sign in sign up
<<
Home , Choana, Nasal cavity, Nasal placode, Nasal septum, Nose, Vomeronasal organ

AM. ZOOLOCIST, 7:397-413 (1967).

Evolution of the Nasal Structure in the Lower

THOMAS S. PARSONS

Department of Zoology, University of Toronto, Toronto, Ontario, Canada

SYNOPSIS. The gross structure of the nasal cavities and the distribution of the various types of lining them are described briefly; each living order of and possesses a characteristic and distinctive pattern. In most groups there are two sensory areas, one lined by with libers leading to the main and the other by vomeronasal epithelium Downloaded from https://academic.oup.com/icb/article/7/3/397/244929 by guest on 04 October 2021 with fibers to the accessory bulb. All except turtles have the vomeronasal epithelium in a ventromedial outpocketing of the , the Jacobson's organ, and have one or more conchae projecting into the from the lateral wall. Although urodeles and turtles possess the simplest nasal structure, it is not possible to show that they are primitive or to define a basic pattern for either amphibians or reptiles; all the living orders are specialized and the nasal of extinct orders is unknown. Thus it is impossible, at present, to give a convincing picture of the course of nasal evolution in the lower tetrapods.

Despite the rather optimistic title of this (1948, squamates), Stebbins (1948, squa- paper, I shall, unfortunately, be able to do mates), Bellairs and Boyd (1950, squa- iittle more than make a few guesses about mates), and Parsons (1959a, reptiles). Most the evolution of the nose. I can and will of the following descriptions are based on mention briefly the major features of the these works, although others, specifically nasal anatomy of the living orders of cited in various places, were also used. amphibians and reptiles, but all of these Further references may be obtained from are more or less specialized. Truly primitive the bibliographies given by Matthes (1934), amphibians and reptiles are known only Parsons (1959a), and the other papers as fossils and, since the nasal capsules are cited. virtually never ossified in them, their nasal In the following descriptions and discus- anatomy cannot be studied directly. It is sion I shall limit myself to a consideration possible to attempt working back to a com- of the structure of the nasal cavities with mon and hopefully primitive pattern from brief comments on their innervation and the living forms, but such an attempt is development. The structure of the nasal dangerous since it generally involves un- capsule and the nasal physiology with its likely assumptions, e.g., that the loss of behavioral aspects are also important for structures is uncommon and that secondary any complete understanding of nasal evo- simplification is rare. lution, but time does not permit their con- The literature on the nose is extensive. sideration here. Before describing the vari- The most important review is that by ous groups I should first review the major Matthes (1934). Other major works include subdivisions and structures of the those of Sarasin and Sarasin (1887-1890, nose. gymnophionans), Seyclel (1895, amphi- bians; 1896, turtles), Hoppe (1934, rhyn- MAJOR NASAL STRUCTURES chocephalians), Schuch (1934, urodeles), The nasal cavities of most tetrapods may Bertau (1935, crocodilians), Helling (1938, anurans), Malan (1946, squamates), Pratt be divided into three main parts. In most cases the largest is the central cavum nasi I wish to thank Dr. Margaret C. Parsons for proprium, a variably shaped but typically her assistance in the preparation of this manuscript. enlarged cavity which is lined, in part, by Some of the work reported here was supported by Grant A-1724 from the National Research Council the sensory olfactory epithelium. Anterior of Canada. to this, between it and the external naris, (397) 398 THOMAS S. PARSONS

there may be a more or less tubular the nasal cavity which forms a distinct vestibulum. The latter is often lined by sensory structure; in many cases it becomes squamous epithelium resembling the epi- associated with the oral cavity of the adult dermis and is sometimes said to be formed and loses its connection with the nasal from epidermal epithelium rather than cavity. Other workers, most notably Seydel from epithelium derived from the nasal (1895 and 1896), have considered not only placode of the embryo, but these characters such distinct outpocketings to be the are variable or difficult to determine; thus, Jacobson's organs, but also areas of sensory epithelium presumably homologous to

the term vestibulum is here used in a gross Downloaded from https://academic.oup.com/icb/article/7/3/397/244929 by guest on 04 October 2021 morphological referring to any more- them in forms that lack such outpocketings. or-less tubular connection between the ex- Their homology is recognized by three ternal naris and the main portion of the main characters: first, the sensory epitheli- nasal cavity. Posterior to the cavum nasi um of Jacobson's organ in the broad sense proprium there is commonly another tu- lacks Bowman's glands while these glands bular portion, the nasopharyngeal duct, are present in the olfactory epithelium of leading to the or internal naris. virtually all tetrapods (neotenic urodeles Some authors restrict the term naso- and one genus of sea are the only pharyngeal duct to the channel dorsal to a known exceptions); second, the olfactory well-formed secondary , but I will nerve fibers from jacobson's organ tend to here use it for any posterior tubular lead to the accessory olfactory bulb while portion of the nasal cavity and will not, at those from the olfactory epithelium lead to this time, consider the very extensive the main olfactory bulb; and, third, Jacob- literature on the types of in reptiles son's organ tends to be ventrally located and their relationships to the nasal cavities while the olfactory epithelium lies mainly in the dorsal parts of the cavum nasi pro- (for reviews of this topic see the papers of prium. Although in this paper I shall use Fuchs, 1915, and Parsons, 1959a). Either the first and more restricted definition of the external naris or the choana may enter Jacobson's organ, I wish to emphasize the the cavum nasi proprium directly in which homology of that organ with those regions case the vestibulum or the nasopharyngeal lined by sensory epithelium which meets duct is not present. the three criteria presented above. Such epi- Projections of the lateral nasal wall of thelium is here termed vomeronasal epithe- the cavum nasi proprium (or, more rarely, lium in contrast to the remaining types of of the vestibulum) into the nasal cavity are epithelia found in the cavum nasi propri- termed conchae. Although Gegenbaur um, sensory olfactory epithelium and non- (1873) and de Beer (1937) have proposed sensory . In no case different and even mutually exclusive does a tetrapod possess both a true Jacob- definitions in an attempt to give greater son's organ and other areas of vomeronasal precision to this term, I prefer to use epithelium within the nasal cavity. concha in a general sense. Neither of the restricted definitions assists in the re- SARCOPTERYGIAN cognition of homologies nor even of gen- eral structural similarities and the more Since the sarcopterygian fishes are uni- general usage of the term is convenient versally accepted as the ancestors of the (see Parsons, 1959a, for a discussion of tetrapods, it seems logical to start a discus- these definitions). Conchae are found only sion of the evolution of tetrapod with in amniotes. a few words on their nasal anatomy. Un- The Jacobson's or is fortunately, however, such a procedure tells frequently an important accessory olfactory us essentially nothing about the nasal organ in tetrapods. It has been defined in cavities of the most primitive land forms. two different ways. Some authors restrict The best known sarcopterygians are the the term to a ventromedial outpocketing of ; they have been described most TETRAPOD NASAL STRUCTURE 399 recently by Bertmar (1965 and 19666) and presence and, to a certain extent, the shape Thomson (1965). In the Dipnoi the nasal of a Jacobson's organ in some rhipidistians, cavities are basically like those of most but, on the basis of Thomson's study, I feel other and show few, if any, tetrapod that any such identification of Jacobson's characters. They are relatively simple sacs organ must be considered extremely ten- with numerous olfactory folds or lamellae. tative although quite possibly correct. In Although the posterior does open any event the uncertainty is great enough into the oral cavity, there is general agree- to make further speculation on rhipidistian

