Pouches, pockets and sacs in the heads, necks and chests of mammals, part VI: guttural pouches, false nostrils and preorbital fossae in horses, tapirs and rhinos


Back to the series on pouches, pockets and sacs (for previous articles see links below). The previous article finished by looking at the guttural pouches present in the Mongolian gazelle Procapra gutturosa. This links us nicely to the select group of mammals - perissodactyls, hyraxes, bats and rodents - that possess air-filled structures (called guttural pouches) located in the upper respiratory tract, pressed up close to the tympanic region at the back of the skull. In this article, I'm only going to deal with perissodactyls: the other groups can wait to later.


Guttural pouches are best known in the Horse Equus caballus where they contain 300-500 ml of air (accounting for up to 4% of the head's entire volume) and consist of both lateral and medial compartments. The size and shape of the pouches changes as the animal contracts and relaxes surrounding musculature. The medial guttural pouch compartment is virtually wrapped around the internal carotid artery; this close proximity suggests that the pouches might play a role in selective brain-cooling. In other words, they might help to cool blood running from the hot body core, up the internal carotid and past the brain. Many mammals - including bovids, deer, dogs and cats - keep brain temperature low by virtue of a carotid rete (a heat-exchange network where carotid arterial blood is cooled by venous blood drained from the nasal region), but horses don't have this [the diagram below, showing the guttural pouches and other details of horse cranial anatomy, is from the Forever Horses blog. The guttural pouches are the cream-coloured areas at the back of the skull].


In a brilliant case of logical deduction preceding discovery, Keith Baptise (and colleagues) suggested a cooling role for the guttural pouches of horses as early as 1993 (e.g., Baptiste et al. 1993), and were later able to demonstrate exactly this somewhat later (Baptise et al. 2000). However, other possible functions have been suggested over the years, including an immunological one and a role in swallowing. So far as I can tell, these are more speculative.

Having mentioned the comparatively large size of the pouches in the domestic horse, note that they're rather smaller in Przewalski's horse. In fact, domestic horses bred for racing or use as pack animals have the biggest pouches, suggesting that selective breeding has encouraged an increase in the size of the pouches concomitant with an increase in exercise and activity.

The literature reporting, describing and interpreting horse guttural pouches is fairly substantial. Donkeys definitely have them too, and their presence in tapirs was demonstrated during the 1850s. A result of this distribution is that it's sometimes been assumed that they're ubiquitous in perissodactyls, and hence also present in rhinos. However, Endo et al. (1998) showed that guttural pouches were absent in the white rhinos Ceratotherium and their absence was also later demonstrated in a juvenile Indian rhino Rhinoceros indicus (Endo et al. 2009). Some sources do state that they are present in other rhino species, however, and in these species we await modern study that confirms or refutes their possible presence.

False nostrils

There are a few other weird pockets and cavities present in the heads of perissodactyls. Horses have paired, cone-shaped, fur-lined, blind recesses - the false nostrils (properly called nasal diverticula) - that open dorsal to the true nostrils and extend posteriorly for a short distance on the upper surface of the snout [the extent of the false nostril is shown below, in a diagram from Witmer et al. (1999). dn = nasal diverticulum, in' = position of incisura nasoincisiva, pa' = plica alaris, with lamina of alar carilage dorsal to it, cac' = lateral ala of nostril overlying cornu of cartilago alaris].


It's been thought that both tapirs and rhinos possess the same structures, and that these are particularly well developed in tapirs. However, the supposed 'false nostrils' of tapirs are completely different to those of horses: the tapir structures are not external, cutaneous pouches "but rather... intracapsular, mucocartilaginous tube[s] that maintain communication with the nasal cavity at both ends" (Witmer et al. 1999, p. 263). Exactly what's going on with the rhino nostril cavity isn't entirely clear, as detailed descriptions of rhino noses don't seem to exist (though I confess that I haven't seen Bordoloi & Kalita (1996)). Of incidental interest to this section of the article (but of relevance to this series of articles as a whole) is that all species of living rhino possess an epipharyngeal diverticulum (a pouch-like recess) on the dorsal wall of the pharynx (Cave 1973, 1974a, b). We'll return to epipharyngeal diverticula in a later article.

