The Maxillary Sinus in Three Genera of New World Monkeys: Factors That Constrain Secondary Pneumatization



The air filled cavities of paranasal sinuses are thought by some to appear opportunistically in spatial “gaps” within the craniofacial complex. Anthropoid primates provide excellent natural experiments for testing this model, since not all species possess a full complement of paranasal sinuses. In this study, two genera of monkeys (Saguinus and Cebuella) which form maxillary sinuses (MS) as adults were compared to squirrel monkeys (Saimiri spp.), in which a MS does not form. Using microCT and histomorphometric methods, the spatial position of paranasal spaces was assessed and size of the adjacent dental sacs was measured. In Saguinus, secondary pneumatization is underway perinatally, and the sinus extends alongside deciduous premolars (dp). The MS overlaps all permanent molars in the adult. In Saimiri, the homologous space (maxillary recess) extends no farther posterior than the first deciduous premolar at birth and extends no farther than the last premolar in the adult. Differences in dental size and position may account for this finding. For example, Saimiri has significantly larger relative dp volumes, and enlarged orbits, which encroach on the internasal space to a greater degree when compared to Saguinus. These factors limit space for posterior expansion of the maxillary recess. These findings support the hypothesis that secondary pneumatization is a novel, opportunistic growth mechanism that removes “unneeded” bone. Moreover, paranasal spaces occur in association with semiautonomous skeletal elements that border more than one functional matrix, and the spatial dynamics of these units can act as a constraint on pneumatic expansion of paranasal spaces. Anat Rec, 2010. © 2009 Wiley-Liss, Inc.


Paranasal sinuses are mucosa-lined, air-filled sacs that are associated with intracranial cavities. The function or biological role of the paranasal sinuses has been a subject of investigation for centuries (Blanton and Biggs,1969; Rae and Koppe,2004). Although no consensus has been reached, a recent flurry of research, fueled in part by the availability of noninvasive imaging methods, has renewed progress (see Rae and Koppe,2004).

Recent attempts to explain paranasal sinus function can be crudely divided into “physiological” versus “structural” or “architectural” hypotheses (Rae and Koppe,2004; Márquez and Laitman,2008a). The latter have fared particularly badly in recent years, but this may be partly due to the difficulty of devising valid tests. The notion that sinuses play an architectural or structural role rather than performing an active physiological function has been espoused many times (Proetz,1922; Weidenreich,1924; Moss and Young,1960), but the precise meaning of the proposition is not always clear. For example, the sinus spaces may develop as an incidental byproduct of cranial growth, or they may actually play a more active role in facilitating the growth. In either case, the proposed structural role of the sinuses relates to the growth of the skull, and attempts to test it with recourse to adult morphology may be at a disadvantage.

Recent inferences on the relationship between paranasal sinuses and primate craniofacial form have been drawn mainly from comparative samples of adults (Koppe et al.,1999b; Rae and Koppe,2000), while only a few postnatal growth studies have been conducted (Koppe et al.,1995,1999c; Koppe and Nagai,1997; Rossie,2006). Because the “structural” or “architectural” hypotheses are explicitly developmental, it is necessary to improve our understanding of sinus formation within the context of the developing cranium. A thorough documentation of this relationship requires study of the cranium from the earliest phases of sinus growth, that is, when the sinuses first expand beyond the limits of the nasal capsule (i.e., secondary pneumatization). This time-frame is well documented in humans, where secondary pneumatization begins prenatally, in the maxillary sinus (Sperber,2000). Its onset marks the beginning of an exponential increase in sinus cavity volume (Koppe et al.,1994; Smith et al.,1997,1999). Moreover, studies of human fetuses with cartilaginous dysmorphologies, such as in cleft lip and palate, clearly illustrate that neighboring elements affect at least sinus symmetry in these early stages (Smith et al.,1997). Prenatally, sinuses develop “within the constraints of adjacent growing tissues” that may influence their shape (Smith et al.,1997, p 488).

With few exceptions (Rossie,2006; Smith et al.,2008) it is uncertain how much of the time frame of secondary pneumatization has been captured in previous studies on nonhuman primates. Thus, a comparative perspective on this formative stage of sinus development may be lacking. This is particularly true for the most precocious paranasal space, the maxillary sinus. Indeed, growth studies of the maxillary sinus in macaques have assessed development beginning at the emergence of the first permanent molar (M1; Koppe and Nagai,1997; Koppe et al.,1999c), which could be far too late to assess influences on the sinus by growth of neighboring structures.

Using a comparative sample of nonhuman primates, our ultimate aim is to illuminate the dynamic spatial relationships among the sinuses and the other functional units of the skull (e.g., orbital contents, dentition, oral cavity, brain). In the present study, we take a first step in this direction by investigating the spatial relationship between the developing dentition and maxillary recess and sinus in three monkey genera. The specific aims of the present study are twofold. First, the development of paranasal spaces from fetal (Saguinus) or perinatal (Cebuella, Saguinus, Saimiri) to adults stages is compared among two genera that do undergo pneumatization of the maxillary bone (Saguinus, Cebuella) and one that does not (Saimiri). Second, a perinatal sample of two species of primates is used to test the hypothesis that size and position of the deciduous dentition explains the magnitude of paranasal pneumatization.