ment that this opening does not correspond noses unprofitable for the purposes of this Downloaded from https://academic.oup.com/icb/article/7/3/397/244929 by guest on 04 October 2021 to the choana of tetrapods, a structure com- paper; for detailed descriptions, specula- pletely lacking in the lungfish. Various tions, and references, see the papers by workers have described different parts of Jarvik (1942) and Thomson (1964). the dipnoan nasal cavity as a Jacobson's organ, but Bertmar (1965) has shown that URODELA there is no evidence for the existence of any The nasal cavities of many urodeles are such organ in lungfish except, possibly, for very simple (Fig. 1). There is little or no Rudebeck's (1944) report of a rudimentary vestibulum and no nasopharyngeal duct. accessory olfactory bulb. I am unconvinced In most forms, such as Triturus and Sala- by Rudebeck's description and doubt that mandra, the cavum nasi proprium consists an accessory bulb is present. of a rather large, dorsoventrally-flattened, The crossopterygians include the actual ovoid chamber. Along much of its lateral ancestors of tetrapods and are thus of great or ventrolateral margin, there is a lower, interest. The only living representatives of lateral, groove-like extension (Figs. 1A and this order are , a highly diver- B); this extension is termed the "seitliche gent group in which the nasal cavities lack Nasenrinne" in most of the German litera- choanae and bear little resemblance to ture and the lateral nasal sinus in the Eng- those of tetrapods. Rhipidistians, which lish. The lachrymal duct enters the presumably did resemble primitive tetra- anterior part of this sinus. Its posterior pods, are extinct and therefore only their half bears vomeronasal epithelium and thus skeletal anatomy can be studied. Jarvik is often called the Jacobson's organ. Ol- (1942), in a very detailed study of the factory epithelium covers much of the snout of these forms, believed that some of dorsal and ventral walls plus the anterior them, the Osteolepiformes of his classifi- part of the medial wall of the main portion cation, showed marked resemblances to the of the cavum nasi proprium. Vomeronasal modern Anura while others, the Porolepi- epithelium occurs laterally and ventro- formes, were more similar to the Urodela. laterally in the posterior half of the lateral However Thomson (1964) has questioned nasal sinus and the remaining parts of the the validity of Jarvik's conclusions and cavum are lined by respiratory epithelium. does not accept the division of the rhi- Although the pattern just described may pidistians into osteolepiforms and porolepi- be considered typical for fully metamor- forms in Jarvik's sense. Both authors re- phosed forms, considerable variation is ported certain ridges and depressions in the seen when the neotenic urodeles are con- bony wall of the nasal capsule; although sidered. Some, such as neotenic Ambystoma Jarvik tried to correlate these with soft (axolotls), resemble the metamorphosed structures of the nose, Thomson stated (p. forms closely and others, such as Proteus, 347) that "It is not possible to decide with have much simplified nasal cavities with no any degree of certainty whether or not the lateral nasal sinus. Siren (Fig. 1C) is quite apparent morphological subdivisions of the distinctive, having an inverted T-shaped nasal capsule represent any functional ana- "lateral nasal sinus" which lies ventral to tomical features of the nasal sac." Jarvik the medial half of the main portion of the believed that he could demonstrate the cavum nasi proprium. In many neotenic 400 THOMAS S. PARSONS

CNP

LNS Downloaded from https://academic.oup.com/icb/article/7/3/397/244929 by guest on 04 October 2021

CNP

CNP—1 I— LNS FIG. 1. A. Dorsal view of the ventral half of thelium is cross-hatched, and non-sensory epithelium the nasal cavity of Triturus (mainly after Matthes, is solid black. C. Transverse section through the 1934, and Schuch, 1934). B. Transverse section nasal cavity of Siren (after Seydel, 1895). D. Dia- through the nasal cavity of Triturus (after Seydel, gram of the nasal area of a urodele in transverse 1895). In this and subsequent figures of sections, section showing the nasal innervation (after Par- olfactory epithelium is lined, vomeronasal epi- sons, 19596). urodeles the nasal epithelium is thrown or ventral to the nasal cavity and tend to into moderately high folds, quite similar to enter the main bulb (Fig. ID). Herrick the olfactory folds of fish, and Bowman's (1921) gave the most detailed description glands are often completely lacking in the of the olfactory and bulbs in a olfactory epithelium. Not all the neotenous urodele and stated that the separation of forms have been studied in detail and the the fibers from the olfactory and vomerona- extent of the variation in them cannot be sal epithelia was not complete. stated. There is a very extensive literature on The olfactory and vomeronasal epithelia the of the nasal cavities of give rise to nerve fibers which extend urodeles. Most of it concerns the method of posteromcdially to enter the main and ac- formation of the choanae, which is quite cessory olfactory bulbs, the latter being a different from that seen in amniotes and rather indistinct posterior prolongation of has caused some workers to deny the the lateral side of the main olfactory bulb. homology of the choanae in the two groups. All of the fibers from the vomeronasal Bertmar (1966a) has restudied this prob- epithelium pass ventral to the nasal cavity lem and concluded that the choanae are in their path to the bulb and tend to enter indeed homologous in all tetrapods; the the accessory bulb while those from the ol- reader is referred to his paper for a discus- factory epithelium may pass either dorsal sion of the earlier work and many refer- TETRAPOD NASAL STRUCTURE 401