But to return to false nostrils... no obvious function has ever been discovered for these structures. Because their floors flutter during exhalation, I wonder if they might play a role in amplifying or otherwise enhancing the loud snorting noises that horses use in communication.

Preorbital cavities in horses, rhinos and tapirs


I think - rightly or not - that false nostrils are weird. But they're not the only weird, poorly understood 'pockets' known to be present in horse heads. A number of fossil horses - in particular members of the equine clade Hipparionini - possess a roughly oval bony recess on the side of the skull, located about half-way between the bony nostril and eye socket. It's typically termed the dorsal preorbital fossa (maxillary fossa is also sometimes used) and is highly variable in size and shape [skull of Hipparion gracile with obvious preorbital fossa shown here; image by Ghedo, from wikipedia]. In some specimens of some taxa (Cormohipparion for one), its posterior part excavates parts of the nasal, lacrimal and maxillary bones. This structure superficially resembles the antorbital cavity you might recognise from the skulls of dinosaurs and other archosaurs, so here is obvious evidence that horses are not mammals, but actually archosaurs. I'm joking: regular readers will recognise one-character reclassifications of this sort as characteristic of the Mihalda School.

Preorbital fossae are not unique to horses among perissodactyls: they're also present in various fossil rhinos, and in some living and fossil tapirs. As we'll see, however, these various structures are highly distinct, non-homologous and probably associated with very different bits of soft-tissue biology.


Among rhinos, preorbital fossae are present in the amynodontids Rostriamynodon and Amynodon, and in Forstercooperia, Juxia and Paraceratherium. The last three taxa have often been grouped together in a small group termed Indricotheriinae, itself 'traditionally' contained within the more inclusive Hyracodontidae (e.g., Prothero et al. 1986, Prothero 1998); the presence of a preorbital fossa has been regarded as an indricotheriine synapomorphy. However, the preorbital fossae of these various 'indricotheriines' aren't that similar after all. The structures in Juxia are shallow, positioned high up on the sides of the maxillae, and not all that different from the fossae seen in various tapiroids [the adjacent Juxia skull photo shown here is by Captmondo and is from wikipedia], those in Forstercooperia are larger and partially roofed by the nasals (an amnynodontid-like condition), while those in Paraceratherium [skull shown below: by Ryan Somma, from wikipedia] are very dorsal in position and located immediately posterior to the enormous narial incision.


Holbrook (1999, 2001) included data from these characters and many others and didn't support indricotheriine monophyly, nor did he find these taxa to be within Hyracodontidae. Instead, Paraceratherium and Forstercooperia were successive sister-taxa to an amnyodontid + rhinocerotid + Uintaceras clade. While the preorbital fossae present in the taxa discussed here look rather distinct, the peculiar thing is that - under Holbrook's (2001) topology - it looks most likely that fossae evolved once in rhinos (early on in the clade that includes amynodontids, Paraceratherium and rhinocerotids), but were then lost a few times independently.


A slightly different structure - a semi-circular depression present closer to the eye socket - is present in the tapiroids Helaletes, Colodon, Plesiocolopirus, Protapirus and Tapirus (Holbrook 1999, 2001) [in the Tapirus skull shown here, the structure we're interested in is the rounded concavity directly above the eye socket and at the back of the deep nasal cavity. The nasal bones ordinarily roof the whole of the nasal cavity, but are very obviously well retracted in tapirs]. However, while this concavity is sometimes called a preorbital fossa, it's very different from the same-named structures of horses and rhinos and could well have a separate origin.

Facial glands? Elaborate facial muscles? What gives??

So what were, and are, these bony 'pockets' for? Did/do they house soft structures of some sort? Several roles for the fossae present in fossil horses have been suggested. One is that they housed inflatable diverticula that were used in somehow enhancing the volume or resonance of vocalisations (Gregory 1920, MacFadden 1984). Another is that they housed glands like the superficially similar preorbital pockets present in deer (Meladze 1967) [and, no, I don't think hippariorine horses had inflatable sacs on the sides of their faces - as shown below - but I wanted to know what a horse would look like with them].