In this study, three genera of New World primates were examined. In two of these genera, Saguinus and Cebuella, a maxillary sinus is present postnatally (Nishimura et al.,2005; Rossie,2006). In the third, Saimiri, no secondary pneumatization occurs, and no maxillary sinus develops (Rossie,2006). Twenty-one specimens were studied (Table 1), including 10 tamarins (Saguinus geoffroyi: 1 fetal, 2 perinatal, 1 53-day old; S. oedipus: 4 perinatal; 2 adults); three pygmy marmosets (Cebuella pygmaea: 1 perinatal, 1 one-month old, 1 adult); eight squirrel monkeys (Saimiri boliviensis: 6 perinatal, 1 adult; Saimiri sciureus: 1 adult). The fetal Saguinus geoffroyi was previously estimated to be in a fetal stage based on body size relative to “perinatal” specimens (see below) and based on maturation of the nasal capsule (Smith et al.,2008). Other subadult specimens ranged from P0 to P53 in age.

Table 1. Age and sexes of the sample under study
Specimen numberSpeciesAgeSex
So104Saguinus oedipusPerinatal?
So108Saguinus oedipusPerinatalm
So109Saguinus oedipusPerinatalf
So110aSaguinus oedipusPerinatalm
So2Saguinus oedipusAdult (3.5 years)m
USNM #306848Saguinus oedipusAdultm
MM105 (SG1)Saguinus geoffroyiInfant (1 month, 23 days)m
SG3Saguinus geoffroyi0f
SG4Saguinus geoffroyi0f
SG10Saguinus geoffroyiFetalf
Ss2662Saimiri boliviensisPerinatal (stillborn)?
Ss2671Saimiri boliviensisPerinatal (stillborn)m
Ss2672Saimiri boliviensisPerinatal (P0)m
Ss4238Saimiri boliviensisPerinatal (P0)m
Ss4267aSaimiri boliviensisPerinatal (P10)m
Ss4269Saimiri boliviensisPerinatal (P0)m
90620Saimiri boliviensisAdultf
USNM#518545Saimiri sciureusAdultm
CP1Cebuella pygmaeaPerinatal (P0)f
CP3Cebuella pygmaeaAdult (1 year, 1 month)f
CP4Cebuella pygmaeaInfant (one month, 11 days)f

Except for two skeletal specimens [Table 1, United States National Museum(USNM)], all specimens were acquired from cadaveric remains preserved in 10% formalin. The “perinatal” sample included animals that were stillborn or died within 10 days of birth in captivity at the New England Primate Center, the University of South Alabama Primate Research Center, or the Cleveland Metroparks zoo. Two adult cadaveric heads were acquired after the animals were euthanized at the New England Primate Center and the University of South Alabama Primate Research Center. When possible, both males and females of each genus were acquired. However, only male perinatal Saimiri (and one of unknown sex) were available for study. As adults of the genus Saimiri are described to have only minimal dimorphism (Corner and Richtsmeier,1992) it is not expected that Saimiri boliviensis is dimorphic at birth.

Excluding the fetal Saguinus, the earliest aged specimens represent a mix of body sizes. They are not all assumed to be neonatal in stage of somatic maturity, and likely represent a mix of premature and early postnatal stages (thus, these are referred to as the “perinatal” sample). Perinatal specimens of known age died at P0 except one of the Saimiriboliviensis, which died at P10. As this specimen fell within the length range of both the cranium and palate of P0 specimens, it was treated as part of the perinatal sample. The 1-month-old Cebuella and 53-day-old Saguinus are referred to as “infants” throughout the text.

Eight of the perinatal specimens (4 Saimiri boliviensis; 4 Saguinus oedipus) were scanned using micro CT (19 μm voxel size; VivaCT40 scanner, Scanco Medical). High resolution X-ray and CT scans of two adult monkeys (previously acquired by Rossie,2006) were also analyzed. Fetuses, perinatal specimens, infants, and two adults were dissected and processed for paraffin embedding. Before embedding, cranial length (prosthion-inion) and palatal length (prosthion-posterior mid-palatal point) were measured with digital calipers to the nearest 0.01 mm. Paraffin blocks were serially sectioned at 10–12 μm and stained with hematoxylin-eosin and Gomori trichrome procedures for histomorphometric analysis using Scion Image or ImageJ software (NIH). All heads were sectioned in the coronal plane except one of each genus. In the latter specimens, half the head was sectioned coronally while the contralateral head was sectioned in the sagittal plane. Every 10th section, at least, was mounted on glass slides with serial numbers for staining. Intervening sections were saved for future histochemical and immunohistochemical analysis.

The perinatal sample of Saguinus and Saimiri was analyzed in two dental regions regarding palatal, nasal, and dental variables (Table 2). Following Rossie's (2006) analysis of adult monkeys, we selected the mid-level of dp4 to measure variables related to palatal and nasal cavity breadth. These variables were also measured at M1, as the maxillary sinus extends beyond this in postnatal Saguinus (Rossie,2006). To acquire data, each stained section in the regions from dp4 to M1 was photographed using a Leica DMLB microscope and Cat-Eye digital camera. Measurements were then taken from digital micrographs of histological sections using ImageJ software (NIH). First, palatonasal index was measured. As per Rossie (2006), nasal width was measured as the distance between the right and left sides of lateral nasal wall at the base of the nasal fossa (at P4 this corresponds to the wall of the inferior meatus). As the medial margin of alveolar bone was difficult to isolate from the palatine process of the maxillary bone in perinatal animals, alternative landmarks were used: the medial margin of the right and left dental follicles. The distance between right and left sides was measured as palatal width. Average palatonasal indices for each species were calculated as interdental width/nasal width. In specimens where only half of the face was sectioned in the coronal plane, the ratio was calculated by measuring distances from the dental sac or lateral limit of the nasal fossa to the midline (using the center of the vomer as the midline).

