the structure, function, and evolution of primate dental enamel

ARTICLES

Built to Last: The Structure, Function, and Evolutionof Primate Dental EnamelMARY CAROL MAAS AND ELIZABETH R. DUMONT

THE STRUCTURE ANDCOMPOSITION OF ENAMEL

Enamel is the strongest tissue in thebody. It is not only hard (resistant topermanent surface deformation), but

tough (resistant to crack propagationand brittle fracture). Its durabilitymeans that enamel can potentially bepreserved, unaltered, for literally mil-lions of years. Moreover, because it isnot remodeled during life, enamel isunique among mammalian tissues inthat it preserves a permanent recordof its ontogeny.

Enamel is composed of fiber-likemineral crystals and a small nonmin-eral fraction of water and protein thatholds the mineral fibers together. Theresulting composite material is muchtougher than mineral alone. The min-eral and nonmineral components areorganized in a complex fabric thatdissipates forces traveling throughteeth and protects them from fracture.

Mineral is the dominant compo-nent of enamel, which is much morehighly mineralized than the twoother calcified dental tissues, dentineand cementum. The enamel mineralis a form of carbonate hydroxyapatite[Ca10(PO4)6(OH)2]. The Ca:P ratio ofenamel apatite is slightly lower than thatof hydroxyapatite.1–3 In human teeth,enamel is densest (96% apatite) close tothe outer surface and less dense (84%apatite) near the enamel-dentine junc-tion.4

Though water and protein compriseonly a minor part of mature enamel,

they are crucial to its development(see Box 1) and are important in under-standing its structural organizationand physical properties. Abundant hy-drophobic proteins called amelogeninsdominate developing enamel, but asthe mineral crystals grow in size dur-ing maturation, amelogenins disap-pear so that only the less abundant,acidic proteins, enamelins and tufte-lins remain.5–8 This remnant proteinfraction is contained in the spacesbetween mineral crystals, where itserves as a ‘‘glue’’ between crystallites.9

Most of the water within enamel istightly bound to the mineral phase10

and serves to influence enamel’s com-pressibility and permeability.11,12 Al-though the high mineral content ac-counts for enamel’s hardness, it is thearrangement and organization of itsmineral and nonmineral constituents(enamel structure) that modulates theway enamel responds to stress.

Enamel structure can be visualizedthrough a simple hierarchical modelof increasing complexity and scale.13–16

The least complex and smallest struc-tural units are apatite crystallites. Inmost mammal teeth, crystallites aregrouped into more complex and larger-scale structures known as prisms. Thearrangement of groups of prisms deter-mines the next most complex unit,enamel types. On a still larger andmore complex scale, the differentenamel types within a tooth define itsschmelzmuster; this German term ispreferred over its literal English trans-lation, ‘‘enamel pattern,’’ which is aspecialized description of the arrange-ment of enamel types on occlusal sur-faces of hypsodont teeth.15 The mostcomplex and largest-scale hierarchicallevel is that of the dentition, whichdescribes the variation of schmelzmus-

Mary Carol Maas is Postdoctoral Fellow inthe Department of Anatomy at NortheasternOhio Universities College of Medicine andResearch Fellow in the Department of Anthro-pology and Laboratory of Vertebrate Paleon-tology at the University of Texas, Austin. Herinterest in the biogeographic origins of mod-ern mammals has taken her to Paleoceneand Eocene fossil localities in North Americaand, most recently, the Eocene of Turkey. Herinvestigations of enamel microstructure andtooth wear have led her to electron micro-scopes in England, Germany, and the UnitedStates. At other times she lives near Merida,Mexico.

Elizabeth (Betsy) Dumont isAssistant Profes-sor of Anatomy at Northeastern Ohio Univer-sities College of Medicine and Research As-sociate at the Carnegie Museum of NaturalHistory, Section of Mammals. After receivinga Ph.D. in anthropology from The State Uni-versity of New York at Stony Brook for herwork on primate tooth enamel, she struck outin a new direction with a new group of mam-mals—bats. She is currently studying the rela-tionships among the craniofacial anatomy,feeding behavior, and community structure ofboth New and Old World fruit bats. Her fieldwork takes her to Central America, Australia,and Papua New Guinea.

Key words: enamel microstructure; enamelprisms; dental function; primate evolution; pri-mate phylogeny

The teeth of every primate, living and extinct, are covered by a hard, durable layer ofenamel. This is not unique: Almost all mammals have enamel-covered teeth. In addition,all of the variations in enamel structure that occur in primates are also found in othergroups of mammals. Nevertheless, the very complexity of enamel and the variation wesee in it on the teeth of living and fossil primates raise questions about its evolutionarysignificance. Is the complex structure of primate enamel adaptive? What, if anything, doesenamel structure tell us about primate phylogeny? To answer these questions, we need tolook more closely at the characteristics of prismatic enamel in primates and at thedistribution of those characteristics, both in relation to our knowledge of primate dentalfunction and feeding ecology and from a phylogenetic perspective.

Evolutionary Anthropology 133

Box 1. Enamel Development

The first stages of tooth develop-ment consist of a series of epithelial-mesenchymal interactions that definethe enamel-dentine junction.111 As isthe case in many other morphologicalsystems, both morphogen gradientsand the products of homeobox genesappear to play significant roles in deter-mining the location and timing of toothdevelopment.112–114 The mesenchymewithin tooth germs is derived from theneural crest and gives rise to cellsknown as odontoblasts. The onset ofdentine secretion by odontoblasts in-duces enamel formation. The epithe-lium within tooth germs differentiatesinto ameloblasts, the cells that secretethe enamel matrix proteins in whichhydroxyapatite crystals grow. It is notentirely clear whether dental epithe-lium is ectodermal or endodermal inorigin.8,111

At some point in the life cycle ofmost mammalian ameloblasts, the se-cretory surface develops a protrusioncalled a Tomes process, which is sur-rounded by a flattened area called anameloblast shoulder (A). Both of thesestructures secrete enamel matrix pro-teins. During enamel mineralization,hydroxyapatite crystals within the se-creted enamel matrix proteins growperpendicular to Tomes processes andameloblast shoulders.115,116 Becauseof the fixed relationship of the crystal-lites growing perpendicular to amelo-blast secretory surfaces, the shape ofthe secretory poles of ameloblastsdetermines the orientation of maturecrystallites and, therefore, the forma-tion of either prismless or prismaticenamel (A). At both the beginning andend of enamel deposition in mam-mals, ameloblasts lose their Tomesprocesses. This results in a very thinlayer (sometimes less than a micron)of parallel crystallite enamel near theenamel-dentine junction and usually athicker layer at the outer surfaces ofteeth. Tomes processes form and as-sume a variety of shapes during themiddle phase of ameloblast secretion,leading to the different enamel prismtypes seen in mid-thickness enamel.117

(see Fig. 1).

The movement of ameloblasts rela-tive to one another and the local me-chanical environment surroundingTomes’ processes determines theenamel type—in primates, either ra-dial or decussating enamel. Thoughall agree that complex movements ofameloblasts result in prism decussa-tion, the specific mechanisms mediat-ing it remain a source of debate.8,23,43,87

The activity of ameloblasts, like thatof other secretory cells, varies in regu-lar cycles. Ameloblasts exhibit bothcircadian (daily) and circaseptan (ap-proximately weekly) variation in therate at which enamel matrix proteinsare deposited. A record of daily andweekly fluctuations in depositional ac-tivity are visible in fully mineralizedenamel as two types of incrementallines: prism cross-striations (B) and(Brown) striae of Retzius, respectively.In most primate teeth (but not neces-sarily all mammal teeth), some striaeof Retzius are expressed at the toothsurface as perikymata (see Fig. 4).

The relationship between cross stria-tions and circadian variation in enameldeposition has been demonstrated ex-perimentally.118,119 Studies document-ing longer depositional cycles arebased on counts of cross striationsbetween adjacent striae of Retzius.These cycles have been called circa-septan, but the seven-day interval im-plied by this name is certainly notuniversal, even among primates. Al-though the interval between striae andtheir expression on the tooth surfaceas perikymata is approximately sevendays in humans,120 it varies bothwithin121–123 and among124–127 spe-cies. Nevertheless, many studies havemade use of the relatively constant(roughly weekly) time intervals repre-sented by striae of Retzius and periky-mata to investigate comparative ratesof growth and development in fossilhominoids and hominids.128–132 Suchstudies may ultimately go a long waytoward illuminating how growth ratesand patterns constrain prism patternsand enamel types and, in doing so,have influenced the evolution of pri-mate enamel microstructures.

A. The relationship between ameloblast shapeand the orientation of hydroxyapatite crystal-lites in parallel crystallite prismless enamel (top)and prismatic enamel (bottom). Hydroxyapa-tite crystals (short dashed lines) always growperpendicular to the secretory faces of am-eloblasts. Thus, parallel crystallite enamel isformed when ameloblasts lack Tomes pro-cesses and prismatic enamel is formed whenthesecretorypolesofameloblastshaveTomesprocesses and ameloblast shoulders.

B. Prism cross striations appear under back-scattered scanning electron microscopy asdark bands (arrows) perpendicular to prismlong axes. The dark bands indicate regions ofincreased carbonate concentration that formduring daily periods of decreased deposition.In the scanning electron microscope, crossstriations usually, but not always,122 appear tobe associated with so-called prism varicosi-ties, or alternating swellings and constrictionsalong prism long axes. Several studies havedemonstrated an association between cross-striation repeat interval and enamel prismtype.9,41

134 Evolutionary Anthropology ARTICLES

ter from tooth to tooth. Some of thesestructural levels are unique to mamma-lian teeth, although many other verte-brates have teeth with complex enamelor enamel-like organization.14,17,18

Enamel CrystallitesThe apatite crystals in enamel are

called crystallites because of their sub-microscopic dimensions. Like all crys-talline structures, enamel crystalliteshave a short dimension, the a-axis,and a longer dimension, the c-axis oroptical axis. Reports of crystallitelength vary from a fraction of a mi-cron to 100 microns,19 but most agreethat they are very long compared tothe short crystals that characterize

bone. Early in enamel development,crystallites are flat and ribbon-like,20

but fully mineralized crystallites arehexagonal19 or rhomboidal21 in crosssection. Complicating their descrip-tion is the likelihood that crystalliteshape differs between species and maybecome more irregular as crystallitesincrease in size during enamel matura-tion.22 Small groups of adjacent crys-tallites are arranged in subparallel crys-tal groups, where both a- and c-axes ofcrystallites are aligned parallel to oneanother.23

Although there is some evidence ofinterspecific differences in crystallitecomposition, size, and shape,22 mostcomparative studies have focused oninterspecific variation in their arrange-

ment and orientation. These criteriaare used to define the two major classesof mammalian enamel: prismless andprismatic enamel. (The terms ‘‘apris-matic enamel,’’ ‘‘nonprismatic enamel,’’and ‘‘preprismatic enamel’’ have beenused extensively in the literature torefer to various types of prismlessenamel, but confusion has arisen frominconsistent application of these con-cepts.14 The most recent synthesis ofenamel terminology recommendsabandoning their use.13). The simplestarrangement of enamel crystallites is atype of prismless enamel called paral-lel crystallite enamel. In parallel crystal-lite enamel, the crystallites are ori-ented with their long axes roughlyparallel to one another and perpen-dicular to the outer tooth surface andto the enamel-dentine junction (Fig.1A). Parallel crystallite enamel is foundin at least some regions of most mam-mal teeth, most prominently close tothe outer surface of the tooth. Somereptiles and early mammals have aslightly more complex type ofprismless enamel. This enamel, termedcolumnar unit enamel, is character-ized by a regular pattern of gentlydiverging crystallites.17 Prismlessenamel (either parallel crystalliteenamel or columnar unit enamel) isthe only type of enamel found in rep-tile teeth, with the single exception ofthe agamid lizard, Uromastyx.14,17,24 Inprimates, as in most mammals,prismless enamel is a minor compo-nent of the schmelzmuster. The bulkof primate enamel has a more com-plex organization, known as prismaticenamel, in which the primary struc-tural units are enamel prisms.

Enamel PrismsThese are both more complex and

larger in scale than crystallites.Whereas crystallites are a fraction of amicron in diameter, the diameter ofprisms ranges from two to ten mi-crons. Enamel prisms are rod-shaped(cylindrical) bundles of crystallites(Fig. 1B–D). The crystallites within abundle, or prism, are oriented roughlyparallel to the prism long axis, thoughin some prism types they fan out fromthe center toward the edges (Fig. 2A).

