is critical information that is often neglected both in terms of data made available in collections and data used by researchers. Where taxonomic designations are out of date, finer-scale geographic information can help identify the subspecies present in the area where a specimen was collected. Although in some cases the nation of origin may be helpful, in other cases the same modern political entity is home to multiple subspecies. Latitude and longitude information from original field notes is not common in many cases, but it is invaluable for maintaining accurate taxonomic designations for those specimens that have it. In addition, detailed geographic information allows for evaluation of the contribution of biogeographic variation to overall morphological variation within a taxon, an important consideration when making comparisons with fossil samples that may be drawn from geographically (and temporally) distant portions of a taxon's range.
  3. Counts of specimens by skeletal element. Many museums provide information on the number of skulls, the number of sets of postcrania, and the number of skins. But it would be more useful to have information about the presence or absence of each skeletal element, for in many cases only portions of the postcranium are preserved, and some, or all, of the teeth may be missing from a cranium or mandible.
  4. Information on the maturity of individual specimens. Museums occasionally provide information on the number of adult and non-adult specimens, although the criteria by which maturity is assessed are usually not indicated. Ideally, each specimen would be associated with a set of developmental variables indicating relative maturity (e.g., epiphyseal fusion of various elements, dental eruption).
  5. Information on the physical condition of specimens/elements. Although an element may be present in the collection, it may be rendered unusable for a given study because of in vivo pathology, post mortem damage, or post mortem destructive sampling. Furthermore, researchers interested in conducting destructive sampling could identify elements that have undergone previous destructive sampling, researchers interested in documenting the frequency or nature of in vivo pathologies would have information on collections and individuals to focus on, etc.
  6. Additional information related to specimen condition at time of collection. There is a variety of other information related to specimens that is useful to know when building a study sample, including information on whether individuals were pregnant or lactating at the time of capture, or whether they were captive or wild-caught. See Borries et al. (2013) for an illustration of how whether or not one takes into account such metadata can dramatically alter the patterns identified in primate life history analyses.
  7. Information regarding previously collected data. Preferably this would be in the form of the data themselves and a record of who collected the data. Given that many standard measurements have been collected multiple times for some collections, such information would make it possible to assess interobserver reliability for those measurements. It would also facilitate a comparison of “standard” measurements that may be measured in slightly different ways by different researchers. In the absence of the data themselves, it would still be useful to have information about the type of data collected (e.g., linear distances, radiographs, 3D surface scans, 3D landmark data, CT scans) and contact information for the researcher(s) who collected those data.
  8. Publications citing the collection. Ideally, this information would also include the types of data published in the articles that make reference to the collection.

We do not claim that the above is an exhaustive list of all of the categories and types of collection information that could potentially be provided online either publicly, securely, or some combination thereof, but we believe this list covers most of the information a researcher would need when preparing a grant proposal or when making plans for a research trip. It would allow researchers to precisely and accurately budget the time required for a visit, and it would enable them to make a better informed assessment of whether the data they require could be obtained from previous researchers, thus minimizing wear and tear on irreplaceable specimens (see below).

As mentioned above, institutions differ in the amount and type of information they make available for researchers. Most of the larger institutions provide some information on several of the points listed above; online information for smaller (and some larger) institutions is minimal or non-existent. In the next sections, we describe a collection at a small private institution with limited financial resources, and then present an online database built by ADG and BW to provide much of the information we refer to above for that collection. We present this as the starting point of a discussion about standardizing specimen information across great ape skeletal collections to ensure the maximum utility and the maximum lifespan for these irreplaceable resources.


We focused on this collection because it is the one ADG and BW know best and because we suspect it is the great ape collection for which the most comprehensive contextual information is available. The collection of great ape skeletons (and skins) is just one component of a large collection of Asian and African mammals within the Powell-Cotton Museum.

History of the Powell-Cotton collection

The Powell-Cotton Museum, which is located in Quex Park on the outskirts of the village of Birchington in Northeastern Kent, England, was founded in 1896 to provide a home for the collections generated by Major Percy Horace Gordon (PHG) Powell-Cotton (1866–1940) (Cooke and Barton, 1957). Percy Powell-Cotton was a keen naturalist, collector, and hunter who between 1887 and 1939 made 28 collecting trips to Africa and parts of Asia; twenty-two of the trips focused on the African continent (see map of expedition routes and area of coverage on p. 276 by Cooke and Barton, 1957). Percy Powell-Cotton's goal in Africa was to acquire representative specimens of all of the major game animals in that continent, and the resulting mammal collection is substantial. The non-human primate collection is particularly extensive and it has long been recognized to be one of the world's largest skeletal collections of wild-shot Western gorillas (G. g. gorilla) and chimpanzees (mainly P. t. troglodytes, but some P. t. schweinfurthii and P. t. ellioti) ([Anonymous], 1940; Cooke and Barton, 1957; Tappen, 1969; Almquist, 1973).

The ape specimens collected by Powell-Cotton came from countries that were, at the time of his various collecting trips, called the Congo Free State, the French Congo, and the British and French Cameroons (Table 2), but he also purchased specimens from collectors operating in what is now Cameroon. Some specimens were purchased from Kurt Zenkerman, a German plantation owner who supplemented his income by collecting natural history specimens that he sold to European museums. However, the majority of the great apes in the Powell-Cotton Collection (P-CC) were acquired by Fred G. Merfield, a British expatriate living in the Cameroons whom Powell-Cotton met in 1926. These acquisitions were made in the name of the Yaounde Zoological Trading Company, which was formed with Merfield as executive director and Major Powell-Cotton and Mrs. Hannah Brayton-Slater Powell-Cotton as co-directors, with the purpose of supplying the Powell-Cotton Museum and other European museums with specimens of African mammals.