Table 2. Quantitative results on perinatal primates
Species (N)dp4/palatal ratioaM1/patatal ratiodp4/cranial ratioM1/cranial ratioPNI: dp4PNI:M1
Mean ± SEMMean ± SEMMean ± SEMMean ± SEMMean ± SEMMean ± SEM
  • PNI, palatonasal index; SEM, standard error of the mean; ?, cranial length was unavailable for this specimen.

  • a

    All ratios calculated as cube root of dental sac volume relative to palatal or cranial length.

Saguinus oedipus (4)0.123 ± 0.0080.072 ± 0.0080.034 ± 0.0020.019 ± 0.0021.56 ± 0.0331.54 ± 0.034
Saimiri boliviensis (6)0.145 ± 0.0040.112 ± 0.0050.032 ± 0.0010.025 ± 0.0021.34 ± 0.0851.35 ± 0.040
Cebuella pygmaea (1)0.1370.173??1.681.85

Using ImageJ software, dental sac volumes were measured for dp4 and M1 on the right side of the nasal fossa, except in two specimens where dp4 or M1 were damaged. Among specimens in which right and left sides were both well-preserved for measurement, asymmetry was not biased toward one side (in about half of the specimens the right dental sacs were larger, and in the other half the left sacs were larger). Therefore, the left side was used for analysis in the two specimens where the right side was damaged. Volumes were acquired by tracing the cross-sectional area of the dental sac as estimated by the outer edge of the outer enamel epithelium. Once all sections containing a cross-section of the tooth were measured, the cross-sectional areas in mm2 were multiplied by the distance between sections for a segmental volume of the tooth. This was repeated for each section of the tooth up to the penultimate section (the last section containing the tooth was regarded as the end). Then, all segmental volumes were summed to obtain total volume of the tooth. The cube root of these tooth volumes was divided by palatal and cranial lengths to obtain relative size of the dp4 and M1 (Table 2).

In Saguinus oedipus and Saimiri boliviensis, one perinatal and one adult specimen of each species was selected for measurement of the mucosal surface area (mm2) in the nasal and paranasal spaces. Using ImageJ software, the data were obtained by tracing the perimeter of the nasal fossa and paranasal spaces in each section beginning at the first coronal section in which the right nasal fossa was enclosed, and ending at the choana. A linear distance was obtained in mm, and this was multiplied by the distance between sections to obtain a segmental surface area. Segmental surface areas were summed to estimate total surface areas for nasal and paranasal mucosa. These data were used for a preliminary evaluation of age related changes in these dimensions in the two species.

SPSS 15.0 (SPSS) was used to analyze the data. Data were compared between groups using a Mann-Whitney U test. All differences were considered significant at P ≤ 0.05.


Position of the Maxillary Sinus Relative to Dentition in Fetal and Early Postnatal Geoffroy's Tamarins (Saguinus geoffroyi)

The anteroposterior spatial relationship of the maxillary sinus to the dentition in Saguinus geoffroyi is shown in Figs. 1, 2. In the fetal specimen the primordial maxillary sinus runs the mesiodistal length of dp2 and most of the length of dp3 (Figs. 1, 2). In the perinatal S. geoffroyi, the maxillary sinus extends posteriorly to overlap all of dp2 and dp3 and a portion of dp4 (Figs. 1, 2). The MS extends over more than one-half the length of dp4 in the 53-day-old S. geoffroyi (Figs. 1, 2).

Figure 1.

Anteroposterior position of the maxillary sinus in the tamarin (Saguinus geoffroyi) at three ages. The graphs are standardized with a start point at the anterior end of the deciduous or permanent canine and end at the posterior limit of M1. Light blue indicates the portion of the sinus that opens into the nasal fossa; dark blue indicates levels at which the sinus is an enclosed sac. Note that in the fetal specimen, the sinus was completely surrounded inferolaterally by cartilage (see Smith et al.,2008). Thus, the sinus in this specimen has not undergone secondary pneumatization.

Figure 2.

Drawings based on computer three-dimensional reconstructions of the teeth (blue) and the sinus mucosa (beige) of three specimens of the tamarin (Saguinus geoffroyi) at different ages (the perinatal specimen is a different individual than that shown graphically in Fig. 1). Structures are seen on their medial surface. Dental sacs are illustrated for fetal and perinatal specimens. In the 53-day old, a mixed set of erupted teeth and dental sacs are illustrated.

In dorsoventral space, the primordial maxillary sinus is well separated from the developing dental sacs in the fetus (Fig. 2). In the perinatal S. geoffroyi the mucosal lining of the maxillary sinus is dorsoventrally larger and is more closely adjacent to the deciduous premolars, though this is largely due to the growth of the teeth. In the 53-day old, the maxillary sinus is positioned medial to dp2 and dp3, and overlaps a portion of dp4 (Fig. 2). The roots of the deciduous premolars lie lateral to the maxillary sinus once they develop.