Yet another hypothesis is that they were the attachment sites for lip and/or nasal muscles that allowed the lips and nasal region to be flared in a manner more elaborate than that seen in any living equids (MacFadden 1994). Possible links with feeding behaviour have been mentioned on occasion: quite how facial fossae relate to ecology has never been explained, but Zhegallo (1978) suggested that the fossae might be correlated with a complex lip and jaw musculature that was itself linked to complicated jaw movements used in browsing.

The structures termed preorbital fossae in tapirs house cartilaginous structures associated with the proboscis, and are better termed meatal diverticulum fossae (Witmer et al. 1999). Similar structures in some fossil tapirs almost certainly served the same role (Colbert 2005), but this doesn't help us in interpreting the preorbital fossae of fossil horses and rhinos as these fossae - particularly those of horses - are in a very different location and don't seem to be associated with nasal cartilages.

For now, the story behind the evolution of preorbital fossae in horses and other perissodactyls remains enigmatic, and we await a key discovery (perhaps a beautifully preserved fossil with some preserved soft tissue). Anyway... few (if any) previous sources have combined discussions of guttural pouches, false nostrils and preorbital fossae: I hope this review has been interesting, and perhaps that it proves useful in future.

For the previous articles on pouches, pockets and sacs in mammal heads, necks and chests, see...

For previous Tet Zoo articles on perissodactyls, see...

Refs - -

Baptiste, K. E., Grandage, J. & Johnson, K. 1993. Is the guttural pouch of the horse a brain-cooling device? Journal of Anatomy 183, 174.

Baptiste KE, Naylor JM, Bailey J, Barber EM, Post K, & Thornhill J (2000). A function for guttural pouches in the horse. Nature, 403 (6768), 382-3 PMID: 10667779

Bordoloi, C. C. & Kalita, H. C. 1996. Gross anatomical studies of nasal cavity of rhino calf (Rhinocerous unicornis). Indian Veterinary Journal 73, 470-472.

Cave, A. J. 1973. The bursa epipharyngea in the Sumatran rinoceros (Didermocerus sumatrensis). Mammalia 37, 654-657.

- . 1974a. The epipharyngeal bursa in the Rhinocerotidae. Journal of Zoology 172, 133-145.

- . 1974b. Bilocular epipharyngeal bursa in Diceros bicornis. Journal of Zoology 174, 159-169.

Colbert, M. W. 2005. The facial skeleton of the Early Oligocene Colodon (Perissodactyla, Tapiroidea). Palaeontologia Electronica 8 (1); 12A:27p; http://palaeo-electronica.org/paleo/2005_1/colbert12/issue1_05.htm

Endo, H., Manglai, Fujisawa, M., Kurohmaru, M. and Hayashi, Y. 1998. The guttural pouch is not present in the white rhinoceros (Ceratotherium simum); morphology of the Eustachian tube and nasopharynx. Anatomia, Histologia and Embryologia 27, 327-330.

- ., Taru, H., Hayashida, A., Kimura, J., Itou, T., Koie, H. & Sakai, T. 2009. Absence of the guttural pouch in a newborn Indian rhinoceros demonstrated by three-dimensional image observations. Mammal Study 34, 7-11.

Gregory, W. K. 1920. Studies in comparative myology and osteology, no. V: On the anatomy of the preorbital fossae of Equidae and other ungulates. Bulletin of the American Museum of Natural History 42, 265-284.

Holbrook, L. T. 1999. The phylogeny and classification of tapiromorph perissodactyls. Cladistics 15, 331-350.

- . 2001. Comparative osteology of early Tertiary tapiromorphs (Mammalia, Perissodactyla). Zoological Journal of the Linnean Society 132, 1-54.

MacFadden, B. J. 1984. Astrohippus and Dinohippus from the Yepómera Local Fauna (Hemphillian, Mexico) and implications for the phylogeny of one-toed horses. Journal of Vertebrate Paleontology 4, 273-283.

- . 1994. Fossil Horses: Systematics, Paleobiology, and Evolution of the Family Equidae. Cambridge University Press, Cambridge.