Merfield made a point of establishing friendly relationships with the heads of many villages throughout parts of what is now southern and central Cameroon, and after doing so, he persuaded them to collect mammal specimens for him. Merfield also collected specimens himself and it is estimated that he personally shot c.10–15% of the specimens he sent to the Powell-Cotton Museum (Malcolm Harman, Personal Communication). Specimens (including apes, monkeys, and other mammals) were numbered in five series of up to 1,000 each. Collection of specimens from the regions of Yaounde, Batouri, Yabassi, and Lomie in Cameroon began in 1927 and continued up to the beginning of WWII in 1939. All of the material was sent to the Powell-Cotton Museum where PHG Powell-Cotton also acted as a distribution agent. Specimens were sold to museums, to Sir Francis Collier, and to dealers such as Rowland Ward Ltd. and Edward Gerrard & Sons. The arrangement was that the Powell-Cotton Museum was allowed to keep one specimen for every one sold on to museums or dealers; the museum still has detailed records about all of the specimens that were sold.

The existence and significance of the wider P-CC was well known long before data from the great apes in the collection began to be used in the scientific literature. Major Powell-Cotton's collecting expeditions were reported on regularly in Nature and Science (e.g., [Anonymous], 1899a,b, 1903b–f, 1904a,b, 1905, 1906, 1907a), and the sale and contribution of specimens to the British Museum (now the Natural History Museum) and to collections at other institutions such as the Royal College of Surgeons of England were noted in the scientific press (e.g., [Anonymous] 1903, 1907b–d, 1909, 1928, 1934, 1936a, b, 1938). Major Powell-Cotton was well known in natural history circles, and many species and subspecies of non-hominoid mammals were given the eponymous name “cottoni.” For example, Lydekker and Matschie note “it is a great pleasure to name this interesting new species after Major P.H.G. Powell-Cotton”, and in many places in Lydekker's (1913–1916) Catalogue of Ungulate Mammals in the British Museum it states “type in Major Powell-Cotton's collection in Quex Park, Birchington, Kent” (Colin Groves, Personal Communication). Mammalian comparative morphologists were certainly aware of the importance of the broader P-CC by the early 1950s, and it was at this time that it apparently came to the attention of primatologists and students of human evolution.

African ape component

The first report of research that used data collected from the great apes in the P-CC was by Ashton and Zuckerman (1951). It is a reasonable inference that Solly Zuckerman was among the first, if not the first, to recognize and realize the potential for comparative research of the African apes in the collection. During WWII the part of Kent in which Quex Park is situated was effectively one large military base with most people in the area involved in the war effort in one way or another, so by the end of the war the majority of the great ape skeletons in the P-CC had not yet been fully macerated. Some anatomical regions (e.g., hands, feet) were still articulated with remnants of flesh and ligaments adhering to them. Solly Zuckerman arranged for the postcranial material in the P-CC to be carefully macerated and curated at the Anatomy Department in Birmingham, where in 1939 he had been appointed to the Sands Cox Chair of Anatomy. The arduous and time-consuming work of maceration and curation was undertaken by Ernie Sims, Bill Pardoe, and Tom Spence, under the supervision of Eric Ashton (Charles Oxnard, Personal Communication). Apparently, a quid pro quo of the arrangement between Zuckerman and the Powell-Cotton Museum was that the material was to be kept in Birmingham for a period so that it could be used by Zuckerman and his colleagues for their research. However, it was also part of the agreement that while in Birmingham the material would be made accessible to anyone else who wanted to work on it. Zuckerman and his colleagues honored this agreement. Charles Oxnard recalls “Colin Groves coming with huge calipers to measure long bone lengths,” and in 1966 BW was warmly welcomed in Birmingham when he was working on talus morphology (Day and Wood, 1968, 1969; Wood, 1973). Zuckerman left Birmingham in 1968 and not long thereafter the ape postcrania were returned to the Powell-Cotton Museum in surplus ammunition boxes. Zuckerman almost certainly had access to these because from 1960 he was Chief Scientific Adviser to the UK Ministry of Defence, and in 1964 he was appointed the first UK Government Chief Scientific Adviser. Publications on comparative great ape morphology in the context of human evolution that used the P-CC began appearing in the late 1960s, and one or more such articles have appeared nearly every year up to the present (Fig. 2; for a bibliography of works referencing the collection go to

Figure 2.

Time versus cumulative number of research publications in peer-reviewed journals in which the Powell-Cotton Collection is explicitly cited as a source of primate data. The master table from which these data were extracted, which is based on a literature search, is available at the Human Origins Database,

Nature and taxonomy

The P-CC is one of the world's largest collections of Western gorillas (G. gorilla) and Central African chimpanzees (P. t. troglodytes). Western gorillas are represented by 148 complete or nearly complete skeletons (most with skins) as well as six sets of postcrania without skulls, 49 skulls without postcrania, 16 crania without mandibles, and three mandibles without crania. Central African chimpanzees are represented by 146 complete or nearly complete skeletons (most with skins) in addition to 27 skulls without postcrania and 11 partial or complete crania without mandibles. There are also six complete skeletons and one unassociated skull that appear to have been collected from the range of P. t. ellioti, and four complete skeletons and two unassociated skulls of chimpanzees from the Ituri Forest (P. t. schweinfurthii). The P-CC provides an excellent ontogenetic sequence for both genera. The gorillas range in maturity from specimens with only a few deciduous teeth erupted to adults with a completely erupted dentition and complete epiphyseal and cranial suture fusion. Chimpanzees expand this range by including a specimen with no erupted teeth.

