Comparative Ethology of the Large-spotted Genet (Genetta tigrind) and Some Related Viverrids CHRISTEN M. WEMMER m SMITHSONIAN CONTRIBUTIONS TO ZOOLOGY ? NUMBER 239 SERIES PUBLICATIONS OF THE SMITHSONIAN INSTITUTION Emphasis upon publication as a means of "diffusing knowledge" was expressed by the first Secretary of the Smithsonian. In his formal plan for the Institution, Joseph Henry outlined a program that included the following statement: "It is proposed to publish a series of reports, giving an account of the new discoveries in science, and of the changes made from year to year in all branches of knowledge." This theme of basic research has been adhered to through the years by thousands of titles issued in series publications under the Smithsonian imprint, commencing with Smithsonian Contributions to Knowledge in 1848 and continuing with the following active series: Smithsonian Contributions to Anthropology Smithsonian Contributions to Astrophysics Smithsonian Contributions to Botany Smithsonian Contributions to the Earth Sciences Smithsonian Contributions to the Marine Sciences Smithsonian Contributions to Paleobiology Smithsonian Contributions to Zoology Smithsonian Studies in Air and Space Smithsonian Studies in History and Technology In these series, the Institution publishes small papers and full-scale monographs that report the research and collections of its various museums and bureaux or of professional colleagues in the world cf science and scholarship. The publications are distributed by mailing lists to libraries, universities, and similar institutions throughout the world. Papers or monographs submitted for series publication are received by the Smithsonian Institution Press, subject to its own review for format and style, only through departments of the various Smithsonian museums or bureaux, where the manuscripts are given substantive review. Press requirements for manuscript and art preparation are outlined on the inside back cover. S. Dillon Ripley Secretary Smithsonian Institution S M I T H S O N I A N C O N T R I B U T I O N S T O Z O O L O G Y ? N U M B E R 239 Comparative Ethology of the Large-spotted Genet Genetta tigrina and Some Related Viverrids Christen M. Wemmer SMITHSONIAN INSTITUTION PRESS City of Washington 1977 A B S T R A C T Wemmer, Christen M. Comparative Ethology of the Large-spotted Genet (Genetta tigrina) and Some Related Viverrids. Smithsonian Contributions to Zoology, number 239, 93 pages, 37 figures, 18 tables, 1977.?The study describes and analyzes the behavior of the large-spotted genet (Genetta tigrina) in captiv- ity, and treats selected aspects of behavior in eight viverrid genera occupying five different subfamilies. The natural history of the genus Genetta is reviewed, and the relationships of the living members of various subfamilies are discussed. The behavior of solitary animals is treated by considering locomotion, stances, and object- and self-oriented contact patterns. Scent-marking with glandular products, urine, and feces is considered with reference to the communicative properties of the marking behavior and the substance. Sonographic and con- textual analyses are used to classify vocalizations and serve to establish hom- ologies between the calls of several species. Further analyses of social behavior in Genetta and Civettictis are based on encounters within all possible pairs of animals. A small number of behavior patterns predominates during most inter- actions between animals of both species. The primary differences in the communi- cation systems of these two species are probably related to semi-arboreal as o p posed to terrestrial habits. OFFICIAL PUBLICATION DATE is handstamped in a limited number of initial copies and is recorded in the Institution's annual report, Smithsonian Year. SERIES COVER DESIGN: The coral Montastrea cavernosa (Linnaeus). Library of Congress Cataloging in Publication Data Wemmer, Christen M. Comparative ethology of the large-spotted genet (Genetta tigrina) and some related viverrids. (Smithsonian contributions to zoology ; no. 239) Bibliography: p. 1. Genetta tigrina?Behavior. 2. Genetta?Behavior. 3. Viverridae?Behavior. I. Title. II. Se- ries: Smithsonian Institution. Smithsonian contributions to zoology ; no. 239. QL1.S54 no. 239 [QL737.C28] 591'.08s [599'.74422] 76-15195 Contents Page Introduction 1 Background 1 Intra-viverrid Affinities 2 Behavioral Classification and Description 4 Natural History of the Genus Genetta 5 Materials and Methods 6 Acknowledgments 9 Behavior of Solitary Animals 9 Primary Senses 9 Locomotion and Its Ontogeny 10 Shaking and Twitching Movements 16 Stationary Body Positions 17 Stretching Movements 18 Object-oriented Contact Patterns 20 Object-oriented Contact Patterns in Other Genera 23 Communicative Function of Object-oriented Contact Patterns in Genetta 26 Self-oriented Contact Patterns 29 Self-oriented Contact Patterns in Other Genera 32 Ontogeny of Shaking, Stretching, and Self-directed Behaviors in Genetta 32 Elimination 36 Communicative Functions of Elimination in Genetta 36 Some Intergeneric Comparisons and Derivational Schemes 37 Behavior of Associated Animals 42 Companion-oriented Locomotion 42 Companion-oriented Stances 43 Noncontact Body Movements 44 Companion-oriented Contact Patterns 47 Head 47 Limbs 49 Trunk 50 Compound Contact Patterns 50 Companion-oriented Contact Patterns in Other Genera 50 Vocalization 51 Vocalization in Other Genera 54 Analyses of Interaction 59 Materials and Methods 59 Results 60 Effects of the Encounter Introduction Schedule on Behavior in Genetta 60 General Features of Encounters 61 Interaction through Sniffing and Head-darting: the Sniff-Dart Axis . . 74 Discussion 76 Vocalizations among Viverrids 76 Nonvocal Interaction in Genetta and Civettictis 79 Addendum 80 Appendices: Encounter Schedules and Statistical Procedures 81 Literature Cited 90 iii Comparative Ethology of the Large-spotted Genet {Genetta tigrina) and Some Related Viverrids Christen M. Wemmer Introduction BACKGROUND The genus Genetta occupies the niche of the small, semiarboreal, solitary, nocturnal, insectivore- carnivore. The dozen or so species comprising the genus are structurally coherent, exploit nearly all natural and cultivated habitats over most of Africa and southwest Europe, and in many areas are a numerically dominant small carnivore. This success may be considered exceptional, for Genetta, like Didelphis and Tupaiia, is believed to be a relatively unmodified derivative of an ancestral form (Gregory and Hellman, 1939; Colbert, 1961). The view that the viverrines are the most structurally conservative viverrids, however, rests on the premise that tribosphenic molars, attenuated body form, and relatively unrestricted rotatory ability of manus and pes constitute features of the ancestral morphotype (Gregory and Hellman, 1939; Colbert, 1961; and Romer, 1945). Since both fossil viverrines, such as the Oligocene and Miocene genera Paleo- prionodon and Stenoplesictis, and extant civets such as Prionodon, Viverricula, Poiana, and Genetta more or less share these features, the subfamily Viverrinae is considered to be the most central and basal viverrid group (Winge, 1941; Thenius and Hofer, 1960; Gregory and Hellman, 1939; Colbert, Christen M. Wemmer, Conservation and Research Center, Na- tional Zoological Park, Smithsonian Institution, Front Royal, Virginia 22630. 1961; and Romer, 1945). Whether the lineage of the more terrestrial civets {Viverricula, Viverra, Civettictis) or the semiarboreal civets (Poiana, Genetta, Prionodon) represents the least altered derivative of the common ancestor is debatable, for the fossil record is inadequate. What little is known about Poiana richardsoni, though, suggests that a semiarboreal habitus may more closely approximate the ancestral viverrine condition. Besides displaying more cranial resemblance to Paleoprionodon than other extant genera (R. H. Tedford, pers. comm.), Poiana is purported to build leafy arboreal nests (Walker, 1968), a trait interpretable as primitive by comparative carnivore standards. Furthermore, Booth (1954) has inter- preted Poiana and other West African species having disparate distributions to be stabilized forest forms that reached an evolutionary climax before the appearance of the Dahomey Gap. Genetta's capacity for radiation and dispersal into woodland and savanna is a progressive feature that presum- ably emerged early in its lineage. Contrary to the prevailing view of the genet, it is herein acknowledged as a conservative and gen- eralized solitary predator, but not as a model antecedent of more specialized contemporary viver- rids; rather, the genus displays a complex of mor- phological and behavioral characteristics some of which are considered to be fundamental to the evolution of more advanced or specialized forms. With this perspective in mind, the present study attempts (1) to establish qualitative and quantita- 1 SMITHSONIAN CONTRIBUTIONS TO ZOOLOGY tive behavioral characteristics, and to outline the principle features of the communication system in captive genets, and (2) to compare certain behaviors of selected viverrids and interpret the results in light of known taxonomic affinities. INTRA-VIVERRID AFFINITIES Because modern taxonomy strives to arrive at evolutionary classification by delimiting taxa on the basis of common ancestry and subsequent diver- gence (Mayr, 1968), behavioral evolution must ulti- mately derive from morphologically based appraisals of systematic position. Though behavioral criteria are often useful in elucidating relationship (Lorenz, 1953; Lawrence and Novick, 1963; and McKinney, 1965, 1966), their role in clarifying the affinities of higher taxa (families, orders) is minor. Unfortu- nately, the Viverridae have received so little atten- tion that it is questionable whether the existing sys- tematic schemes serve the practical needs of a sound classification. Many genera are rare or con- fined to remote tropical regions, resulting in a scarcity of study animals as well as few preserved specimens. In addition, a number of genera and perhaps one subfamily exhibit character traits that defy a simple scheme of filial relationship. While the pretense of a solution might be made by refer- ring only to the most recent synopses of taxa (Anderson and Jones, 1967; Wenzel, 1972), it would ignore the problematic status of several genera dealt with herein. Though a large number of viverrid species was available to me for comparative study, it became increasingly apparent, as it had to others (Diicker, 1971; Ewer, pers. comm.), that the mongooses (Her- pestinae) exhibited a more uniform and distinctive level of behavioral organization than other viver- rids. Similar observations were made by Pocock (1916b) and Gregory and Hellman (1939) who had advised elevation of the mongooses to familial status on the basis of their uniformity in foot, dental, and glandular characteristics. This sugges- tion was subsequently endorsed by Wurster and Benirschke (1968:374), who found that eight species and seven genera of mongooses examined karyo- tvpicallv were "all nearly identical to one another and without exception, they bear no satellited marker chromosome." In addition, Ewer (1973), noting that all four known herpestine genera had horizontal oval pupils, stated that this structure may be characteristic of the Herpestinae as a whole. A vertical slit pupil is found in all the other known viverrids with the exception of Paradoxurus whose pupil is a horizontal slit (Table 1; cf. Walls in Ewer, 1973). Lastly Radinsky (1971) has shown that in 25 of the 36 living genera representing all sub- families, Crytoprocta, the herpestines, and the gal- idiines all possess a well-developed cruciate sulcus, while the structure is small or absent in viverrines, paradoxurines, and hemigalines. The existence of the Madagascan Galidiinae accounts for some of the difficulty in distinguishing the mongooses (Herpestinae) from the bulk of the viverrids. Both cranially and in gross body form the subfamily shows a strong affinity with the herpes- tines, but unlike the latter they possess the carnivore satellite chromosome (Table 1). Carlsson (1910) showed that Galidia elegans shared 14 anatomi- cal features with the other Malagasy civets (in- cluding Ciyptoprocta and Eupleres), 10 with the mongooses, and 7 with the viverrines, but he con- cluded that relegation of the Galidiinae to either the viverrids or herpestids would obscure its posi- tion as a stem form that differentiated prior to the evolution of either the herpestine or viverrine grade of organization. Gregory and Hellman's decision to place the galidiines within the Viverridae was consistent with their implicit belief that the rate of evolution in viverrids (particularly viverrines) has been much slower than in herpestids. Though the fossa (Cryptoprocta ferox) of Mada- gascar is distinctive at the subfamilial level, it too displays features of conflicting affinity. Although Pocock (1916a) gave the fossa familial status, Gregory and Hellman placed it as a subfamily within the Felidae, believing it to be an excellent structural link between the primitive Viverrinae and the Felidae. This opinion was predicated on skeletal similarities. While its auditory region and dentition resemble the viverrines and felids respec- tively, the presence of an anal pouch (and the absence of the perineal gland) prompted Mivart to ally this species with the mongooses (in Gregory and Hellman, 1939). Karyotypically, however, the fossa has much in common with the paradox- urines, hemigalines, and galidiines (Wurster and Benirschke, 1968). Thus, without interpreting the fossa's affinity, it appears to be an early viverrid NUMBER 239 TABLE 1.?Comparisons of major characteristics among viverrid subfamilies and genera ( - = absence within taxon, + = presence within taxon but not neces- sarily universal, ? = present in some members of taxon, when no character alternatives listed) Characteristics t - T3 Skin Glands Perineal pouch ? Preputial pcoich... Anal gland , Anal pouch Feet Plantigrade Sub-digitigrade... Digitigrade Digit reduction... Retractile claws.. Teeth Sectorial Tuberculate Molar reduction... Food Habits Carnivorous Omnivorous , Piscivorous Insectivorous Myrmecophagic Coloration Spotted Striped Banded Uniform Misc. Features Marker chromosome Vertical pupil Distribution Africa Madagascar Asia Europe survivor displaying convergent or possibly archaic incipient felid features. The African palm civet, Nandinia binotata, re- sembles the paradoxurines in both its gross body form and its skull. On the basis of its more trenchant carnassials and cartilaginous auditory bulla Gregory and Hellman surmised, however, that this civet branched off from the paradoxurine SMITHSONIAN CONTRIBUTIONS TO ZOOLOGY stem prior to frugivorous specialization. The exist- ence of the subossified or cartilaginous ento- tympanic bone of the auditory bulla is generally regarded as a secondarily derived feature rather than a vestige of a miacid ancestry. The extra- ordinary nature of this structure and the presence of a preputial scent gland prompted Pocock (1915b) to separate Nandinia as a family. Ironically, the species strongly resembles the viverrines in its possession of a subtelocentric marker and a small acrocentric Y-chromosome (Wurster and Benirschke, 1968). The affinity of Asian linsangs (Prionodon) to the viverrines resembles that of Nandinia to the para- doxurines. While sharing features of body form with Genetta and karyotype with the the viverrines in general, Prionodon lacks the metatarsal pads, the perineal gland, and the second upper molar. These differences were interpreted by Pocock (1915a) as specializations justifying a distinct subfamilial sta- tus. Similarly, Gregory and Hellman relegated the linsang to a viverrine tribe (Prionodontini). The