Figure 1. Scanning electron micrographs of primate prism morphology. A: Longitudinal sectionthrough a lower incisor of Varecia variegatus showing inner layer of radial prismatic enamel(PE) and outer layer of prismless parallel crystallite enamel (PCE). Enamel-dentine junction atbottom. B: Surface-tangential section through a Loris tardigradus molar showing the arrange-ment of prisms (P) and interprismatic enamel (IP). In some primates and other mammals,crystallites within the prisms and contiguous interprismatic enamel converge, forming a lineardiscontinuity called an enamel seam (arrows). C: Longitudinal section through a Loris tardigra-dus molar shows that crystallites in the prisms (P) are oriented roughly parallel to the prism longaxis and at an angle to crystallites in the interprismatic (IP) enamel. D: Surface-tangentialsection through a Tarsius syrichta molar showing cross sections of enamel prisms demarcatedfrom interprismatic enamel by the prism sheath (arrows). The change in crystallite orientationbetween prismatic and interprismatic regions is gradual at the bottom (cervical direction) ofmost prisms, so the prism sheath is arc-shaped. However, some prisms have closed prismsheaths, and there is considerable variation in prism shape and their arrangement. Scalebars 5 10 µm.

ARTICLES Evolutionary Anthropology 135

The long axes of individual prismsextend from the enamel-dentine junc-tion to the outer surface of a tooth.The crystallites that lie outside theprisms are called interprismaticenamel. Prismatic and interprismaticenamel differ only in the arrangementand orientation of its crystallites; theircomposition is identical. The bound-aries of a prism are defined by submi-croscopic gaps formed by the changein orientation between prismatic andthe interprismatic crystallites. Be-cause protein and water accumulatein these gaps, prism boundaries ap-pear as distinct structures called prismsheaths (Fig. 1D).

Prism patterns are two-dimensionaldescriptions of cross-sectioned prisms(sectioned perpendicular to their longaxes). They are defined by the shapesof prism sheaths and differences inprism packing, or the way prisms andinterprismatic enamel are arrangedrelative to one another. Boyde9,25–27

proposed an influential scheme ofenamel classification based on obser-vations of the developing enamel sur-

face that correlated with differences inthe size, shape, and arrangement ofmature enamel prism cross-sections.Although many other workers haveinvestigated interspecific variation inprism structure,28–33 no other descrip-tive scheme for tooth microstructurehas been so broadly used in primatedental studies.

Boyde’s scheme distinguished threemajor classes and several subclassesof prisms in mammalian enamel. (Fig.3). He demonstrated that many spe-cies are characterized by the predomi-nance of a single prism pattern. Heand others went on to characterizeentire families and even orders ofmammals by prism pattern.14,26,27,34 Inreality, prism cross sections show farmore variability in shape and align-ment than is evident from schematicdiagrams of prism packing arrange-ments (Fig. 1D). Indeed, many speciesshow a combination of all three prismpatterns. Moreover, the appearance ofprism cross sections can change dra-matically depending on the angle atwhich they intercept tooth surfaces

(Fig. 2B). Because of this, confusionarises over distinguishing prism pat-terns in primate teeth, especially vari-ants of the basic patterns.35,36 Someresearchers have advocated a simplerclassification that distinguishes onlybetween ‘‘closed’’ prisms (as in Pattern1) and ‘‘open’’ or ‘‘arc-shaped’’ prisms(as in Patterns 2 and 3).14,37–39 Despitethe practical drawbacks to applyingBoyde’s prism classification schemewithin Primates, that scheme has beenalmost universally adopted in the lit-erature on primate enamel. We reportthose designations here (Table 1 andTable 2), with the important caveatthat they have not necessarily beenapplied consistently among differentstudies.

Enamel Types

Much of the variation in enameloccurs not at the level of crystallites orprisms, but at the more complex levelof enamel type. Enamel types are inde-pendent of prism patterns, and thusrepresent a separate class of charac-ters as well as a distinct level of organi-zation.15 The type of prismless enamelfound in primate teeth, like that of allmammals, is parallel crystalliteenamel. The two prismatic enameltypes that occur in primate enamel areradial enamel and decussating enamel(Fig. 4). In radial enamel, all prismsare roughly parallel to one another asthey extend out from the enamel-dentine junction to the surface. In

Enamel types areindependent of prismpatterns, and thusrepresent a separateclass of characters aswell as a distinct level oforganization. The type ofprismless enamel foundin primate teeth, like thatof all mammals, isparallel crystalliteenamel.

Figure 2. Schematic block diagrams of prismatic enamel (above), showing the relationship ofprismatic (P) and interprismatic (IP) crystallites. Crystallites in interprismatic regions are orientedat an angle to crystallites within prisms. In this prism pattern, the change in crystallite orienta-tions between prisms and IP enamel is abrupt adjacent to the prism sheaths, but gradual at theopen, cervical ends of the prisms. If enamel is sectioned at different angles, as a result of eitherartifacts produced during specimen preparation or the formation of differently angled wearfacets during normal tooth use, this will produce different orientations of prismatic andinterprismatic crystallites relative to the sectioned surface (above, right) and differences in thesize and shape of prism cross-sections108 (below).


decussating enamel, prisms are ar-ranged in regularly organized, alternat-ing layers or groups that extend outfrom the enamel-dentine junction tothe surface at different orientations.Decussating enamel produces the opti-cal phenomenon referred to as Hunter-Schreger bands, which are a result ofthe different refractive properties ofdifferently oriented prisms in adjacentlayers. Other prismatic enamel typesthat do not occur in primates includeirregular or 3-D enamel, which consistsof complexly interwoven bundles ofprisms,16 and tangential enamel, whichconsists of prisms with a strong lateraldeviation, so that prism axes are ori-ented tangential to the tooth sur-face.15,40

Radial enamel is the prismaticenamel type most common amongmammals.40 In simple radial enamel,the orientation of interprismatic crys-tallites deviates only slightly from thatof prismatic crystallites. Simple radialenamel is the only type of prismaticenamel that occurs in the teeth of theearliest mammals and many small-bodied living mammals, including

many of the smaller primates. A de-rived variant of radial enamel, modi-fied radial enamel, is characterized byvery high angles between prismaticand interprismatic crystallites (closeto 90° in some cases) and by interpris-matic crystallites that are arranged insheets as thick as the prisms them-selves. Modified radial enamel occursin many hypsodont ungulates andmany marsupials. It has not beenfound in primate teeth. Although someprimate enamels show high anglesbetween prismatic and interprismaticcrystallites, the interrow sheets arenot as thick as those in the modi-fied radial enamel of ungulates andtend to anastomose between adjacentrows rather than maintaining thesheet-like arrangement of the interpris-matic crystallites of modified radialenamel.

Decussating enamel, (or Hunter-Schreger bands) like radial enamel, iscommon in primates. In decussatingenamel, layers of prisms extend out-ward from the enamel-dentine junc-tion to the tooth surface at an angle toprisms in adjacent layers. In primates

and most other mammals, the layersare usually several prisms thick(known as multiserial enamel or pluri-serial enamel), but in some rodentseach layer is only one prism thick(uniserial enamel). The most commontype of decussating enamel is horizon-tal decussation, in which Hunter-Schreger bands are stacked on top ofone another from crown to root, withlong axes of prisms in adjacent hori-zontal layers of prisms extending to-ward the outer enamel surface at differ-ent angles (Fig. 5). This is the type ofdecussating enamel found in primateteeth.32,35,38,41,42

‘‘Hunter-Schreger bands’’ and ‘‘prismdecussation’’ (literally, ‘‘crossing ofprisms’’) are sometimes used synono-mously. However, complete crossingof prisms at 90° to those in adjacentgroups (‘‘true decussation’’43) does not

occur in all species. More typically, theangle between prisms in adjacentbands is less than 90°, and changes asprisms pursue a slightly sinuous coursefrom the enamel-dentine junction tothe outer tooth surface. There is oftena gradual change in the orientation ofprisms from the center of one band tothe center of the next, giving the im-pression of a transitional zone be-tween bands (Fig. 4E). Most descrip-tions of Hunter-Schreger bands arebased on the appearance of prisms insectioned teeth. Bands of prisms sec-tioned more or less parallel to theirlong axes are called parazones; thosesectioned more or less perpendicularto their long axes are called diazones.

There is considerable interspecificvariation in Hunter-Schreger bands,including the numbers of prisms in

Simple radial enamel isthe only type of prismaticenamel that occurs inthe teeth of the earliestmammals and manysmall-bodied livingmammals, includingmany of the smallerprimates.

Figure 3. Diagram of Boyde’s prism patterns.9,26 Pattern 1 prisms are usually small (3 to 5 µm) andhave complete, roughly circular, boundaries. Prisms are completely separated by interpris-matic enamel and are arrayed in offset horizontal rows. Close to the outer enamel surface, theprisms of most mammals are Pattern 1. Boyde’s Pattern 2 and Pattern 3 defined two majorclasses of prisms with incomplete prism boundaries and arc-shaped prism sheaths. In both ofthese prism patterns, the open side of the prism is toward the cervix (root-crown junction) of thetooth. Pattern 2 prisms are small 2 to 4 µm in diameter and are arranged in longitudinal columnsfrom the apex to the cervix. Pattern 2 variants (Patterns 2, 2A, and 2B) differ in the amount anddistribution of interprismatic enamel and angles between prisms in adjacent rows. Pattern 3prisms are larger 5 to 8 µm in diameter and packed in horizontally offset rows, like Pattern 1prisms. The variants (Pattern 3, 3B, and 3C) differ in the shape of the prism sheath and theamount and distribution of interprismatic enamel. Although Boyde’s scheme has been widelyused in primate enamel studies, prism patterns in many primate species are so variable that itmay be more practical to distinguish only between ‘‘closed’’ (Pattern 1) and ‘‘open’’ (Pattern2 and 3) prisms.


TABLE 1. Summary of Prism Patterns and Enamel Types for Living Primates

TaxonMass (gm)96

(F/M)PrismPatterns

Enamel Types(From enamel-dentine junction to surface)

LemuridaeHapalemur griseus 670/748 1, 338 Radial, prismless38

Lemur sp. HSB, radial32

Lemur catta 2,210 1, 338 HSB, radial, prismless38

Eulemur fulvus 2,250/2,180 1, 338 HSB, radial, prismless38

Eulemur macaco 1,760/1,880 1, 338 HSB, radial, prismless38

Varecia variegata 3,520/3,630 1, 338 HSB, radial, prismless38

IndriidaePropithecus verreauxi 2,950 1, 3 HSB, radial, prismless

CheirogaleidaeMicrocebus murinus 63/59 1, 3 Radial, prismlessCheirogaleus medius 282/283 1, 3 Irregular HSB, radial, prismless

DaubentoniidaeDaubentonia madagascariensis 2,490 280 HSB72,80

LorisidaeNycticebus coucang 626/679 1, 3 Radial, prismless54

Irregular HSB (canine only)Loris tardigradus 269/264 1, 3 Radial, prismlessPerodicticus potto 836/830 1, 341 Radial, prismless41

GalagidaeOtolemur garnetti 734/794 1, 3 Radial, prismlessGalago senegalensis 199/227 1, 341 Radial, prismless41

Otolemur crassicaudatus 1,110/1,190 1, 337 Radial, prismless37

TarsiidaeTarsius syrichta 117/114 1, 3 Radial, prismlessTarsius spectrum 108/125 335 Radial, prismlessTarsius bancanus 117/128 339 Radial, prismless

AtelidaeAlouatta sp. 2A;81 1, 336 Radial, weak HSB, prismless, radial32

Alouatta seniculus 5,210/6,690 1, 2, 335 Radial54

Alouatta fusca 4,350/6,730 1, 2, 335

Ateles sp. 2A;81 1;35 1, 3B41,82 irregular HSB, radial, prismless32,41

Ateles belzebuth 7,250/8,290 HSB54

Lagothrix lagotricha 7,020/7,280 1, 3;36 1, 2, 335 HSB54

Brachyteles arachnoides 8,070/9,610 335 HSB54

Callicebus moloch 956/1,020 2B;81 2, 3;36 1, 2, 335 Radial, prismless54

Cacajao melanocephalus 2,710/3,160 342 HSB, prismless42,54

Chiropotes satanus 2,580/2,900 2;35 3;42 1, 336 HSB, prismless42

Pithecia sp. 1, 336

Pithecia pithecia 1,580/1,940 335

Pithecia monachus 2,110/2,610 335 HSB, prismless, HSB54

CebidaeAotus trivirgatus 736/813 2B;81 3C;82 1, 335 Radial, prismless54

Cebus sp. 2B;81 1, 336

Cebus apella 2,520/3,650 1, 342 Radial, HSB, prismless42,54

Cebus capucinus 2,540/3,680 1, 335 Radial, HSB, prismless42

Cebus albifrons 2,290/3,180 Radial, HSB, prismless42

Saimiri sciureus 668/779 1, 3C;41,82 235 Radial, prismless46,54

Callithrix sp. 1;9,81 3C;82 236 HSB9

Callithrix argentata 360/333 HSB46

Callithrix humeralifer 472/475 HSB46

Callithrix jacchus 324/317 1, 2, 349 HSB, prismless32,46,49

Cebuella pygmaea 122/110 3;35 1, 2, 349 HSB, prismless46,49

Leontopithecus rosalia 598/620 1, 2, 3;35 1, 2, 349 Radial, prismless46,49,54

Saguinus sp. 1;81 3C;82 Radial, HSB82

Saguinus fuscicollis 358/343 1;35 1, 2, 349 Radial, irregular HSB, prismless46,49

Saguinus geoffroyi 502/482 1, 2, 349 Radial, irregular HSB, prismless49

Saguinus imperator 475/474 1, 2, 349 Radial, irregular HSB, prismless46,49

Saguinus inustus ?803/585 Radial, irregular HSB, prismless46

Saguinus labiatus 539/490 1, 2, 349 Radial, irregular HSB, prismless46,49

Saguinus leucopus 490/494 Radial, irregular HSB46

(Continued)


each band, the width of the transi-tional zone, the extent of the bandsfrom the enamel-dentine junction, andthe orientation of bands. However,there are problems in documentingthis variation. In many primates,Hunter-Schreger bands vary in differ-ent parts of the teeth, such as cusptips, basins, and near the root-crownjunction. In addition, the appearanceof Hunter-Schreger bands is stronglyinfluenced by the plane of section, afactor that is almost impossible tocontrol because there is no precisecorrelation between orientation of themicroscopic, subsurface structure ofenamel and the gross features of teeth.Some attempts have been made atsystematic documentation of interspe-cies variation in horizontal Hunter-Schregerbandsusingquantitative32,42,44,45

or qualitative assessments,38,45–50 but nogenerally accepted scheme has beendevised. One important distinction isbetween enamel in which discreteparazones and diazones can be dis-tinguished and enamel in which itis impossible to identify discreteHunter-Schreger bands, although theprisms are not parallel to one an-other as in radial enamel. Instead,sectioned teeth show poorly demar-cated, irregular clusters of differentlyoriented prisms. This type of enameloccurs in several primate speciesand has been called irregular decussa-tion49 or irregular Hunter-Schregerbands (Fig. 4B). This type of enameldiffers from irregular (3-D) enamel(3-D enamel) because the clusters ofprisms are not interwoven in a com-plex fashion.