Additional features of the African apes in the P-CC make it a special resource. First, a large number of individuals have most of the postcranial skeleton preserved, including most with the hand and foot bones—the latter thanks to the careful preparation undertaken in Birmingham. The hyoid is also preserved for many individuals. Second, the greater part of the collection is backed up by field notes that include the date of collection, the name of the locality, and its latitude and longitude in degrees and minutes. Third, most of the specimens are from a restricted geographical area, providing a large sample in one collection that records local geographic variation. Fourth, many have linear measurements of the cadaver, and some have information about body mass.


For nearly all of the mammal specimens in the P-CC, basic locality information, specimen sex, and general preparation information (i.e., skin, skull, and/or skeleton) are available upon request from the museum in spreadsheet format. Presently, Ph.D. student Jaimie Morris of Canterbury Christ Church University is converting these locality data to a digital format and entering them into ArcGIS, with the eventual goal of providing these data to researchers. However, until recently, one would have to visit the collection to get any more specific information regarding the number of individuals preserving specific elements, the number of adults vs. juveniles, or whether specimens were damaged. In an attempt to make these important data available, for each African ape specimen ADG conducted a systematic inventory of all skeletal elements (down to individual teeth and bones) in which the presence or absence was recorded along with notes regarding any in vivo pathologies, presence of connective tissue on bones that may obscure some measurements, and post mortem damage. Developmental stage was also recorded for some elements (as detailed further below). These data were then incorporated into an online relational database and made available as specimen datasheets through the Human Origins Database (

These data are searchable online in a variety of ways, allowing users to identify whether, given a set of selection criteria, a suitable sample is available within the P-CC. In addition, users can identify the individual specimens that meet those criteria. In all cases, the results of a database search can be exported from the Human Origins Database as a text file and then opened in a spreadsheet program such as Excel.

Using the various searches, a researcher anywhere in the world with a computer with internet access can identify specific samples from the P-CC that would be appropriate for a particular study. For example, if one were interested in studying a sex-balanced sample of right tali from adult P. t. troglodytes, the Human Origins Database could be used to generate a list of all specimens at the P-CC that could be included in such a study.

Developmental sequence and ontogenetic series

As an example of what can be achieved with the types of data we have recorded on the P-CC, we present below an analysis of the sequence of developmental markers in P. troglodytes and G. gorilla based on information recorded in the Human Origins Database. We also provide the resulting lists for both species of all specimens in the collection sorted by relative maturity along with each specimen's developmental data (Appendices A and B). The annotated inventory of African ape skeletal material in the P-CC includes data about (A) the degree of fusion of sixteen long bone epiphyses, (B) the eruption status of all of the teeth, and (C) the general state of fusion of all of the main cranial sutures (Tables 4 and 5 and see below). These annotated inventories plus the growth data have been incorporated into the online database, both as searchable values in the relational database and as a series of data sheets provided for each specimen in the collection.

Table 4. Counts of Pan troglodytes specimens that have initiated and completed each developmental marker (dental eruption or epiphyseal fusion)
Pan troglodytesAbbr.MFTotalMFTotal
  1. Counts are for specimens which have complete data for all developmental markers. Dental markers are italicized, postcranial markers are not. Developmental markers are given in order of decreasing number of specimens that have initiated that marker; i.e., developmental markers that initiate (begin eruption or begin fusion) earliest are presented first, those that initiate latest are presented last. Some markers appear to develop out of sequence; i.e., they initiate later than other markers but also complete earlier than those markers. The markers that are completed for a higher number of specimens than markers that are initiated for a higher number of specimens (roughly translating to markers that complete earlier than markers which initiated earlier) are shown in bold. For example, more chimpanzee specimens have at least initiated eruption of the second mandibular incisor (i2) than have initiated eruption of the first maxillary incisor (I1) (i.e., i2 begins to erupt earlier on average than I1). However, eruption is complete for I1 in more specimens than is the case for i2, indicating that despite the earlier initiation of i2 eruption, I1 tends to be fully erupted earlier on average than i2.

deciduous first incisor, maxillarydI156891455689145
deciduous second incisor, mandibulardi256891455689145
deciduous first incisor, mandibulardi156881445688144
deciduous second incisor, maxillarydI256881445688144
deciduous third premolar, mandibulardp356881445688144
deciduous third premolar, maxillarydP356881445688144
deciduous fourth premolar, mandibulardp456881445688144
deciduous fourth premolar, maxillarydP456881445688144
deciduous canine, mandibulardc56881445688144
deciduous canine, maxillarydC56881445688144
first molar, mandibularm152841365184135
first molar, maxillaryM152841365184135
first incisor, mandibulari139771163676112
second molar, mandibularm240751153273105
second incisor, mandibulari237771143273105
first incisor, maxillaryI137751123474108
second molar, maxillaryM236751113070100
second incisor, maxillaryI235751103374107
third premolar, maxillaryP330731033072102
fourth premolar, maxillaryP431711023170101
fourth premolar, mandibularp43171102297099
third premolar, mandibularp33071101297099
canine, maxillaryC286997236891
canine, mandibularc276895236891
humerus, distalHD266995216687
humerus, epicondyleHE246791206585
pelvis (os coxa)OC246791206484
third molar, mandibularm3246690226486
ulna, proximalUP236689206585
third molar, maxillaryM3226587226486
femur, lesser trochanterFeLT216687206484
femur, greater trochanterFeGT216586206282
femur, headFeH206585206181
radius, proximalRP206484196180
fibula, distalFiD206282185775
tibia, distalTD206181185775
tibia, proximalTP206181174966
fibula, proximalFiP205979175370
femur, distalFeD196079185068
humerus, proximalHP195877165167
radius, distalRD195877174865
ulna, distalUD195675184866
Table 5. Counts of Gorilla gorilla gorilla specimens that have initiated and completed each developmental marker (dental eruption or epiphyseal fusion)
Gorilla gorilla gorillaAbbr.MFTotalMFTotal
  1. Counts are for specimens which have complete data for all developmental markers. Formatting and abbreviations follow Table 4.