Meladze, G. K. 1967. Gipparionovaja fauna Arkneti I Bazaleti. Izdatel'stvo "Metsnierebl" 1-168.

Prothero, D. R. 1998. Hyracodontidae. In Janis, C. M., Scott, K. M. & Jacobs, L. L. (eds) Evolution of Tertiary Mammals of North America. Volume 1: Terrestrial Carnivores, Ungulates, and Ungulatelike Mammals. Cambridge University Press, pp. 589-594.

- ., Manning, E. & Hanson, C. B. 1986. The phylogeny of the Rhinoceratoidea (Mammalia, Perissodactyla). Zoological Journal of the Linnean Society 87, 341-366.

Witmer, L. M., Sampson, S. D. & Solounias, N. 1999. The proboscis of tapirs (Mammalia: Perissodactyla): a case study in novel narial anatomy. Journal of Zoology 249, 249-267.

Zhegallo, V. I. 1978. Gippariony Tsentral'noj Azii. Sovmestnaya Sovetsko-Mongol'skaya Paleontologicheskaya Ekspiditsiya, Trudy 7, 1-152.


More like this

Over the past several months months and months I've been trying to complete a series of articles on the various sac- and pocket-like head and neck structures that have evolved in such diverse mammals as apes, horses, camels and baleen whales (links to the previous articles in the series are…
Everybody knows that camels are weird. As you'll know if you've been keeping an eye on SV-POW! lately, we've recently been quite taken with their necks. But it's not just camel's necks that are weird. Here, we embark on another look at the sometimes bizarre pouches, pockets and sacs present in…
I really must get this series on pouches, sacs and pockets finished. Last time, we looked at baleen whales (and then I got distracted by Caperea): in these animals, a large, inflatable laryngeal sac is used in producing loud, resonating noises (though roles in gas storage or the mechanics of…
Before I start: TIANYULONG TIANYULONG TIANYULONG TIANYULONG TIANYULONG. Ok, moving on... It is the contention of some that the field of Mesozoic reptile research is plagued with bizarre, nonsensical hypotheses. You may or may not agree with me that skim-feeding giant pterosaurs, wind-surfing sail…

Sorry Darren, that is spam in Turkish about an Islamic search engine?

[from Darren: thanks, now removed]

Great post Darren. The false nostrils are totally new to me.

Is it just a coincidence that horses and hyraxes contain both guttural pouches and dorsoventrally deepened posterior portions of the dentary bones? Tapirs seem to fit this bill too (rhinos not so much), although their dentary bones are not quite as expanded. I'm guessing with guttural pouches one needs more room?

Besides the antorbital fossae... I note that an accessory recess (associated with the anterior tympanic recess) around the internal carotid canal is found in many theropods (e.g. Piatnitzkysaurus, Allosaurus, Tyrannosaurus), so may have had a similar cooling function.

I thought that hyraxes were distantly related to horses through the Phenacodontine lineage?

By Tim Morris (not verified) on 07 Dec 2010 #permalink

I thought that hyraxes were distantly related to horses through the Phenacodontine lineage?

Nope. Eocene hyraxes look considerably less perissodactyl-like than extant hyraxes, and there is morphological evidence for affinities of the hyraxes to some or all of the other afrotherians.

By David MarjanoviÄ (not verified) on 07 Dec 2010 #permalink

IMO the common ancestors of the rhino/horse/tapir and of the (separate) elephant/hyrax/manatee were both wetland foragers eating near-surface vegetation (beaver-muskrat style), with dentition and facial morphology reflective of that, but with more prehensile upper lip. It isn't surprising that the descendants would greatly diverge when their habitats changed (drying). Does the guttural pouch have inlet/outlet for air exchange?

One thing to note with horses is that they aren't just indiscriminate grazers; they're capable of quite a lot of fairly delicate manipulation using the lips and jaws. One horse my mother owned had a taste for rose hips, the fruit of dog-rose bushes. These roses are equipped with particularly ferocious spines, so to pick these from the bush the horse would roll its lips right back off its teeth, and delicately nip the fruits off one by one. Quite a feat for an animal that couldn't easily see where its teeth were when it was close to the plant.