More derived variants of decussat-ing enamel have evolved in somegroups of mammals, but not primates.Some bone-feeding Carnivora, such ashyenas, have evolved zig-zag Hunter-Schreger bands, with the undulationsof prism layers being so extreme thatthese layers are vertical in some re-gions.51 In vertical Hunter-Schregerbands, layers of alternately orientedprisms are oriented vertically fromcrown to root so that, in transversesection, zones appear to radiate out-ward from the central axis of the tooth.Such bands occur in many hypsodontungulates.23,52

Schmeltzmuster and DentitionThe enamel of most mammalian

teeth consists of more than a single

TABLE 1. (Continued)

TaxonMass (gm)96

(F/M)PrismPatterns

Enamel Types(From enamel-dentine junction to surface)

Saguinus midas 575/515 Radial, irregular HSB, prismless46

Saguinus mystax 539/510 1, 2, 349 Radial, irregular HSB, prismless46,49

Saguinus nigricollis 484/468 Radial, irregular HSB, prismless46

Saguinus oedipus 404/418 1, 2, 349 Radial, irregular HSB, prismless46,49

Callimico goeldii 468/499 1, 2, 349 Radial, irregular HSB, prismless46,49

CercopithecidaeCercocebus torquatus 6,230/11,000 2, 335 HSB, radial, prismlessCercopithecus mona ?/5,100 2, 335

Cercopithecus neglectus 4,230/7,350 2, 335

Cercopithecus pygerythus 3B41,81 HSB, radial, prismless41

Chlorocebus aethiops 2,980/4,260 2, 383

Colobus angolensis 7,570/9,680 1, 335

Colobus polykomos 8,300/9,900 1, 335,83

Erythrocebus patas 5,770/10,600 1, 2;9 3B82

Lophocebus alibgena 6,020/8,250 2, 335 HSB, radial, prismlessMacaca fasciculata 3,590/5,360 341 HSB, radial, prismless32,41

Macaca mulatta 5,370/7,710 3;41,83 1, 2, 39,25,35 HSB, radial, prismless41

Macaca nemestrina 6,500/1,120 2, 335

Macaca speciosa 341 HSB, radial, prismless41

Macaca sylvanus 11,000/16,000 2, 335

Nasalis larvatus 9,820/20,400 383

Papio anubis 1,330/25,100 2, 383

Papio cynocephalus 12,300/21,800 3;41 2, 335 HSB, radial, prismless41

Papio hamadryas 9,900/16,900 383

Papio sphinx 2, 335

Presbytis obscura 1, 235

Semnopithecus entellus 9,890/13,000 1, 2, 335,83

Trachypithecus cristatus 5,760/6,610 1, 335

HominoideaGorilla gorilla .70,000 1;90 1, 2, 39,35,85 Radial;41 weak HSB, radial, prismless35

Homo sapiens .40,000 1, 2, 39,35,85 HSB, radial, prismless35

Pan troglodytes .33,000 1;90 1, 2, 39,35,85 HSB, radial, prismless35

Pongo pygmaeus 35,600/77,900 1;90 1, 2, 39,35,85 HSB, radial, prismless32,35

Hylobates sp. 5,500–11,900 1, 2, 39 HSB32

Bold type indicates the predominant prism pattern or enamel type reported in the literature; contradictory or alternativeinterpretations are discussed in the text. Unreferenced data are from unpublished work of the authors. Prism patterns follow theconvention established by Boyde,25–27 but descriptions such as ‘‘closed’’ (as in Pattern 1) or ‘‘open’’ (as in Patterns 2 and 3) moreaccurately reflect the variability of prism patterns in many primate species. HSB 5 Hunter-Schreger bands.


TABLE 2. Prism Patterns and Schmelzmuster for Fossil Primates and Plesiadapiformes

TaxonEstimatedMass96 (gm)

PrismPattern

Schmelzmuster(from enamel-dentinejunction to surface)

ORDER PLESIADAPIFORMESPlesiadapidae

Chiromyoides sp.39 150–300 1, 3 HSBNannodectes intermedius39 221 1, 3 Radial, prismlessPlesiadapis cookei39 3,055 1, 3 HSB, prismlessPlesiadapis rex39 506 1, 3 HSB, prismlessPlesiadapis tricuspidens 759 1, 3 HSB, prismlessPurgatorius unio15 92 1, 3

ParomomyidaeIgnacius frugivorus39 96 1, 3 Radial, prismlessIgnacius graybullianus39 152 1, 3 Radial, prismless

ORDER PRIMATESAdapidae

Adapis parisiensis 1,300 1;39 3 Radial, prismlessAdapis sudrei 1,400 2, 3 Radial, prismlessCantius ralstoni45 1,300 1, 3 Radial, prismlessCantius mckennai45 1,600 1, 3 Radial, prismlessCantius trigonodus45 2,000 1, 3 Irregular HSB, radial, prismlessCantius abditus45 3,000 1, 3 HSB, radial, prismlessCantius venticolis45 3,000 1, 3 HSB, radial, prismlessCopelemur praetutus45 1,300 1, 3 Radial, prismlessEuropalemur dunfei 1,360 1, 3 Radial, prismlessLeptadapis sp. 1,300–4,000 2, 3 RadialPericonodon jaegeri 920 1, 3 Radial, prismlessNotharctus nuniensus45 2,000 1, 3 Irregular HSB, radial, prismlessNotharctus robinsoni45 4,700 1, 3 HSB, radial, prismlessNotharctus sp.45 3,000 1, 3 HSB, radial, prismless

OmomyidaeAnaptomorphus sp. 200–500 1, 3 Radial, prismlessOmomys sp. 300 1, 3 Radial, prismlessShoshonius sp. 150 1, 3 Radial, prismlessTeilhardina americana39 120 3Tetonius sp.39 100–300 3Washakius sp.39 150 1, 3 Radial, prismlessHemiacodon sp. 1,005 3

Parapithecidae78

Apidium moustafai 850 1, 3 Radial, prismlessApidium phiomense 1,600 1, 3 Radial, prismlessParapithecus grangeri 3,000 1, 3 HSB, radial, prismless

OligopithecidaeCatopithecus browni97 900 1, 3 Radial, prismless

PropliopithecidaeAegyptopithecus zeuxis78 6,700 1, 3 HSB, radial, prismless

CercopithecoideaMesopithecus pentelicus84 8,000 3

HominoideaOreopithecus sp.35 30,000 3Proconsul sp.93,94 .17,000 3AProconsul major99 50,000 1 HSBProconsul africanus99 27,400 1 HSBDryopithecus sp.99 .20,000 3Otavipithecus sp.99 14,000–20,000 1, 3Sivapithecus sp.93,94 .40,000 3AGigantopithecus sp.93,94 .190,000 3AAustralopithecus africanus35,92 .30,000 3Paranthropus robustus35,92 .32,000 3

Bold type indicates the predominant prism pattern or enamel type; contradictory or alternative interpretations are discussed in thetext. Unreferenced data are from unpublished work of the authors. Prism patterns follow the convention established by Boyde,25–27

but descriptions such as ‘‘closed’’ (as in Pattern 1) or ‘‘open’’ (as in Patterns 2 and 3) more accurately reflect the variability of prismpatterns in many primate species. HSB 5 Hunter-Schreger bands.


enamel type. As noted, schmelzmusterrefers to the arrangement of theseenamel types within a tooth (Fig. 4F).In molars of small-bodied primates,

for example, the schmelzmuster typi-cally consists of an inner layer ofradial enamel and a much thinnerouter layer of prismless parallel crystal-

lite enamel. In larger primates, theschmelzmuster typically includes aninner layer of horizontal Hunter-Schreger bands, a middle radial layer,and an outer prismless layer. Thor-ough characterization of the schmelz-muster is facilitated by examinationby light microscopy, since the differ-ences in prism orientation and ar-rangement that characterize the dif-ferent enamel types are sometimesmost easily distinguished by theirdifferent refractive properties. Theconsiderable variation in schmelz-muster among primate species is ex-pressed as presence, absence, or differ-ences in proportions of some enameltypes, but has not been studied system-atically.

The dentition, or variation in schm-elzmuster among different teeth in thetooth row, is the highest level of enamel

Figure 4. Scanning electron micrographs of prismatic enamel types that occur in primate teeth. A:Longitudinal section through a cusp of a Nycticebus coucang molar showing radial enamel. Inradial enamel, prisms extend outward from the enamel-dentine junction (bottom) in straight,curved, or sinusoidal paths, but are always parallel to one another. Scale bar 5 10 µm. B: Longitudinalsection through lateral enamel of a Cheirogaleus medius molar, with enamel-dentine junction atbottom. Irregular Hunter-Schreger bands appear as clusters of prisms (cl) oriented at angles toadjacent prisms, but do not form discrete zones. Scale bar 5 10 µm. C: Longitudinal section throughlateral enamel of a Cebus apella molar showing broad Hunter-Schreger bands extending in asinusoidal course from the enamel-dentine junction (right) to the outer surface. Adjacent Hunter-Schreger bands are sometimes referred to as diazones and parazones. Prisms in diazones (dz) aresectioned perpendicular to their long axes; prisms in adjacent parazones (pz) are sectioned parallelto their long axes. Scale bar 5 100 µm. D: Longitudinal section through lateral enamel of a Chiropotessp. molar showing narrow, well-organized Hunter-Schreger bands extending straight from the enamel-dentine junction (bottom right) to the tooth surface (left). Perikymata (left arrows) visible on the outertooth surface are the surface manifestations of the incremental striae of Retzius (horizontal arrows),which mark the place at which the ameloblasts along the developing enamel front appear to haveundergone decreased activity.1,122 Scale bar 5 100 µm. E. Longitudinal section through lateralenamel of a Cacajao molar showing gradual change in the orientation of prisms between adjacentHunter-Schreger bands. Because the change in prism orientation between parazones (pz) anddiazones (dz) is gradual, there appears to be a transitional zone (tz) where prisms are sectionedoblique to their long axes. Striae of Retzius are visible in the outer, parallel crystal enamel (right). Theenamel-dentine junction is at left. Scale bar 5 100 µm. F: Longitudinal section through buccalenamel of a Lemur catta premolar with the enamel-dentine junction at bottom. The schmelzmusterconsists of Hunter-Schreger bands, radial enamel, and prismless parallel crystal enamel (PCE) at theouter surface (top). Relative proportions of enamel types may vary in different regions of a toothcrown, such as at the tip of a cusp and the root-crown junction. Scale bar 5 100 µm.