Perinatal and Postnatal Position of the Maxillary Sinus or Recess Relative to Dentition in the Tamarin (Saguinus), Pygmy Marmoset (Cebuella), and Squirrel Monkey (Saimiri)

As the findings on the position of the sinus or recess were similar within genera, the genus alone is used in most of the ensuing text. Figures 3–7 show the position of the maxillary sinus, specifically the mucosal sac, relative to dental level in Saguinus, Cebuella, and Saimiri at different postnatal ages. In all species, the maxillary sinus or recess overlaps more posterior teeth in adults than in perinatal specimens (Figs. 3, 4). The most extreme posterior expansion of the sinus relative to the teeth is in the adult Saguinus, where the sinus completely overlaps M1 (Fig. 3; the sinus actually extends posterior to M1). In Cebuella and Saimiri, the sinus/recess does not extend beyond the level of P4 in adults. However, in Cebuella the sinus extends more anteriorly compared to the recess of the adult Saimiri (Figs. 3, 4). In all species, the anterior limit of the maxillary sinus or recess shifts posteriorly relative to the maxillary tooth row in adults compared to perinatal specimens (Figs. 3, 4).

Figure 3.

Anteroposterior position of the maxillary sinus in the tamarin (Saguinus oedipus) and maxillary recess in the squirrel monkey (Saimiri boliviensis) at perinatal and adult ages. Spatial overlap with dentition is emphasized. The graphs are standardized with a start point at the anterior end of the deciduous or permanent canine and end at the posterior limit of M1. Note that the sinus continues beyond the posterior end of M1 in the adult tamarin (see text). Light blue indicates the portion that opens into the nasal fossa; dark blue indicates levels at which the sinus or recess is an enclosed sac.

Figure 4.

Anteroposterior position of the maxillary sinus in the pygmy marmoset (Cebuella pygmaea) at three ages. The graphs are standardized with a start point at the anterior end of the deciduous or permanent canine and end at the posterior limit of M1. Light blue indicates the portion of the sinus that opens into the nasal fossa; dark blue indicates levels at which the sinus is an enclosed sac.

Figure 5.

The position of the maxillary sinus or recess (open arrows) is shown in three perinatal primates. Each plate shows the approximate anteroposterior mid-level of a tooth in the coronal plane. In the tamarin (Saguinus oedipus; 5ac), the sinus cavity completely overlaps dp3 and is found between the tooth and nasal fossa (NF) at this level. The sinus ends before the mid-point of dp4. Note the relatively small size of M1. In the pygmy marmoset (Cebuella pygmaea; 5df), a relatively small sinus is found superomedial to dp3 (add arrow) and does not overlap dp4. Note the close proximity of dp4 to the NF in the marmoset compared to the tamarin. In the squirrel monkey (Saimiri boliviensis; 5gi), the maxillary recess does not reach the anteroposterior midpoint of dp3. Note the relatively large size of the dentition in the squirrel monkey, and close proximity to the nasal fossa. Or, orbit. Scale bars = 1 mm.

Figure 6.

The position of the maxillary sinus (open arrows) is shown in two infant monkeys. Each plate shows the approximate anteroposterior mid-level of a tooth in the coronal plane. In the tamarin (Saguinus geoffroyi; 6ac), the sinus cavity overlaps dp4 but does not extend alongside M1 (see Fig. 1). Note the dorsoventrally expansive sinus cavity adjacent to dp3. In the pygmy marmoset (Cebuella pygmaea; 6df), a relatively small sinus is found superomedial to dp3 and overlaps most of dp4. Note the relatively large size of M1 and it close adjacency to the nasal fossa. NF, nasal fossa; Or, orbit; double arrows, dental sacs of permanent teeth. Scale bars = 1 mm.

Figure 7.

The position of the maxillary sinus or recess (open arrows) is shown in three adult primates. Each plate shows the approximate anteroposterior mid-level of a tooth in the coronal plane. In the tamarin (Saguinus oedipus; 7ac), the sinus overlaps all permanent premolars and molars in anteroposterior space. Note the relatively greater magnitude of inferior expansion of the sinus at P3 in Saguinus compared to Cebuella (7df). The sinus of Cebuella overlaps P4, but does not extend alongside M1 (see Fig. 4). In the squirrel monkey (Saimiri boliviensis; 7gi), the maxillary recess overlaps P3 and P4. The recess is an enclosed sac posteriorly, but is separated from the nasal fossa by mucosa, not bone. NF, nasal fossa; Or, orbit. Scale bars = 1 mm.

In perinatal specimens, the anterior part of the sinus or recess is open to the middle meatus via an ostium. Posteriorly, the mucosal sac of the sinus or recess is an enclosed cul-de-sac. In adults, an enclosed portion of the sinus or recess also extends anterior to the ostium (Figs. 3, 4).

Figure 5 shows the nasal fossa at coronal mid-levels of deciduous premolars and M1 in perinatal specimens. Saguinus is distinguished by a broad separation of the dental sacs from the nasal fossa and by a comparatively diminutive M1. Cebuella possesses larger dp4 and M1 compared to dp3 (Fig. 5d–f). In Saimiri, all dental sacs are cross-sectionally large and closely approximated to the nasal fossa (Fig. 5g–i). In both Saguinus and Cebuella, the sinus is directly adjacent to or posterior to the mid-point of dp3, whereas in Saimiri the maxillary recess ends anterior to this level (Fig. 5). The cul-de-sac of the sinus ends at or anterior to the mid-point of dp4 in Saguinus, and does not overlap dp4 at all in the perinatal Cebuella under study (Figs. 3–5).