Horses, especially ponies (which are smarter than larger breeds) are noted escape artists, and quite capable of manipulating stable door bolts to unfasten them and get out. This is why stable doors commonly have a top bolt and a lower rotating catch, intended to be operated by a human foot; this lower catch is there as a safeguard to prevent such escape attempts. Seriously bored horses are even quite capable of untying knotted ropes and similar items.

By Dan Holdsworth (not verified) on 07 Dec 2010 #permalink

Re equine nasal diverticula, the Wikipedia article 'Respiratory system of the horse' states that they "filter[s] debris with the help of the hairs lining the inner nostril", but cites no specific supporting reference.

By Terry Hunt (not verified) on 07 Dec 2010 #permalink

I was wondering if there's any chance that the nasal diverticula of horses play a glandular and/or sensory role. Horses routinely greet one another by puffing air into each other's nostrils. Could they be exchanging scent information?

Hi Lynda. What you suggest is plausible, but why would the false nostrils (and not the 'normal' nostrils alone) be required for this? Part of the false nostrils are lined with epithelium, but I don't think there's any evidence showing that they serve a sensory role.

Re Lynda's suggestion: maybe the false nostrils are not sensory, but odoriferous? Like axillary hair? Puffing could exchange heavily-scented air from the false nostrils - to be sampled maybe by each other's Jacobson's Organs?

By Graham King (not verified) on 08 Dec 2010 #permalink

Re the preorbital fossae. My first thought was that it was much more likely that they were used to house preorbital glands as in deer and also in many bovids.
I've been trying to find out whether the fossae, and the glands themselves, are homologous structures within artiodactyls (or even within bovines). Certainly there's a lot of disparity in the structure of the glands, with one set of authors preferring not to dignify the cervid 'preorbital sac' with the name 'gland'. Also it seems that most (all?) artiodactyls have preorbital glands of some kind (though I'd be pushed to find an actual reference) but they don't all have preorbital fossae (of that, I'm sure). Do living perissodactyls have preorbital glands?

By farandfew (not verified) on 09 Dec 2010 #permalink

without meaning to sound too snarky, the discussion in MacFadden 1984 about the function of the nasal diverticulae in horses comes from conversations with me back in the 1970s (when he was a grad student at the AMNH and I was visiting). The nasal diverticulae don't connect to the nasal cavity, so there is no way that they can be involved in filtering air, etc.

However, if you stuff cotton balls up the diverticulae (or cotton wool if you're British, and yes I have tried this), the horse will be unable to snort, and will attempt to remove them by trying to snort, finally blowing them out with a loud trumpet. They're definitely involved with this sort of communication in modern horses. Note that the reduction of the fossae supposedly housing the diverticulae in Equus is seen at the point of the reduction of the number of genera of equids to one -- and probably only one kind of horse at any given locality (as with equids today, more or less). My guess is that they were used for vocal communication involved in species recognition in the days when there were 6 - 8 different kinds of fossil horses at any one locality (e.g., late Miocene).

The fossae in perissodactyls look nothing like the fossae housing preorbital gland in artiodactyls, where the bone is sort of fenestrated (not really what I mean -- looks sort of "lacey", not sure if there's a technical term for that). Complex facial muscles might be part of the deal, but muscles don't originate from deep fossae (and hard to imagine how the upper lip of an extinct horse could be *more* mobile than a living one --- remember Mr. Ed the talking horse).

By Christine Janis (not verified) on 12 Dec 2010 #permalink

The fossae in perissodactyls look nothing like the fossae housing preorbital gland in artiodactyls, where the bone is sort of fenestrated

I'm going to have to have to wait till I get to an appropriate museum before I know what you mean - I guess I haven't seen enough perissodactyl skulls. I can't see any difference which shows up in photos. But you are thinking antelopes, as well as deer, when you talk about Artiodactyls, right?

My guess is that they were used for vocal communication involved in species recognition in the days when there were 6 - 8 different kinds of fossil horses at any one locality

Not that it affects your other arguments but this would also apply to the preorbital gland theory if you replace 'vocal' with 'olfactory'.

By farandfew (not verified) on 12 Dec 2010 #permalink