Figure 5. A: Simplified mechanical model ofloads and stresses in a mammal tooth cusp,adapted from Pfretzschner.69 Vertical chew-ing loads (vertical arrows at top) producehorizontal tensile stresses (horizontal arrows)that increase in magnitude toward the baseof the cusp. Tensile stresses can pull prismsapart and thus form vertical cracks (wavylines). These cracks are resisted by the alter-nating orientation of prism long axes in adja-cent Hunter-Schreger bands. B: Simplified dia-gram of prism decussation, adapted fromRensberger.44 Prism long axes have differentorientations in adjacent Hunter-Schregerbands. Vertical cracks created by tensilestresses generated during chewing follow theplanes of least resistance, which are parallelto prism long axes and perpendicular to thehorizontal plane of decussation. Extension ofthe cracks in a vertical direction is resisted bythe change in prism orientation between ad-jacent Hunter-Schreger bands, which thusserve as ‘‘crack stoppers.’’


complexity. Dentition-level differencesin schmelzmuster may be expressed asmarkedly different proportions of thesame enamel types, different arrange-ments of the same enamel types, or theabsence of some enamel types in someteeth. Dentition-level variation inenamel microstructure has receivedlittle attention. Most rodents have dif-ferent incisor and molar schmelzmus-ter, as do rhinoceroses and some insec-tivores; in some rodents, schmelzmustereven differs between upper and lowerincisors.15 Some of these differencesare clearly related to tooth function,but in other cases the explanation isnot obvious. Among primates, there

appears to be relatively little variationin enamel microstructure at the denti-tion level. We have observed that le-murs, lorises, and tarsiers all havesimilar schmelzmuster in incisors, ca-nines, and cheek teeth, despite thevery different gross morphologies,enamel thicknesses, and functions ofthose teeth. Baboon incisors, canines,and premolars also show similar schm-elzmuster, despite differences in toothmorphology and enamel thickness.53

Ceboid microstructure also is reportedto be consistent throughout the denti-tion.42,54 Callitrichids have the sameschmelzmuster in incisors and cheekteeth despite the pronounced func-

tional differences between thoseteeth.46,49

TOOTH ENAMEL AND DENTALFUNCTION IN PRIMATES

Most nonhuman primates acquiretheir permanent teeth early in life.Enamel functions to control toothwear, thereby maintaining the func-tional shape of tooth cusps and crests,and to protect teeth from catastrophicdamage so that they can continue toprocess food throughout an animal’slife. Enamel structure facilitates thesefunctions in two important ways. First,it affects the way that chewing sur-faces wear by mitigating the process

of abrasion. Second, it enhances toothdurability by inhibiting brittle frac-ture of the enamel cap.

The resistance of enamel to abra-sion and brittle fracture depends onstructural features expressed at twodifferent scales. Tooth abrasion, theloss of enamel by food-on-tooth wear,55

occurs at a microscopic scale and isinfluenced by small-scale structuralelements, especially the orientation ofcrystallites. In contrast, brittle frac-ture, failure of a material resultingfrom crack propagation, is a larger-scale process and is mitigated by

Among primates, thereappears to be relativelylittle variation in enamelmicrostructure at thedentition level. We haveobserved that lemurs,lorises, and tarsiers allhave similarschmelzmuster inincisors, canines, andcheek teeth, despite thevery different grossmorphologies, enamelthicknesses, andfunctions of those teeth.

Figure 6. Scanning electron micrographs of Lemur catta molar showing experimentally air-polished wear facets in functionally distinct regions of molars (A and B) and the enamelmicrostructure of those regions etched with dilute HCl (C and D) to show the underlyingdifferences in crystallite orientation relative to the surface. Air polishing highlights the differ-ences in wear resistance between a cuspal facet (A) where interprismatic enamel has beenpreferentially removed and an adjacent chewing facet (B) where there is no difference in rateof removal of prismatic and interprismatic enamel. C: At the cuspal facet, prisms intercept thewear surface so that the interprismatic crystallites (IP) are nearly parallel to the surface andprismatic crystallites (P) are nearly perpendicular; the latter are therefore more resistant toabrasion. D: In contrast, at chewing facets prisms (P) and interprismatic (IP) crystallites inter-cept the surface nearly perpendicular to their long axes, and thus are equally resistant to wear.Scale bars 5 10 µm.


larger-scale structural elements, espe-cially enamel types.

Enamel Structure and AbrasionExperimental studies show that

abrasion, or small-scale wear, removesenamel in units of crystallites or, per-haps, subparallel crystal groups, andthat the primary structural constrainton rate of wear is the orientation ofcrystallites.56–58 Crystallites that areparallel to a wear surface and forcevector are less resistant to abrasionthan are those perpendicular to thesurface. This can be explained by thefact that perpendicularly interceptedcrystallites are largely enclosed by ad-jacent crystallites and present only asmall area for abrasive contact,whereas those intercepted parallel totheir long axes are less protected andpresent a larger area for contact. Invitro studies also show that enamel isleast resistant to abrasion in areaswhere crystallites are parallel to thewear surface, but only if the abrasivevector also is directed parallel to thetooth surface.23,52,53,56,57 A similar rela-tionship between crystallite orienta-tion and wear resistance also has beendemonstrated for naturally worn toothsurfaces.59–61

This relationship between crystal-lite orientation and abrasion has impli-cations for dental microwear studies.The shape of microwear features—that is, scratches vs. pits62,63—mostlydepends on the direction and magni-tude of chewing forces. There is littleevidence that enamel microstructureinfluences the shape of microwear fea-tures. However, there is evidence thatthe size of some microwear features,especially the width of microscopicscratches, is affected by the orienta-tion of underlying crystallites.57,60 Thismeans that, all other things beingequal, abrasion of predominantly sur-face-parallel crystallites will producebroader striations than will abrasionof surface-perpendicular crystallites.Though this phenomenon is unlikelyto overpower the signal of diet in mostinstances, it probably accounts for thelack of discriminatory power of somemicrowear variables.

Crystallite density, or how closelycrystallites are packed, also influences

the way enamel responds to abrasion.Prismless enamel is more resistant toabrasion than is prismatic enamel.This is because prismatic enamel crys-tallites are less densely packed alongprism boundaries. In vitro abrasion ofprismatic enamel has demonstratedthat low-energy mechanical etchingpreferentially removes crystallites fromprism boundaries.56,23 This same phe-nomenon can be observed on natu-rally worn teeth. It is especially pro-nounced if interprismatic crystallitesare parallel to the surface and crystal-lites within prisms are perpendicularto the surface (Fig. 6). However, ifocclusal forces are higher or if foodcontains hard, abrasive particles, thesestructurally constrained patterns willbe swamped. This was illustrated inthe seminal study of microwear inbrowsing and grazing hyraxes64: In thebrowsing species, low-energy polish-

ing wear preferentially removed theinterprismatic crystallites, but in thegrazing species abrasion by harderphytoliths obscured this structurallydependent wear.

Of course enamel structure is notthe only thing that affects how teethwear. Tooth wear is a complex interac-tion among chewing forces, toothshape, food properties, the quantity ofenamel (enamel thickness), andenamel structure. Enamel may wearat different rates depending on howhard food particles are, how muchchewing force is generated, and thedirection of chewing force relative toenamel crystallites and the tooth sur-face. Because of this, enamel maywear at different rates in differentparts of teeth, even though the enamelstructure is the same, or wear at thesame rate in different parts of teeth,

even though its structure is differ-ent.23,37,59,61,65 (See Box 2.)

Enamel Structure and FractureWear also takes place on a much

larger scale. Large-scale enamel losscan occur if a crack passes through theenamel and fractures the tooth crown,and this can have catastrophic effectson tooth function. Most mammalshave powerful chewing muscles andcan generate enough force during nor-mal food processing to fracture abrittle material such as enamel. Theextent of the damage depends on themagnitude of force an animal gener-ates (and thus on the animal’s muscu-lar capacity), the physical propertiesof the food that the animal is chewing,and the structure of its enamel.

Prism boundaries represent naturalplanes of weakness in enamel andconstrain the path through enamelthat cracks will travel. This is becausethe energy required to generate a crack(work of fracture) is lower along prismboundaries, where crystallites areless densely packed, than throughprisms.66,67 At the same time, becauseprism boundaries form a convolutedpath, more total energy is needed for acrack to travel between two pointsthan if it could follow a straight path.This is one way that the structure ofprismatic enamel inhibits crack prop-agation. Many mammals, however,generate enough force to overcomethe crack resistance offered by prismstructure.47 In these mammals, prismdecussation, and thus structural varia-tion at the level of the enamel type, caninhibit crack propagation47,66,67–71

(Fig. 5).Prism decussation is common in

mammal teeth. Enamel with Hunter-Schreger bands first appears in themammalian fossil record in the Pa-leocene, soon after the beginning ofthe major Cenozoic radiation of mam-malian herbivores. The mechanical re-lationship between stress and crackpropagation suggests that prism decus-sation evolved in response to increasedchewing stresses associated with largerbody size, and thus larger chewingmuscles.47 Horizontal Hunter-Schregerbands occur in some species of almostevery family of living primates (Table1).32,37,38,41,43,46,54,72 They occur in teeth

Tooth wear is a complexinteraction amongchewing forces, toothshape, food properties,the quantity of enamel(enamel thickness), andenamel structure.


of hard object and soft-object feedersand in primates with thick and thinenamel. Although, for the most part,they are found only in the teeth ofprimates weighing more than about2,000 gm, there are some puzzlingexceptions.32,35,49 (Fig. 7).

The occurrence of Hunter-Schregerbands in primates with thick and thinenamel also raises questions regard-ing the functional interpretation ofenamel thickness. Thick enamel hasbeen invoked as part of a suite ofmorphological features that includes

tooth occlusal shape and cranial archi-tecture and that is associated withhard-object feeding in primates.73–76

However, the relationship betweenenamel thickness and diet seems to bemore complex. While there is somesupport for the hypothesis that pri-

Box 2. Enamel Microstructure and Wear Facets

Do differences in enamel structurecorrespond to differences in the grossmorphology of primate teeth? Varia-tion in the arrangement of prisms andcrystallites across the surface of someprimate molar teeth suggests that thismay be the case. The three function-ally important and mechanically dis-tinct areas of primate molars are cusptips, cusp slopes—including both morevertically oriented shearing surfacesand more horizontally oriented grind-ing surfaces—and crushing basins.Cusp tips are especially important dur-ing initial puncture-crushing of food.All three functional regions are in-volved during the subsequent chew-ing cycle. Different occlusal forces,such as shearing (surface-parallelforce) and crushing (surface-perpen-dicular force) dominate at differentfunctional regions because of theirdifferent geometries and differences injaw movements during differentphases of the chewing cycle.133

Many mammals with low-crownedtribosphenic molars develop flat hori-zontal wear surfaces at cusp tips (api-cal facets) and more vertically ori-ented Phase I and Phase II chewingfacets along cusp slopes. Becausethese surfaces are angled differentlyrelative to enamel prisms, they inter-cept prismatic and interprismatic crys-tallites at different angles as well, al-though there are no enamel structuraldifferences between the two regions(A). Scanning electron microscopy ofnaturally worn cheek teeth of thegreater galago shows differences incrystallite orientations between thesefunctionally distinct wear facets.37 Atapical facets, interprismatic crystal-lites intercept the wear surface parallelto their long axes and prismatic crystal-lites intercept the surface more perpen-dicular to their long axes. According to

the mechanics of enamel wear, thisarrangement should be resistant tocompressive force (forces perpendicu-lar to the wear surface) generated duringpuncture-crushing, but less resistantto abrasion by shearing forces thatalso act at cuspal facets. In contrast tothe surface of apical facets, interpris-matic and prismatic crystallites inter-cept the Phase I and Phase II facetsslightly oblique to the wear surface,and thus should be optimally resistantto shearing abrasion. This same con-figuration has been observed in somelemur molars (those of Lemur, Hapale-mur, and Eulemur but not Varecia) andin molars of the marmosets Callithrixand Saguinus49 (Fig. 6).

Can this arrangement of crystallitesand wear surfaces be considered anadaptation in these primates? The evi-dence is equivocal. Mammals as func-tionally diverse as the opossum,65

musk shrew,102 and koala59,60 showa similar pattern of different orienta-tions of crystallites relative to cuspaland shearing facet surfaces. In theseanimals, differential resistance to abra-sion between the two regions ac-counts for the creation and mainte-nance of sharp cutting edges alongthe leading edges of crests at theinterface between apical facets andshearing facets. In fact, this potentialfor differential wear conferred by prismor crystallite orientation is thought bysome to be the major factor driving theinnovation of prismatic enamel amongthe earliest mammals.102 But while theprimates studied thus far resembleprimitive mammals with respect to thedifferences in crystallite and prism ori-entation between apical and chewingfacets, they do not develop sharp cut-ting edges between cusp tips andslopes. This suggests that the me-chanical association between prism

A. Schematic diagram of a longitudinalsection through a primate molar cusp tipand adjacent wear facet showing the ori-entation of enamel prisms and crystallitesrelative to functionally distinct wear sur-faces. Prisms extend outward from theenamel-dentine junction at a uniformangle, but because the two wear surfacesare oriented differently, the prismatic crys-tallites (black lines) and interprismatic crys-tallites (gray lines) intercept the surface atdifferent angles in the two regions. At theapical facet, formed by abrasion of thecusp tip during puncture-crushing, interpris-matic crystallites lie parallel to the surfaceand prismatic crystallites intercept the sur-face nearly perpendicular to their longaxes. At the shear facet, formed by abra-sion along cusp slopes during chewing,both the interprismatic and prismatic crys-tallites intercept the surface nearly perpen-dicular to their long axes.

and crystallite orientation and abra-sion has been exapted to enhancetooth durability in mammals such asprimates, in which the functionally di-verse bunodont molars are very differ-ent from those of the early mammals.In these teeth, the surface-parallelcrystallites at surfaces of apical facetsand more perpendicular crystallites atchewing facets are both optimally ar-ranged to resist the different loadingregimes that predominate in each re-gion.37


mate hard-object feeders have rela-tively thicker enamel than do closelyrelated soft-object feeders, there is noevidence that a threshold value ofenamel thickness separates hard-ob-ject feeders from soft-object feeders.77

Other dietary factors, including acid-ity and the abrasiveness of foods, needto be considered in testing functionalexplanations of enamel thickness inprimates. Another important factormay be schmelzmuster. Schmelzmus-ter differences among thick-enameledprimate species demonstrate that allthick enamels are not mechanicallyidentical.78 It may be that increase in

thickness of certain enamel types suchas Hunter-Schreger bands may bemore important than increase in totalthickness for protecting the teeth ofsome hard-object feeders.