The 53-day-old Saguinus geoffroyi and 1-month-old Cebuella reveal positional differences of the sinus compared to perinatal specimens (Figs. 1, 4, 6). In both species, the sinus extends to partially overlap dp4. A prominent difference between the infant primates is observed in terms of the secondary dentition. In Saguinus the dental sacs of I1, I2, and the permanent canine are prominently visible (Fig. 2) whereas postcanine teeth (not shown) are relatively small and undifferentiated, without a clear enamel organ. In the 1-month-old Cebuella, posterior permanent premolars at the cap stage are prominently positioned adjacent to the maxillary sinus (Fig. 6).

In the adult Saguinus, the mucosal boundaries of the maxillary sinus or recess completely overlap (and pass posterior to) M1 (Figs. 3, 7). In the adult Cebuella and Saimiri, the sinus/recess completely overlaps P4 but does not reach the level of M1 (Figs. 3, 4, 7). The mucosal boundaries of the maxillary sinus in Saguinus and maxillary recess in Saimiri are shown at higher magnification in Fig. 8. In each, there is an enclosed mucosal pocket anterior to the ostium. Medially, this region is enclosed by bone in Saguinus (Fig. 8a) but is bounded by mucosa alone in Saimiri (Fig. 8d). The ostium opens into the middle meatus in both monkeys (Fig. 8b,e). The posterior extension of the maxillary sinus in Saguinus is medially bordered by bone. In Saimiri the enclosed posterior portion of the mucosal sac is bordered medially by mucosa alone (Fig. 8f).

Figure 8.

The mucosal boundaries of the maxillary sinus in the tamarin (Saguinus Oedipus; 8ac) and the maxillary recess in the squirrel monkey (Saimiri boliviensis; 8df). In Saguinus, the sinus is enclosed by bone anterior to the ostium (8a). The ostium is shown in 8b. Posteriorly, the sinus is enclosed by bone (8c). In Saimiri, although there are enclosed portions of the recess anterior (8d) and posterior (8f) to the ostium (8e), these regions are not enclosed medially by bone. IM, inferior meatus; NA, nasal airway; Or, orbit; double arrows, medial bony wall of maxillary sinus. Scale bars: a, c, d, f = 0.5 mm, b, d = 1 mm.

Interrelationship of the Orbit, Nasal Fossa, and Dentition in Tamarins (Saguinus) and Squirrel Monkeys (Saimiri)

Figures 9–11 show the spatial relationship of the orbit, nasal fossae, and dentition relative to the maxillary sinus or recess in perinatal Saguinus and Saimiri. Based on three-dimensional reconstructions in the frontal view, inter-orbital distance is relatively wider in Saguinus at all maxillary tooth levels. In both primates, the orbits encroach on the superior portion of the nasal fossa, especially at posterior levels (Figs. 9, 10a,b). The palate appears proportionally wider relative to bizygomatic width at all levels in Saguinus (Fig. 9). The orbits appear to be more convergent in Saimiri compared to Saguinus (Fig. 10a,c).

Figure 9.

Micro CT reconstruction of the perinatal tamarin (Saguinus oedipus) and squirrel monkey (Saimiri boliviensis) at different dental levels in a frontal view. The position of the sinus or recess cavity is indicated (red arrows). Top: inferior view of the specimens with dental levels indicated. Scale bars = 1 mm.

Figure 10.

Micro CT reconstruction of the perinatal tamarin (Saguinus oedipus) and squirrel monkey (Saimiri boliviensis) from a superior perspective (a,c) and a posterior-oblique view (b,d). The position of the sinus or recess cavity is indicated (arrows). Top: inferior view of the specimens with dental levels indicated. ET1, first ethmoturbinal; N, nasal bone; NC, nasal cavity; V, vomer bone; Z, zygomatic bone. Scale bars = 1 mm.

Figure 11.

Parasagittal sections of the nasal fossa of a perinatal tamarin (S. oedipus) and squirrel monkey (Saimiri boliviensis). Insets are the contralateral side of the head, sectioned in the coronal plane. The position of the maxillary sinus or recess (red arrows) is indicated. Three different cross-sectional levels are shown. At the top (a,d) the parasagittal plane which crosses the middle of the sinus or recess cavity at dp2 is shown. 11b and 11e show a parasagittal plane that transects the lateral limit of the sinus (11b) and recess (11e). 11c and 11f show a parasagittal plane through the entire maxillary tooth row to show its relationship to the orbit. The insets show the approximate location of the parasagittal plane relative to a coronal section through dp2. c, deciduous canine; NLD, nasolacrimal duct; Or, orbit. Scale bars = 1 mm. Serial sections (every 10th–20th section) of these specimens may be viewed as movie files at or at

The position of the sinus or recess cavity is emphasized in Fig. 10b,d. In each monkey the cavity of the sinus or recess is wedged between the orbit and deciduous premolars. Fig. 11a–b shows the maxillary sinus extending ventral to the orbit in Saguinus. The deciduous premolars are positioned lateral to the sinus (Fig. 11c). In Saimiri, the maxillary recess is more restricted to the region anterior to dp3 (Fig. 11d,e). Figure 11f emphasizes the extreme proximity of the relatively large orbit and posterior maxillary teeth (dp3-M1) in Saimiri (also see Fig. 10d).

Figure 12 shows the position of the maxillary sinus or recess in an adult Saguinus oedipus and Saimiri sciureus relative to the orbit and dentition. The spatial relationship of the paranasal cavities to the maxillary dentition agrees with results based on histology (Fig. 3). The maxillary sinus of Saguinus lies medial to all teeth from P2 to M2 (Fig. 12a–e), whereas the maxillary recess lies medial to only P3 and P4 in Saimiri (Fig. 12g,h).