fanaloka (Fossa fossa) in its possession of vestigial metatarsal pads and a genetlike hair pat- tern is superficially somewhat viverrine; karyologi- cally and cranially it is distinctly hemigaline, while in gross body form it is remarkably foxlike. It is probably an early Asian hemigaline waif to Mada- gascar that has convergently acquired a foxlike habitus. BEHAVIORAL CLASSIFICATION AND DESCRIPTION Ethological studies have traditionally relied upon both functional and motivational criteria in cate- gorizing a species' behavioral repertoire. Many behavior patterns, however, can be assigned to several categories, others can be accommodated by none, and new criteria are invoked in order to categorize special cases (e.g., displacement activities). To design an objective, systematic scheme with mutually exclusive categories is both difficult and overreductionistic, because in most animals a large number of biological roles must be served by a number of movement patterns restricted by the mechanical and physiological properties of the spe- cies' body structure. As much as functional and motivational considerations may enhance a de- scriptive work, an accurate understanding of these factors requires both experimental validation and a knowledge of the species in its natural realm. Because these labors exceed the requirements of faithful description, it seems more objective to classify behavior on observable and measurable characteristics of form, rather than on deduction about motivational state, for instance. It will be noticed that though some new terms are used for descriptive categories, implicitly and explicitly func- tional terms, such as grooming, scent-marking, and play, are retained within the descriptions. Behavioral description has for the most part been an empirical process; the events ethologists recog- nize and interpret as true behavioral units (action patterns) are assumed to represent fuctionally mean- ingful acts unabstracted by the observer's conceptual processes. But in mammals, at least, it is question- able whether the discrete phasic properties of action continua accurately represent all the meaningful dimensions of animal conduct. Symposia titles such as "The Search for Natural Units of Behavior" x convey the distinct impression that (1) recognized behavioral units may be abstractions rather than true or complete representations of natural pheno- mena, (2) the types of units currently used have limited analytical utility, and (3) the elucidation of "natural units" will hopefully permit more insight into the meaning of animal behavior. With the exception of locomotion, which received early and exacting description by virtue of historical and esthetic interest (Muybridge, 1957), most forms of mammalian behavior have not received the analyt- ical descriptions exemplified by the Baerends' (1950) cichlid work or Dane, et al. (1959) and McKinney's (1965, 1966) studies on duck behavior. In particular, there have been few critical attempts to establish either indices of variability or typical intensity (Morris, 1957) of mammalian displays (Le Boeuf and Peterson, 1969, is one exception). The Eshkol-Wachmann Movement Notation Method (Golani, 1969), which permits fine-grain temporal and spatial resolution of behavior, has remained an unexploited descriptive tool. With rigorously objective description, the method pro- vides a valuable alternative to the interpretive methods that have prevailed up to now, though from the standpoint of comparative ethology it 'A lecture presented by S. A. Altmann at a symposium titled "Analytic Problems in the Study of Animal Behavior," 67th Annual Meeting of the American Association for the Advancement of Science (1970). NUMBER 239 requires considerable time. Useful descriptions can be achieved more expediently. Not infrequently, however, descriptions are made in such general terms that they have limited utility: it is often necessary to know the postural and locomotory details as well as the basic mechanics of a motor pattern in order to appraise it comparatively. For this reason, photography is often as integral to the analysis and description of body motion as sono- grams are to the description of sound. The classification used in this study is a phenetic one; patterns are lumped together on the basis of overall similarity in form. The system replaces functional and motivational pigeonholing. For example, the patterns usually ascribed to the main- tenance (or grooming) behavior category are herein subdivided into classes of self- and object-oriented contact patterns. The latter category also encom- passes behaviors that are otherwise classed as com- fort movements (e.g., body rubbing) and scent- marking. As Mayr (1968) has remarked, no classification is absolute because the structuring of character sets always involves a subjective element. The present system does not claim to be consistent in its criteria. Although the orientation of the animal's body parts to one another and to the environment is a primary consideration, certain behaviors (e.g., vocalization) do not fit the scheme. The utility of the system, however, lies in bringing to attention certain elemental similarities in be- havioral form. NATURAL HISTORY OF THE GENUS Genetta The separation of the genus Genetta into three subgenera is based upon characters of the skull, teeth, and foot pads. The majority of species are placed within the subgenus Genetta and these are distinguished by the patterning of spots, spinal stripes, and tail rings (in addition to cranial and dental criteria). Forest species are on the whole darker in color and have more spots than savanna forms (Matschie, 1902). Genetta victoriae appears to be a somewhat aberrant species tending towards large size and terrestriality. Its coloration shows considerable resemblance to that of the African civet (Civettictis civetta) with which it is sympatric (Thomas, 1901). Members of the subgenus Pseudogenetta (villiersi and abyssinica) are characterized by relatively small size and the lack of hair in the depression between the interdigital and metacarpal pads. Paragenetta, represented by one species (johnstoni), is known only from skulls. The delicate dental structure of this species suggests its diet may consist largely of arthropods. Genets occupy nearly all African habitats, but the range of habitat tolerance varies between species. Genetta victoriae, servalina, angolensis, johnstoni, and mossambica are primarily forest dwellers, while G. pardina occurs in forest-savanna ecotone, woodland, and Guinea savanna. Drier nonforest areas are occupied by G. genetta, tigrina (sensu latu), villiersi, and abyssinica (Bigalke, 1968; Coetzee, in press). Although Dorst and Dandelot (1970) state that G. villiersi inhabits forests, a large number of this species in the National Museum of Natural History was collected from Sudan and Guinea savannas in Ivory Coast and Upper Volta. At least in Botswana, G. genetta penetrates con- siderably drier areas than tigrina, though both species are often found together (R. H. N. Smithers, pers. comm.). Displaying a tolerance for habitats ranging from low forest to woodland and savanna, G. tigrina has been collected from elevations as high as 2560 m (8500 ft) on the slopes of Mt. Kenya (Hollister, 1918). It often appears to be a "culture- follower," inhabiting the vicinity of villages and cultivated land (Rahm, 1966). There appears to be a distinct sexual dimorphism as to size in some populations of G. tigrina (i.e., stuhlmanni, pers. observ.), but data are inadequate to make conclusive statements for other species. J. C. De Meneses Cabral (in Coetzee, in press) has found that, in areas of low rainfall, skull length in G. rubiginosa (= G. t. rubiginosa, sensu lato) varies from an average of 87.5 mm in females to 90 mm in males, while in high rainfall areas skulls of both sexes average 90 mm. Tables of the genet cranial measurements in Allen (1924) suggest sexual differ- ences in skull dimensions may exist in G. victoriae and servalina as well. Melanistic genets are sometimes seen in museum collections and zoos (Sanderson, 1940; Maliniak, pers. comm.). The natural frequency of the mela- nistic morph in the wild is unknown. At the Na- tional Zoo a wild-type female mated by a melano produced two wild-type offspring (G. Maliniak, pers. comm.). All known species appear to be nocturnal and SMITHSONIAN CONTRIBUTIONS TO ZOOLOGY solitary, though pairs and family groups are occa- sionally seen. Evidence at present indicates that the most complex social unit is the mother family; its integrity is probably maintained only until the young are weaned. Ducker's (1957) observation of a parturient female's antagonism towards its male cage mate suggests that the sexes remain separate during the rearing of young. Verheyen's (1951) statement that the male participates in providing food for the young is doubtful. The exact nature of the social system, however, is unknown. Presumably a land tenure system exists, but information is sparse regarding activity in and defense of the home area. In the course of capture and release work with genets inclined toward poultry thievery, Carpenter (1970) recaptured four of six female G. rubiginosa ( = G. t. rubiginosa sen- su lato) at their original capture points within days after release from as far away as 20 miles. Six of 11 males were taken again at poultry runs, but none was recaptured at the original trapping site. He attributes the fact that most poultry thieves are males to their greater mobility, and suggests that females, unlike males, may be territorial. Cobb (1970), on the other hand, found that in Turkana, Kenya genets frequently crossed paths; he is of the opinion that in this area, at least, genets are not strictly territorial. Locus-specific defecation, a trait often correlated with the territorial syndrome, also characterizes G. tigrina (Verheyen, 1951; Rahm, 1966; and Carpenter, 1970). The tentative and somewhat contradictory nature of all these findings clearly indicates that more work must be done to elucidate this aspect of the genet's natural history. Breeding in G. genetta and G. tigrina (sensu lato) occurs during the wet seasons in both Kenya (Taylor, 1969) and Botswana (Smithers, pers. comm.). Taylor (1969) collected lactating and preg- nant females in May and from September to Decem- ber, and believes the timing of reproduction during these rainy periods to be linked with an abundance of insects. Pregnant females from Botswana were also taken during the warmer wet period from October to February (Smithers, pers. comm.). Whether wild females breed during both wet sea- sons is uncertain; however, a breeding pair at the National Zoological Park has regularly produced a litter of two in April or May and again in July or August, for the past six years. Females have two pairs of inguinal mammae, and the number of young born in the wild varies from two to four (Hollister, 1918; Shortridge, 1934; Ansell, 1960; Copley, 1950; Roberts, 1951; and Verheyen, 1951). The types of resting, parturition, and rearing sites are probably influenced by habitat. Holes in trees, in fallen logs, and burrows in kopjes and cliff areas seem to be usual locations (Shortridge, 1934). Rowe-Rowe (1970) located three young G. genetta in a disused crow's or hawk's nest about 6 m above ground. Feeding appears to be opportunistic with a strong predisposition for animal matter. In a sample of 78 G. genetta from Botswana Smithers (pers. comm.) found that the three animal groups represented most frequently in stomachs were (in decreasing order) rodents, Orthoptera, and scorpions; in a sample of 30 G. tigrina from Botswana, rodents ranked first, followed by Coleoptera and Orthop- tera. Cobb (1970) reported that on the basis of stomach contents, fecal samples, and observation, the genets of Turkana, northern Kenya fed pri- marily on birds which roost in the dominant shrub, Salvador a persica. Nothing is known about the natural predators of genets; they are presumably preyed upon by larger carnivores and birds of prey, and are sought by humans for food and fur. According to Dekeyser (1956), G. villiersi is killed by the Bassari of Itiolo, Senegal, for its tail, used in ceremonial rituals. Genetta's climbing ability probably provides an important niche parameter, not only in facilitating escape into arboreal refuge, but also in making accessible sources of food and cover that are denied to strictly terrestrial animals. MATERIALS AND METHODS Table 2 presents personal history information concerning the 11 genets on which the results of this study are based. Eight of these animals were born in the Small Mammal House at the National Zo- ological Park, Washington, D.C. For convenience they are referred to in the text by the following abbreviations: K (Kitzi), C (Clawdina), I (Ivy), Mo (Monika), M (Marvin), O (Oscar), E (Elsie), Fr (Froda), and F (Felix). The origin of the parental pair is unknown, though comparison of skins and skulls of the previous offspring of this pair with specimens at the Smithsonian Institution reveals similarity to Genetta tigrina stuhlmanni, a sub- NUMBER 239 TABLE 2.?Caging history (cage locations, cagemate(s), periods of confinement, cage dimensions in meters) of study animals (Genetta tigrina) Subject Parent (P) Parent (P) Kitzi (K) Clawdina (C) Ivy (I) Monika (Mo) Marvin (M)1 Oscar (0) Elsie (E)' Froda (Fr) Felix (F)2 Sex 0* ? o" 5 ? ? i (OH 1 | UJUJ ? LU LU ft z O Q_ I/) LU OC LLO o: ID 30 LU I o LU t- Z LU U A ? 0 Og ?A*-'- r ! B o ?g o ? -c ??io.O O ^^ 0 ? ?? 0 O o?o o ?.o?o . :D o? - E O 0 O j F ? # o? : o - G oo . o ? ? o ? o o 0 :? ? ? ? ? ii-,?.i.- H - o 0 ? o o o o a o o ? o ? ? o? ? ? o ?? ? o o ? ?o? 0 . O ? " ? ? ? o 00 ? 0 0 ? ? -?- v o ?# . o 1 ? ? ? ? 0 ? ? ^ 0 O O Q 0 ? ? "o ? .9.9i. .9. i i. .9,^ . o o ? . ? o o ? o ? o ? ? ? * o ? o ??? , ?*'* ?* o ? ? 