CNP

B Downloaded from https://academic.oup.com/icb/article/7/3/397/244929 by guest on 04 October 2021 CNP

CH

FIG. 2. A. Ventral view of a model of the left the nasal cavity of a larval Ichthyophis (after nasal sac of Ichthyophis (after Sarasin and Sarasin and Sarasin, 1887-1890). The midline is to Sarasin, 1887-1890). B. Transverse section through ihe left; the section to the upper left is the most the nasal cavity of Ichthyophis (after Sarasin and anterior and that to the lower right the most Sarasin, 1887-1890). C. Transverse sections through posterior. ences. It appears most probable that there medial three-quarters of the cavum and are no basic differences between the nasal respiratory epithelium the lateral quarter. cavities of urodeles and those of other The external naris enters the anterior end tetrapods, but that the urodeles have con- of the nasal cavity and the choana enters siderably modified (specialized) the devel- the oral cavity from the posterolateral cor- opmental processes leading to the adult ner of the cavum nasi proprium. Anterior configuration. to the choana and ventral to the cavum there is a diverticulum lined by respiratory GYMNOPHIONA epithelium. Just dorsal to this diverticu- Ichthyophis is the best known gymno- lum, a tubular structure enters the medial phionan and almost all of the following choanal wall. The tube immediately turns description is based on it; the other genera anteriorly and then laterally so that it ex- which have been studied appear to be tends transversely beneath the lateral half basically similar. of the cavum; the lachrymal duct enters its There is essentially no vestibulum and, lateral end. Vomeronasal epithelium forms although the posterior connection to the the ventral wall of this tube which is, there- choana is slightly drawn out, no well- fore, generally termed the Jacobson's marked nasopharyngeal duct. The cavum organ. nasi proprium is a large, dorsoventrally- The innervation of the gymnophionan flattened chamber (Figs. 2A and B), more nose is not well known but appears to re- or less triangular in shape with the apex semble that of urodeles. A small accessory anterior. Its ventral wall bulges dorsally olfactory bulb is present laterally, and the into the cavity, forming an olfactory emi- vomeronasal epithelium sends nerve fibers nence resembling somewhat that found in to the ventral branch of the ; anurans. Olfactory epithelium lines the whether or not the fibers entering the ac- 402 THOMAS S. PARSONS

A.

MC LD Downloaded from https://academic.oup.com/icb/article/7/3/397/244929 by guest on 04 October 2021

FIG. 3. A. Posterior view of the anterior half the nasal cavity of Alytes (after Helling, 1938). of a model of the left nasal cavity of Bufo (after D. Transverse section through the nasal cavity of Matthes, 1934). B. Anterior view of the posterior Pipa (after Matthes, 1934). E. Diagram of the half of a model of the left nasal cavity of Bufo nasal area of an anuran in transverse section show- (after Matthes, 1934). C. Tranverse section through ing the nasal innervation. cessory olfactory bulb are those from the ters the medial side of the choana at all vomeronasal epithelium has not, to my stages. Transverse sections through the knowledge, been clearly demonstrated, but anterior part of the nasal region of larvae it appears likely judging by the situation show a structure similar to that of urodeles, in other tetrapods. while posteriorly they resemble sections of The embryology of the nasal cavity of embryos (Fig. 2C). gymnophionans is of interest because in some ways it resembles that of amniotes ANURA and in others that of urodeles and anurans. Anurans possess the most complex nasal The basic pattern is like that of am- cavities of any amphibians. There is no niotes: the becomes indented nasopharyngeal duct and little or no ves- to form a nasal pit, the pit lengthens and tibulum, but the cavum nasi proprium is extends posteriorly into the roof of the oral typically subdivided into a series of three cavity, and lateral and medial nasal pro- chambers or cavities (Fig. 3). cesses grow ventrally and fuse ventral to Dorsally and somewhat medially there is the center of the pit, thus separating the a large, nearly spherical, principal cavity. anterior external naris from the posterior In most cases the floor of this cavity forms choana. According to Sarasin and Sarasin an olfactory eminence projecting into the (1887-1890), Jacobson's organ first appears cavity; the eminence may be either a as a ventral or even ventromedial out- simple, more-or-less hemispherical swelling growth of the nasal cavity which later or a thinner, lamellar projection, the becomes displaced laterally although it en- dorsal end of which frequently curves TETRAPOD NASAL STRUCTURE 403 either medially or laterally and thus re- ple, in Pipa the cavum nasi proprium is a sembles the conchae found in amniotes. relatively simple, dorsoventrally-flattened The external naris may enter the principal chamber which has a narrow lumen lined cavity directly (Fig. 3A) or there may be a by respiratory epithelium along its center, vertical flap of tissue separating off a vesti- but wider lumina and olfactory epithelium bulum, the latter sometimes entering the both medially and laterally (Fig. 3D). The more ventral middle cavity (Fig. 3C). The vomeronasal epithelium is restricted to a choana enters the oral cavity from the small diverticulum that projects antero- posterolateral corner of the floor of the laterally from the posterolateral part of the principal cavity. nasal cavity. Downloaded from https://academic.oup.com/icb/article/7/3/397/244929 by guest on 04 October 2021 Ventral to the anterolateral portion of Anurans possess a small accesssory ol- the principal cavity there is a small middle factory bulb located on the lateral surface cavity (all of these parts of the nose have of the forebrain posterior to the main ol- been given many names; I shall not at- factory bulb. According to McCotter tempt to mention the numerous synonyms (1917) and others, nerve fibers from the here). This cavity, which is dorsoventrally vomeronasal epithelium run to the ac- flattened, is connected to the principal cessory bulb, and fibers from the olfactory cavity dorsolaterally and to the inferior epithelium to the main bulb with little cavity ventromedially. The lachrymal duct mixing although they are in contact for enters its posterolateral end. As noted part of their course (Fig. 3E). above, if a vestibulum is formed it may en- Most studies of anuran nasal embryology ter the middle cavity rather than the prin- have been concerned largely with the devel- cipal cavity. The third and most complex opment of the choanae. In general, the chamber is the inferior cavity. This is a process is similar to that seen in urodeles large, dorsoventrally-flattened chamber ly- and, although anurans have not been ing ventral and ventrolateral to the princi- much studied recently so definite state- pal cavity and posteroventral to the middle ments cannot be made, it seems probable cavity. Anteriorly it is connected to the ven- that Bertmar's (1966a) findings on urodeles tromedial end of the middle cavity, and are also applicable to anurans and that, farther posteriorly it is narrowly connected despite the peculiarities in their develop- to the principal cavity and enters the ment, anuran choanae are homologous choana. The inferior cavity is commonly with those of other tetrapods. divided into two main parts, a smaller medial recess and a much larger lateral TESTUDINES recess. The nasal cavities of most turtles are Olfactory epithelium lines almost all of relatively simple. There is a short tubular the principal cavity except for its lateral vestibulum leading posteriorly from the side where it is connected to the external external naris to the cavum nasi proprium naris, choana, and other chambers. Vo- and, posteriorly, a nasopharyngeal duct of meronasal epithelium is found only in the variable length connecting the cavum and medial recess of the inferior cavity which the oral cavity. At the posterior end of this is thus frequently termed Jacobson's organ. duct there is frequently a flap or one or The remainder of the nasal cavity, the more papillae along the lateral margin of lateral portion of the principal cavity, all the choana. In some forms the naso- of the middle cavity, and most of the in- pharyngeal duct has a small anterolateral ferior cavity including all of its lateral recess. The cavum nasi proprium may be recess, is lined by respiratory epithelium. divided into two parts, a posterodorsal ol- There is considerable variation among factory region and a ventral intermediate the anurans, but most of it is in details of region. Both the vestibulum and the naso- the pattern described above. However, pharyngeal duct are connected to the latter some forms are quite different. For exam- region (Fig. 4A). 404 THOMAS S. PARSONS