ENAMEL STRUCTURE IN LIVINGAND FOSSIL PRIMATES

Primate dental enamel was first sur-veyed in the 1920s by Carter.29 Al-though Carter reached some intrigu-ing conclusions, for example that theadapids Notharctus and ‘‘Pelycodus’’were linked to the Malagasy prosim-ians and the omomyid Hemiacodon to

Tarsius, his work generated little inter-est among anthropologists. Althoughsome primates were included in broadcomparative surveys of mammalianenamel,31,32 the enamel microstruc-ture of primates was generally ignoreduntil the late 1970s, when Boyde’s26,27,43

pioneering application of scanningelectron microscopy to enamel devel-opment and structure heralded a re-newed interest in mammalian enamel.Over the last 20 years, the enamelmicrostructure of many living and fos-sil primates has been sampled (Table 1and Table 2) as a result of increasedinterest in functional as well as phylo-genetic aspects of enamel structure.

ProsimiansRecent studies of prosimian

enamel37–39,79 have demonstrated that

the schmelzmuster of almost all smallprosimians (species weighing less thanabout 2,000 gm), including tarsiers,consists of radial enamel overlaid by avariably thick outer prismless layer(Table 1). There are two known excep-tions: We have observed irregularHunter-Schreger bands in upper andlower canines of the Nycticebus andthroughout the dentition in Cheiro-galeus medius (Fig. 4B). In contrast,all larger prosimians have horizontalHunter-Schreger bands overlaid by ra-dial and prismless enamel, though theHunter-Schreger bands are more dis-tinct and more extensive in some spe-cies than in others.38

We now know that most prosim-ians, including Tarsius, have predomi-

. . . the schmelzmusterof almost all smallprosimians (speciesweighing less than about2,000 gm), includingtarsiers, consists of radialenamel overlaid by avariably thick outerprismless layer. There aretwo known exceptions.

Figure 7. Predominant prismatic enamel types for some living primates. Body masses96 for dimorphicspecies are for the larger sex, and average weights are used for congeneric species with the samedominant enamel type. The dashed line indicates the size (about 1,500 to 2,000 gm) above whichmost herbivorous mammals have Hunter-Schreger bands; radial enamel is dominant in teeth ofsmaller herbivores.47 The strong Hunter-Schreger bands of callitrichine primates are a puzzlingexception to this ‘‘rule.’’


nantly Pattern 3 prisms but that, as isthe case for many mammals, Pattern 1prisms appear throughout the enamelfrom the enamel-dentine junction tothe outer surface, and are commonclose to the tooth surface. This clari-fies the sometimes conflicting descrip-tions of earlier studies35,39 (see Table1). The aye-aye (Daubentonia madagas-cariensis) appears to be unique amongprosimians in that its incisors are char-acterized by distinctive Pattern 2prisms with very broad interrow sheetsand pronounced, narrow Hunter-Schreger bands.72,80 It is not knownwhether this prism pattern and schm-elzmuster are also found in its cheekteeth, though dentition-level variationdoes not appear to be characteristic ofprimates.46,49,53,54

New World MonkeysThese monkeys, like prosimians, are

characterized by a variety of prismpatterns and enamel types. The enamelliterature abounds with contradictorydescriptions of platyrrhine enamel, es-pecially prism patterns (Table 1).Gantt81 thought that platyrrhines re-sembled Old World monkeys in havinga predominance of Pattern 2 prisms,but Shellis41,82 interpreted New Worldmonkey prisms as a variable combina-tion of Pattern 1 and Pattern 3. Propst49

documented all three major prism pat-terns among callitrichines and con-cluded that prism patterns were notunique to genera or species. Likewise,Martin and colleagues35 reported thatmost platyrrhines they studied showsome combination of all three prismpatterns. They also noted unusual ‘‘V-shaped’’ prism cross-sections in aspecimen of Leontopithecus, but mostprisms are the more typical arc-shape.The V-shaped prisms also occasionallyoccur in other callitrichines.49

While prism type does not appear tocharacterize subgroups of platyrrhines,variation in enamel types and schmel-zmuster may be distributed in a moremeaningful way. All of the larger-bodied species have Hunter-Schregerbands (Fig. 7), but the arrangementand appearance of the bands can bestrikingly different (Fig. 4C–E). Pitheci-ines, though not very large, all showwell-defined decussation zones thatextend straight from the enamel-den-tine junction to the outer enamel sur-

face. Cebus has well-defined bandswith a more sinusoidal course to theouter surface; the much larger Al-ouatta has Hunter-Schreger bands, butthey are difficult to distinguish. Asexpected on the basis of their bodysize, the small cebids Saimiri, Aotus,and Callicebus have only radial andprismless enamel. But remarkably,some of the tiny callitrichines weigh-ing less than 500 gm have well-devel-oped prism decussation. Hunter-Schreger bands are well-developed inall teeth of both Callithrix32,35,41,46,49,54

and the pygmy marmoset, Ce-buella.46,49,54 Irregular Hunter-Schregerbands occur in Saguinus and Cal-limico. Contrary to a previous report36

of pronounced Hunter-Schreger bandsin Leontopithecus, lion tamarins arethe only callitrichine that shows theexpected small-mammal schmelzmus-

ter of radial and prismless enamel.46,49

Propst49 offered two hypotheses thatare not mutually exclusive to accountfor the presence of decussation in mar-mosets: one, that they retain prismdecussation from a larger-bodied an-cestor (e.g., the phyletic dwarfing hy-pothesis) and the other that prismdecussation is part of an adaptive-functional complex related to tree-gouging and exudate feeding.

Old World MonkeysEnamel microstructure has been de-

scribed for relatively few of the morethan 80 species of living Old Worldmonkeys (Table 1).32,35,36,41,81,83,84 Mostof these reports have focused exclu-sively on prism patterns, but a fewinclude brief descriptions of enameltypes and schmelzmuster.

Cercopithecoids all appear to have acombination of radial and decussatingenamel with a thin but variableprismless layer. This is not surprising,given their relatively large body sizes.Because schmelzmuster has been de-scribed for so few species, we do notknow whether Old World monkeys,like New World monkeys, show anyvariation in the appearance of Hunter-Schreger bands that might be of eitherfunctional or phyletic importance.Variation in enamel among differentparts of the dentition also has beenneglected. However, it is likely that theschmelzmuster is similar throughoutthe dentition, given that this seems tobe the case for other primate species.Moreover, the canines, premolars, andincisors of Papio all show well-devel-oped Hunter-Schreger bands.53

Initial descriptions of cercopithe-coid prism patterns raised hopes thatprism morphology could be used todistinguish this group of primates,and perhaps to determine affinities ofproblematic species. Large areas ofPattern 2 prisms were reported to pre-dominate in many cercopithecid spe-cies, though Patterns 1 and 3 also wereobserved.25,35,85,86 But, as is the casefor other primate groups, there hasbeen disagreement about the interpre-tation of prism patterns (see Table 1).Shellis and Poole41 argued that despiteits superficial resemblance to ‘‘Type II’’(Pattern 2), Macaca enamel lacked in-terrow sheets. Shellis82 later concludedthat all cercopithecines are character-ized by Pattern 3 prisms. Similarly,Gantt’s81 descriptions of Pattern 2Aprisms in Macaca and Papio enamelwere later interpreted as Pattern 3.35

Detailed studies of colobine and cerco-pithecine prism patterns83,84 illustrateboth Pattern 2 and Pattern 3 arrange-ments, but predominantly the latterfor cercopithecines and colobines. Dos-tal83,84 pointed out that most cercopi-thecine prisms are slender and elon-gate with shapes varying from pointedto parallel-sided with flat, truncatedtops, whereas colobine prisms charac-teristically are broader and morerounded. (The broad, rounded prismsof Papio hamadryas and the the paral-lel-sided truncated prisms of Nasalisare exceptions). These interesting ob-servations have never been followedup, though metrical studies suggestthat both Pattern 2 and Pattern 3

While prism type doesnot appear tocharacterize subgroupsof platyrrhines, variationin enamel types andschmelzmuster may bedistributed in a moremeaningful way.


prisms of cercopithecoids are moreelongate than those of hominoids.35

HominoideaRelative to their taxonomic diver-

sity, hominoids have received a dispro-portionate amount of attention withregard to their enamel microstructure.In a way, this is understandable, formuch of what we know about enamelstructure, function, and developmenthas been driven by interest in humandental health.87,88 Evolutionary stud-ies of nonhuman hominoid enamel,however, have largely focused on whatprism patterns can tell us about homi-noid phylogeny.

Although all three major prism pat-terns can be recognized in some re-gions of human teeth,35 there is gen-eral agreement that human enamel ischaracterized by the predominance ofkey-hole-shaped prisms (Pattern 3B)and well-developed Hunter-Schregerbands.89 Characterization of apeenamel has been more controversial.Shellis and Poole41 thought that Go-rilla enamel, but not that of otherhominoids, consisted entirely of Pat-tern 1 prisms. Gantt90,91 initially ar-gued that recent and fossil hominidshad Pattern 3 prisms and thus weredistinguished from other hominoidswith Pattern 1 prisms. But, respond-ing to concerns about preparationmethods and adequacy of sam-pling,9,85,92 he modified his conclu-sions and reported that extant pongidshave Pattern 3A prisms, in contrast tothe predominantly keyhole-shaped Pat-tern 3B hominid enamel prisms.93,94

Others, however, maintain that thereis no evidence for a hominid-pongiddichotomy based on prism patterns.35

In any event, the preponderance ofevidence indicates that all hominoidshave a combination of prism patterns,dominated by Pattern 3.

The schmelzmuster of most homi-noids is also similar, consisting of ra-dial enamel and Hunter-Schregerbands, usually with an outer prismlesslayer of varying thickness; the reportthat Gorilla lacks Hunter-Schregerbands41 is incorrect.9,85,86 Though somevariations in the pattern and extent ofHunter-Schreger bands have beennoted, the functional or phyletic sig-nificance of this variation has yet to beinvestigated.

Enamel Structure of FossilPrimates

Because of its high mineral content,the structure of enamel is preservedvirtually unaltered in the fossil record.Therefore, we can directly observe theenamel of many of the most primitiveprimates and even compare the enamelstructures of problematic fossil formsto those of living species. The availabil-ity of fossil material not only affordsinsights into the evolution of primateenamel, but provides samples withwhich to test functional hypothesesgenerated from the study of livingspecies.

Plesiadapiformes, though no longerclassified in Primates, are almost cer-tainly close to the origins of the order.The earliest plesiadapiform, Purgato-rius, has radial enamel with some

Pattern 1 but mostly Pattern 3 prisms,all widely separated by interprismaticenamel.39,95 Later plesiadapiforms (mi-crosyopoids, paromomyids, carpoles-tids, plesiadapids, and picrodontids)also have both Pattern 1 and, predomi-nantly, Pattern 3 prisms.39 Most plesi-adapiforms have radial enamel, butHunter-Schreger bands have been re-ported in molars of some species39 andin the enlarged incisors of at least oneplesiadapid, Plesiadapis tricuspidens93

(Table 2). Interestingly, these speciesprobably were smaller96 than mostliving mammals with Hunter-Schregerbands.

The enamel of the earliest primates,like that of the plesiadapiforms, con-sists of Pattern 3 prisms, with Pattern1 prisms largely confined to surfaceenamel (Table 2). The North American

adapids have been studied in detail. Inthese species, including Cantius ral-stoni, one of the earliest and mostprimitive primates, the prisms are rela-tively small and widely separated byinterprismatic enamel.39,45 The Euro-pean adapids also have mostly Pattern3 prisms, though at least one species,Adapis parisiensis, appears to have Pat-tern 1 prisms throughout its enamel.39

The early Eocene North Americanomomyids also show a combination ofPattern 1 and, most commonly, Pat-tern 3 prism cross sections. The distri-bution of enamel types among theEocene adapids and omomyids largelyfollows the predictions of the body-size dependent model (Table 2). Thesmallest adapids and omomyids haveradial enamel and prismless enamel,but Hunter-Schreger bands occur inthe larger species.