5 '5 25 35 45 55 AGE (DAYS) succeeding four days suggest that responses to localized tactile stimulation may well be more varied during early infancy than at later stages. None of the remaining evoked behaviors contri- buted more than 1% to the total number of responses. They can be classified into those that were distributed (1) sporadically during the entire 50-day period (change body position, twitch legs, manus to side of head and ear, synchronous fore- limb face-wiping, stretching); (2) predominantly during the first 25 days of testing (extend hindleg, pes to neck, yawning); and (3) predominantly during the last 25 days of testing (purring). Though scratching can be induced by repetitive stimulation as early as four days after birth, under these testing conditions it was often manifested in an inaccurate or incomplete form. An anterior ex- tension of the hindlimb, which normally precedes scratching, occurred most frequently during the first 7 to 21 days and only occasionally thereafter (Figure 18F). Scratching of the neck (Figure 18E) was irregularly performed by I during the first 20 test days but was an uncommon event in C. While the ear was understandably the most common target of scratching (Figure 18D), the side of the head was occasionally scratched throughout the course of testing (Figure 18c). Ear-twitching and head-shaking commonly oc- curred as part of larger sequences (Figure 19). Table 6 shows that ear-twitching was more frequent over- all than head-shaking, more often a single act, and more common as a serial precedent. In Figure 19 the following points are apparent. First, of those series that began with ear-twitching (A) over half (58.8%) of animal C's and over a quarter (28.5%) of I's re- sponses contained head-shaking (B) as the second serial element. Secondly, seven dyads alone ac- counted for the majority (87.2% in I; 80.5% in C) of all serial combinations that began with ear- FIGURF 18.?Responses to localized tactile stimulation in two infant genets (I and C) over a 50-day period (ages 5-55 days; hollow circles = I, solid circles = C): A, mean number of responses per stimulation bout (total number of responses divided by the number of stimulation bouts completed on that day); B, response diversity (number of response types divided by the total number of responses per test; because the first two values for I are .82, they are omitted to con- serve space); C-H, percentage of the total number of responses per test for the following: c, scratch side of head; D, scratch ear; E, scratch neck; F, extend hindleg; c, shake head; and H, twitch ear. NUMBER 239 35 TABLE 6.?Numbers and frequency (percentage in parentheses) of two responses following bouts of localized tactile stimulation in genets I and C between 5 and 55 days of age Circumstance Single act Preceding act... Succeeding act.. Total Ear 393 434 91 917 Animal twitch (42.7) (47.3) (9.9) I Head 165 56 283 504 shake (32.7) (11.1) (56.2) Ear 299 529 36 864 Animal twitch (34.6) (61.2) (4.2) C Head 76 14 179 269 shake (28.3) (5.2) (66.5) o FIGURE 19.?Percentage occurrence in genets I and C of all behavioral dyads, triads, and tetrads that began with ear- twitching as responses to localized tactile stimulation of the ear (values for I lie below the lines and to the right of the line between A and B; the total of all the figures for each animal equals 100%; values between A and the inner ring of letters represent the percentage occurrences of dyads, between the inner and the second ring of letters, triads, and beyond the second ring, tetrads; A = ear twitching, B = head shaking, c = head jerking, D = move head and replace in original position, E = change head position, H = wipe ear with manus, K = scratch ear with pes, N = extend hind- leg, o = all other behaviors). twitching (A). Thirdly, though moving the head (D) was the most common succedent to ear-twitching in animal I, on only a few occasions did this pair of patterns give rise to triads. In both animals, on the other hand, the sequence of ear-twitching and head- shaking (AB) predominantly gave rise to triads ter- minating with a change of head position (E) and ear-scratching (K). Data not included in Figure 19 shows that when head-shaking was a dyadic ante- cedent for animal I (N = 52), the most common second elements were also change of head position, 26.9%, and ear-scratching, 21.1%. Extending the hindleg accounted for an additional 13.4%, and of the remaining 7 dyad types none was more than 5.7% of the total. Animal C displayed only 12 dyads beginning with a head-shake. The following conclusions may be made. First, the most common responses were movements of the body region most proximal to that of the stimu- lated area (the ear); these proximal responses sub- jectively appeared easier to execute than less com- mon self-directed patterns such as scratching. Sec- ond, the two predominating movements displayed somewhat reciprocal frequency changes over the 50-day test period, but they often occurred together in sequences. The most common triad contained a low effort antecedent (ear twitch) and moderate effort succedents (head-shake, ear-scratch, or change of head position). Third, scratching was the most complicated type of response and though it was presumably an effective means of alleviating the stimulation, it was far less common than proximal responses (ear-twitch and headshake). Lastly, the form of the scratching response was sometimes incomplete and inaccurate. It is difficult to say whether inaccuracy is a developmental characteris- tic, because responses to known sources of tactile stimulation were not observed in adult animals. The intention movement of scratching (anterior extension of the hindleg), however, has not in my recollection been seen during grooming sessions in 36 SMITHSONIAN CONTRIBUTIONS TO ZOOLOGY adults. The pattern is probably an ontogenetic peculiarity as it was seen in Diicker's (1957) young animals and in all of those I hand reared. It more likely results, however, from conditions of the affer- ent or efferent nervous pathways rather than from mechanical restrictions of body form. ELIMINATION URINATION.?The three squatting postures as- sumed during perineal marking are also employed by both sexes during urination, but the crotch does not contact the substrate. Both hindlegs are com- monly on the same level, but if the stream of urine is directed to low-lying objects one leg may be cocked over the object or rested upon it. Lateral swaying of the hindquarters sometimes accompanies urination in both sexes. As urine is expelled, the force and volume seems to vary rhythmically, sug- gesting that flow is controlled by a muscular con- traction rhythm in the urethra. In estrous females, on the other hand, small amounts of urine are often dribbled upon objects. In both sexes urination is usually locus-specific, and animals can easily be trained when young to urinate in a box of litter. Some genets acquire the habit of urinating in their water bowl. Once mature, however, they tend to direct the urine to the elevated edges of the litter box and other low objects on the cage floor. The stance is usually stationary, but occasionally, when urinating on a low-lying object, several short steps may be taken forward. Animal F was seen to urinate in this manner on elevated branches, but in the other animals it was almost always a terrestrial activity. If frightened or traumatized, however, an animal will void urine wherever it may be, and, of course, this often occurs in the inferior animal during fighting. DEFECATION.?In adult animals of both sexes defecation occurs in a squat position with both legs on the same level. The tail is usually slightly elevated and may be pumped up and down once or twice upon completion. The pelvis is also some- times jerked forward at the termination of the squat and this presumably coincident with closing the anal sphincter. In infants, both urination and defecation can occur spontaneously or can be evoked in reclining body postures. ELIMINATION IN OTHER GENERA.?A squatting quadrupedal stance is assumed (with minor varia- tion) during elimination by Civettictis, Nandinia, Paguma, Paradoxurus, Hemigalus, and Fossa. Nan- dinia was frequently seen urinating while slowly walking forward in a squatted posture. Urination in Nandinia, and to a lesser extent in Paradoxurus and Paguma, occurred in and on top of nest boxes, from elevated branches, and directly on the floor. Nandinia also defecated indiscriminately about the cage, but Paradoxurus and Paguma often left feces on the tops of boxes and shelves. Bartels (1964:198) reports that in Western Java Paradoxurus excre- ment is found in elevated positions on top of rocks, on fallen logs, and "strung out on slim branches . . . high above ground . . . as well as in collections on open ground areas." Because "the toddycat keeps going while relieving itself . . . ," the scats have a characteristic "strung-out" (rather than heaped) appearance (1964:198). The same often holds true for Nandinia. We were unable to train the hand- reared Fossa to use a litter box, and even when the animal had free run of a large room it did not dis- play what could be considered locus-specific elimi- nation. The two Civettictis urinated and defecated in the end of the cage opposite the nest box; the male frequently urinated and defecated in his water bowl. In South Cameroun and Senegal, Civettictis often defecates along trails in small openings devoid of grass (Gaillard, 1969). Both Hemigalus and Galidia at the National Zoo urinate in the water trough. Urination in the male Paguma and male Civet- tictis was sometimes closely associated with hindleg movements. The Paguma, for example, would drib- ble a small amount of urine on the floor or its nest- box and then scuff its metatarsus in it. On three occasions during social interaction between the male and female Civettictis the male was seen urinating while walking and scratch-kicking. The urine was spurted to the rear at about 30?-40? to the horizontal as each hindfoot was alternately scratched backward with a kicking motion. The movement resembles the ground-scratching of dogs except that in the civet it occurs during urination as the animal walks, and the kicking stroke is shorter and briefer. COMMUNICATIVE FUNCTIONS OF ELIMINATION IN Genetta Table 5 shows that compared with the various NUMBER 239 37 types of scent deposits resulting from object-oriented contact patterns, urine and feces have a strong and enduring scent value. It is obvious that as odorous chemical mixtures both substances have an intrinsic value as media for information about individual identity, and sexual status and condition (Hediger, 1944). Defecation in the genet is locus specific and probably communal. Presumably it allows solitary animals in the wild to detect the existence (through a chemical symbol) of another animal without con- frontation (Mykytowycz, 1969). In captivity both urine and feces are deposited in a single litter box, but when two boxes are provided, feces are usually restricted to one box, while urination often takes place in both. Anal dragging, a response to voiding unusually soft feces, probably has only incidental signal function as it is infrequent and the re- maining olfactory trace probably fades rapidly. Genet urine is apparently chemically distinct though it has a pungence which characterizes many viverrine and paradoxurine carnivores. Crawhall and Segal (1965) found that Genetta tigrina5 excretes unusually large amounts of sulfocysteine, and concluded that the substance is neither ab- sorbed from the diet nor results from intestinal microbial metabolism. As amino acids lack the aromatic properties of cyclic hydrocarbons, it is doubtful that the compound contributes signifi- cantly to the olfactory quality of urine. This notable feature does suggest, however, that what- ever the odorous components in the urine, they may likewise be distinct and species typical. SOME INTERGENERIC COMPARISONS AND DERIVATIONAL SCHEMES There can be little doubt that body form is a dominant variable limiting behavioral expression. Intertaxa behavioral comparisons usually reveal a small number of fundamental patterns whose ubiquity probably results from a similar (possibly primitive) nervous mechanism and a capacity for movement that is relatively uninfluenced by the structural divergence exhibited by terrestrial mam- malian orders. Definitive adult patterns that fall into this category are sniffing, licking, and nibbling movements, scratching with the hindfoot, head and ?These specimens were incorrectly identified by the Na- tional Zoological Park (cited Washington Zoo) as Genetta genetta neumanni. body shaking, generalized pilo-erection, squatting eliminative stance, tripedal stance, and the walking gait. Licking movements (nose, lips, and body), for example, are generalized motions performed by the majority of terrestrial species that have spatulate intra-oral tongues (Type I tongues of Dor an and Baggett, 1971). Though coordination of fundamental patterns is usually similar between species, differences are often clear reflections of structural adaptations for speed or arboreality. Both Fossa and Nandinia, for in- stance, can touch the pes to the inside of the opposite leg, but this scratching orientation is less commonly displayed by the former than the latter. The lack of an extensive rotatory capacity in Fossa's hindlegs may be among the factors accounting for this difference. The dexterity of Nandinia's hind leg on the other hand probably reflects a primary adaptation for arboreal locomotion, especially head- first vertical descent. As adaptations for climbing and speed imbue differing mechanical properties to the skeleton and muscles, it is not surprising that behavioral differences between species do not always parallel taxonomic distinctions. This contention is supported by Table 7, comparing the distribution of several behavior patterns between members of four subfamilies. I have included the first two behavioral items in Table 7 (bipedal stance and headfirst vertical descent) to give an indirect com- parative measure of arboreal specialization. The assumption is that civets, unable to descend trunks headfirst or stand bipedally, exhibit morphological correlates of at least incipient cursorial specializa- tion, namely digitigrady and suppression of trans- verse and rotatory limb motion (Howell, 1965). In the following discussion, based in part on Table 7, similarities and differences should be viewed both in terms of locomotory adaptation and systematic status. Both the long- and walk-stretch are probably the most common and frequently performed stretching patterns in carnivores as well as many other mam- mals. In their complete form, both of these patterns consist of two primary and independently per- formed components, namely stretching of the fore- quarters and stretching of the hindquarters. The differences between the patterns result from the association of the walk-stretch with forward pro- gression. In the typical walk-stretch the forelegs are not uniformly extended as the forequarters are 38 SMITHSONIAN CONTRIBUTIONS TO ZOOLOGY TABLE 7.?Comparisons of behavior among selected viverrid genera ( + = present, ? ? absent, + ? or ? ? = probably present or absent but substantiation needed, blank = uncertain, a = personal observation, b = Vosseler 1929, c = Ducker 1957, d = Huf 1965, e = Diicker 1971, f = Albignac 1970a, g = Albignac 1970b) Stationary stances Bipedal stance.. Head first vertical descent Stretching Long-stretch Walk-stretch Arched-back stretch Object-oriented contact patterns Metatarsus scuffing Head and neck rubbing Neck sliding Scent-marking patterns Quadrupedal squat Perineal drag... Normal quadrupedal... Reversed upright quadrupedal... Upright quadrupedal... Self-oriented movements Bilateral head wiping References Viverrinae Paradoxurinae Hemigalinae Cryptoproc- tinae a a ac ad a ae _? afg depressed and the hindlegs take one to several stiff- legged steps. In the typical long-stretch, the forelegs are equally extended in the first phase, and the hindlegs equally extended in the second phase. The variation of these patterns within a species or individual, however, does not suggest a clear evolutionarily interpretable trend. Genetta, for ex- ample, performs walk-stretches with the forelegs equally or subequally extended, and partial walk- stretches in which variation in forelimb positions NUMBER 239 39 are manifested without the hindquarter stretch (stiff-legged walk). The same holds true for Nan- dinia and Paradoxurns. Civettictis, on the other hand, stretches like a dog: the forebody is depressed as the forelegs are equally extended forward. Never- theless, both the civet and the domestic dog occa- sionally terminate such a forequarter stretch with several short steps of the posteriorly extended hindlegs. Quantitative data are clearly needed to establish relative frequencies of the different stretching coordinations. The categorization of complicated and compo- nent coordinations (such as fore- and hindquarter stretches) into walk- and long-stretches obscures the possibility that these component coordinations were the precursor elements that combined in time to produce the walk- and long-stretch. Alternatively it is possible that the widespread walk-stretch pat- tern was the prototype that, through fragmenta- tion and changes in form, gave rise to fore- and hindquarter stretches, and evolved into the long- stretch through loss of the association with walking. The arched-back stretch is equally enigmatic, occur- ring in erinaceids (Poduschka, 1969) as well as felids and viverrids. Bilateral face-wiping is an almost identical move- ment in the Viverrinae and Paradoxurinae exhib- iting it (Table 7); Nandinia, which differs slightly from Genetta, holds its forepaws in a semiprotracted