A. Downloaded from https://academic.oup.com/icb/article/7/3/397/244929 by guest on 04 October 2021

FIG. 4. A. Medial view of the lateral wall of the lateral wall of the nasal cavity of Chelonia the nasal cavity of Emys (after Seydel, 1896). (after Parsons, 1959a). D. Diagram of the nasal B. Transverse section through the nasal cavity area of a turtle in transverse section showing the of Emys (after Seydel, 1896). C. Medial view of nasal innervation (after Parsons, 1959b). The olfactory region is a large, relatively only a single sulcus along the medial simple cavity, roughly hemispherical but wall of the intermediate region. In sea flattened lateromedially. Its ventral side turtles the intermediate region is a nar- is open to the intermediate region from rower, almost duct-like structure with which it is normally separated by low prominent dorsal and ventral recesses an- ridges. On the lateral wall of the olfactory teriorly (Fig. 4C). region there may be a rather low and Almost all of the olfactory region is lined generally quite indistinct projection into by olfactory epithelium. Vomeronasal epi- its lumen; this has been called a concha, thelium lines the sulci of the intermediate but it does not closely resemble the conchae region or, in sea turtles, the recesses of that of other forms and I prefer to call it the region. The ridges between the olfactory Muschelwulst. and intermediate regions and between the The intermediate region varies. In most sulci (or in sea turtles the walls of the turtles it is a large and quite simple cham- tubular portion of the intermediate re- ber with various low ridges along its walls. gion) bear respiratory epithelium. The These ridges separate a series of shallow vestibulum is lined mainly by stratified sulci. In emydines and Chelydra there is a squamous epithelium but may have res- crescentic anterior sulcus and, posterior to piratory epithelium posteriorly, and the that, three or four longitudinally arranged nasopharyngeal duct has respiratory or sulci (Fig. 4B), but tortoises appear to have other non-sensory epithelium. TETRAPOD NASAL STRUCTURE 405