Figure 12.

CT slices, in the coronal plane, of an adult tamarin (S. oedipus) and squirrel monkey (Saimiri sciureus). Cross-sectional levels from P2 to M2 are shown. Note that the extent of the maxillary recess cavity (open arrows) in the squirrel monkeys concurs with findings on the adult Saimiri boliviensis (see Fig. 3). Similarly, the cavity of the maxillary sinus (open arrows) in the tamarin agrees with results seen in the serially sectioned adult (Fig. 7). Scale bars = 5mm.

In both adult monkeys, the orbits are highly approximated and encroach on the dorsal part of the nasal fossa at posterior levels (Fig. 12c–e,h–j). Figure 12c,h emphasize the relatively broader palate in Saguinus (as discussed by Rossie,2006). This disparity is equally pronounced at M1 (Fig. 12d,i). The proximity of the orbit to the maxillary teeth and hard palate is greater in Saimiri compared to Saguinus (Fig. 12a–j).

Quantitative Results

Relative to palatal length, dp4 and M1 are larger in Saimiri and Cebuella compared to Saguinus (Fig. 13; Table 2). MannWhitney U tests show that the ratio of cube root of tooth volumes to palatal length is significantly different between species for both dp4 (P < 0.05) and M1 (P < 0.05). In both cases, Saimiri has significantly larger dental volume relative to palatal length. Relative to cranial length, dp4 is not significantly different between species (P > 0.05), but M1 was significantly (P < 0.05) larger in Saimiri.

Figure 13.

(a) Comparison of the ratio of cube root of dental volumes to palatal length in the tamarin (Saguinus oedipus) and squirrel monkey (Saimiri boliviensis). (b) Comparison of the ratio of cube root of dental volumes to cranial length in Saguinus oedipus and Saimiri boliviensis. (c) Palatonasal index interdental width/nasal width) in Saguinus oedipus and Saimiri boliviensis.

Palatonasal indices are lower in Saimiri at both dp4 and M1 levels (Fig. 13). Mann Whitney U tests reveal a significant difference at the level of M1 (P < 0.05) but not at the level of dp4 (P > 0.05).

Age comparisons of mucosal surface areas between individual adult and perinatal specimens indicate marked species differences. In a perinatal Saguinus, total nasal fossa area measures 78.12 mm2, and the maxillary sinus is 9.81 mm2 in surface area. Total nasal fossa surface area in the adult Saguinus is 336.7 mm2. For this specimen, the maxillary sinus measures 66.85 mm2, and the frontal sinus surface area is 51.98 mm2. In a perinatal Saimiri, total nasal fossa area is 160.11 mm2, and the maxillary recess mucosal surface area measures 2.12 mm2 in surface area. In the adult Saimiri, the nasal fossa area is 530.0 mm2, and the maxillary recess is 9.74 mm2. An age/species comparison of internal nasal surface area (nasal fossa and sinus or recess areas), expressed in percentages, is shown in Fig. 14. Compared to the perinatal specimen, sinus area in the adult Saguinus comprises 15% more mucosal area. In contrast, the percentage surface area of the maxillary recess is nearly equal when comparing adult and perinatal Saimiri.

Figure 14.

Surface areas of the nasal mucosa, expressed as percentages, in the tamarin (Saguinus oedipus) and squirrel monkey (Saimiri boliviensis). Note that the sinus mucosa areas in Saguinus show an increase in proportions relative to the remainder of the nasal fossa (NF), whereas the maxillary recess surface area of Saimiri increases isommetrically with the nasal fossa. FS, frontal sinus; MS, maxillary sinus.


Debates concerning the function of the sinuses have yielded little consensus (e.g., Koppe et al.,1999a; Preuschoft et al.,2002; Rae and Koppe,2008). Several authors have suggested an alternative explanation for paranasal sinuses which treats the sinus cavities as “spandrels,” or by-products of craniofacial growth processes (Rae and Koppe,2008; Zollikofer and Weissman,2008). Such a concept has roots in Weidenreich's view that sinuses have an “architektonisch und funktionell passiven Charakter” (1924, p 91). Zollikofer and Wiessman (2008) pair the spandrel hypothesis with the “invasive tissue hypothesis.” The latter hypothesis invokes properties of the mucosal lining of sinus cavities as an explanation for their expansion, and relates to Witmer's (1997) conception of sinus expansion as an “opportunistic” process. The invasive tissue hypothesis pairs equally well with the notion that the growth of the sinuses actually aids in the growth of the skull (e.g., Proetz,1922), and this might begin to address the reason for the expansive tendencies of the sinus mucosa.

As secondary pneumatization is a prolonged process spanning prenatal and postnatal periods; the exploration of these concepts demands examination of a broad age range. To that end, the present study explores the spatial relationships of paranasal maxillary spaces within the midface in a sample that spans prenatal and postnatal stages. Recently, it was demonstrated that in Saguinus geoffroyi, the onset of secondary pneumatization is a perinatal event (Smith et al.,2008). Thus, our study centers on the earliest formative stages of a “true” sinus in at least one genus. Cartilaginous remnants observed in all perinatal specimens suggest that capsule breakdown is occurring at this time in other marmosets and tamarins as well (Smith et al.,2005; unpublished data).