deciduous first incisor, mandibulardi171741457174145
deciduous first incisor, maxillarydI171741457174145
deciduous second incisor, mandibulardi271741457174145
deciduous second incisor, maxillarydI271741457174145
deciduous third premolar, mandibulardp371741457074144
deciduous third premolar, maxillarydP371741457074144
deciduous fourth premolar, mandibulardp470741446872140
deciduous canine, mandibulardc69741436872140
deciduous canine, maxillarydC70731436872140
deciduous fourth premolar, maxillarydP470731436872140
first molar, mandibularm163691326266128
first molar, maxillaryM161701316166127
first incisor, mandibulari157611185461115
first incisor, maxillaryI154611155356109
second incisor, mandibulari254601145458112
second incisor, maxillaryI254601145353106
second molar, maxillaryM253581115349102
second molar, mandibularm253571105354107
third premolar, maxillaryP35350103504999
fourth premolar, maxillaryP45251103504797
fourth premolar, mandibularp45250102494796
humerus, distalHD494998384583
third premolar, mandibularp3504797494796
canine, mandibularc484694414283
canine, maxillaryC494594394382
third molar, mandibularm3494493453782
pelvis (os coxa)OC474693374279
third molar, maxillaryM3494493433679
ulna, proximalUP444791374481
humerus, epicondyleHE434588374481
femur, lesser trochanterFeLT404686344377
femur, headFeH404585334073
femur, greater trochanterFeGT384583334073
radius, proximalRP354378323870
tibia, distalTD333972293564
fibula, distalFiD333972283462
femur, distalFeD313869273562
tibia, proximalTP313869273158
fibula, proximalFiP303868283361
radius, distalRD293564262955
humerus, proximalHP293564262753
ulna, distalUD293564242751

Epiphyseal fusion

For all of the long bones in the collection, epiphyseal fusion at each epiphysis was scored as follows:

Unfused (U)

Epiphysis is distinct from diaphysis. There is no bony connection, but there may be other connective tissue between the epiphysis and diaphysis.

Partially fused (P)

Epiphysis and diaphysis are connected by some bone, but the growth plate is still present.

Fully fused (F)

Epiphysis and diaphysis are continuous. A line may mark the previous site of the growth plate.

The degree of fusion of the ilium, ischium, and pubis was scored as follows:

Unfused (U)

Ilium distinct from ischium and pubis. Ischium and pubis may, or may not, be fused.

Partially fused (P)

Ilium and ischiopubis are connected by some bone, but fusion is not complete.

Fully fused (F)

The components of the pelvic bone are completely fused with no gaps between bones.

Dental eruption

The deciduous teeth are referred to by their developmental origin rather than by their morphology (i.e., we refer to the deciduous second “molar” as the deciduous fourth premolar). This allows for a scoring system that refers to the eruption stage at a particular tooth position rather than the eruption stage of a specific tooth in the dental succession.

Deciduous unerupted (dU)

Deciduous tooth is in the crypt, or if partially emerged, the crown does not project above the alveolar margin.

Deciduous partially erupted (dP)

Deciduous tooth crown projects above the alveolar margin, but it is not in full occlusion with occluding dentition. Or, if the corresponding tooth in the other jaw is not fully erupted, the crown does not project above the alveolar margin at the level of other fully erupted teeth.

Deciduous erupted (D)

Deciduous tooth crown projects above the alveolar margin.

Unerupted (U) (for teeth without deciduous antecedents)

Tooth is in crypt, or if partially emerged, the crown does not project above the alveolar margin.

Partially erupted (P)

Tooth crown projects above alveolar margin, but it is not in full occlusion with occluding dentition. Or, if the corresponding tooth in the other jaw is not fully erupted, the crown does not project above the alveolar margin at the level of other fully erupted teeth.

Fully erupted (F)

Tooth crown is in full occlusion, or if the corresponding tooth in the other jaw is not fully erupted, the crown projects above the alveolar margin at the level of other fully erupted teeth.

Cranial fusion

Cranial sutures were evaluated as follows:

Unfused (U)

All cranial sutures are open.

Partially fused (P)

At least one cranial suture is partially fused.

Fully fused (F)

Complete fusion of the sutures of the neurocranium and face (i.e., not including sutures of the basicranium).

In order to assess the sequence of developmental events recorded in this data set and to assess the relative maturity of each specimen, left and right scores on all dental and postcranial variables were reduced to a single score for each individual. If the scores differed between antimeres, the more advanced developmental stage was chosen (but both sides are reported independently in the online database). Because fusion of the cranial sutures occurs after dental and postcranial maturity in all of the crania inspected, cranial data were not included in this analysis (although they are included in the Appendices A and B; see below). The number of specimens in the P-CC of known sex with complete information regarding dental eruption and postcranial epiphyseal fusion is the same for chimpanzees and gorillas (P. troglodytes N = 145, G. gorilla N = 145). For each epiphysis, counts were taken of the number of complete individuals in each species that A) had begun the process of fusion at that epiphysis and B) were completely fused at that epiphysis. Likewise, for each tooth (deciduous and permanent), counts were taken of the number of complete individuals in each species that A) had at least begun to erupt that tooth, and B) had fully erupted that tooth. These counts were then sorted in descending order (Tables 4 and 5, Fig. 3). In general, most developmental events were completed in the same order in which they initiated; the exceptions are shown in bold in Tables 4 and 5. The sequence of dental development identified here for P. troglodytes and G. gorilla is generally similar to the patterns reported by earlier workers (e.g., Schultz, 1935; Dean and Wood, 1981; Bolter and Zihlman, 2011). There is little published information available on the combined relative timing of dental eruption and epiphyseal fusion for African apes (but see Zihlman et al., 2007). We report these data for both Western gorillas (G. g. gorilla) and all chimpanzees in the collection, which are primarily Central chimpanzees (P. t. troglodytes) (Tables 4 and 5).