position and consequently after each descending stroke they do not meet on their medial surfaces. According to Diicker (1965), the pattern in Viver- ricula also resembles that of Genetta, except that it is performed in a reclining position. Civettictis on the other hand lacks this behavior, though I once saw the female make strained but synchronous forepaw contact to the nose as she reclined in a lateral extended position. The ability to touch the head simultaneously with both forepaws, however, is widespread within the Carnivora; even domestic dogs will eventually remove a rubber band placed about the snout with simultaneous use of the dew claws. Though bilateral face wiping has not been seen in paradoxurines other than Nandinia, the forelimbs appear to have ample dexterity. Uni- lateral face wiping, on the other hand, was uncom- monly seen in Paradoxurus, Paguma, and Arctictis, and rarely seen in Civettictis and Fossa. With the exception of Burger's study (1959), bilateral face-wiping has been overlooked as one of the most complex and stereotyped of mammalian behaviors. The resemblance in the coordination and form of this pattern between Insectivores, Tupaiids, rodents, marsupials, and carnivores implies either remarkable convergence in neuro- physiological evolution, or retention among disjunct lineages of a primitive and formerly widespread pattern. The genera I have observed6 exhibited these similarities: (1) bipedal stance, (2) forelimb motions with progressively increasing amplitude, and (3) forepaw licking strongly linked to the com- pletion of each wiping motion (at least initially dur- ing the bout). The primary differences between species relate to the posture and action of the manus, the phase relationships of the forelimbs to one another during wiping, and the speed with which the movement is executed. No combination of behavioral, morphological, or ecological criteria can be used to predict accurately which species possess bilateral face wiping within their repertoires, and consequently it has little utility as a taxonomic character. Judging subjectively, the pattern seems to occur only in species with either (1) lax pelage, or (2) a tactile receptive field on the face composed of long vibrissae and possibly scattered tylotrichs, and (3) a body form sufficiently generalized that the limbs have considerable freedom of movement in several planes. This grooming ability is displayed by non- brachiating arboreal and semiarboreal species, and those terrestrial species having plantigrade feet. The existence of the pattern in carnivores is highly unusual and suggestive of a more widespread occur- rence among the extinct viverrines that presumably gave rise to modern forms. This is confirmed by the occurrence of unilateral face-wiping in many felids (Leyhausen, 1956) and its resemblance to bilateral wiping in the form of the motion and the alter- nating lick and wipe pattern. Furthermore, no movements resembling bilateral face-wiping are known in the most convergently similar arctoid carnivores, such as Potos and Bassaricyon (Poglayen- Neuwall, 1962, 1965) and Bassariscus (H. ?Viverridae: Nandinia, Genetta, Prionodon; Didelphidae: Didelphis virginianus, Marmosa robinsoni, Caluromysiops ir- rupta, Caluromys lanatus, Chironectes minimus; Phalanger- idae: Phalanger gymnotus, P. orientalis, Petaunis breviceps; Dasyuridae: Dasyuroides byrnei, Sminthopsis macrura; In- sectivora: Sorex sp., Blarina brevicauda, Microgale talazaci, M. dobsoni, Echinosorex gymnurus. 40 SMITHSONIAN CONTRIBUTIONS TO ZOOLOGY Egoscue, pers. comm.). On the other hand, this behavior is exhibited by several mustelids (Mustela erminea, M. frenata, Spilogale putoriw: Egoscue, pers. comm.). The bilateral forepaw patting motion that Enhydra lutra directs to its head superficially resembles face wiping, but because the movement lacks an association with paw licking and is directed to the entire body, it probably derives from a be- havior pattern other than the one in question. In the felids and viverrids where uni- or bilateral face-wiping is a regular constituent of the toilet that precedes activity or follows feeding, the pat- tern's appearance seems to be largely self-generated, or at least responsive to afferent input from the stomach or the mouth, lips, and vibrissae. Ewer (1967) has postulated a scheme which simulates the sequential and repetitive character of grooming sessions based on the interaction of the excitatory levels of the component actions. In those species which only on occasion employ unilateral face wiping, its appearance seems largely governed by external stimuli, and the causal linkage of uni- lateral wiping to other self-directed patterns seems to be weak or lacking. Fossa fossa, for example, was rarely seen to wipe its snout, but a loose, deciduous premolar precipitated a prolonged series of wiping motions that eventually dislodged the tooth. Head- and neck-rubbing, and neck- and shoulder- sliding (Table 7) are basically similar in that largely overlapping anterior parts of the body are rubbed. The former pattern, however, is generalized in the sense that it is evoked by a variety of circumstances and its component motions exhibit variable forms, duration, and temporal organization. On the con- trary, neck and shoulder sliding seems to be evoked only by certain aromatic substances and consists of one motion directed to alternate sides of the head and neck. Though it is tempting to interpret neck and shoulder sliding as a simplified and exaggerated version of Genetta'% rubbing movement, two observations suggest a more logical alternative: (1) when thoroughly wet, domestic dogs, mustelids (M. frenata, M. erminea: pers. observation), and viverrids (Genetta, Nandinia, Fossa) sometimes pro- pel themselves forward with the hindlegs while sliding on the side of the head, neck, and, to a lesser extent, the shoulder; (2) sliding movements, in which substrate contact coincides largely with the extension rather than retraction phase of the neck motion, occur in several species either as a response to a strong odor field (canids, Eira barbara, the three viverrids listed above), and/or in species that have a dermal scent-producing gland in the cheek, throat, neck, or sternal region {Trichosurus vulpecula, Helogale undulata: Zannier 1965, prob- ably Herpestes auropunctatus: D. Nellis, pers. comm.). In dogs neck-sliding directed to animal dung may be repeated several times on alternate sides of the neck, or it may occur only as a brief preliminary to rolling about on the sides and back. When in heat, female domestic cats slowly glide the neck to the shoulder on the substrate while standing (Leyhausen, 1960). This behavior is apparently spontaneous. In comparison, neck and shoulder sliding as a prefatory component of body rolling and rubbing appears to be an incipient expression of the defini- tive pattern seen in civets. According to Diicker (1957) neck rubbing (Kinn- und Halsscheuern) in Viverricula is followed by vigorous bouts of body rolling and rubbing. The factors that have selected for the abandonment of body rolling in favor of only neck sliding are obscure; however, the simplifi- cation of the movement may be linked to intensifica- tion of the sniffing role; brief sniffing bouts are fre- quently interposed between each sliding motion during the onset of this activity. Considering other patterns of object-oriented contact, Table 7 attests to the intra- and inter- specific variation in the methods by which the selected viverrids appress the pregenital or perineal gland to inanimate objects. The most common marking methods, namely the perineal drag and quadrupedal squat, have a distinct resemblance to anal dragging and differ from it chiefly in the angle of the pelvis and so the emphasis of contact. The physical similarity of these three movements, and the fact that anal dragging is a prevalent response among morphologically generalized mammals to moisture, feces, or paniculate matter on the anus or hindquarters, suggests that both perineal and anal dragging primitively shared common causal features related to elimination. (For the sake of clarity in this discussion, the terms anal drag and perineal drag will be used to designate respectively the probable primitive cleaning movement and the scent deposition movement assumed to be its deriva- tive.) The hypothesized significance of autonomic elimination for the evolution of many types of NUMBER 239 41 anogenital scent glands is founded on the following observations. In many mammals reflex discharge of urine and feces is most frequently caused by hostile inter- and intraspecific interaction, often resulting in anogenital contamination and conse- quently followed by anal dragging and self-directed licking and nibbling. Contained within the auto- nomic elimination reflex and the circumstances of its occurrence are behavioral "building blocks" or preadaptations for specialized forms of scent- marking. The building blocks include two distinct modes of scent deposition (urine and feces), and a potential social referent and/or message receiver (the animal evoking the autonomic response). To elucidate, both autonomic defecation and evacua- tion of the anal gland share common features of agonistic contexts and protrusion and eversion of the anus. The implication is that anal protrusion was a possible source of anal scent emission. Anal dragging may likewise have served as a behavioral vehicle that potentiated the development of glan- dular fields in the anal, perineal, and pregenital regions. The remaining three viverrid marking patterns (Table 7; Figure 20) differ from the perineal drag in that the association with forward progression has been relinquished and glandular deposition is medi- ated by a simple pressing motion (quadrupedal stance in Civettictis), a uni-directional wipe (re- versed upright quadrupedal stance in Genetta), or a rubbing motion (antero-posterior in Cryptoprocta, or lateral in Genetta). A fundamental similarity of the two upright quadrupedal patterns (reversed and normal) to the perineal drag can be appreciated by visualizing the necessary adjustments in posture that would accompany rotation of the marking log seen in Figure 20 to a horizontal plane. The implication that both patterns arose as concomitant expressions of arboreal locomotory abilities need not be true. The postural properties of the re- versed quadrupedal upright position, for example, characterize two situations that could conceivably have served as transitional activities as denned by Lind (1959). As an animal steps from an elevated position to a lower level, the effective angle of descent and center of gravity are decreased by depression of the hindquarters and the acquisition of a spinal concavity (Figure 20). The loss of this hy- pothetical stepping-down component, which would convert the pattern to its definitive form, could POSSIBLE PRIMITIVE SOURCES: NORMAL AND AUTONOMIC ELIMINATION & ASSOCIATED ANAL-PERINEAL DRAGGING gland origins(anal, perineal. & preputial) GLAND DRAGGING TRI-& QUADRUPEDAL SQUAT (loss of locomotory component) FIGURE 20.?A derivational scheme of some stances employed by viverrid carnivores during scent-marking (arrows point to specific stances exhibited by extant species and a hypothetical intermediate stage; black arrow = logical evolutionary path- ways of stances in which glandular contact with the substrate is mediated by variations on the squatting theme; white arrow = two possible (but rather tentative) routes for the development of the normal quadrupedal marking stance from a squatting stance). 42 SMITHSONIAN CONTRIBUTIONS TO ZOOLOGY have occurred through an increased dependence upon olfactory releasers and an escape from the transitional action. The transitional act hypothesis is supported by the observation that in Crypto- procta the upright marking stance is frequently preceded by substrate clawing accompanied by a lordotic spinal curvature; this pattern may have likewise undergone similar changes. Alternatively and perhaps more feasibly, marking in the reversed upright quadrupedal stance may have evolved as an exaggeration of the tendency to place a slightly elevated mark by raising a hindleg over or by resting it upon the marking object (e.g., Genetta tigrina, Suricata suricatta, Herpestes edwardsi: Ewer, 1968). Cricetomys gambianus (Ewer, 1967) and female Speothos venaticus (Kleiman, 1966) achieve more elevated marks by walking back- wards up inclined surfaces; such a method is feasi- bly intermediate to the technique used by Genetta in which the hindquarters are raised up to the marking position via a quick handstand rather than climbing up backwards. The simple pressing glandular contact and the absence of the squat component indicate a minimal resemblance of the quadrupedal marking pattern to the perineal drag. Derivation from the perineal drag or squat presumably involved a concurrent posterior shift in glandular position for, other than a slight spreading of the hindlegs and lifting of the tail, the quadrupedal marking pattern exhibits no other postural adjustments. De novo origin of this pattern is tenable only if it is assumed that perineal glands evolved independently in the semiarboreal and terrestrial viverrines after divergence from the common ancestor. The remote resemblance of marking with the perineal gland in a normal quadrupedal stance is perhaps a manifestation of Civettictis' particular subcursorial habitus. Behavior of Associated Animals Patterns of behavior performed by solitary ani- mals also occur in the presence of other animals. A number of behaviors, however, seem to occur predominantly in the presence of conspecifics be- cause their execution requires an animal referent, or because stimuli and contexts associated with interaction are usually absent in solitary situations. The wisdom of classifying some of these behavior patterns as either solitary or social is debatable. Generalized pilo-erection, for example, is mediated through any or all of the sensory systems; the characteristics of the stimuli are not specifically limited to social contexts. Under the conditions of this study, though, certain patterns of pilo-erection were seen primarily in social contexts, hence their inclusion in this section. The descriptions in the first part of this study defined types of body motion as they were identified and recorded during social encounters and subse- quently analyzed. With the exception of vocaliza- tion that is distinguished by broadcast transmission, the behaviors in this section are classified on the basis of orientation and the spatial characteristics of their occurrence. It is assumed that these patterns have a communication function. Both the percep- tual effect and, presumably, the message of each pattern is or may be modified when executed con- currently with other patterns; thus, the behaviors described under each category represent components that may be combined within and between cate- gories and modified by independent variables such as velocity of movement and posture. The second part of this section analyzes the organization of signals in time and the dynamics of social inter- action in Genetta and Civettictis. COMPANION-ORIENTED LOCOMOTION An animal's locomotion is often affected by the presence of another animal. The criteria used to distinguish different types of companion-oriented locomotion are listed below. APPROACHING.?Technically, any movement of one animal that decreases the distance between it and a companion may be defined as an approach. Such a definition would include movement tan- gential to, but not specifically oriented to the other animal. As any movement of one or two animals will involve spatial changes that may have no im- mediate social significance, the term approach is used herein only to describe those cases in which a decreasing distance