Turtles possess a large accessory olfactory greater part of its length, into the oral bulb that lies posterodorsal to the main cavity through the long choana; there is no olfactory bulb. The nerve fibers from all of nasopharyngeal duct (Fig. 5A). Although the vomeronasal epithelium plus those the cavum nasi proprium opens into the from the olfactory epithelium of the me- mouth, the opening is not straight (Fig. dial wall of the olfactory region form the 5B). The ventral end of the medial trunk of the olfactory nerve most of projects laterally forming a vomerine whose fibers lead to the accessory bulb. cushion and, just ventral to that, choanal Fibers from the dorsal and lateral walls of folds project medially from the lateral the olfactory region form the lateral trunk margins of the palate. This arrangement Downloaded from https://academic.oup.com/icb/article/7/3/397/244929 by guest on 04 October 2021 of the olfactory nerve which enters the presumably prevents food particles from main bulb (Fig. 4D). There is some dispute becoming lodged within the nasal cavities. over the exact arrangement of the nerve The large cavum nasi proprium has two trunks and the degree of separation be- incompletely separated conchae projecting tween the fibers from the different regions into it from the lateral wall. The anterior of the nasal cavity, but most workers have concha is attached along a line from just thought that all fibers from the vomero- posterior to the external naris to the mid- nasal epithelium reach the accessory bulb dle of the cavum and slopes from antero- and all those from the olfactory epithelium dorsal to posteroventral. From its posterior the main bulb (see Parsons, 1959a, for fur- end, the posterior concha continues pos- ther discussion and citation of varying opin- terodorsally, extending nearly to the pos- ions on these points). terior end of the cavum. The free margins The nasal embryology of turtles is quite of the conchae, especially the posterior one, simple. The nasal placode becomes in- tend to curl ventrally. Several different dented to form a nasal pit. This pit elon- parts of the cavum are defined on the basis gates so that its posterior end comes to of their relationships to the conchae, but lie in the roof of the oral cavity. Medial they need not be considered here. and lateral nasal processes develop, ex- Jacobson's organ is a tubular structure tending ventrally on either side of the lying along the nasal septum and opening nasal pit and eventually fusing ventral to into the anteroventral part of the medial the middle of the pit. This fusion separates wall of the cavum nasi proprium through a the anterior external naris from the narrow connection near its anterior end. posterior choana and converts the nasal The lachrymal duct enters the lateral wall cavity into a chamber with a restricted of the cavum opposite it. Some authors opening at each end. The various ridges have disagreed on the exact nature of the along the surface of the cavum nasi pro- connections of these structures, but the prium form gradually in place until the statements just given represent the most adult configuration is reached. In some widely held opinions. turtles there is considerable development of a (in a broad sense) Olfactory epithelium lines the dorsal and hence elongation of the nasopharyngeal half of the cavum nasi proprium, roughly duct. the dorsal surfaces of the conchae and those parts dorsal to the conchae, and vomerona- sal epithelium is restricted to the roof of KHYNCIIOCEI'IIALIA Jacobson's organ. The ventral half of the The nasal anatomy of Splienodon has cavum and the floor of Jacobson's organ been studied by few workers and many bear respiratory epithelium and the vesti- details of its structure are not well known. bulum is lined by stratified squamous epi- The vestibulum is a very short tube leading thelium. medially from the external naris to the Splienodon possesses a small accessory ol- large cavum nasi proprium. The latter factory bulb located well posteriorly on chamber opens ventrally, throughout the the dorsomedial wall of the main olfactory 406 THOMAS S. PARSONS Downloaded from https://academic.oup.com/icb/article/7/3/397/244929 by guest on 04 October 2021

FIG. 5. A. Medial view of the lateral wall o£ Thamnophis shortly before the fusion of the nasal the nasal cavity of Sphenodon (after Hoppe, 1934). processes (after Parsons, 1959a). D. Transverse sec- B. Transverse section through the nasal cavity of tion through the nasal area of an embryo of Tham- Sphenodon (after Hoppe, 1934). C. Transverse sec- nophis shortly after the fusion of the nasal pro- tion through the nasal area of an embryo of cesses (after Parsons, 1959a). bulb. The nerve fibers from Jacobson's organ of the adult and its size and rate of organ mingle with those from the olfactory development reflect its adult condition; it epithelium and have, to my knowledge, is very small in forms, such as crocodilians never been traced to their connections in and , which lack this organ as adults the (Parsons, 1959a). Presumably, as and very large in those, such as most in other forms, fibers from Jacobson's squamates (Fig. 5C), in which it is pro- organ enter the accessory bulb while those minent in adults. At about the same time from the cavum nasi proprium lead to the or very shortly thereafter, the concha first main bulb. appears as an inpocketing of the lateral The nasal embryology of Sphenodon nasal wall (Fig. 5D). In Sphenodon resembles that of all amniotes with the ex- Jacobson's organ retains its embryonic ception of turtles. The external naris and position as an outpocketing of the ven- choana are separated as in turtles, but two tromedial nasal wall, but the development further processes occur at roughly the same of the concha is more complex. Certainly time. Shortly before the lateral and medial the posterior concha of the adult forms nasal processes fuse, an outpocketing of the from the original embryonic concha. Since ventromedial wall of the nasal pit appears. the two conchae are always continuous, the This outpocketing becomes the Jacobson's anterior one also probably develops from TETRAPOD NASAL STRI'CTURE 407 Downloaded from https://academic.oup.com/icb/article/7/3/397/244929 by guest on 04 October 2021