Figure 3.

Developmental markers organized by decreasing number of individual specimens that have initiated that marker (i.e., dental eruption or epiphyseal fusion has begun). Dotted lines track the number of individuals that have initiated each marker and solid lines track completion of each marker. Represented in gray is the difference between the number of individuals that have completed eruption of the maxillary third molar (generally the last tooth to fully erupt in both species in this data set) and the number of individuals that have completely fused the distal ulna (generally the last epiphysis to completely fuse in both species in this data set).

Regarding relative dental and postcranial maturity, c.17% of all chimpanzee specimens (24 of 145) and c.19% of all gorilla specimens (28 of 145) were dentally but not postcranially mature (as approximated by the gray boxes in Fig. 3) at the time of death. The extent of the problem of assuming that dentally mature specimens are “adult” becomes evident when the specimens referred to above are considered as a percentage of dentally mature specimens. In the chimpanzee sample c.28% (24 of 86) of dentally mature specimens are not skeletally mature, and in the gorilla sample c.38% (28 of 74) of dentally mature specimens are not skeletally mature. Only two skeletally mature gorillas had not completed eruption of their third molars; all of the skeletally mature chimpanzees were also dentally mature. Thus, in both taxa approximately one-quarter to one-third of dentally mature individuals are unlikely to have achieved their full adult size. For example, facial growth has been shown to continue past dental maturity at different sex-specific rates in great apes and other primates; thus, depending on the criteria used to define adulthood, this will affect estimates of mean adult facial size and sexual dimorphism (Wang et al., 2007; Balolia et al., 2013). For some studies the difference between dental and skeletal maturity may not be a concern, whereas for others it might represent a significant source of error due to additional variation in the traits under consideration.

It is reasonable to assume that other skeletal collections have similar proportions of gorillas and chimpanzees that are dentally, but not skeletally, mature. Therefore, we extend the caution that a quarter or more of African ape specimens in museum collections that are identified as adult based on the eruption of the third molar may not be fully skeletally mature, and that this may be an important consideration for some investigations. Similar cautions may apply to collections of other higher primate taxa, and we are exploring how this may relate to subtle and not-so-subtle differences in life history within and among great ape species.

In addition to identifying the sequence of developmental events, we also identified the relative maturity of each chimpanzee and gorilla in the P-CC collection. For each complete specimen, we calculated a “maturity score,” which was calculated as the proportion of the permanent dentition that is completely erupted and the proportion of the postcranium that is completely fused. See King (2004) for a similar approach applied to 16 non-human primate species. For each specimen, the number of permanent teeth that are completely erupted (on one side) were added to the number of epiphyses that are completely fused (on one side), as well as 0.5 times the number of teeth that are partially erupted and 0.5 times the number of epiphyses that are partially fused. The resulting value was divided by the total number of developmental markers exclusive of cranial suture fusion (i.e., 32) to produce a score between zero and one.

In Appendices A and B of this article, we list all of the complete chimpanzee and gorillas specimens sorted by maturity score. In the case of specimens with no permanent dentition, specimens are ordered by the number and type of deciduous teeth present. Specimens in which fusion has occurred at all sutures except those of the basicranium (which are rarely fused in this collection) are always completely mature both postcranially and dentally, thus they are placed later in the sequence than otherwise fully mature specimens for which cranial fusion is not complete. We estimated the maturity score for specimens that are damaged or incomplete by matching them as closely as possible with complete specimens. Individuals with fully fused crania that lack postcranial evidence are listed last because all of the complete specimens in our sample with fused cranial sutures were postcranially mature at the time of death.

Thus, in addition to the adult skeletons, the P-CC preserves a remarkable ontogenetic series of Western gorillas and Central chimpanzees. For the gorillas (all of which appear to have been collected within the range of G. g. gorilla), among the specimens that preserve all developmental markers there are 47 fully adult complete or nearly complete skeletons, and 101 complete or nearly complete skeletons in various stages of development. In addition, there are many skulls without associated postcrania, and some postcrania without associated skulls. With respect to the latter, at least one of the “missing” gorilla skulls is in the collection of the Hunterian Museum at the Royal College of Surgeons of England (postcranial specimen P-CC M21 is from the same individual as the skull specimen RCSOM/A 64.21). For the chimpanzees (including three subspecies, but nearly all of which are P. t. troglodytes), among the specimens that preserve all developmental markers there are 62 fully adult complete skeletons and 84 complete skeletons in various stages of development, again in addition to several skulls without associated postcrania.


The temporal and paleoecological contexts of the hominin fossil record are important for its interpretation, but the primary data used to investigate human evolutionary history are morphological. Morphological evidence is the currency of hominin paleobiology; it is the raw material for all analyses, be they taxonomic, phylogenetic, functional, or developmental. These morphological data are analogous to sequence data in molecular biology, or to seismic records and temperature proxy data from piston cores in the earth sciences. Yet, any testable hypotheses about taxonomy, evolutionary relationships, and functional morphology are dependent on having substantial collections of extant higher primates for comparison. Some museums provide helpful information available online about their collections (e.g., the Smithsonian Mammal collections database [] often includes in a notes field a list of major elements present for individual great ape specimens), but for most of the collections it is difficult for researchers to assess how much of a seemingly large collection will be of any use for their particular research purposes.