Computer modeling supports the view that sinus form is largely dictated by the form of the space available for pneumatization (Zollikofer and Weissman,2008). Although individual cranial bones are pneumatized by sinus cavities, these bones are formed from ossification centers that abut more than one craniofacial region, such as soft tissue (“functional”) matrices, dental follicles, or the developing sinus itself (Fig. 15; Moss and Young,1960; Moss and Greenberg,1967). The potential influence of neighboring elements on the maxillary sinus or recess is assessed and discussed below.

Figure 15.

Top: Three-dimensional reconstruction of a fetal tamarin (Saguinus geoffroyi) in a postero-oblique view, with the septum removed. The position of the primordial maxillary sinus (gold) before secondary pneumatization is emphasized. Note the position of the deciduous premolars (dp) and eye relative to the sinus, which is still surrounded by an inferior scroll of the nasal capsule. Bottom: A serial section of the same specimen, just posterior to the maxillary sinus (*, glands at posterior limit of sinus). The cutting plane, at the posterior end of dp3, is indicated to the left of the section. Pneumatization proceeds posteriorly between the orbit, dentition, and nasal fossa in this species. Dc, deciduous canine; Di1, deciduous central incisor; dp2, deciduous premolar 2 (or 1st deciduous molar); ET1, ethmoturbinal I; F, frontal bone; M, maxilla; M1, 1st permanent molar; NA, nasal airway; OB, olfactory bulb; ONL, orbitonasal lamina; TN, tectum nasi. Scale = 1 mm.

Patterns of Pneumatic Expansion

Current researchers appear to agree on the definition of a paranasal sinus (Witmer,1999; Rae and Koppe,2000; Rossie,2006). The identification of sinuses can be complicated by the presence of recesses that may persist even in absence of secondary pneumatization. Thus, the very definition of a sinus is critical (Rae et al.,2003; Rossie,2006). “Recesses” of the nasal cavity are most clearly defined prenatally, when they are still encapsulated by nasal capsule cartilage (much of which later ossifies into the ethmoid). “True” paranasal sinuses form by an invasive process wherein mucosa-lined recesses of the nasal fossae escape the confines of these recesses and expand within the body of the maxillary, palatine, ethmoid, frontal, or sphenoid bones (Witmer,1999). This expansion is termed secondary pneumatization, in contrast to primary pneumatization. The latter is an unfortunate misnomer referring to the formation of recesses via growth and folding of prenatal cartilaginous walls of the nasal cavity (e.g., recessus maxillaris).

Primates that lack maxillary sinuses postnatally, such as most cercopithecoids, nonetheless possess the natal maxillary recess (Maier,2000). The failure of secondary pneumatization to occur results in the absence of a paranasal cavity outside the limits of the nasal capsule. This is also the case in at least two New World monkeys, including Saimiri (Rossie,2006). The posterior portion of the maxillary recess in Saimiri is separated from the nasal cavity by mucosa alone. This mucosal sac also typifies the primordial maxillary sinus of other taxa prior to pneumatization, after which it is mostly enclosed by bone (Fig. 1; Smith et al.,2008).

Preliminary data on internal nasal surface areas provided herein, if typical for the species, emphasizes the different magnitude of growth in paranasal spaces that do pneumatize versus those that do not. The proportional difference in sinus mucosal area between the perinatal and adult Saguinus is striking, especially when compared to the same age comparison in Saimiri. In the latter, preliminary data (Fig. 14) suggest that the mucosa of the maxillary recess simply follows a growth trajectory in common with the remainder of the nasal fossa. In Saguinus, expansion of “pneumatic” mucosa outpaces that of the nasal fossa.

Rossie (2006) noted the presence of diplöic bone adjacent to the maxillary recess in adult Saimiri. In this study, such bone is seen posterior to the enclosed mucosal recess (e.g., Fig. 7i). Our findings support Rossie's hypothesis that size-related crowding by alveolar skeletal units limits potential space of pneumatization; the maxilla of perinatal Saimiri is relatively crowded by dental sacs compared to Saguinus (see Figs. 10, 11). A critical question is why the diplöic bone that persists adjacent to the maxillary recess is not subject to “opportunistic” pneumatic expansion. Elsewhere in the cranium, diplöic bone scales positively in relation to age and body mass in humans (Lynnerup et al.,2005; Hatipoglu et al.,2008). These regions of bone may be physiologically or architecturally essential in the context of dental crowding within the maxilla. It may be that there is a temporal window within which secondary pneumatization can commence if conditions permit, and only the spatial relations during this time are relevant. In this light, it is worth noting that the diplöic space found adjacent to the maxillary recess in adult Saimiri specimens is occupied by the developing permanent premolars in younger specimens (see Fig. 9b in Rossie,2006).

The small sample of adults and infants available in this study indicate subtle species differences in pneumatic expansion between Saguinus and Cebuella. In S. geoffroyi, three-dimensional reconstructions of the fetal specimen provide the initial context of sinus/dental spatial relationships. The mucosal sac of the primordial sinus is spatially separated from the dental follicles, although the ostium bears the same anteroposterior position relative to dp2. Older specimens suggest that the sinus expands posteriorly at a greater rate than the deciduous premolars grow in mesiodistal length. This expansion is accompanied by a posterior displacement of the sinus ostium, which may relate to displacement by anterior teeth. An increase in vertical dimensions of the sinus is evidenced in the perinatal specimen (Fig. 2). In the infant, the sinus is expanded vertically and posteriorly relative to the deciduous premolars (Fig. 2).