Nothing is known of enamel struc-ture of the earliest anthropoid fossils,from the middle Eocene of China, butthe more diverse and better knownradiation of late Eocene to early Oli-gocene primates from the Fayum ofEgypt has been studied in detail (Table2). Like the North American and Euro-pean adapids and omomyids, all fivespecies studied thus far (Apidium phio-mense, A. moustafai, Parapithecusgrangeri, Aegyptopithecus zeuxis, andCatopithecus browni) have mostly arc-shaped prisms with fairly broad inter-prismatic regions (comparable to Pat-tern 3A), but also some Pattern 1prisms, especially in the outerenamel.97,98 The larger species haveextensive Hunter-Schreger bands,whereas the smaller ones have onlyradial and prismless enamel. Apidiumphiomense represents a unique combi-nation of characters, at least amongprimates. Its molar enamel is amongthe thickest, relative to body size, ofany living or fossil primate, but, unlikeother primates with thick enamel, itlacks Hunter-Schreger bands. This ex-tremely thick radial enamel, along withanalysis of microwear patterns, hasled researchers to reconsider ideasabout the diet of this enigmatic anthro-poid.96

Enamel has been described for afew fossil members of later primateradiations, mostly with reference tophyletic relationships. Thus, the pre-dominance of Pattern 3 prisms in theproblematic Miocene Oreopithecus

The availability of fossilmaterial not only affordsinsights into the evolutionof primate enamel, butprovides samples withwhich to test functionalhypotheses generatedfrom the study of livingspecies.


may support its affinity with Hominoi-dea rather than Cercopithecoidea,though only if the absence of Pattern 2prisms is derived.35 Prism morphologyof the Miocene cercopithecoid Mesopi-thecus supports its systematic positionwithin the Colobinae.83 In the case ofthe Miocene hominoid Otavipithecus,confocal microscopy revealed a pre-dominance of Pattern 1 prisms,99 lead-ing to speculation that Otavipithecusmight be more closely related to ex-tant African apes than to otherMiocene hominoids. However, restric-tion of the sample to surface enameland the contradictory reports of thepredominant prism patterns of extantapes (see Table 1) make this conclu-sion problematic. Gantt93,94,100 arguedthat predominance of Pattern 3Aprisms, which he found in the Miocenehominoids Proconsul, ‘‘Ramapithecus,’’Sivapithecus, and Gigantopithecus, andin the living apes, are primitive forhominoids, whereas the fossilhominids Australopithecus and Homoerectus share with humans the derivedkeyhole-shaped Pattern 3B prisms.Others, however, dispute this distinc-tion.35 More recent investigations ofthe enamel structure of fossil homi-noids have moved away from phyloge-netic interpretations of prism pat-terns. These studies have focusedalmost exclusively on the functionalimplications of enamel thickness andHunter-Schreger bands101 and espe-cially on the ontogenetic evidence forthe evolution of dental developmentpatterns contained in microstructuralfeatures such as incremental lines (seeBox 1).

FUNCTIONAL ANDPHYLOGENETIC SIGNAL IN

PRIMATE ENAMELAt the beginning of this paper, we

posed two questions: Is the structureof primate enamel adaptive? What, ifanything, does enamel structure tellus about primate phylogeny? The evo-lution of primate enamel, like that ofother morphological systems, has beendriven by a combination of develop-mental constraints, functional influ-ences, and phylogenetic history. Much,though not all of the developmentalbasis of enamel structure, is well un-derstood,1,8,22 but to decipher the func-tional and phylogenetic signals, a grasp

of two key elements is essential. Thefirst is an understanding of the levelsof structural complexity at which bothinfluences are played out. The secondis an understanding of the levels of thetaxonomic hierarchy where they aremanifest.

It is clear that components of pri-mate enamel structure representingseveral levels of structural complexityhave important functional associa-tions. At the smallest scale, crystallitesintersect functionally distinct toothsurfaces at angles that enhance resis-tance to wear. At the level of the prism,prismatic enamel inhibits crack prop-agation, but only under relatively lowloads. At the level of the enamel type,

radial and prismless enamel optimizeabrasion resistance (a larger-scale ex-pression of the crystallite-level phe-nomenon). Also at the level of theenamel type, Hunter-Schreger bandsfunction as a crack-stopping mecha-nism in cases where occlusal loads arehigh. This typically occurs in large-bodied species weighing more thanabout 2,000 gm, but among primatesHunter-Schreger bands also are foundin some small exudate feeders, per-haps because they experience high oc-clusal loads during bark gouging. Atthe level of the schmelzmuster, enameltypes combine to optimize functionand protection for the whole tooth.

Change in either total enamel thick-ness or in the thickness of one or moreenamel types, also acts at the schmelz-muster level to protect teeth. What issignificant from a phylogenetic per-spective, however, is that none of thefunctionally important enamel fea-tures at any structural level is uniqueto primates.

In all mammals with prismaticenamel, crystallites and prisms inter-sect different tooth surfaces at differ-ent angles. This is, at least in part, adirect result of the way in whichenamel is deposited during develop-ment. Also, among mammals in gen-eral Hunter-Schreger are a commonsolution to the problem of dealingwith increased occlusal loads. Like-wise, the schmelzmuster variation inproportions of enamel types andenamel thickness that we see in pri-mates duplicates mechanical solu-tions in other mammals. In otherwords, primate enamel is not charac-terized by a key innovation that pro-moted the evolution of the variety oftooth morphologies within the order.Indeed, the earliest primates had rela-tively simple enamel composed mostlyof Pattern 3 and some Pattern 1 prisms,both with fairly broad interprismaticregions, and a schmelzmuster consist-ing of radial enamel with a prismlesssurface layer. This enamel structure istypical of many primitive mam-mals.15,102 We see similar enamel pat-terns in many extant lipotyphlans andchiropterans,40 as well as many livingsmall primates. Early in primate evolu-tion, Hunter-Schreger bands appear inconjunction with increased body size.Again, this echoes evolutionary pat-terns in other mammals. Essentially,the evolution of primate enamel, likethat of the enamel of many othermammals, was a process of using exist-ing structures and developmental path-ways and, perhaps, making relativelyminor adjustments to them.

The distribution across Mammaliaof the functionally important featuresalso expressed in primate enamel, aswell as the variations in these features,suggest that they have been acquiredindependently many times, are rela-tively plastic, and thus are most likelyto carry phylogenetic informationacross limited clades and at low levelsof the taxonomic hierarchy. There issome evidence that this is the case for

. . . to decipher thefunctional andphylogenetic signals, agrasp of two keyelements is essential. Thefirst is an understandingof the levels of structuralcomplexity at whichboth influences areplayed out. The secondis an understanding ofthe levels of thetaxonomic hierarchywhere they are manifest.


primates. For example, Hunter-Schreger bands have clearly been ac-quired many times within primates.The different patterns of decussationthat seem to characterize pitheciines,cebines, and atelines may reflect theirindependent origin within each groupand, at that taxonomic level, carrysome phylogenetic signal. On a stilllower taxonomic level, enamel thick-ness varies among species within gen-era in conjunction with dietary adapta-tions for soft or hard food items77 and,perhaps, other dietary specializa-tions.98

In sum, the answer to the question,‘‘Is the structure of primate enameladaptive?’’ is a qualified ‘‘yes.’’ Someelements of primate enamel structureserve specific functional roles, but arebest understood as exaptations, ormorphological innovations that havebeen co-opted for some new function.This is certainly the case for the small-scale structural variation in crystallite-to-surface orientation. This phenom-enon, common among mammals, islargely a by-product of the way thatmammalian enamel develops, but it isa functionally plastic feature that hasbeen co-opted in some species to en-hance differential wear and in othersto resist wear. Other elements ofenamel structure, including larger-scale features such as Hunter-Schregerbands and variation in both enamelthickness and schmelzmuster, requiresignificant alterations of primitive de-velopmental pathways, and probablyoriginated independently in differentprimate taxa in association with evolu-tion of special adaptations in thoseclades.

To fully answer the question of whatenamel structure tells us about pri-mate phylogeny we need to considervariation in prism cross-sectional mor-phology, the only enamel feature thathas been investigated from a phyloge-netic perspective but that is withoutclear functional correlates. Until themid 1980s, most phylogenetic studiesof primate enamel focused on placingprisms into discrete categories basedon qualitative, descriptive criteria.More recent work has incorporatedquantitative criteria. The impetus forthis shift came from three differentarenas. The first was the developmentof efficient, high-magnification elec-tron and confocal microscopy, which

increased potential sample sizes andmade statistical assessments of varia-tion feasible. The second was the dem-onstration that enamel prisms ofclosely related species (domestic sheepand goats) could be distinguished bymetrical criteria so long as compari-sons were made at homologous depthswithin the enamel.103 The third devel-opment, the most intriguing to system-atists, was the successful use of prismsize, shape, and spacing to test phylo-genetic hypotheses of relationshipsamong multituberculates, an extinctorder of mammals.104,105 Unfortu-nately, the attempts to use either quali-tative or quantitative descriptions ofprisms to answer questions about pri-mate phylogeny were stymied by their

inherent variation and homoplasy.Rather than providing clear solutionsto phylogenetic questions, these stud-ies have primarily served to point outthe practical limitations of enamelprism data at a variety of taxonomiclevels.

One of the first modern attempts touse prism patterns in primate system-atics focused on hominoid phylogenyand evolution, a family-level taxo-nomic problem. The initial report thathominids could be distinguished fromother hominoids on the basis of prismmorphology90 generated considerablepolemics.9,35,85,92,94 But it is now clearthat enamel prisms and patterns donot distinguish among hominoids, liv-

ing or extinct; they are not phylogeneti-cally informative. Attempts to applyqualitative prism morphology criteriato questions of primate subordinalsystematics were equally unsuccess-ful. Initial reports that lemurs werecharacterized by Pattern 1 prisms ledsome to speculate that the presence ofprism patterns other than Pattern 1might characterize haplorhines (an-thropoids plus tarsiers).9,35 In fact, Pat-tern 3 prisms are common among allprimates, both haplorhines and strep-sirhines,32,37–39,41,82,106 and prism pat-terns simply do not contribute mean-ingful information to the debateregarding primate subordinal relation-ships.79

These studies, however, stimulatedcareful assessments of preparation andsampling techniques, including appro-priate etching regimes for differenttypes of specimens,85,86,107 control forsectioning artifacts,37,38,108 and controlfor enamel depth to reduce the prob-lem of within-sample variation.101 As aresult, most recent studies have em-ployed a rigorous protocol and pro-duced thoroughly documented, repli-cable results.37,39,40,49,79 But even whenmethodological factors are taken intoaccount, attempts to distinguish spe-cies, genera, and subfamilies on thebasis of prism patterns have beenunsuccessful. In lemurids,38 callitrichi-nes,49 and cercopithecids,83 prism mor-phology failed to distinguish unequivo-cally among taxa.

Even at the ordinal level, prism mor-phology has proved an intractablemeans of distinguishing primate phy-letic relationships. Dumont106 investi-gated the utility of enamel prism size,shape, and spacing for testing themonophyly of Archonta (primates, treeshrews, bats, and flying lemurs). Thisstudy, implementing careful controlsfor preparation artifacts and intra-and interindividual variation, codedprism metrical and shape variables for17 archontan and insectivoran (theoutgroup) species into character states,but the results were equivocal. Parsi-mony analyses using species as termi-nal taxa demonstrated that homoplasywas rampant; orders and even familieswere consistently rendered paraphy-letic. In other words, as for lower-leveltaxonomic analyses, enamel prismsare phylogenetically uninformative.

The failure of these studies of prism

. . . enamel prismsand patterns do notdistinguish amonghominoids, living orextinct; they are notphylogeneticallyinformative. Attempts toapply qualitative prismmorphology criteria toquestions of primatesubordinal systematicswere equallyunsuccessful.


patterns to elucidate phylogenetic sig-nal in primate enamel highlights acritical flaw in their approach: ignor-ing other levels of structural complex-ity (crystallite, enamel type, and schm-elzmuster) obscures the fundamentalnature of prism variation, and thusobscures enamel’s phylogenetic signal.In other words, one ‘‘can’t see theforest for the trees’’: youcannotcharacter-ize enamel solely on the basis of prismcross-sections any more than you cancharacterize a forest on the basis of theshapes of tree trunks. Investigations ofenamel structure in other groups ofmammals that have adopted a more ho-listic concept of enamel have producedsome intriguing functional and phyloge-netic insights.16,40,51,109,110 It is not yetclear whether this will be the case forprimates.

After almost 80 years of primateenamel studies, we conclude that thequestion of what, if anything, enamelstructure tells us about primate phylog-eny, has yet to be answered fully. Ourcurrent understanding is largely lim-ited to what the functional features ofenamel can tell us about primate evo-lution; to date, phylogenetic assess-ments of the nonfunctional enamelcharacters, prism shape and size, havenot been successful. While the techni-cal constraints on preparing and view-ing enamel and the presence of varia-tion in its structure contribute to thisfailure, it primarily reflects the lack ofa clear phylogenetic signal in primateprism pattern and size. We suspectthat phylogenetic analyses of primateenamel, and perhaps even most mam-malian enamel, must take into ac-count the functional influences at thelevel of the enamel type and schmelz-muster if they are to meet with suc-cess. It may be that enamel will onlybe useful in defining clades that arecharacterized by common functionaladaptations. In the case of enamel,functional and phylogenetic signal areso strongly associated that perhapsthey are one and the same.