The ability to deliver specimen- and element-specific information through a relational database that allows users to search through that information in a variety of ways has many implications for our field, and below we draw attention to two of them. One of the realities we face as a discipline is that financial resources for comparative morphological analyses are decreasing and are unlikely to increase in the foreseeable future. In such an economic climate (and indeed, even in a climate where research money is plentiful), it is particularly useful to be able to identify the specimens that meet study criteria before arriving at a museum so that researcher's time and money can be put to maximum use, as opposed to spending a significant portion of that financial and temporal effort to identify the study sample on site. Developing databases such as the one we have assembled for the Powell-Cotton Collection would be a great help to researchers.

A second point relates to the ultimately ephemeral nature of the collections themselves. Higher primate skeletons are vulnerable to damage, and given enough wear and tear over time, they will eventually disintegrate beyond the point of yielding useful morphological information. The Powell-Cotton Collection is a case in point. When BW compared observations he made in 1973 with those made three years ago, several specimens are now missing teeth that were present in 1973, and dimensions that could be measured then cannot be measured now. Although there will always be a need to maintain access to the original specimens for specialist examination, the original specimens could be afforded a substantial degree of protection if most of the information needed by most of the people who presently examine them was made available in the form of an electronic relational database, or at the very least if those original data were published in ways that identified data with individual specimens (e.g., through online supplemental tables linked to journal articles, or through other permanently maintained online data depositories such as Dryad []).

It is widely recognized that fossils are irreplaceable, but this is also the case for museum collections of extant higher primates. Many of these collections include the skeletons of animals collected in locations where wild higher primates have locally been driven to extinction. These important comparative collections are irreplaceable, and there is an urgent need for the raw, non-aggregated data obtained from them to be placed in the public domain so that the collections can be preserved for future generations. With that in mind, we encourage morphologists to publish their raw data with their analyses whenever possible, and we encourage journal editors to accommodate such efforts whenever they can.

So what specifically can be done to address these points, and what has already been done? In April 2007, a workshop designed to address these questions in paleoanthropology in general was organized by Eric Delson, Stephen Frost, Will Harcourt-Smith, and Christopher Norris, and was co-funded by the Wenner-Gren Foundation and the National Science Foundation (a summary of the workshop can be found in the work by Delson et al., 2007). An international group of participants (including ADG and BW) brought together researchers involved in database creation and museum representatives to initiate discussion on issues such as database standards, data standards, and museum involvement in these initiatives. The workshop brought to light many different perspectives on these issues, and here we outline specific proposals for great ape skeletal collections that follow from our thoughts on those discussions.

Regarding the first point, over the past decade an international standard has emerged for organismal biology. The Darwin Core ( is a database standard for biodiversity information that was originally developed in the late 1990s and was last updated and ratified in October 2009. It sets out specific database field names for data related to specimen identification, taxonomic and biogeographic information, and many other types of metadata. The adoption of such a standard by collections/research databases has allowed for the development of web portals, which are online aggregators of information derived from simultaneous queries of multiple online databases using a common set of user-specified search terms. Such searches work because the various databases either use the standard data fields specified by the Darwin Core, or they have fields that can be translated to match the standards. Examples of web portals in disciplines related to paleoanthropology are the Global Biodiversity Information Facility (GBIF, the Mammal Networked Information System (MaNIS,, and the Paleontology Portal ( These websites and others like them typically have additional content, but the portal portion of the site allows the user to build a query that is sent to some or all of a predetermined set of databases for which the portal has a communication protocol in place. For example, the Paleontology Portal can query from among several private and university-affiliated natural history museums as well as MorphoBank.

Web portals provide the ability to query many collections at once and thus have the potential to aggregate, for example, all information on world-wide holdings of P. t. schweinfurthii into the results of a single query. Individual apes collected from the same time and place are often distributed among several museums, and it would be useful to know where those specimens are. For example, the populations represented by the Merfield material in the P-CC are also distributed across Europe and the US through the sale of half of the specimens Merfield collected, and the Asiatic Primate Expedition material is split between the Museum of Comparative Zoology at Harvard and the Department of Anthropology at the University of Zurich. In addition, in some cases elements from single individuals may be split between different collections. As mentioned above, at least one of the skulls that are missing from some African ape skeletons in the P-CC is in the Royal College of Surgeons in London; it also appears that some limb elements missing from bonobo skeletons at the Royal Museum of Central Africa in Tervuren, Belgium probably ended up in Belgian university collections during the partitioning of those samples in the mid-20th century (Wim Wendelin, Personal Communication). Queries to a web portal such as the GBIF that aggregates all of the information from great ape collections would allow researchers to identify where the components of these various populations currently reside, with one large caveat: web portals are reliant on both the existence of databases containing the information desired and standard database fields from which the information can be queried. Leaving aside for the moment the issue of the existence of a database for a specific collection and focusing on database standards, the Darwin Core already has such standards for metadata such as subspecies name and collection locality for individual specimens, but does not specify a set of standard osteometric measurements.

The Darwin Core retains a great deal of flexibility regarding measurements derived from specimens. Without going into too much detail on database structure, rather than defining fields with specific names such as “right maxillary canine height” or “femoral head superoinferior diameter,” the Darwin Core specifies fields such as “measurementID,” “measurementMethod,” “measurementType,” and “measurementValue.” This allows database developers to name and define their own sets of measurements that can either be queried specifically by measurement name or returned as part of a set of all measurements for an individual specimen.

The measurement flexibility within the Darwin Core leads into a discussion of the second issue raised above: standards for the data collected (as opposed to database structure). Although comparative primate morphologists ostensibly share a common set of “standard” measurements (e.g., buccolingual molar width or femoral length), in practice two researchers may use slightly different definitions of the “same” measurement, or may even collect the same measurement using the same definition in two slightly different ways. This is why it is important to spend time in a manuscript describing measurement techniques, and why one of us devoted 27 pages of the monograph on the Koobi Fora cranial remains to describing the 394 measurements used (Wood, 1991).