FIG. 6. A. Medial view of the lateral wall of D. Lateral view of the dissected nasal cavity of the nasal cavity of Lacerta (after Leydig, 1872). Callisaurus (after Stebbins, 1948). E. Diagram of B. Transverse section through the Jacobson's or- the nasal area of a squamate in transverse section gan of Anguis (mainly after Bellairs and Boyd, showing the nasal innervation (after Parsons, 1950). C. Transverse section through the nasal 19596). cavity of Anguis (after Bellairs and Boyd, 1950). the original concha, but it is possible that prominent concha projecting medially the anterior concha is a new and separate from its lateral wall (Fig. 6C). As in structure that is found only in Sphenodon. Sphenodon, numerous parts of the cavum may be recognized, but they are not im- SQUAMATA portant for this study. Posteriorly, in snakes The Squamata are by far the largest and and some , there may be a short naso- most diversified order of living reptiles and pharyngeal duct, but in most lizards the their nasal anatomy, as might be expected, cavum nasi proprium enters the oral cavity shows considerable variation. There is, directly. The choana may either be short however, a general pattern which, despite and lie ventral to the posterior end of the numerous modifications in various forms, cavum or it may be a long slit-like opening does appear to be typical for the group. resembling that of Sphenodon; in either This may be seen in Leydig's (1872) often- case its structure is similar to that of copied drawing of Lacerta (Fig. 6A). Sphenodon with a vomerine cushion at the The vestibulum is a relatively small but base of the nasal septum and a medially distinct chamber entered anterolaterally by projecting choanal fold ventrolateral to it. the external naris and connecting pos- Jacobson's organ is a separate, roughly teriorly with the cavum nasi proprium. spherical structure lying ventral to the The latter is a large chamber with a single posterior part of the vestibulum or the an- 408 THOMAS S. PARSONS terior part of the cavum (Fig. 6B). Its ven- Stebbins (1948) has studied this variation Lral side is normally invaginated to form a in a series of lizards and shown that vesti- large mushroom body which fills most of bular modification is often related to the the organ and leaves only a narrow lumen. prevention of sand entering the nasal cav- From the posteroventral part of Jacobson's ities in desert-inhabiting forms. Iguanids organ there is a very narrow duct leading and agamids show the greatest range of to the palate and entering the oral cavity vestibular structures, but members of sev- anterior to the choana; there is, in the eral other families may also be adult, no direct connection between Jacob- greatly modified. Downloaded from https://academic.oup.com/icb/article/7/3/397/244929 by guest on 04 October 2021 son's organ and the remainder of the nasal Squamates typically have a very large ac- cavity. The lachrymal duct enters the duct cessory olfactory bulb which lies posterior of Jacobson's organ, the choana or other and medial or dorsomedial to the main adjacent parts of the nasal cavity or palate, olfactory bulb. The nerve fibers from or both. Jacobson's organ form a quite sharply In forms such as that described, olfactory separate part of the olfactory nerve, the epithelium lines the dorsal or posterodorsal vomeronasal nerve, leading to the acces- parts of the cavum nasi proprium, roughly sory bulb, and fibers from the olfactory the dorsal surface of the concha and parts epithelium of the nasal cavity form another dorsal or posterior to the concha. Vo- trunk leading to the main bulb (Fig. 6E); meronasal epithelium is restricted to the there appears to be very little if any inter- roof and sides of Jacobson's organ. The change of fibers between the two trunks. floor of Jacobson's organ, the ventral and The degree of development of the accessory anterior parts of the cavum nasi proprium, and main bulbs is closely correlated with and the nasopharyngeal duct (if present) the amounts of vomeronasal and olfactory all bear respiratory epithelium. Stratified epithelia. squamous epithelium typically lines the The nasal embryology of squamates is vestibulum, but that region is quite vari- essentially as described in the section on able and other epithelial types may occur Sphenodon. Only one concha, developed there. from the embryonic concha, is present. The Although the pattern just described does only complication is that the anterior part seem to be typical for squamates as a of the embryonic choana becomes closed in whole, there are numerous variants. Snakes such a way that a small anterior remnant are apparently a rather homogeneous is left as the duct of Jacobson's organ group; the vestibulum is short, Jacobson's which is thus separated from the adult organ is very well developed, and a short choana. The posterior extent of such nasopharyngeal duct is present. In aquatic closure or fusion varies, causing the vari- forms the concha may be lost and the ation in the size and position of the adult amount of olfactory epithelium greatly choanae already mentioned. reduced. Lizards are more variable. The olfactory areas, the concha, and Jacobson's organ may be poorly developed or even, as Crocodilians possess exceedingly complex in some chameleontids, completely absent; nasal cavities with a large number of reces- such a general reduction in the olfactory ses and accessory sinuses opening off the apparatus seems most frequently to be main chamber. All of the genera which associated with arboreal habits although have been studied are very similar and one many arboreal forms have very well devel- description will suffice for the Order. The oped nasal organs. The greatest variation vestibulum is a short vertical tube leading occurs in the vestibulum which may be- ventrally from the external naris to the come greatly elongated and even overlie the anterior end of the large cavum nasi pro- cavum nasi proprium and enter the pos- prium. From somewhere near the middle terior end of the latter chamber (Fig. 6D). of the floor of the latter chamber a very TETRAPOD NASAL STRUCTURE 409

l—ECR

EN Downloaded from https://academic.oup.com/icb/article/7/3/397/244929 by guest on 04 October 2021

EN

— PRC PTC—I -CO FIG. 7. A. Medial view of the lateral wall of of a crocodilian showing the various recesses and the nasal cavity of Alligator (after Parsons, 1959a). sinuses (mainly after Bertau, 1935). B. Diagrammatic dorsal view of the nasal cavity long and narrow nasopharyngeal duct eral to the concha, the postconchal cavity leads posteriorly to the choana. Croco- within the postconcha, the postturbinal si- dilians have a fully formed secondary nus connecting the extraconchal recess and palate so that the choanae enter the pos- the postconchal cavity, and the posterolat- terior end of the oral cavity. Jacobson's eral recess lying ventrolateral to the post- organ is completely lacking in adults and concha. The last recess is found in Alliga- the lachrymal duct enters the lateral wall tor and Melanosuchus but not in Crocody- of the cavum nasi proprium near the an- lus. terior end of the latter. Olfactory epithelium lines the parts of There are three conchae projecting into the cavum nasi proprium dorsal and the cavum from its lateral wall (Fig. 7A). posterior to the conchae, including the The preconcha and the concha are elon- dorsal surfaces of the conchae and the dor- gated and partially connected while the sal parts of the preconchal and extracon- postconcha is an isolated oval structure chal recesses. Respiratory epithelium lines bulging into the lumen. The preconcha the more anterior and ventral parts of the and concha are separated by a preconchal cavum and all of the accessory sinuses. recess and the concha and postconcha by an There is no vomeronasal epithelium. Strati- extraconchal recess; both are really part fied squamous epithelium is found in the of the cavum nasi proprium. All of the vestibulum, and some non-sensory epi- other spaces are outgrowths of the cavum thelium, not to my knowledge well de- comparable to the accessory sinuses of mam- scribed, lines the nasopharyngeal duct. mals (Fig. 7B). They are the anteroventral Correlated with the absence of Jacobson's , the caviconchal recess lat- organ is the lack of any accessory olfactory 410 THOMAS S. PARSONS bulb in crocodilians. Nerve fibers from the organs, but in most the nasal cavities are olfactory epithelium lead, as in all cases, to well developed. The vestibulum is defined the olfactory bulb. histologically and is rarely a grossly separ- The nasal embryology of crocodilians is able chamber. Within the large cavum nasi complex (Bertau, 1935). The embryonic proprium there are a number of conchae: outpocketing of the nasal pit which in the anteroventral maxilloturbinal which other amniotes forms Jacobson's organ is appears to represent the embryonic concha, poorly developed and quickly disappears. a more dorsal nasoturbinal, and a series of more posterior ethmoturbinals. There are