between two animals is accom- panied by at least one animal maintaining a frontal orientation to some part of the other animal's body (Figure 23A). DEPARTURE.?The term is used to describe the net movement of one or both animals away from the immediate vicinity of the companion. In most cases a departure is defined as the separation of NUMBER 239 43 animals engaging in contact, or the movement of one animal out of a one body length perimeter of space surrounding the companion's head and body. FOLLOWING.?An animal is described as following when its locomotion and spatial orientation are closely correlated with and apparently guided by the companion's movements. Following is usually a distinct action with the animals trotting briskly in close proximity, the follower at the rear or side of the leader. CIRCLING.?When the locomotion of each of two animals is oriented toward the side or hindquarters of the partner, the course of movement is circular. This is seen during mutual sniffing of the hind- quarters or when both animals attempt nasal con- tact with the partner's rear. This definition of mutual orientation to the companion's body excludes following in which the path may also be circular. ENCIRCLING.?Occasionally as one animal remains stationary or moves slowly, its companion walks or trots about it in a tight circle. JUMPING OVER.?Though this pattern may in- volve an approach and departure, it is best described as a separate case in which the body of one animal passes over the companion's body. During bouts of interaction involving extensive body contact, an animal may jump over another without leaving the partner's proximity. It may also be executed by an animal traversing a tree limb, in order to negotiate the obstacle presented by a resting companion's body. GAIT AND POSTURAL VARIABLES.?Speed of move- ment and postural variants of the walking gait add additional though somewhat overlapping dimen- sions to simple spatial changes resulting from the locomotory movement of two animals (see "Gait- associated Postures" under "Locomotion and Its Ontogeny"). While approaching, following, and departure may involve any gait, circling and en- circling are performed only in a walk or trot. The association of postural configurations with types of companion-oriented locomotion has not been consistently pursued, but it appears that nearly all the variants diagrammed in Figure 5B-F are exhib- ited by approaching animals, while departing animals rarely exhibit postural variants involving crouched forequarters (Figure 5B,D). When a di- rected gait is interrupted by a stance, the postural configuration of the spine is usually retained. As TABLE 8.?Associations of postural variants (companion- oriented stances of tripedal and quadrupedal types, char- acterized by protracted visual fixation) and torso orienta- tion to companion in Genetta tigrina (+ + = most frequent association, + = less frequent association, ? = uncommon association) Postural variant Full Crouch Normal Crouched fore- and semi-raised hind- quarters Semi-raised fo re - and hindquarters. Maximum raised fore- and hind- quarters Frontal Tangential Lateral there is no discrete distinction between a slow walk and brief alternating tri- and quadrupedal stances, the stances described in the following section represent the stationary extreme of a graded con- tiuum of noncontact-oriented behavior. COMPANION-ORIENTED STANCES Visual fixation of a companion requires only that the head be oriented in the companion's direc- tion, but in a number of circumstances, particularly when a stranger is encountered, relatively long periods of time may be occupied by staring. The stances assumed during visual fixation' are charac- terized by (1) rigid-appearing immobility, (2) inter- ruption by brief periods of walking but often unbroken visual contact, and (3) the maintenance of a particular body orientation to the companion irrespective of the companion's movement. Table 8 presents a rough approximation of the way these stationary body positions are associated with the orientation of the torso relative to the companion's body. It is clear that with the exception of the normal quadrupedal stance there is a tendency for the body to be low during frontal orientation and high during lateral or tangential orientation (Fig- ure 23A). This is in part a result of the association of the low or crouched stances with approach toward the companion whereas the high or arched- back stances are usually seen in a context of avoidance. 44 SMITHSONIAN CONTRIBUTIONS TO ZOOLOGY 12 FIGURE 21.?Bipedal pivot of the African palm civet (Nan- dinia) based on tracings of 16 mm film (24 fps), in which the animal took several bipedal steps. The bipedal stance and the elongate quadrupedal crouch are usually maintained for relatively brief periods, the latter being associated with ap- proaching a companion and establishing nasal contact. SIMILAR STANCES IN OTHER GENERA.?Civettictis, Paradoxurus, Paguma, Nandinia, and Fossa are capable of assuming a crouched posture while sta- tionary or walking. Though not all these genera display equal facility in maintaining this position, crouching is fairly widespread among civets. With the exception of Paguma, the posture involving semicrouched forequarters and semiraised hind- quarters (Figure 5D) is also seen in these genera as a companion-oriented stance. This body position can be converted to a low-intensity arched-back posture by straightening the forelegs and slightly arching the back. This has been seen in Paradox- urus, Nandinia, and Fossa. In the male Nandinia, this half crouched, half raised body carriage was associated with slow and deliberate walking on branches, accompanied by visual fixation of the observer or another animal. The male would reverse the direction of his move- ment by rising "stiff bodied" into a bipedal position and pivoting on his hindlegs. Eye contact with the other animal during this maneuver usually was not broken. An unusual form of this bipedal pivot was displayed by the male toward a strange female (Figure 21). The animal quickly approached the female, rose into a high bipedal position, and pivoted to one side while holding the forelegs up and tightly flexed to the sides of the body. It can be seen in Figure 21 that the animal actually took several steps while standing upright. Intense arched-back postures have been seen only in Fossa. I am reasonably certain that it does not occur in Civettictis and Nandinia, but neither Paradoxurus nor Paguma was seen under conditions conducive to this behavior (e.g., fighting). Intense arching in Fossa occurs as a laterally oriented sta- tionary stance, and the tail may be held vertically erect. In Civettictis tangential or lateral orientation of the head, neck, and body to the companion is an outstanding aspect of social interaction. There is no simple way this can be described because the position of the head and neck is clearly dependent to a large extent upon the attitude, proximity, and orientation of the companion. In general, the neck is held at the horizontal or slanted down- wards while the head is maintained in the same plane (Figure 22A) or deflected downwards to as much as 60? (Figure 22B). The head is usually rotated slightly upwards in the partner's direction (Figure 22A,D) and may be turned slightly toward or away from and out of the companion's sight. Though it is possible for an animal to display a lateral head and neck orientation to a companion while the trunk has a frontal attitude (Figure 22B), it is more common for the trunk axis to be tan- gential or lateral to the companion's head and neck (Figure 22A,C,D). Tangential or lateral body or head and neck orientations to a partner occur during slow walking in a normal posture, in a normal or crouched quadrupedal stance (Figure 22c), or in reclining positions (Figure 22A,D). NONCONTACT BODY MOVEMENTS ERRATIC HEAD MOVEMENTS.?The simplest form of this movement involves a rapid vertical bobbing motion of the head with the neck contributing minimally. This is most frequently seen when an animal is approached or followed. Such erratic motions result from a rapid twisting of the neck, and lateral, circular, and vertical jerking move- ments of the head, often associated with a brisk trotting gait, or brief bipedal stances. ALTERNATING EAR MOTIONS.?As described under NUMBER 259 45 FIGURE 22.?Lateral and tangential orientations of the head, neck, and body in the African civet (Civettictis): A, head and neck deflection in a reclining animal; B, lateral head and neck orientation by an approaching animal (male); c, assumption of a reclining position by an animal receiving head-darts; D, head deflection by a recumbent animal towards the superior companion. the primary senses, this behavior may be directed toward another animal or to an area of interest. It is difficult to discern whether the movement itself or the independent perception of sound triggers the same response in companion animals. TURNING TOWARD.?Any motion of the body or its parts resulting in a frontal orientation of the face toward the companion is denned as turning towards. Facial exposure can be achieved by either turning the neck or pivoting the entire body in the companion's direction. This pattern may initiate a long period of visual fixation accompanied by a companion-oriented stance or it may involve only a brief gaze in the other animal's direction. TURNING AWAY.?Any movement that terminates facial exposure to a companion is denned as turning away. TAIL WIGGLING.?An erratic and rapid bending and extension of the tail along its length produces a spasmodic jerking motion that in its most intense form produces a rattling sound against wooden floors or a rustling sound in leaf-litter substrates. In my animals, tail wiggling was always performed in either a ventral retracted limb or an extended 46 SMITHSONIAN CONTRIBUTIONS TO ZOOLOGY body reclining position, but according to Leyhausen (pers. comm.) a violent thrashing motion of the tail accompanies arched-back rubbing in Genetta riippelli. ANAL GLAND SCENT EMISSION.?The emission of anal gland scent is associated with fighting, intense attempts to avoid or escape from another animal or painful stimuli. Hence, it is not accompanied by any particular stance or movement. On the one occasion I observed scent emission at close range, the anus was briefly opened several millimeters and the fluid ejected as a barrage of about two dozen small drops that traveled about 2 decimeters before striking a surface. The force of the emission is weak, and the associated circumstances are such that scent is usually deposited close to the inter- acting animals. The scent almost always contami- nates the fur of the tail and that surrounding the anus; animals usually spend considerable time licking these regions following scent release. Anal gland scent appears to arise from two spherical structures lying beneath the tail on either side of the anus. The scent is a thin translucent coffee-colored fluid that is highly volatile and has a slightly noxious musky odor. The odor lingers in the air for several minutes after release; to the human nose it is detectable within a radius of about 3-4 meters in a room with minimal air movement. Even after several weeks the scent residue emits a mild musky but not unpleasant odor that is detect- able at close range; washing with detergent and water will not completely remove it from clothing. PILO-ERECTION PATTERNS.?The longer hair on the head, trunk, upper legs, and tail is erected in a variety of circumstances that are not always related to the presence of another animal. The degree of erection is variable and not necessarily uniform on the body surface. Disturbance of the nap of a pelage often occurs during scratching and object- oriented contact, and uniform erection of the hair seen in resting animals is presumably associated with thermoregulation. In addition, certain patterns of pilo-erection are exhibited in circumstances in- volving other animals. Vibrissae: Movement of all the facial vibrissae from depressed to an erect position is presumably under the control of pili-arrectores muscles and superficial facial musculature. Maximal mobility is displayed by the mystacial vibrissae. The anterior movement of these hairs, whether or not associated with gaping of the jaws, always involves a squaring of the rostral contour (Figure 26). Erection and depression can be either gradual (and arrested at a certain degree of erection) or sudden. Sudden changes in vibrissae position (<1 sec) seem to be maximal changes involving erection. Depression is usually a gradual process. The mobility of the other facial vibrissae seems to be less developed and less noticeable. Spinal Crest Hair: The black hair of the spinal crest, extending from immediately behind the shoulder to the base of the tail, slightly exceeds the length of the surrounding hair (Table 9). Thus, when there is uniform erection of hair on the torso, the crest is slightly elevated (Figure 23B); more- over, it appears that the spinal crest itself can be TABLE 9.?Average hair lengths (in mm; range in parentheses) of spinal crest and tail in a sample of 18 Genetta tigrina skins in NMNH (sides = 3 cm from dorsal midline; tail base = third white tail ring to rump; tail tip = distance from end equal to tail base distance from rump) 99 dorsum side diff. dtf dorsum side diff. 99 + oV dorsum side diff. Shoulder 17.7 17.1 0.6 19.2 18.8 0.4 18.4 17.9 0.5 (16-20) (16-19) (0- 2) (16-22) (16-23) (-2- 4) (16-22) (16-23) (-2- 4) Spina1 Crest Midback 21.7 19.4 2.3 24.0 18.8 5.2 22.8 19.1 3.7 (20-24) (19-22) (0- 4) (20-31) (17-24) (0- 9) (20-31) (17-24) (0- 9) Rump 25.8 18.8 7.0 26.1 20.1 6.0 25.9 19.4 6.5 (22-30) (16-20) (5- 9) (22-31) (17-24) (2-11) (22-31) (16-24) (2-11) 24.8 24.8 0 25.5 25.5 0 25.1 25.1 0 Tail Base (21-27) (21-27) (22-29) (22-29) (21-29) (21-29) 20.4 20.4 0 21.6 21.6 0 21.0 21.0 0 Tip (16-23) (16-23) (16-26) (16-26) (16-26) (16-26) NUMBER 239 47 raised to a greater extent than surrounding hair. Maximal and prolonged erection of the crest, however, seems to be linked with a uniform level of erection in the torso hair. When the body hair is depressed, there is no indication of a crest. Tail Hair: The degree of erection may be uni- form or may vary along the length of the tail. A uniform level of erection seems to be the most com- mon condition (Figure 23c). Lengthwise pilo- erection gradients are usually unidirectional with either the terminal black segment or the first 2-4 white rings being the expanded portions. The velocity of hair erection and compression (hair flux) varies from moderately rapid (<2 sec) to gradual, but cine film is necessary for accurate description of this aspect. While gradual expansion of the tail hair is more or less continuous, rapid expansion or compression is often, perhaps most frequently, stepwise. That is, rapid flux in erection or com- pression is followed by a period of very slow change in either direction. NONCONTACT MOVEMENTS IN OTHER GENERA. With the exception of alternating ear motions and tail wiggling, Civettictis exhibits all the noncontact body movements described for Genetta. The Afri- can civet is reputed to expel an anal scent (Walker, 1968), but though my animals fought on several occasions this was not detected. Unlike Genetta, the hair on the tail is not erected in stepwise incre- ments. This pattern of hair flux is seen, however, in the well-developed spinal crest. Erection of the crest results in an obvious increase in lateral body dimension and adds a light-colored stripe to the contour of the back above the level of the spine; this is not visible when the crest is relaxed because the light-colored basal segments of the hair are covered by the longer black distal segments. COMPANION-ORIENTED CONTACT PATTERNS Head SNIFFING (NASAL CONTACT).?Sniffing