Cebuella shows less evidence of early vertical expansion, by comparison. The small sample studied suggests that mesiodistal expansion may be completed early in postnatal ontogeny, since pneumatization does not extend beyond P4. Cebuella, like Saguinus, has a pneumatic diverticulum that expands anterior to the level of the ostium.

Interrelationship of Sinus and Adjacent Midfacial Structures and Spaces

On the basis of patterns of osteoclastic activity, Smith et al. (2005) asserted that deciduous dentition have “morphogenetic primacy” over the developing maxillary sinus. In marmosets and tamarins, the alveolar wall of the maxillary sinus drifts medially in locations where a single plate of bone separates the sinus from deciduous premolars. The findings of the present study are consistent with this view and strongly suggest that the posterior dentition can limit the extent of posterior pneumatization. The lack of expansion of the maxillary recess in Saimiri coincides with the significantly larger relative size of dp4 and M1 in these monkeys compared to Saguinus at birth. The even larger relative size of M1 in the single perinatal specimen of a Cebuella is suggestive of the same size-related constraint as seen in Saimiri, especially given the absence of pneumatization beyond P4 in the older Cebuella. The wide-set palate presumably mitigates this effect, and allows space for pneumatic expansion anterior to M1. This emphasizes that competing factors may dictate the course and extent of pneumatization in a species.

The influence of these competing factors requires a clear understanding of the position of the maxillary sinus before secondary pneumatization. Figure 15 illustrates the position of the primordium of the maxillary sinus relative to adjacent structures in a fetal Saguinus. The primordial sinus orients in a postero-inferior direction and ends at about the level of dp3. This is a level at which pneumatization proceeds parallel to the palate in perinatal Saguinus (see Fig. 11a,b). In perinatal Saimiri, the orbit appears relatively larger than in Saguinus and is more closely adjacent to the posterior dentition (dp3-M1). This is consistent with Hartwig's (1995) observation that orbital approximation is extreme in newborn and adult Saimiri, and also emphasizes that the encroachment of the nasal and paranasal spaces by highly convergent orbits increases posteriorly. Together, the contents of the orbit and maxillary dentition from dp3 to M1 conspire to limit potential for posterior extension of the maxillary recess in the perinatal period (Figs. 10d, 11e illustrate this constraint).

In summary, observations on the primates under study support the hypothesis that the size and position of the deciduous dentition constrains secondary pneumatization of the maxillary sinus (Smith et al.,2005). Observations on the position of the developing dentition relative to the nasal fossae and orbits prompt the following hypothesis that must be tested on a broader range of primates: large relative size of the posterior maxillary dentition, a high degree of orbital approximation, and low palatonasal indices each act to constrain pneumatization from the maxillary recess. Examining the combination of these measurements in individual species may afford a nuanced view of the ultimate extent of pneumatization.

An ontogenetic approach to such comparisons will be beneficial to account for effects of the transient deciduous dentition. While dental constraint appears to easily explain sinus extent at birth, the deciduous teeth are transient elements. A noteworthy difference between the 1-month-old Cebuella and 53-day-old Saguinus is the more advanced development of permanent premolars in Cebuella (Fig. 6). The dental sacs for the replacement premolars emerge lingually (e.g., Fig. 6d,e), directly adjacent to the sinus. Variation in the schedule of development of the replacement dentition in relation to pneumaticity requires further study. An ontogenetic approach will also be needed to assess the changing proximity of soft tissue matrices (and the skeletal units that surround them) to the sinus during growth of the anthropoid face. Noteworthy in this regard is the increasing depth of the growing midface in Saimiri (Corner and Richtsmeier,1992) which may underlie the positional change in the maxillary recess relative to the dentition in the adult.


Rossie (2006) revealed major ontogenetic variations in secondary pneumatization of the maxillary sinus in New World monkeys. Our results provide a glimpse of variation beginning at the earliest formative stages of a “true” sinus, as compared to an unpneumatized recess. Magnitude is the single most distinguishing aspect of the postnatal trajectory of these paranasal spaces in Saguinus and Cebuella compared to Saimiri. Our results indicate that the maxillary sinus grows anteroposteriorly, and probably in surface area, at a disproportionately higher rate compared to the nasal fossa. In contrast, the maxillary recess in Saimiri appears to grow at a rate similar to the nasal cavity itself. These observations support the hypothesis that secondary pneumatization is an opportunistic growth mechanism (Witmer,1997) that may be constrained by adjacent elements (Smith et al.,1997,1999,2005). While the reasons for maxillary sinus suppression in Saimiri may be complex, the present data strongly suggest that the size and position of the postcanine dentition, as well as the encroachment of the orbits play a substantial role (cf. Rossie,2006).

During facial morphogenesis, paranasal sinus expansion may provide a mechanism for resolving spatial shifts among skeletal elements of the nasal capsule, orbit, and dentition (Proetz,1922; Shea,1985; Enlow and Hans,1996; Zollikofer and Weissman,2008). In this view, variations in the nature of sinus expansion are related to specific craniofacial growth patterns (Enlow and Hans,1996). Even if the occurrence of paranasal sinuses is best explained by such a “structural” hypothesis, this does not preclude the acquisition of secondary functions such as those affected by climate (e.g., Rae et al.,2003; Márquez and Laitman,2008b). If secondary functions reshape sinuses to physiological demands, inferences drawn from adult crania may fail to ascertain the validity of any “structural” hypothesis. In other words, if the process of pneumatization serves to remove unneeded bone (Proetz,1922; Weidenreich,1924; Blaney,1986), evidence of this is better sought during, rather than after, the pneumatic process.