Both the preconcha and concha develop Downloaded from https://academic.oup.com/icb/article/7/3/397/244929 by guest on 04 October 2021 from the single early embryonic concha, also several accessory sinuses within the but the postconcha is a separate structure bones surrounding the nose. The naso- at all stages and forms as an inpocketing pharyngeal duct is long and the secondary of the posterolateral nasal wall; it has no palate well developed. Jacobson's organ is apparent homologue in other reptiles. The usually present as a quite small tubular various accessory sinuses and recesses are structure lying along the nasal septum. In outpocketings of the nasal cavity and tend the adult, it may be connected to the nasal to appear only in quite late embryonic cavity, to the oral cavity, or to both. Nerve stages, slowly enlarging to reach the adult fibers from Jacobson's organ lead to an condition. accessory olfactory bulb, while those from the nasal cavity lead to the main olfactory HIGHER TETRAPODS bulb as in other forms. Although birds and fall out- DISCUSSION: TETRAPOD NASAL EVOLUTION side the scope of this paper, both are descended from reptiles and hence a brief From the descriptions just given, it seems survey of their nasal structure may be help- clear that most tetrapods have two quite ful before considering the evolutionary distinct sensory areas in the nose. One is history of reptilian noses. However, no at- lined by olfactory epithelium with Bow- tempt will be made to show the range of man's glands, typically sends nerve fibers to variation or to describe any details of their the main olfactory bulb, and is more or less nasal anatomy. dorsally located; the other, Jacobson's or- Birds commonly possess a relatively large gan in a broad sense, is lined by vomero- vestibulum which is not grossly distinct nasal epithelium without Bowman's glands, from the cavum nasi proprium. There is typically sends nerve fibers to the accessory no nasopharyngeal duct and the cavum olfactory bulb, and is more or less ventrally, enters the oral cavity directly. As in croco- most commonly ventromedially, located. dilians there are three conchae along the Although some forms have lost one or both lateral nasal wall. Although the concha and sensory areas secondarily, this division does postconcha are essentially the same in the seem to be primitive for tetrapods. Fish, two groups, the preconchae are very dif- on the other hand, do not show any com- ferent; as already noted the crocodilian parable separation of sensory areas. No preconcha forms as part of the concha and living fish, to my knowledge, has any Bow- lies in the cavum nasi proprium, but the man's glands nor is there ever a division of avian preconcha is an independent struc- the olfactory bulb into main and accessory ture of vestibular origin. Jacobson's organ bulbs (as already noted, I do not accept is lacking and the olfactory epithelium is Rudebeck's, 1944, suggestion of such a generally reduced in extent. The kiwi is an separation in lungfish). Presumably, there- exception in having a well developed ol- fore, the distinction between olfactory and factory sense and a series of olfactory con- vomeronasal areas first appeared either in chae in place of a single postconcha. primitive amphibians or in their rhi- Mammals are very variable. Some, such pidistian ancestors. The first alternative is as and , have reduced olfactory suggested by the incomplete separation of TETRAPOD NASAL STRUCTURE 411 the areas in modern urodeles in which the real evidence that any of them does. accessory olfactory bulb is poorly devel- Neither do turtles have a true Jacobson's oped, and the second by Jarvik's (1942) organ; rather the vomeronasal epithelium studies on crossopterygians; none of the lines a sizable portion of the main nasal evidence is really convincing, and the time chamber. I have, therefore, suggested of origin of Jacobson's organ or equivalent (Parsons, 1959£>) that turtles diverged from structures cannot at present be determined. the primitive reptilian stock before any Much of the more recent history of these of the other surviving groups of amniotes. Such a suggestion agreed well with certain areas seems to be almost as obscure as their Downloaded from https://academic.oup.com/icb/article/7/3/397/244929 by guest on 04 October 2021 origins, but there does seem to be a com- opinions then current on the early evo- mon pattern with modifications in all lution of reptiles (Olson, 1947) although it amniotes with the exception of turtles; in was in marked disagreement with other all these forms Jacobson's organ appears as theories (e.g., those of Watson, 1957). A a ventromedial outpocketing of the early more recent paper by Olson (1965) has embryonic nasal cavity and, at an almost shown fallacies in the earlier work on equally early stage, a single lateral concha reptilian phylogeny, but has not clarified is formed. Thus, it seems logical to pos- the phylogenetic position of the turtles. It tulate an ancestral form whose nasal ana- is possible that turtles are descended from tomy resembled closely that of Sphenodon a stock possessing the normal amniote nasal except that it had only a single concha characters and that they have lost the resembling that of a squamate. The changes concha and secondarily replaced Jacobson's from such a form that may have occurred organ with a simpler structure, but I find in the evolution of the living groups can this hard to believe. However no theories quickly be summarized. In rhynchocepha- based on the changes in a single organ are lians a second or anterior concha developed reliable and, until further palaeontological either by the splitting of the original evidence is available, the phylogenetic posi- concha or as a new projection from the tion of turtles and the evolutionary history lateral nasal wall. In squamates the an- of their nasal cavities will probably remain terior part of the choana became partially unknown. closed isolating Jacobson's organ and its The situation with regard to the three duct from the rest of the nasal cavity; in orders of living amphibians is similar. some forms the fusion was extensive enough There are several theories concerning their to produce a short secondary palate (the phylogeny (Parsons and Williams, 1963), last in a broad sense). Archosaurs lost but none of them is supported by enough Jacobson's organ, developed a new pos- evidence to be really convincing. Each of terior projection of the lateral nasal wall, the three orders has a distinctive pattern of the postconcha, and then divided into sev- nasal anatomy, and, in the absence of evi- eral lines, two of which survive. In one, the dence showing their probable relationships, crocodilians, the original concha became it hardly seems worthwhile to try to imag- divided into a preconcha and a concha, and ine intermediate morphological stages. The a large secondary palate was formed; in the nasal history of the amphibians is further other, the birds, a vestibular concha complicated by the occurrence of neoteny, evolved. Finally in the -like rep- with consequent aquatic life and the re- tiles more conchae, the nasoturbinal and lated nasal modifications, in many groups ethmoturbinals, were formed and a secon- of urodeles and by the considerable differ- dary palate developed. ences in the early nasal embryology of This scheme leaves the position of the anurans and urodeles as compared with all other tetrapods. turtles in doubt. They do not have any normally developed concha; although the Thus, I am really unable to add any- Muschelwulst and various ridges have been thing important to the discussion of nasal thought to represent the concha, there is no evolution in primitive tetrapods that I 412 THOMAS S. PARSONS wrote seven years ago. Certainly urodeles and vomeronasal organs. Proc. Zoo!. Soc. (Lon- and turtles have the simplest nasal cavities don) 120:269-310. Bertau, M. 1935. Zur Entwicklungsgeschichte des and are very similar. One possible theory is Geruchsorgans der Krokodile. Z. Anat. Errtwickl. that they are primitive and that all other 104:168-202. groups have gradually added further com- Bertmar G. 1965. The olfactory organ and upper plexities. Such a theory is easy to diagram in Dipnoi, an embryological study. Acta and does not necessarily conflict with Zool. 46:1-40. Bertmar, G. 1966a. On the ontogeny and homology phylogenetic schemes based on the fossil of the choanal tubes and choanae in Urodela. record; unfortunately there is no evidence Acta Zool. 47:43-59. Downloaded from https://academic.oup.com/icb/article/7/3/397/244929 by guest on 04 October 2021 to support it. It is, of course, equally pos- Bertmar, G. 19666. The development of skeleton, sible to take a more complex starting point blood-vessels and nerves in the dipnoan snout, and either add further complications or with a discussion on the homology of the dipnoan posterior . Acta Zool. 47:81-150. lose them in different lines; again there is de Beer, G. R. 1937. The development of the no evidence and, as a further problem, no . Oxford Univ. Press, Oxford, real reason to select any particular pattern xxiv -f 552 p. as the starting point. We know that paral- Fuchs, H. 1915. Ober den Bau und die Entwicklung lelism can occur in nasal evolution: a des Schadels der Chelone imbricata. Ein Beitrag preconcha in both birds and crocodilians zur Entwicklungsgeschichte und vergleichenden Anatomie des Wirbeltierschadels. Erster Teil: and a secondary palate in both crocodilians Das Primordialskelett des Neurocraniums und and mammals. We know that structures can des Kieferbogens. In A. Voeltzkow, Reise in be lost: the Jacobson's organ of crocodilians Ostafrika in den Jahren 1903-1905, Wissenschaft- and others and the concha of various liz- liche Ergebnisse 5:1-325. ards. We do not know the nasal morphol- Gegenbaur, C. 1873. Ober die Nasenmuscheln der Vogel. Z. Naturwiss. (Jena) 7:1-21. ogy of extinct groups or the phylogenetic Helling, H. 1938. Das Geruchsorgan der Anuren, history of several of the living orders. verleichendmorphologisch betrachtet. Z. Anat. Therefore we can only make guesses about Entwickl. 108:587-643. the early stages in the nasal evolution of Herrick, C. J. 1921. The connections of the tetrapods. vomeronasal nerve, accessory olfactory bulb and in Amphibia. J. Comp. Neurol. 33:213- 280. ABBREVIATIONS USED ON FIGURES Hoppe, G. 1934. Das Geruchsorgan von Hatteria AC, Anterior concha; AOB, Accessory olfactory pmictata. Z. Anat. Entwickl. 102:434-461. bull): CCR, Caviconchal recess; CD, Choanal di- Jarvik, E. 1942. On the structure of the snout of verticulum; CF, Choanal fold; CH, Choana; CNP, crossopterygians and lower gnathostomes in Cavum nasi proprium; CO, Concha; DR, Dorsal re- general. Zool. Bidrag 21:235-675. cess of intermediate region; ECR, Extraconchal re- Leydig, F. 1872. Die in Deutschland lebenden Arten cess; EN, External naris; IC, Inferior cavity; IR, In- der Saurier. H. Laupp'sohen Bu