of another animal does not necessarily involve contact, but if not repelled an animal will almost always bring its nose into contact with the fur of its companion. The nose may lightly contact the tips of the com- panion's hair or may be pushed through the hair to the skin. The rhinarium, lips, and the lightly furred zone between them are often rubbed through the underhair and against the skin (Figure 23D). A variety of facial expressions may be displayed as an animal sniffs and contacts a companion with its nose. When sniffing without contact the eyes are usually wide open and the ears erect and directed forward. If nasal contact is involved, there is a tendency for the eyes to be squinted and the ears oriented laterally (Figure 23D). When the perineal gland, vagina, urine, or an estrous female are smelled and/or licked by a male, the upper lips may be lifted slightly and the corners of the mouth retracted to expose the canines and premolars. This activity, Flehmen, is presumably associated with olfactory and gustatory perception via Jacob- sen's organ. The mouth is usually slightly opened, the eyes squinted, vibrissae retracted, and the orbits of the ears directed laterally in a semierect position. LICKING AND NIBBLING.?As described for the solitary animal, these patterns may also be directed to a companion. Both types of mouth contact are usually associated with a relaxed facial expression. The eyes may be squinted or closed and the ears oriented laterally in a semierect position. BITING.?Duration of contact, force of compres- sion, and head movements during contact are three biting parameters that probably form a graded continuum. Mouthing occurs when an animal's jaws engage with another animal's body but there is little or no compression (Figure 23E); biting occurs if the force is strong. Duration of biting and mouthing is highly variable. Facial characteristics of hard biting include closed eyes, depressed mysta- cial vibrissae, and laterally directed ear orbits that are folded posteriorly above the notch. During soft biting or mouthing the eyes may be wide open or squinted; the ears are usually held with the orbits oriented laterally, but they are not folded above the notch (Figure 23E). HEAD-DARTING (Figure 24).?This is a highly vari- able movement consisting of several components and involving several levels of integration. It is a rapid form of head extension which may or may not be followed immediately after by head retrac- tion. It may be executed with a closed mouth, with an open mouth, or with an open mouth and an explosive hiss emitted at the end of the exten- sion phase. Figure 25 displays the variability in the duration of extension and retraction phases and the way an open mouth position may be associated with this activity. Thus, head-darting may or may 48 SMITHSONIAN CONTRIBUTIONS TO ZOOLOGY FIGURE 23.?Patterns of companion-oriented locomotion, movement, and contact in the genet: A, crouching animal approaching a companion exhibiting an arched-back stance; B, low in- tensity arched-back stance exhibiting general pilo-erection and the spinal crest; c, maximal pilo-erection of the tail; D, juvenile genet sniffing the spinal crest hair of another animal (note facial expression); E, juvenile animal mouthing its companion's tail; F, pawing by an animal engaged in wrestling. NUMBER 239 49 1.3 4 1 3 4 5 8 10 10 8 FIGURE 24.?Extension and retraction phases of head-darting in the genet (animals E and Fr; based on 16 mm film, 24 fps; shaded images represent last frame in series; in this sequence recipient began to respond in less than i/A second). EXTENSION/RETRACTION FRAMES 12 10 8 4 0 4 8 10 12 FIGURE 25.?Duration of extension and retraction phases in eight samples of genet head-darting recorded on 16 mm film (24 fps) of E head-darting at Fr (none involved contact with Fr and the occurrence of hissing is not indicated; black bars = extension and retraction of the head, clear bars = opening of the mouth). not involve contact with a companion. When it does involve contact, the companion is most com- monly struck with the incisors and canines of the partly opened jaws. Brief snapping bites are de- livered less often. Associated with such contact is a momentary squinting or closing of the eyes (Fig- ure 24). A backward deflection of the ears is a far less frequently associated change. The mystacial vibrissae are often extended anteriorly preceding a head-dart and during the early part of neck exten- sion, but as the subject is approached, they are retracted at least momentarily. The vibrissae may retain an extended position if the velocity of the extension phase is relatively slow. JAW-GAPING STARE.?The mouth may be gaped at 20?-30? for varying amounts of time as an animal stares at a companion. Extended vibrissae are usually associated with a widely gaping jaw (ca. 40?) and a strong tendency to head-dart at a companion. The mouth may be opened briefly and repeatedly while the animal is staring and as it inhales between growls. ERRATIC HEAD MOVEMENTS (see "Noncontact Body Movements").?When executed in the close proximity of another animal virtually any part of the head or neck may touch the companion's body. Limbs Three kinds of forelimb movement directed to- ward other animals can be discerned. Whether in- volving contact or not, pawing motions are usually brief and similar in form to those directed at inani- mate objects (Figure 23F; cf. "Object-Oriented Con- tact Patterns"). Stepping and standing on another animal is distinguished from pawing because the force of the contact is presumably greater and the animal establishing contact is usually in a superior or elevated position. Clasping of another animal's body occurs in the same manner described for an inanimate object. As a partner's body is clasped the hindlegs may also be pumped rapidly back and forth with the major thrust occurring during the backstroke. Though both legs may be pumped concurrently in a reciprocal manner, usually only one leg pumps at a time while the other is pressed against the companion's body or held free of con- tact in a flexed position. Kicking, the hindleg analog of pawing, consists of an often rapid extension of the limbs as the ani- mal reclines on its side or back. When contact oc- 50 SMITHSONIAN CONTRIBUTIONS TO ZOOLOGY curs it may be forceful and brief or light and of variable duration. Trunk SQUATTING OVER.?Individuals of both sexes are sometimes seen to straddle the body of another ani- mal with the hindlegs, and then depress the pelvic region over some part of its body. This may or may not involve contact; when it does, it is not a force- ful pressing movement and there is no indication that the penis is extended or the perineal gland everted. Contact of the inside of the upper legs and genital area seems to be light and intermittent. The position may be retained without movement or an animal may creep forward over the other's body with small steps. STEPPING OVER.?When one animal steps over the body of another, it often briefly touches some part of the latter with the ventrum and or hindlegs. Sim- ilar fleeting contact may also occur in jumping over another animal. BRUSHING AGAINST.?An animal may briefly brush any part of its body against that of another animal in passing. SOMERSAULTING AGAINST.?Occasionally as two animals engage in long bouts of contact behavior, one may place the side of the head on the ground and perform a lopsided somersault by rolling over onto the neck, shoulder, and then the side. Compound Contact Patterns The behaviors in this category differ from those described heretofore in being composed of several independent but concurrent or simultaneous move- ments. The movements of the two animals may be alike and reciprocal or different and complementary. This category is used for practical reasons because these behaviors are more easily recognized and de- scribed at a higher level of organization. The de- scriptions are simplifications which outline basic elements of complicated dynamic events. HEAD-SPARRING.?While frontally oriented, a pair of animals may mutually swing and toss their heads in apparent efforts to avoid mouth striking or bit- ing by the companion. The character of the move- ment resembles that of erratic head movements except that there is an evident action-reaction rela- tionship between the animals' motions. GRAPPLING.?Standing bipedally upright, the ani- mals paw at and clasp one another about the head, neck, shoulders, or chest while head-sparring or head-darting and biting each other's head, neck, forelimbs, and shoulders. WRESTLING.?Animals engaged in this activity maintain a broad zone of body contact in a head to head, head to rear, or cross-body orientation. One or both animals may clasp the other with the fore- and/or hindlegs, and paw, pump, bite, or head-spar. When in body contact, one animal's belly contacts the other's side, back, or belly. FIGHTING.?Combat between two animals re- sembles wrestling but it almost always involves a cross body or head-to-head orientation and is usu- ally accompanied by screaming, urination, and anal scent emission of one animal. Wrestling may grade into fighting, but if fighting movements are more rapid, body contact is usually briefer, and usually the initiator repeats the attack. Animals matched in size, fighting in a head to head orientation, bite each other primarily on the head, neck, and breast, and usually manage only to pull out each other's hair; wounding does not occur. When a considerable size disparity exists between combatants, the larger one subdues the smaller by clasping and may bite it severely. COPULATION.?The male clasps the female from behind about the groin and rests his chest and belly upon her lumbar and sacral region. Intromission is achieved by the male's pelvic thrusting and the as- sumption of a lordotic spinal curvature in the fe- male. In one out of three copulations seen, the male (K) bit the hair on the female's neck (C) during the terminal seconds and pulled backwards apparently counterbalancing the precarious position that ac- companied final thrusting. In the other two in- stances (F and I), biting of the female's nape was not seen. Companion-oriented Contact Patterns in Other Genera The two Civettictis were not seen squatting over, pumping with the hindfeet, or grappling. Clasping with either the fore- or hindlimbs is relatively ineffectual because the wrists and ankles have very limited powers of rotation. Consequently the Afri- can civets are unable to effectively restrain one an- other by clasping during wrestling or fighting. In addition, the following contact patterns merit spe- cial attention. NUMBER 239 51 FLEHMEN.?In Civettictis this is an elaborate and time-consuming pattern usually performed in a quadrupedal stance during and after sniffing and/or licking the female's perineal gland, perineal scent, or urine. The first phase consists of sniffing or lick- ing the scent source; the corners of the mouth may be slightly retracted, the ears directed laterally, and the eyes squinted; this is equivalent to Flehmen seen in Genetta. In the second phase, the head is raised away from the scent to approximately a horizontal level and then slowly depressed. If the accompany- ing facial expression is not assumed during sniffing or licking, it is assumed during the nonsniffing phase. Occasionally the eyes are closed and the tip of the tongue extends over the lower incisors. At close range a soft, brief, slowly repeated "puff" is audible, suggesting that the breath is held after sniffing and expired in small amounts at periodic in- tervals. The nonsniffing phase is terminated by licking the nose and lips and sometimes a brief head shake. This may be followed by another ac- tivity or resumption of sniffing. PERINEAL GLAND SCENT-MARKING.?On one occa- sion the male Civettictis was seen to evert the perineal gland, back up, and lightly press it against the female's neck as she was sniffing the wall of the cage. HEAD- AND NECK-WIPING.?While walking past the female the male Civettictis, on a number of occa- sions, extended his head or neck toward her and, with a brief extension movement, wiped some part of her body. VOCALIZATION Table 10 compares the physical characteristics of the various genet vocalizations described in the fol- lowing pages. COUGHING (Muckern: Herter, 1952).?The cough, a brief sound pulse having a broad energy distribu- tion and a narrow intensity latitude (Figure 27A), can be mistaken for a sneeze. A true sneeze, how- ever, has a longer duration and is often accompanied by a characteristic head motion. Coughs may be phonetically rendered, "uff-viff" (Ducker, 1957), and are produced singly, in couplets or triplets, or may be repeated in long series with intervals of various duration. It could be evoked in all individuals at the day of birth and continues throughout life. Production of the sound is preceded by a bumplike expansion of the throat. Expiration of air through the glottis and out the nostrils is accompanied by the disappearance of the bump and a single jerking compression of the rib cage. Spectrograms were occasionally obtained from animals O, K, and C showing a weakly harmonic pulse and a wide noise band that seem to fit the apparent glottal and nasal events, (Figure 27A). HISSING (Fauchen, Explosive Hiss, Spucken).? This vocalization, phonetically rendered as a noisy "hcchhh," is another brief duration call having a broad energy distribution. It is produced by the rapid expiration of air through the opened mouth, and the extent of the mouth opening seems to be correlated with the intensity of the call. Dur- ing violent hissing the mouth assumes an angle of about 35?-40? and the jet of air from the glottis blows saliva out the mouth. Most of the time ex- plosive hissing is accompanied by head-darting with the glottis opened at or near the end of the exten- FIGURE 26.?Facial expression during the panting-hiss in the genet (animal ?; 24 fps; shaded and clear columns = expira- tion and inspiration respectively, estimated from rib cage movements; circles = measurements of the jaw-gape angle in degrees estimated from the film; the curve was fitted by eye). 