ACKNOWLEDGMENTSWe thank John Fleagle for the oppor-

tunity to present our ideas about pri-mate enamel structure, and our manycolleagues in England, Germany, In-dia, Japan, and the United States who,over the past 15 years, have shared

their insights about this complex tis-sue. We thank the Duke UniversityPrimate Center and the SmithsonianInstitution for use of specimens intheir care. This work has been sup-ported by NSF grants BNS 9020788and SBR 9713148 to MCM, and IBN9507488 to ERD.

REFERENCES1 Boyde A. 1976. Amelogenesis and the structureof enamel. In: Cohen B, Kramer I, editors. Scien-tific foundations of dentistry. London: Wm. Heine-mann Medical Books. p 335–352.2 Waters NE. 1980. Some mechanical and physi-cal properties of teeth. In: Vincent JF, Curry JD,editors. The mechanical properties of biologicalmaterials. Cambridge: Cambridge UniversityPress. p 99–135.3 Elliott JC. 1997. Structure, crystal chemistryand density of enamel apatites. In: Chadwick DJ,Cardew G, editors. Dental enamel. Ciba founda-tion symposium 205. New York, John Wiley &Sons. p 54–72.4 Robinson C, Kirkham J, Brookes SJ, BonassWA, Shore RC. 1995. The chemistry of enameldevelopment. Int J Dev Biol 39:145–152.5 Deutsch D, Palmon A, Fisher LW, Kolodny N,Termine JD, Young MF. 1991. Sequencing ofbovine enamelin (‘‘Tuftelin’’): a novel acidicenamel protein. J Biol Chem 266:16021–16028.6 Deutsch D, Catalono-Sherman J, Dafni L, DavidS, Palmon A. 1995. Enamel matrix proteins andameloblast biology. Connect Tissue Res 32:97–107.7 Deutsch D, Dafni L, Palmon A, Hekmati M,Young MF, Fisher LW. 1997. Tuftelin: enamelinmineralization and amelogenesis imperfecta. InChadwick DJ, Cardew G, editors. Dental enamel.ciba foundation symposium 205. New York: JohnWiley & Sons. p 135–155.8 Moss-Salentijn L, Moss ML, Sheng-Tien YuanM. 1997. The ontogeny of mammalian enamel.In: Koenigswald Wv, Sander PM, editors. Toothenamel microstructure. Rotterdam: Balkema. p5–30.9 Boyde A, Martin LB. 1984. The microstructureof primate dental enamel. In: Chivers DJ, WoodBA, Bilsborough A, editors. Food acquisition andprocessing in primates. New York: Plenum Press.p 341–367.10 Carlstrom D, Glas J-E, Angmar B. 1963. Stud-ies on the ultrastructure of dental enamel, V. thestate of water in human enamel. J Ultrastruct Res8:24–29.11 Poole DFG, Newman HN, Dibdin GH. 1981.Structure and porosity of human cervical enamelstudied by polarizing microscopy and transmis-sion electron microscopy. Arch Oral Biol 26:977–982.12 Kirkham J, Robinson C, Weatherell JA, Rich-ards A, Fejerskov O, Josephsen K. 1988. Matura-tion in developing permanent porcine enamel. JDent Res 67:1146–1160.13 Koenigswald Wv, Sander PM. 1997. Glossaryof terms used for enamel microstructures. In:Koenigswald Wv, Sander PM, editors. Toothenamel microstructure. Rotterdam: Balkema. p267–280.14 Carlson SJ. 1990. Vertebrate dental structures.In: Carter JG, editor. Skeletal biomineralization:patterns, processes and evolutionary trends, vol I.New York: Van Nostrand Reinhold. p 531–556.15 Koenigswald Wv, Clemens WA. 1992. Levels ofcomplexity in the microstructure of mammalian

enamel and their application in studies of system-atics. Scan Microsc 6:195–218.

16 Koenigswald Wv, Martin T, Pfretzschner HU.1992. Phylogenetic interpretation of enamel struc-tures in mammalian teeth: possibilities and prob-lems. In Szalay FS, Novacek MJ, McKenna MC,editors. Mammal phylogeny: placentals. NewYork: Springer-Verlag. p 303–314.

17 Sander PM. 1997. Non-mammalian synapsidenamel and the origin of mammalian enamelprisms: the bottom-up perspective. In: Koenig-swald Wv, Sander PM, editors. Tooth enamelmicrostructure. Rotterdam: Balkema. p 40–62.

18 Smith MM. 1989. Distribution and variationin enamel structure in the oral teeth of sarcopte-rygians: its significance for the evolution of aprotoprismatic enamel. Hist Biol 3:97–126.

19 Daculsi G, Menanteau J, Kerebel LM, Mitre D.1984. Enamel crystals: size, shape, length andgrowing process: high resolution TEM and bio-chemical study. In: Fearnhead RW, Suga S, edi-tors. Tooth enamel IV. Amsterdam: Elsevier. p14–18.

20 Warshawsky H. 1987. External shape of enamelcrystals. Scan Microsc 1:1913–1923.

21 Apkarian RP. 1990. High resolution SE-I SEMstudy of enamel crystal morphology. J ElectronMicrosc Technol 14:70–78.

22 Sasaki T, Goldberg M, Takuma S, Garant PR.1990. Cell biology of tooth enamel formation.functional electron microscopic monographs.Basel: Karger.

23 Boyde A, Fortelius M. 1986. Development,structure and function of rhinoceros enamel.Zool J Linnaeus Soc 87:181–214.

24 Cooper JS, Poole DFG. 1973. The dentitionand dental tissues of the agamid lizard, Uromas-tyx. J Zool, London 169:85–100.

25 Boyde A. 1964. The structure and develop-ment of mammalian dental enamel. Ph.D. disser-tation, University College.

26 Boyde A. 1965. The structure of developingmammalian dental enamel. In: Stack MV, Fearn-head RW, editors. Tooth enamel. Bristol: Wrightand Sons. p 163–167.

27 Boyde A. 1971. Comparative histology of mam-malian teeth. In: Dahlberg A, editor. Dental mor-phology and evolution. Chicago: University ofChicago Press. p 81–93.

28 Tomes CS. 1906. On the minute structure ofthe teeth of creodonts, with especial reference totheir suggested resemblance to marsupials. ProcZool Soc, London i:45–58.

29 Carter JT. 1922. On the structure of the enamelin the Primates and some other mammals. ProcZool Soc, London 1922:599–608.

30 Korvenkontio VA. 1934. Mikroskopische unter-suchungen an nagerincisiven unter hinweis aufdie schmelzstruktur der backenzahn. Ann ZoolSoc Zool-Bot Fennicae Vanamo 2:1–274.

31 Shobusawa M. 1952. Vergleichende untersu-chungen uber die form der schmelzprismen derSaugetiere. Okaj Folia Anat Jpn 24:371–392.

32 Kawai N. 1955. Comparative anatomy of thebands of Schreger. Okaj Fol Anat Jap 27:115–131.

33 Moss ML. 1969. Evolution of mammaliandental enamel. Am Museum Novitates 2360:1–39.

34 Kozawa S. 1984. Development and the evolu-tion of mammalian enamel structure. In: Fearn-head RW, Suga S, editors. Tooth enamel IV,Amsterdam: Elsevier. p 437–441.

35 Martin LB, Boyde A, Grine FF. 1988. Enamelstructure in primates: a review of scanning elec-tron microscope studies. Scan Microsc 2:1503–1526.

36 Boyde A, Martin L. 1988. Tandem-scanningreflected-light microscopy of primate enamel.Scan Microsc 1:1935–1948.


37 Maas MC. 1993. Enamel microstructure andmolar wear in the greater galago, Otolemur crassi-caudatus (Mammalia, Primates). Am J Phys An-thropol 92:217–234.

38 Maas MC. 1994. Enamel microstructure inLemuridae (Mammalia, Primates): assessment ofvariability. Am J Phys Anthropol 95:221–242.

39 Dumont ER. 1995. Enamel prism morphologyin molar teeth of small eutherian mammals. ScanMicrosc 10:349–370.

40 Koenigswald Wv. 1997. Brief survey of enameldiversity at the schmelzmuster level in Cenozoicplacental mammals. In: Koenigswald Wv, SanderPM, editors. Tooth enamel microstructure. Rotter-dam: Balkema. p 137–161.

41 Shellis RP, Poole DFG. 1977. The calcifieddental tissues of primates. In: Lavelle CLB, ShellisRP, Poole DFG, editors. Evolutionary changes tothe primate skull and dentition. Springfield:Charles C. Thomas. p 197–279.

42 Maas MC. 1986. Function and variation ofenamel prism decussation in ceboid primates.Am J Phys Anthropol 69:233–234.

43 Boyde A. 1969. Electron microscopic observa-tions relating to the nature and development ofprism decussation in mammalian dental enamel.Bull Group Int Rech Sci Stomatol 12:151–207.

44 Rensberger JM. 1993. Adaptation of enamelmicrostructure to differences in stress intensityin the Eocene perissodactyl Hyracotherium. In:Kobayashi I, Mutvei H, Sahni A, editors. Struc-ture, formation and evolution of fossil hard tis-sues. Tokyo: Tokai University Press. p 131–145.

45 Maas MC, O’Leary MO. 1996. Evolution ofmolar enamel microstructure in North AmericanNotharctidae (Primates). J Human Evol 31:293–310.

46 Nogami Y, Natori M. 1986. Fine structure ofthe dental enamel in the Family Callitrichidae(Ceboidea, Primates). Primates 27:245–258.

47 Koenigswald Wv, Rensberger JM, Pfretzs-chner HU. 1987. Changes in the tooth enamel ofearly Paleocene mammals allowing increased dietdiversity. Nature 328:150–152.

48 Maas MC, Thewissen JGM. 1995. Enamelmicrostructure of Pakicetus (Mammalia: Archaeo-ceti). J Paleontol 69:1154–1162.

49 Propst K. 1995. Enamel microstructure, func-tion, and phylogeny in marmosets and tamarins(Callitrichidae, Primates). Ph.D. Dissertation, In-diana University. Bloomington.

50 Maas MC. 1997. Enamel microstructure innotoungulates. In: Kay RF, Madden RH, CifelliRL, Flynn JJ, editors. Vertebrate paleontology inthe neotropics: the Miocene fauna of La Venta,Colombia. Washington, DC: Smithsonian Institu-tion Press. p 319–334.

51 Stefan C. 1997. Differentiation in Hunter-Schreger bands of carnivores. In: KoenigswaldWv, Sander PM, editors. Tooth enamel microstruc-ture. Rotterdam: Balkema. p 123–133.

52 Rensberger JM, Koenigswald Wv. 1980. Func-tional and phylogenetic interpretation of enamelmicrostructure in rhinoceroses. Paleobiology6:477–495.

53 Walker A. 1984. Mechanisms of honing in themale baboon canine. Am J Phys Anthropol 65:47–60.

54 Nogami Y, Yoneda M. 1983. Structural pat-terns of enamel in the Superfamily Ceboidea.Primates 24:567–575.

55 Gordon KD. 1982. A study of microwear onchimpanzee molars: implications for dental mi-crowear analysis. Am J Phys Anthropol 59:195–215.

56 Boyde A. 1984. Dependence of rate of physicalerosion on orientation and density in mineralisedtissues. Anat Embryol 170:57–62.

57 Maas MC. 1991. Enamel structure and mi-crowear: an experimental study of the responseof enamel to shearing force. Am J Phys Anthropol85:31–49.

58 Maas MC. 1994. A scanning electron micro-scope study of in vitro abrasion of mammaliantooth enamel under compressive loads. ArchsOral Biol 39:1–11.

59 Young WG, McGowan M, Daley TJ. 1987.Tooth enamel structure in the koala Phascolarc-tos cincereus: some functional interpretations.Scan Microsc 1:1925–1934.

60 Young WG, Douglas WB, Robson SK. 1990.Differential wear resistance as a function of toothenamel structure in the koala Phascolarctos ci-nereus. In: Lee AK, Handasyde KA, Sanson GD,editors. Biology of the koala. Sydney: Surrey,Beatty and Sons. p 45–50.

61 Ferreira JM, Phakey PP, Palamara J, Rach-inger WA. 1990. Electron microscopic investiga-tion relating the occlusal morphology to theunderlying enamel structure of molar teeth of theWombat (Vombatus ursinus). J Morphol 200:141–149.

62 Teaford MF. 1988. A review of dental mi-crowear and diet in modern mammals. ScanMicrosc 2:1149–1166.

63 Teaford MF. 1994. Dental microwear and den-tal function. Evol Anthropol 3:17–30.

64 Walker A, Hoeck HN, Perez L. 1978. Mi-crowear of mammalian teeth as an indicator ofdiet. Science 201:908–910.

65 Stern D, Crompton AW, Skobe Z. 1989. Enamelultrastructure and masticatory function in mo-lars of the American opossum, Didelphis virgin-iana. Zool J Linnaeus Soc 95:311–334.