Com p. Zool. Harvard Coll. 120:101-277. Seydel, O. 1895. Uber die Nasenhohle und das Parsons, T. S. 19596. Nasal anatomy and the jacobson'sche Organ der Amphibien. Eine ver- phytogeny of reptiles. Evolution 13:175-187. gleichend-anatomische Untersuchung. Morphol. Parsons, T. S., and E. E. Williams. 1963. The re- Jahrb. 23:453-543. lationships of the modern Amphibia: A re- Seydel, O. 1896. uber die Nasenhohle und das examination. Quart. Rev. Biol. 38:26-53. Jacobson'sche Organ der Land- und Sumpf- Pratt, C. W. McE. 1948. The morphology of the schildkroten. Festchr. 70 Geburtst. C. Gegenbaur ethmoidal region of Sphenodon and lizards. Proc. 2:385-486. Zool. Soc. (London) 118:171-201. Stebbins, R. C. 1948. Nasal structure in lizards with Rudebeck, R. 1944. Does an accessory olfactory reference to olfaction and conditioning of the bulb exist in Dipnoi? Acta Zool. 25:89-96. inspired air. Am. J. Anat. 83:183-221. Sarasin, P., and F. Sarasin. 1887-1890. Zur Entwick- Thomson, K. S. 1964. The comparative anatomy Downloaded from https://academic.oup.com/icb/article/7/3/397/244929 by guest on 04 October 2021 lungsgeschichte und Anatomie der ceylonischen of the snout in rhipidistian fishes. Bull. Mus. Blindwiihle Ichthyophis glutinosus, L. In Ergeb- Comp. Zool. Harvard Coll. 131:313-357. nisse naturwissenschaftlicher Forschung auf Thomson, K. S. 1965. The nasal apparatus in Ceylon. Kreidels, Wiesbaden. 2:263 p. Dipnoi, with special reference to Protopterus. Schuch, K. 1934. Das Geruchsorgan von Triton Proc. Zool. Soc. (London) 145:207-238. alpestris. Eine morphologische, histologische und Watson, D. M. S. 1957. On Millerosaurus and the entwicklungsgeschichtliche Untersuchung. Zool. early history of the sauropsid reptiles. Phil. Jahrb., Abt. Anat. 59:69-134. Trans. Roy. Soc. (London), B, 240:325-400. Downloaded from https://academic.oup.com/icb/article/7/3/397/244929 by guest on 04 October 2021