52 SMITHSONIAN CONTRIBUTIONS TO ZOOLOGY NUMBER 239 53 sion phase. Eye closure may or may not coincide with the production of this sound but often occurs if the mouth contacts the recipient. PANTING HISS.?This call shares many of the phys- ical characteristics of the explosive hiss, but its temporal organization is a function of accelerated respiration (Figure 26). The expiration phase of this sound produces a broad frequency range with bands of energy concentration resembling dissipated harmonics, while inspiration produces low fre- quencies with lower intensity (Figure 27D). The facial expression of this call is characterized by retracted lips and a jaw-gape that varies with respi- ratory phase. In highly aroused animals the position of the vibrissae during the production of the pant- ing hiss changes the contour of the muzzle (Figure 26); thus, an intense form of the call has an ex- piration phase accompanied by a widening of the jaws and an extension of the vibrissae. The retrac- tion of the lips exposes only the lower cheek teeth, canines, and incisors. The call was first observed at the age of two days (Fr), five days (M and E), 13 days (O, I, and C), and 16 days (K). Both forms of hissing may occur with growling and screaming. WHINING (Nestgezwitscher: Goethe, 1952; Quicken: Diicker, 1957).?This vocalization varies widely within and between individuals in har- monics, intensity, and duration (Figures 28A-C). The leading and/or trailing segment of a whine may be frequency modulated. Harmonic form often varies between bouts, and is probably a function of the animal's activity and general muscle tonus. Calls of moderate duration have the largest number of har- monics while very brief and very long calls have fewer harmonics. Whining is phonetically rendered "wee" or "wew," and is first produced within min- FIGURE 27.?Vocalizations of the genet, African civet (Civet- tictis), and the lesser oriental civet (Viverricula): A, sequence of three coughs and a whine in Genetta (O); B, series of eight coughs in Viverricula at the age of 18 days; c, sequence of calls in the female Civettictis: snort (0.2 sec), growl (0.6- 1.0 sec), and two screams (1.0-1.7, 1.9-2.4 sec); D, panting hisses in Genetta (M), high frequency hisses occur during expiration, low frequency hisses during inspiration (six res- piratory cycles are displayed); E, purring in Genetta (Fr), the breaks at 0.8 and 1.7 sec correspond to changes in the flow of glottal air; F, a growl grading into a scream and returning to a growl in Genetta (O), from 0.7 to 1.1 sec the scream is marked by the appearance of a 0.5 kHz frequency band; c, a growl occurring during a single protracted expiration in Genetta (M). utes after birth; it acquires a lower frequency as the animal matures, and is rarely produced after the animal becomes independent. Whining is almost always repetitive and performed with an open mouth, the extent of the gape varying with the intensity of the call. SCREAMING (Schreien).?The scream is essentially a very high intensity whine but having a greater number of overtones (Figure 28c). The distribution of the overtones suggests that two or more mem- brane surfaces are vibrating at different rates. Though the call is produced in situations of high autonomic arousal, the range of intensity subjec- tively seems to be moderate. Softer calls have fewer harmonics and, consequently, a narrower range of energy distribution (Figure 28E, left). The sound was infrequently heard, but it could sometimes be evoked by rough handling in animals that had not yet been weaned. Young animals seem to be unable to respond to pain or trauma with a sound more forceful than a loud whine. An open mouth, ex- tended mystacial vibrissae, retracted lips, and di- lated pupils compose the facial expression that accompanies this call. Very high intensity screams display degeneration of harmonic structure and broad noise distribution (Figure 28E, right). In such cases screams appear to grade into noisy ex- plosive hisses. PURRING (Schnurren).?Purring is composed of continuous and rapid volleys of low intensity clicks produced during both inspiration and expiration (Figure 27E). Usually clusters of sound pulses with similar interval length are separated from other such clusters by longer intervals; these breaks in the sound undoubtedly correspond to the change from inspiration to expiration. Purring occurs with the mouth closed or while the animal sucks milk; the eyes may or may not be closed. It was first heard at various ages: 2 days (Fr), 5 days (M, E), 13 days (O, I, C), and 16 days (K). The sound is very quiet during the first week; this probably explains why it was not detected in my first animals. Some genets continued to purr until they were six months old, but these were raised singly and were exceptionally tame. As noted by Diicker (1957), purring probably does not occur in adults. GROWLING (Knurren).?Growling is a noisy, usu- ally continuous, moderate intensity vocalization with a repetitive pulse. As in purring, the rhythm of growling conforms to the period of the respiratory 54 SMITHSONIAN CONTRIBUTIONS TO ZOOLOGY TABLE 10.?Characteristics of genet vocalizations (frequency in kHz, duration in seconds, duration range in parentheses) Type and number of vocalizations Cough (0) (K) Explosive 28 12 hiss (0) 7 Panting hiss (M) 12 Whine (0) Scream (0 Purr (K) (F) Growl (M) (0) 94 9 6 6 6 8 Frequency Range Total 0.85-8.0 0.85->8.0 0.85-7.5 1.4^>8.0 0.5--7.0 0.85->8.0 0.35--1.5 Fundamental Dominant frequency 0.85- 7.0 0.35->8.0 0.35-4.0 -1.0-2.0 1.4--6.0 -1.4 0.5--4.0 ~0.7 0.35-~3.0 0.85--1.5 Duration Smallest discrete elements < 0.07 -- x=0.18 (0.11-0.27) -- ? <0.007 <0.007 Call < 0.07 x=0.13 (0.07-0.27) variable >1.0 x=Q.76 (0.11-3.59) x=0.86 (0.42-1.5) variable >1.0 variable >1.0 cycle, but the sound is only produced during ex- piration. The duration of each uninterrupted pe- riod of sound seems to be inversely related to the level of the animal's excitation. Periods of growling in a moderately excited animal may last 10 seconds between breaths, while a greatly agitated animal growls for about two seconds before inhaling. In- tensity of the sound is similarly related to period. Growling is often, but not always, accompanied by an alert facial contour; the ears are directed an- teriorly, the pupils dilated, and the mystacial vibris- sae extended. It is often associated with hissing and may grade into a scream (Figure 27F). Though first heard at the age of 80 days (O), it can probably be evoked earlier. VOCALIZATION IN OTHER GENERA The occurrence of call types within selected viver- rid genera is summarized in Table 11. COUGHING.?Unlike the coughing of Genetta, the call in Viverricida malaccensis 7 has a clicking qual- ity and is produced in bouts composed of several cough elements (Figure 27B). Of seven recorded ' A recording of coughing and whining in an 18-day-old male born at the Chicago Zoological Park, Brookfield, Illinois, was kindly made available to me by Dr. Robert Hoi with. coughing bouts, two were composed of six and two of seven elements, and three bouts contained eight elements. The intervals separating the cough ele- ments varied from 0.03-10.46 sec. The majority of intervals (79%) fell within the range of 0.06-10.12 sec. The coughing vocalization in Civettictis civetta, like that of Genetta, is the most common adult call. It is emitted with the mouth closed and is always a rapid series of sound bursts composed of from 2 to 14 elements (Figure 29). The sound is phonetically similar to, but deeper than coughing in Genetta. Coughing in the fanaloka, Fossa fossa, has a distinc- tive quality rendered phonetically as "chuck." The intervals between coughs are variable; while the sound is most commonly repeated in a rapid series as with Civettictis, is is also emitted singly at longer sporadic intervals. HISSING (SNORTING).?Hissing or snorting was heard in all the observed genera with the exception of Viverricida. Diicker (1957), however, recorded "spitting" (Spucken) in this species (Table 11). Snorting constitutes a specialized variant of the hiss in which the dominant lower frequencies probably result from vibration of the soft palate. Unlike hiss- ing, however, snorting always has a loud, brief, and explosive character. As infants are often unable to produce the most forceful of adult vocalizations, it NUMBER 239 55 FIGURE 28.?Vocalizations of the genet and the lesser oriental civet (Viverricula): \~c, six varia- tions in the whining call of genet K. (not a series); D, four examples of whining in Viverricula (not a series); E, three examples of screaming in the genet O (not a series). 56 SMITHSONIAN CONTRIBUTIONS TO ZOOLOGY IJ U14 LL OlO LJJ ^ cQ 6 i 2 . 9 ? i N=67 I x=6.12 Th n , d N=53 p -| x=4.96 rh 2 6 10 14 2 6 10 NUMBER OF ELEMENTS PER CALL FIGURE 29.?Variation in the number of elements composing coughing series in the male and female African civet (Civet- tictis). is possible that hissing may replace snorting in the early life of Civettictis (Figure 27c; Table 11). Though snorting or hissing may be associated with head-darting in all the observed species, limited ob- servations on the male Paguma revealed that snort- ing was always associated with a modified method of head-darting. The snort was produced as the animal extended and raised its neck while swinging the head upwards. At times the animal would quickly rise up off its forefeet while snorting. WHINING (GROANING, BLEATING) AND SCREAMING. ?Four Viverricula whining calls from a recording of 21 are reproduced in Figure 28D. In form these calls resemble those of Genetta except that the har- monics are more widely spaced and there is rela- tively little variation in call duration. The one kHz TABLE 11.?Distribution of vocalizations in social situations for selected viverrid genera (+ = present, ? = absent, +? or ? ? = probably present or absent but substantiation needed, blank = uncertain; calls divided into 6 major families on basis of overall physical similarity) Vocalization Viverrinae Paradoxurinae Hemigalinae o fc. s- > > I/I ?t? +-> o +J + J 0) > <_J CO c ?9? -o c 0.05, Appendix 3). Tail- flaring often occurred as an animal maintained an arched-back stance, but this relationship is not shown in the table. CONTINGENCIES OF COMMON NONVOCAL INTERAC- TION PATTERNS IN Genetta.?Table 14 gives a rough approximation of some common behavioral associa- tions. The table is an abbreviation of the original matrix, which contains twice as many behavior cate- gories. The consequence of eliminating many of the fine grained behaviors (e.g., closes eyes, ears back, mouth open, etc.) is that some common dyads are obscured. Sniffing and darting is a case in point. This common dyad is not shown in Table 14 be- cause the intention movements of darting were re- corded between the occurrence of the sniff and the head-dart. Several contact patterns are presented as having followed depart (e.g., sniff, dart, bite, paw, and stand-on). This resulted from failure to record all approaches, particularly during fast action. In addition, seemingly incongruous sequences like these occurred when an animal attempted to de- part. For example, as an animal turned to depart, its tail might be sniffed by the companion. The simplest form of interaction during en- counters consisted of movement toward and away from the companion. Companion-oriented move- ment was frequently interposed with sniffing by one or both animals. Sniffing and head-darting often occurred in minimal- and maximal-contact encoun- ters and constituted the most common elements of social exchange. For convenience the association between these two patterns will hereafter be re- ferred to as the sniff-dart axis. The sniff-dart axis and multiple contact ex- changes represent the ends of a complexity contin- uum involving contact. Though it is possible for any two patterns to occur either singly or in re- peating and alternating series, sniffing and darting were by far the most common elements in dyadic interaction. Other types of contact (e.g., pawing and biting) had no dyadic integrity. Biting, pawing, standing-on, and erratic head movements are relatively infrequent behaviors dur- ing most encounters, and their contingencies are by no means fixed. Tail-flaring is almost always seen in minimal-contact situations in association with move- ment away from the companion (Table 13). The arched-back stance tends to repeat itself as an initial response to the visual fixation and approach of an- other animal. It can be seen that noncontact pat- terns such as tail-flaring, the arched-back stance, and sometimes erratic head movements are associated with movement toward or away from the com- panion, but the first two patterns are at times precipitated by fighting. Erratic head movements were subequally contingent upon contact, sniffing, and companion-oriented locomotion, indicating that the pattern was performed during, as well as between, episodes of contact. CONTINGENCIES OF COMMON, NONVOCAL INTERAC- TION PATTERNS IN Civettictis.?The two Civettictis 66 SMITHSONIAN CONTRIBUTIONS TO ZOOLOGY TABLE 14.?Contingency matrix of interaction patterns in Genetta tigrina based on encounters of 12 paired arrangements of 6 study animals; matrix includes behavior patterns from sequences enacted by one animal and/or between two animals (App = walking approach, Appc = crouched approach or stalk, Appr = running approach, SO = stand on or clasp with fore- limbs, EHM = erratic head movements, TF = tail flare, ABS = arched-back stance) Fol 1 ows ^""~-\ App Appc Appr Sniff Dart Bite Paw SO EHM Depart Follow TF ABS TOTALS App 19 18 12 45 2 1 3 84 6 9 1 200 Appc 28 00 3 5 1 45 Appr 6 2 1 9 Sniff 241 10 323 4 1 1 3 6 30 18 6 643 Dart 4 1 5 36 1 6 1 54 Bite 1 4 3 3 3 2 1 17 Paw 3 4 1 1 9 SO 2 1 2 8 3 1 17 EHM 8 1 6 2 1 4 22 Depart 97 12 2 177 8 5 2 2 7 26 3 21 1 363 Follow 8 1 17 3 1 1 9 1 7 48 TF 4 2 2 1 24 7 6 46 ABS 2 1 5 8 TOTALS 423 54 26 595 55 14 5 9 19 181 34 53 13 1481 exhibited behavioral contingencies similar to Gen- etta but different by the incorporation of several other patterns. As with genets, departure frequently terminated sniffing, and sniffing and head-darting were both repeated more often than they were fol- lowed by other patterns (Table 15). Aside from such repetitions, sniffing was most frequently followed by head-darting. Unlike the genets, however, the recipient of head-darting more commonly reclined than departed from the com- panion's vicinity. Under these circumstances, re- cumbancy characteristically involved a lateral or tangential body orientation to the companion. This kind of body orientation was also assumed by stand- ing animals; the pattern was most commonly a response to the head-darting or frontal orientation (turn towards) of a proximate companion. Sparring and clasping were also common antecedents of re- clining, but only in the former was reclining asso- ciated with a lateral body orientation. It was evident in these two animals that the assumption of either a standing or reclining lateral orientation often pre- cipitated further darting. After the darting animal departed or desisted, the reclining one would stand and move away. Approaching, turning away, depart- ing, and following bear a basic similarity to the same patterns in the genet. Erratic head movements were not commonly seen, but occurred most often before an animal moved away from its companion. In both Genetta and the Civettictis sniffing and NUMBER 239 67 TABLE 15.?Contingency matrix of interaction patterns in Civettictis civetta based on nine en- counters of 2 opposite sex study animals (EHM = erratic head movements, LO ? lateral or tangential body orientation, TT = turn towards, TA = turn away, boxed figures = values that account for 20% or more of column totals) ^\Precedes Follows ^ \ App Sniff Dart Spar B1te Clasp EHM Recline Depart Follow Stand LO TT TA TOTALS App Sniff 125 1 108 | 47 6 7 23 5 26 17 2 4 13 17 3 295 27 7 2 7 6 24 96 26 6 37 10 356 Dart 14 140 16 4 7 73 48 19 32 10 15 383 Spar 1 12 5 1 19 21 8 4 1 72 Bite 1 4 2 1 2 2 2 13 2 3 2 34 Clasp 2 4 11 6 1 3 21 8 2 1 1 60 EHM 1 1 1 1 3 2 4 10 3 2 1 29 Reeli ne 4 40 47 8 4 3 5 11 26 I 42 1 15 3 208 Depart 1 32 1 11 13 2 97 Follow 8 4 1 3 1 6 7 7 37 Stand 4 4 13 4 5 4 3 7 65 2 2 10 123 L0 3 21 5 4 18 1 3 3 58 TT 3 13 40 13 3 16 12 3 14 1 5 123 TA 5 4 6 12 7 2 2 38 TOTALS 43 325 364 73 36 54 30 235 353 41 132 71 110 46 1913 head-darting were more commonly contingent upon themselves than upon any other single pattern of contact. At this level, interaction in Civettictis ap- peared to be more complicated because of the as- sumption of postural modifications that permitted further head-darting; proximal interaction was usu- ally terminated when the head-darting animal de- parted. In Genetta head-darting and sniffing were terminated when either animal departed. Lateral body orientation, assumed during back-arching, was characteristically displayed at a distance rather than in the companion's proximity. In addition, Genetta modified its movement toward and away from the companion with postural adjustments such as crouching and tail-flaring. CONTINGENCIES OF VOCALIZATIONS IN Genetta.? As mentioned previously, coughing in Genetta was usually produced in bouts with intervals ranging from one to several seconds. It is evident in Table 16 that more than half of all coughs were preceded or followed by vocalizations, the majority of which were coughs. Locomotory transitions, and move- ment to and from the companion accounted for nearly 30% of all contingent behaviors; about 15% of the calls occurred either before or after the ani- mal sniffed the cage or its companion. Although 68 SMITHSONIAN CONTRIBUTIONS TO ZOOLOGY sometimes a response to another animal's contact, an outstanding feature of this vocalization was that it commonly occurred during transitions in gaits, stances, reclining positions, and sniffing. Once an animal produced the sound, it was likely to do so again before a change occurred in its ongoing be- havior. Hissing also occurred most frequently in series; of the 110 times it was contingent upon vocaliza- tions (Table 16), only 10% involved another call type, i.e., growling. There was a fourfold difference in the percent of the time that hissing was preceded and followed by companion-oriented locomotion; this is because the call was usually produced as a response to a companion's approach. Darting was al- most always performed by the vocalizing animal and was twice as common after the call as before. About 25% of the time either visual fixation or jaw-gaping followed the delivery of a hiss; both of these acts were displayed predominantly by the caller. It is clear that hissing occurred primarily during proxi- mal interaction in which contact was minimal. In the encounter situation, growling was only exhibited by animals that had been defeated in fights. Following defeat, animals often growled merely at the sight of the companion indulging in "neutral" activities, such as self- directed licking and nibbling, cage sniffing, or changing body position. TABLE 16.?Number and percentage (in parentheses) of selected behavioral contingencies with vocalizations in Genetta tigrina Contingent behaviors Sniff cagel Locomotion2 Companion- oriented movements Arched-back stance Visual fixation Jaw-gaping Head-darting... Other contacts Vocalizations.. Stretching, shaking and self-oriented contact TOTALS3 Cough Preceded by 45 97 69 ? -- ? 