66 Rasmussen ST, Patchin RE, Scott DB, HeuerAH. 1976. Fracture properties of human enameland dentine. J Dent Res 55:154–164.

67 Boyde A. 1976. Enamel structure and cavitymargins. Operative Dent 1:13–28.

68 Hassan R, Caputo AA, Bunshah RF. 1981.Fracture toughness of human enamel. J Dent Res60:820–827.

69 Pfretzschner H-U. 1986. Structural reinforce-ment and crack propagation in enamel. In: Rus-sell DE, Santoro J-P, Sigogneau-Russell D, edi-tors. Teeth revisited: proceedings of the VIIthinternational symposium on dental morphology,Paris, 1986. Paris: Mem Museum Nat Hist NatParis (ser C). p 133–143.

70 Pfretzschner HU. 1992. Enamel microstruc-ture and hypsodonty in large mammals. In: SmithP, Tchernov E, editors. Structure, function andevolution of teeth. London: Freund PublishingHouse, Ltd. p 147–162.

71 Rensberger JM. 1995. Determination ofstresses in mammalian dental enamel and theirrelevance to the interpretation of feeding behav-iors in extinct taxa. In: Thomason J, editor.Functional morphology in vertebrate paleontol-ogy. Cambridge: Cambridge University Press. p151–172.

72 Koenigswald Wv, Pfretzschner HU. 1987.Hunter-Schreger bands in the enamel of mamma-lian teeth. Arrangement, orientation of the prisms.Zoomorphology 106:329–338.

73 Andrews P, Martin L. 1991. Hominoid dietaryevolution. Philos Trans R Soc London B 334:199–209.

74 Jolly CJ. 1970. The seedeaters: a new model ofhominoid differentiation based on a baboon anal-ogy. Man 5:5–26.

75 Kay RF. 1981. The nut-crackers—a new theoryof the adaptations of the Ramapithecinae. Am JPhys Anthropol 55:141–151.

76 Martin L. 1985. Significance of enamel thick-ness in hominoid evolution. Nature 314:260–263.

77 Dumont ER. 1995. Enamel thickness and di-etary adaptations among extant primates andchiropterans. J Mammology 76:1127–1136.

78 Teaford MF, Maas MC, Simons EL. 1996.Dental microwear and microstructure of earlyOligocene primates from the Fayum, Egypt: impli-cations for diet. Am J Phys Anthropol 101:527–544.

79 Dumont ER. 1996. Variation in quantitativemeasures of enamel prisms from different spe-cies as assessed using confocal microscopy. ArchOral Biol 41:1053–1063.

80 Shellis RP, Poole DFG. 1979. The arrangementof prisms in the enamel of the anterior teeth ofthe aye-aye. Scan Electron Microsc 11:497–506.

81 Gantt DG. 1980. Implications of enamel prismpatterns for the origin of the New World mon-keys. In: Ciochon RL, Chiarelli AB, editors. Evolu-tionary biology of the New World monkeys andcontinental drift. New York: Plenum Press. p201–217.

82 Shellis RP. 1984. Inter-relationships betweengrowth and structure of enamel. In: FearnheadRW, Suga S, editors. Tooth enamel IV. Amster-dam: Elsevier. p 467–471.

83 Dostal A, Zapfe H. 1986. Zahnschmelzprimenvon Mesopithecus penticulus Wagner, 1839, imvergleich mit rezenten cercopitheciden (Pri-mates: Cercopithecidae). Folia Primatol 46:235–251.

84 Dostal A. 1989. Dental enamel prism patternsof Old World monkeys and man. Fortschr Zool35:251–255.

85 Boyde A, Martin L. 1982. Enamel microstruc-ture determination in hominoid and cercopithe-coid primates. Anat Embryol 165:193–212.

86 Boyde A, Martin L. 1984. A non-destructivesurvey of prism packing patterns in primateenamels. In: Fearnhead RW, Suga S, editors.Tooth enamel IV. Amsterdam: Elsevier. p 417–421.

87 Fearnhead RW, Suga S. 1984. Tooth EnamelIV. Amsterdam: Elsevier.

88 Chadwick D, Cardew G, editors. 1997. Dentalenamel. Ciba Foundation symposium. 205. NewYork: John Wiley & Sons.

89 Osborn JW. 1990. A 3-dimensional model todescribe the relation between prism directions,parazones and diazones, and the Hunter-Schregerbands in human tooth enamel. Arch Oral Biol35:869–878.

90 Gantt DG, Pilbeam DR, Steward GP. 1977.Hominoid enamel prism patterns. Science 198:1155–1157.

91 Gantt DG. 1979. A method of interpretingenamel prism patterns. Scan Electron Microsc1979:975–981.

92 Vrba ES, Grine FE. 1978. Australopithecineenamel prism patterns. Science 202:890–892.

93 Gantt DG. 1983. The enamel of Neogene homi-noids: structural and phyletic implications. In:Ciochon RL, Corruccini RS, editors. New interpre-tations of ape and human ancestry. New York:Plenum Press. p 249–298.

94 Gantt DG. 1986. Enamel thickness and ultra-structure in hominoids: with reference to form,function, and phylogeny. In: Swindler DS, ErwinJ, editors. Comparative primate biology, Vol 1:systematics, evolution, and anatomy. New York:Alan R. Liss. p 453–475.

95 Clemens WA, Koenigswald Wv. 1991. Purgato-rius, plesiadapiforms, and evolution of HunterSchreger bands. J Vertebrate Paleontol 11:23.

96 Fleagle JG. 1998. Primate adaptation and evo-lution, 2nd ed. New York: Academic Press.

97 Maas MC, Simons EL. 1996. Enamel micro-structure of early anthropoids from the Fayum ofAfrica. Am J Phys Anthropol 20(suppl):138.


98 Teaford MF, Maas MC, Simons EL. 1996.Dental microwear and microstructure of earlyOligocene primates from the Fayum, Egypt: impli-cations for diet. Am J Phys Anthropol 101:527–544.

99 Conroy GC, Lichtman JW, Martin LB. 1995.Some observations on enamel thickness andenamel prism packing in the Miocene hominoidOtavipithecus namibiensis. Am J Phys Anthropol98:595–600.

100 Gantt DG, Cring FD. 1981. Enamel ultrastruc-ture and its implications to paleontology. ScanElectron Microsc 1981:595–602.

101 Macho GA, Thackeray JF. 1992. Computertomography and enamel thickness of maxillarymolars of Plio-Pleistocene hominids from Sterk-fontein, Swartkrans and Kromdraai (South Af-rica): an exploratory study. Am J Phys Anthropol89:133–143.

102 Crompton AW, Wood CB, Stern DN. 1994.Differential wear of enamel: a mechanism formaintaining sharp cutting edges. In: Bels VL,Chardon M, Vandewalle P, editors. Advances incomparative and environmental physiology, vol18. Biomechanics of feeding in vertebrates. NewYork: Springer-Verlag. p 321–346.

103 Grine FE, Krause DW, Fosse G, Jungers WL.1987. Analysis of individual intraspecific andinterspecific variability in quantitative param-eters of caprine tooth enamel structure. ActaOdontol Scand 45:1–23.104 Carlson S, Krause D. 1985. Enamel ultrastruc-ture of multituberculate mammals: an investiga-tion of variability. Contrib Museum PaleontolUniv Michigan 27:1–50.105 Krause DW, Carlson SJ. 1987. Prismaticenamel in multituberculate mammals: tests ofhomology and polarity. J Mamm 68:755–765.106 Dumont ER. 1993. Functional and phyleticfeatures of mammalian dental enamel: a key toprimate higher-level relationships. Ph.D. Disserta-tion, State University of New York at Stony Brook.107 Grine FE. 1986. Effects of different etchingagents on bovid tooth enamel. South Afr J Sci82:265–270.108 Dumont ER. 1995. The effects of sectioningangle on measurements of enamel prism size andspacing. Arch Oral Biol 40:959–966.109 Martin T. 1993. Early rodent incisor enamelevolution: phylogenetic implications. J Mamma-lian Evol 1:227–254.

110 Pfretzschner HU. 1993. Enamel microstruc-ture in the phylogeny of the Equidae. J VertebratePaleontol 13:342–349.

111 Lumsden AGS. 1987. The neural crest contri-bution to tooth development in the mammalianembryo. In: Maderson PFA, editor. Developmen-tal and evolutionary aspects of the neural crest.New York: John Wiley & Sons. p 261–300.

112 MacKenzie A, Lemming GL, Jowett AK, Fer-guson MWJ, Sharpe PT. 1991. The homeoboxgene Hox 7.1 has specific regional and temporalexpression patterns during early murine craniofa-cial embryogenesis, especially tooth developmentin vivo and in vitro. Development 111:269–285.

113 MacKenzie A, Ferguson MWJ, Sharpe PT.1992. Expression patterns of the homeobox gene,Hox-8, in the mouse embryo suggests a rolespecifying tooth initiation and shape. Develop-ment 115:403–420.

114 Thesloff I, Vaahtokari S, Vainio S. 1990.Molecular changes during determination and dif-ferentiation of the dental mesenchymal cell lin-eage. J Biol Buccale 18:179–188.

115 Boyde A. 1976. Amelogenesis and the struc-ture of enamel. In: Cohen B, Kramer I, editors.Scientific foundations of dentistry. London: Wm.Heinemann Medical Books. p 335–352.

116 Moriwaki Y, Doi Y, Kani T, Aoba T, TakahashiJ, Okazaki M. 1983. Synthesis of enamel-likeapatite at physiological temperature and pH us-ing ion-selective membranes. In: Suga S, editor.Mechanisms of tooth enamel formation. Tokyo:Quintessence. p 239–258.

117 Wakita M, Kobayashi S. 1983. The threedimensional structure of Tomes’ Processes andthe development of the microstructural organiza-tion of tooth enamel. In: Auga A, editor. Mecha-nisms of tooth enamel formation. Tokyo: Quintes-sence. p 65–89.

118 Schour I, Poncher HG. 1937. Rates of apposi-tion of enamel and dentin measured by the effectof acute fluorosis. Am J Dis Child 54:7575–776.

119 Bromage TG. 1991. Enamel incremental pe-riodicity in the pig-tailed macaque: a polychromefluorescent labeling study of dental hard tissues.Am J Phys Anthropol 86:205–214.

120 Dean MC. 1987. Growth layers and incremen-tal markings in hard tissues; a review of theliterature and some preliminary observationsabout enamel structure in Paranthropus boisei. JHuman Evol 16:157–172.

121 Risnes S. 1986. Enamel apposition rate andthe prism periodicity in human teeth. Scand JDent Res 94:394–404.122 Risnes S. 1998. Growth tracks in dentalenamel. J Human Evol 35:331–350.123 FitzGerald CM. 1998. Do enamel microstruc-tures have regular time dependency? Conclusionsfrom the literature and a large-scale study. JHuman Evol 35:371–386.124 Newman HN, Poole DFG. 1977. Observa-tions with scanning and transmission electronmicroscopy on the structure of human surfaceenamel. Arch Oral Biol 19:1135–1143.125 Dumont ER. 1995. Mammalian enamel prismpatterns and enamel deposition rates. Scan Mi-crosc 9:429–442.126 Dean MC, Scandrett AE. 1996. The relationbetween long-period incremental markings indentine and daily cross-striations in enamel inhuman teeth. Arch Oral Biol 41:233–241.127 Fisher DC. 1987. Mastodon procurement bypaleoindians of the Great Lakes region: huntingor scavenging? In: Nitecki MH, Nitecki DV, edi-tors. The evolution of human hunting. New York:Plenum Press. p 309–421.128 Bromage TG, Dean MC. 1985. Re-evaluationof the age at death of immature fossil hominids.Nature 317:525–527.129 Beynon D. 1986. Tooth growth and structurein living and fossil hominoids. In: Cruwys E,Foley RA, editors. Teeth and anthropology: B.A.R.international series 291. p 23–30.130 Beynon AD, Dean MC, Reid DJ. 1991. Onthick and thin enamel in hominoids. Am J PhysAnthropol 86:295–309.131 Beynon AD. 1992. Circaseptan rhythms inenamel development in modern humans andPlio-Pleistocene hominids. In: Smith P, TchernovE, editors. Structure, function and evolution of teeth.London: Freund Publishing House. p 295–309.132 Dean MC. 1998. A comparative study ofcross-striation spacings in cuspal enamel and offour methods of estimating the time taken togrow molar cuspal enamel in Pan, Pongo, andHomo. J Hum Evol 35:449–462.133 Hiiemae KM, Crompton AW. 1985. Mastica-tion, food transport, and swallowing. In: HildebrandM, Bramble DM, Liem KF, Wake DB, editors. Func-tional vertebrate morphology. Cambridge: BelknapPress of Harvard University Press. p 262–290.

r 1999 Wiley-Liss, Inc.


the structure, function, and evolution of primate dental enamel

Documents