The recognition of differences in the protocols used for “standard” measurements tends to elicit one of two different responses from researchers who would like to use a combined dataset of standard measurements collected by two or more people. The first is to conclude that the uncertainty regarding measurement differences degrades data quality for combined datasets to the point that they are unreliable and should not be used, while the second is to recognize the problem but suspect that inter-observer bias will have minimal effect (e.g., Bailey et al., 2004) on the results of a particular analysis and hope for the best.

We would like to suggest a middle way, in which comparison of repeated measurements on a common set of specimens by different researchers can be used to determine whether a combined dataset is problematic, or not, in a given situation. This suggestion relies on access to researchers' measurements for individual specimens, as well as specific information regarding how measurements were collected. As a discipline we lack a single source of definitions for common measurements used by a majority of practitioners in the field. We are not proposing the prescription of a limited set of “correct” measurements that should be taken on skeletal material, but rather the identification and description of the multiple ways “standard” measurement techniques are interpreted so that researchers can explicitly identify the technique they used when sharing their measurements with others. We develop this suggestion further below.

The final issue raised above is that of museum involvement, and it is a critically important issue. First, there is a considerable investment of trained-personnel time that must go into a skeletal element-level annotated inventory such as the one ADG and BW undertook with the P-CC, as well as into the development of a database, translation into database fields of information from what are often fading or crumbling hand-written field notes, and other associated activities associated with compiling a database along the lines of what we have described here. Typically this investment must come from the museum or institution curating the collection, and as mentioned above, many institutions are already limited in their ability to meet whatever they consider to be core functions of the institution. A challenge we must face as a discipline is how to meet these goals without impinging on museum resources for those collections that cannot prioritize these efforts. This challenge could be met by external research groups taking on components of the work (as ADG and BW have done in the case of the P-CC), by raising funds to support in-house efforts, or by other mechanisms.

Second, we must recognize that museums housing great ape skeletal material vary in their access policies, data sharing policies, whether or not they charge bench fees, etc. Some museums will be comfortable making their collections information available online to the general public for free, while others may prefer to limit access to bona fide researchers and/or to individuals or institutions that pay for access to the information. There are many factors that go into each museum's existing policies, including (but not limited to) their statutory remit, concerns regarding specimen preservation over the long term, funds required to maintain collections, and maintaining institutional control over collection data. Developing databases such as we describe here will ameliorate some of these concerns and aggravate others, with the consequence that it may not be clear to museum curators and administrators whether making such information available would be a net gain from their perspective. Thus, it is imperative that museum representatives be party to discussions and decisions made regarding establishing data sharing standards for data derived from great ape skeletal collections. This brings us to our closing suggestion.

A congress to identify measurement and collections information standards

Over the past few years there have been several workshops and symposia at annual meetings that have brought together representatives from various research groups to talk about the databases they have developed for their various projects. We applaud these initiatives to get research database developers talking to each other about common issues, but we also think that it is time to try something different and more specific.

We propose that the time has come to bring together comparative morphologists working on great ape skeletal material (and some who work on primates in general), regardless of whether they are involved in database projects, and the curators who oversee those collections and who make policy about great ape collections. The purpose of the meeting would be two-fold: first, to identify and define a set of standard measurements typically collected on skeletal material, along with a set of variants with different definitions for each measurement, and second, to identify specific concerns and limitations that those responsible for the collections have regarding making such information available on a searchable online database. Such a meeting could be broken into working groups focusing on specific types of measurements (e.g., craniofacial, dental, axial, long bones, manual, and pedal elements) and specific collection concerns (e.g., intellectual property issues, funding concerns, impact on research visit rates). While ambitious in scope, we believe that such a meeting would be an important step that would A) greatly improve the utility of information collected over the past several decades by the various practitioners in our field, B) help preserve the collections and ensure their utility in the years to come, and C) provide a template for applying the same principles to collections of fossil material.

In closing, we recognize that because of the vagaries of the data collection protocols used in the field sometimes a century or more ago, for some collections not all of the types of data we would wish to have will be available in any form, digital or otherwise. We also recognize that for a variety of reasons, some of which we have touched on above, for some collections it may be a long time before those data that do exist become available online, and we certainly would not suggest that researchers avoid collections that do not have such resources available online. What we would suggest is that developing measurement standards, migrating data to the web where feasible, and encouraging the open sharing of data will improve the utility and increase the lifespan of these essential and irreplaceable resources.


We would like to thank the Trustees of the Powell-Cotton Museum and Malcolm Harman for granting generous access to the African ape specimens and records under their care, and for their warm reception to the efforts to highlight the exceptional nature of the Powell-Cotton Collection. In particular, Malcolm Harman has been very helpful over the years in acquainting us with the history of the collection and its connection to collections at other institutions. We thank David Pilbeam for arranging library access for Emily Marcus. Madison Evans researched the previous names of countries in which African great apes are found. We would also like to thank Trudy Turner and Chris Vinyard for their editorial guidance and the anonymous reviewers whose comments have greatly improved the content and structure of this manuscript. The Human Origins Database and our work on the Powell-Cotton Collection was made possible through a grant to BW from the G. Harold & Leila Y. Mathers Foundation. ADG was partially supported in the writing of this article by a Wenner-Gren Hunt Postdoctoral Fellowship. Finally, ADG would like to extend a special thank you to his good friends Wendy and Rob Teale, whose hospitality and charm is well known to all who have stayed with them over the years while working at the Powell-Cotton Museum.


Developmental information for Pan troglodytes specimens at the Powell-Cotton Museum. Specimens are listed in increasing developmental age as determined by degree of dental eruption and epiphyseal fusion. Developmental marker abbreviations (column headings) follow Table 4 and are ordered within the dentition and postcranium by their developmental sequence as shown in Table 4; developmental data abbreviations as given in the text. Calculation of maturity score is described in the text; estimated maturity scores for specimens missing data are highlighted in bold. Skulls with fully erupted dentition but without postcrania are listed after all other specimens because long bone epiphyses may remain unfused or partially fused after full dental eruption. However, postcranial fusion precedes cranial suture fusion in all cases for which complete skeletons are available, so completely fused crania presumably belong to individuals that were also postcranially mature. Underline indicates a specimen appears to have been collected from the geographic range of P. troglodytes ellioti, bold indicates a specimen belongs to P. troglodytes schweinfurthii. A color version is available at the Human Origins Database,

Table 6. APPENDIX A.
SpecimenSexMaturity scorem1M1i1m2i2I1M2I2P3P4p4p3Ccm3M3HDHEOCUPFeLTFeGTFeHRPFiDTDTPFiPFeDHPRDUDCranium
M 781F0.000UUDUDDUDDDDDdPdPUU                U
M 888M0.016UUDPDDUDDDDDDdPUU                U
N 1F?0.016UPDUDDUDDDDDDDUU                U
FC 70F0.031PPDUDDUDDDDDDDUU                U
CONGO 371M0.063FFDUDDUDDDDDDDUU                U
M 03F0.063FFDUDDUDDDDDDDUU                U
M 133F0.063FFDUDDUDDDDDDDUU                U
M 182M0.063FFDUDDUDDDDDDDUU                U
M 300F0.063FFDUDDUDDDDDDDUU                U
M 27 series 2U0.094FFDPDDPDDDDDDDUU                U
M 891F0.266FFFFPFFPPFDDDDUU                U
M 573F0.359FFFFFFFFPFFFDDUU                U
M 886F0.516FFFFFFFFFFFFFFPU                U
CAM II 15F0.703FFFFFFFFFFFFFFFP                U
CAM 236M FFFFFFFFFFFFFFFF                U
M 134F FFFFFFFFFFFFFFFF                U
M 574F FFFFFFFFFFFFFFFF                U
CAM 200M  F   FFFFF  F  F                P
CAM II 326F  F   F FFF  F                   P
M 803F FFFFFFFFFFFFFFFF                P
M 792F FFFFFFFFFFFFFFFF                P
CAM 11F  F   FFFFF  F  F                F
CAM 26F?  F   FFFFF  F  F                F
CAM 74M  F   FFFFF  F  F                F
CAM II 9F  F   FFFFF  F  F                F
M 02F  F   FFFFF  F  F                F
CAM 199M FFFFFFFFFFFFFFFF                F
CAM 237F FFFFFFFFFFFFFFFF                F
CAM 238F FFFFFFFFFFFFFFFF                F
PCM M 491F FFFFFFFFFFFFFFFF                F
M 814F FFFFFFFFFFFFFFFF                F
M 165F FFFFF F  FFF FFF                F
FC 181U  F   FFFFF  F  F                 
M 04F FFFFFFFFFFFFFFFF                F
CONGO 259M FFFFFFFFFFFFFFFF                 
CAM 201U  F    P        U                U
CAM 67U                                  
CAMEROONS 1929U                         U        


Developmental information for Gorilla gorilla gorilla specimens at the Powell-Cotton Museum. Specimens are listed in increasing developmental age as determined by degree of dental eruption and epiphyseal fusion. Developmental marker abbreviations (column headings) follow Table 5 and are ordered within the dentition and postcranium by their developmental sequence as shown in Table 5; developmental data abbreviations as given in the text. Calculation of maturity score is described in the text. Skulls with fully erupted dentition but without postcrania are listed after all other specimens because long bone epiphyses may remain unfused or partially fused after full dental eruption. However, postcranial fusion precedes cranial suture fusion in all cases for which complete skeletons are available, so completely fused crania presumably belong to individuals that were also postcranially mature. Finally, three isolated mandibles are listed at the very end of the list of specimens. A color version is available at the Human Origins Database,

Table 7. APPENDIX B.
SpecimenSexMaturity scorem1M1i1I1i2I2M2m2P3P4p4p3cCm3M3HDOCUPHEFeLTFeHFeGTRPTDFDFeDTPFiPRDHPUDCranium
M 780M?                UUUUUUUUUUUUUUUU 
M No number - babeM??                UUUUUUUUUUUUUUUU 
M 32 series 2M0.000UUDDDDUUDDDDDDUU                U
CAM II 331M0.000UUDDDDUUDDDDDDUU                U
M 631M0.000UUDDDDUUDDDDDDUU                U
M 532F0.000UUDDDDUUDDDDDDUU                U
M 000 (M)M0.047FPDDDDUUDDDDDDUU                U
M 487M0.063FFDDDDUUDDDDDDUU                U
M 000 series 1F0.063FFDDDDUUDDDDDDUU                U
M 319F0.063FFDDDDUUDDDDDDUU                U
M 33M0.063FFDDDDUUDDDDDDUU                U
M 000 series 2F0.063FFDDDDUUDDDDDDUU                U
M 32 series 1F0.250FFFFFFFFDDDDDDUU                U
CAM II 315F0.313FFDFDFFFFFFFDDUU                P
M 791F0.469FFFFFFFFFFFFPPPP                P
M 03F0.484FFFFFFFFFFFFDFPF                P
CAM I 97F0.641FFFFFFFFFFFFPDUU                P
Z V 91F0.656FFFFF FFFFFFFFPP                P
CAM I 198F0.672FFFFFFFFFPFFFFUU                P
M 21M0.828                PFPPFFFUFFFFFUUU 
M II 23U0.875                FFFFFFF PFPPP PP