49 6 316 25 607 (7.4) (15.9) (11.4) -- -- - (a D (0.9) (52.1) (4.1) Followed by 56 96 73 -- -- ? 45 6 316 9 606 (9.2) (15.8) (12.8) -- ? -- (7 A) (0.9) (52.1) (1.5) Hiss Preceded by 7 -- 33 2 7 14 I 25 1 58 148 (4.7) ? (22.3) (1.4) (4.7) (9.5) (0.7) (16.9) (0.7) (39.2) ? Followed by 0 ? 8 1 16 20 45 1 52 143 ? ? (5.6) (0.7) (11.2) (13.9) (31.5) (0.7) (36.4) - Growl Preceded by 2 1 "3 11 1 5 5 11 2 4 6 60 (3.3) (21.6) (18.3) (1.6) (8.3) (8.3) (18.3) (3.3) (6.6) (10.0) Followed by 3 f. 2 1 8 12 14 2 __ 3 5 56 (5.4) (10.7) (3.6) (1.7) (14.3) (21.4) (25.0) (3.6) (5.4) (8.9) Scream Preceded by ? -- 14 ? ? 3 18 3 38 ? -- (36.8) ? __ ? (7.9) (47.4) (7.9) ? Followed by ? -- 3 ? ? 2 25 3 33 ? -- (9.1) ? - (6.1) (75.8) (9.1) 1 Includes licking and nibbling the cage or scent marks. 2 Including transitions from stances to locomotion. 3 Estimates of degree of ta i l -hair erection were occasionally contingent upon vocalizations The figures results in unequal tota ls. exclusion of these NUMBER 239 69 Defeated animals also growled as they paced in a stereotyped manner in front of the cage. Growling occurred before as well as after animals visually fixated their companions, or assumed an arched-back stance. These three behaviors accounted for 38% of all antecedent acts and 46% of the suc- cedent acts (Table 16); hissing was the only vocal pattern contingent on this call. Screaming was the least frequently heard call dur- ing encounters. It occurred before or after the re- ception of head-darts, but was most frequently emitted as animals engaged in wrestling, fighting, biting, and clasping. Contact patterns accounted for 9 BEFORE 9 a AFTER B 9 d AFTER 9 cf BOTH n _ n n ra dartspar h 5 2 change body pos. g 1 recline f 9 3 4 approach e depart d 4 1 2 4 latorient c 2 1 frontal orient b 8 1 1 no change a 13 3 7 3 2 2 2 41 5 13 3 14 2 2 1 6 0 - 5 0 - 30 20 10- Lm. i I 50%-75% of all antecedent and succedent behav- iors. Approaching was the only form of companion- oriented locomotion that animals responded to by screaming. Screaming then differs from growling and hissing by its association with extensive body contact. The contingencies (Table 16) do not indi- cate that pain inflicted during biting is the most important stimulus evoking this call. Associated patterns such as clasping and wrestling were never seen to elicit the call by themselves. It appears that once an animal received a painful bite, however, more generalized patterns such as approaching be- came effective stimuli. CONTINGENCIES OF VOCALIZATIONS IN Civettictis.? As in Genetta, Figure 34A shows that sniffing the cage and general body movement constituted the vast majority of activities associated with coughing. It can be further seen that in each animal olfactory and (less commonly) gustatory sampling of the cage was quite often an antecedent to vocalization. Less than 20% of all coughing-contingent behavior was nasal contacts. In both animals less than 95% of all nasal contacts was contingent upon coughing. Loco- motory transitions were relatively more frequent succedents than antecedents. More than 70% of the occasions after a call was FIGURE 34.?Behavioral contingencies of vocalizations in the African civet (Civettictis): A, Caller's behavior before, after, and both before and after coughing (the segments of each bar in the graphs represent different behavior patterns: bottom segment = sniffing, lick- ing, and nibbling of the cage (or scent marks); middle seg- ment = sniffing of the companion; top segment = transitions in locomotion and stances; difference between bar and 100% indicates percent of calls with no overt contingent behavior; numerals above each bar represent the number of calls from which percentages were calculated). B, Companion's behavior before, after, and both before and after the caller's coughs (bottom shaded segment ?? nasal contacts; lower clear segment = all other companion-oriented contacts (bite, dart, paw, stand on, etc.); middle shaded seg- ment = approach, turn toward, look at; upper dear seg- ment = depart, turn away, look away; top shaded segment = coughing). c, Behavior patterns antecedent and succedent to growling by the female (the bottom graph displays patterns of the male?white bare?and female?black ban?that were ante- cedent to growling by the female; the succedent events in the male are enumerated above each bar; upper graph presents the relative proportions to those succedent events expressed as percentages of the total number of times the female growled, the lower case letters correspond to the behavior patterns listed at left). 70 SMITHSONIAN CONTRIBUTIONS TO ZOOLOGY given there was no overt change recorded in the companion's behavior (Figure 34B). Antecedent ac- tivities of the companion may have evoked the call on some occasions. It is evident that contact by the female was the preponderant antecedent to the male's coughing; and that moving or looking toward or away from the partner was the chief activity of the male preceding the female's coughing. Differ- ences between animals were also exhibited in types of response. In response to the male's coughs, the fe- male sniffed the male 3%, approached 9%, departed 2% of the time, and never coughed back. After the female called, the male contacted her 15% of the time, approached 5%, and coughed 7% of the time. The contingencies of growling differ considerably from coughing. Of 85 growls by the female, 80% were apparent responses to actions of the male (Figure 34c, bottom). Female-oriented movements and nasal contacts accounted for 81% of all the male's actions that evoked the female's growling; she also growled when the male simply moved. When there was no change in the male's behavior, the female commonly growled when she moved in his direction. It is evident that nearly 40% of the time after the female growled there was no change in the male's conduct (Figure 34c, top). The most common form of response consisted of position or orientation changes (e.g., reclining, lateral and frontal orienta- tion, and departure). Growling led to contact in the form of sparring only 10% of the time. SENSORY CHARACTERISTICS OF SOME NONVOCAL BE- HAVIORS.?As a nocturnal forest species having a retina composed solely of rods (Diicker, 1957), Genetta distinguishes only degrees of brightness and probably cannot discern fine visual detail. In view of these limitations, it is important to consider the visual characteristics of noncontact patterns and how they might be perceived by conspecifics. The patterns discussed below can be recognized only visually. Noncontact patterns involving gaits (ap- proach, departure) may also have important visual components, but concurrent environmental sounds produced by footfalls also provide information con- cerning proximity, and direction and rate of move- ment. Staring is a common activity during encounters between strange animals. If visual fixation provides information only to the animal engaging in it, there would not be selection for behaviors or structures that emphasize this activity to the companion. On the contrary, the patterning of light-colored fields of facial hair provides two characteristics hypothesized as facilitating the companion's recognition of the head and its orientation in dim lighting. (1) The markings present a distinct visual configuration not present elsewhere on the body, and (2) angular de- flection of the head from a frontal position is pre- cisely indicated by relative changes in the number and surface area of the markings. Figure 35 shows the appearance of different head positions and provides measures of the amount of white facial hair exposed through a 180? horizontal arc. It is evident that disruption of pattern symmetry and disparity in spot size are two manifestations of only a 30? deflection from a frontal head position. These changes occur with about a 20% decrease of the area of white hair exposed during frontal orienta- tion. As the eye has relatively little mobility (see "Primary Senses"), visual fixation of movement is achieved primarily by adjustments of head position. A moving animal being stared at therefore prob- ably receives a constant facial image in which four symmetrical light markings are prominent. Iron- ically, a similar effect may also be produced by an animal with closed eyes if the head is directed toward the companion. Staring accompanies crouched and arched-back stances, but the visual characteristics of these pat- terns are distinct. The body axis is frontally oriented during crouching, exposing a relatively small body area to the companion. The tail is con- cealed or minimally visible and the above- mentioned facial characters predominate. On the other hand, a lateral or tangential position maxi- mizes body exposure in the arched-back stance. The light-colored hair on face and tail delimits the animal's dimensions. Differences between these two patterns may be considered antithetic (Darwin, 1872) and involve both quantitative (head elevation, degree of body exposure) and qualitative differences (visibility or nonvisibility of tail). The rapid flaring of hair on the tail produces several changes that may be perceptible to a com- panion. Because the bases of light hairs are lighter than the tips, erection seems to increase the con- trast between the dark and light colored bands. Pilo-erection also noticeably enhances the demarca- tion between bands; in the relaxed state the over- lapping of light and dark hair at band boundaries NUMBER 239 71 FIGURE 35.?Appearance of different head positions in the genet and variation in the relative proportions of the white facial markings through a 180? horizontal arc from a frontal position; (gray horizon delimits the area of exposed white hair as a percentage of the amount visible at 0?; approximate percentages of the muzzle and eyespot markings are indi- cated by the solid and dotted lines respectively, together they equal the percentage expressed by the gray horizon; cal- culations are based on planimeter measurements from en- larged photos of the horizontally resting head of a dead genet). produces a more graded effect. In all probability the visual pattern that is produced?a rapid expan- sion of a series of light-colored bars?draws atten- tion to the tail and the animal. The sensory properties of other behavior patterns are more difficult to enumerate. The properties actually transmitted are not necessarily those per- ceived by the observer and probably vary under different environmental and social circumstances. Head-darting, for example, involves several levels of motor integration. In its simplest form it may be visual or tactile. If the companion is watching the darting animal, the signal can exhibit visual, as well as tactile and auditory properties. The sensory mani- festations of the transmitted signal, therefore, de- pend upon the receiver's body orientation and attentiveness. The redundant character of such a pattern probably compensates for interference of one or more sensory channels. Sudden high stimulus contrast characterizes all the sensory properties of this pattern. It appears that the most obvious properties of certain patterns may not always embody the salient stimuli. Though erratic head movements are usually conspicuous in a lighted room, they are commonly performed in circumstances where visual detection is not possible. The irregular and noisy sequence of footfalls is an outstanding feature of this move- ment detectable irrespective of orientation to the other animal. In a sense the quality of this sound mimics the visual character of the movement. When the movement is performed during following, the head and neck glance and brush against the rump, hindlegs, and tail of the anterior animal. BODY TARGETS OF SNIFFING AND HEAD-DARTING IN Genetta.?Figure 36 reveals a bimodal distribution of sniffing frequencies to the fore and hindbody. While the head, nose, and neck were the most com- monly sniffed forebody targets, the rump, anogenital region, and tail (including the tailbase) received 45.2% of all nasal contacts, the tail and tailbase ac- counting for 29%. These common anterior and posterior targets accounted for exactly two-thirds of all companion-oriented sniffing. The side was the remaining most commonly smelled anterior target region; it also constitutes a relatively large surface area. Forty-six percent of all sniffs were directed to targets between the head and side, while the re- maining parts of the body, composing slightly less than half the available body surface, received 54% of the total number of sniffs. The ratio of the frequency to the total duration of sniffling for any given target varied considerably (Figure 36). Ratios of <1:1.5 were obtained for all targets anterior to and including the side. On the hindbody only the legs and feet had similar ratios. Almost all other posterior body targets had fre- 72 SMITHSONIAN CONTRIBUTIONS TO ZOOLOGY 172 167 196 54 298 300 312 HEAD-DARTING 89 - 3 2 9 4 1 66 70 96 88 1 J i m^ 4_tailbase tail hindtoot hindlegV un-spec FIGURE 36.?Body targets of head-darting and sniffing in the genet (numerals above bars in head- darting graph and top numerals above bars in sniffing graph = number of observations for that target; bottom numerals above bars in sniffing graph = total duration in seconds of sniffing directed to each target; solid bars = frequency percentages based on total number of all ob- servations for specified and unspecified targets; shaded bars = duration percentages based on total duration of sniffing directed to all specified and unspecified targets). quencyrduration ratios of 1:1.5-1.9. The spinal crest had the exceptional ratio of 1:3.75. This clearly demonstrates that duration and frequency of contact are not equally proportional for all targets. Fre- quently sniffed targets such as the head were sniffed briefly (frequency:duration = 1:1.3), while rela- tively long periods were devoted to infrequently sniffed targets such as the spinal crest (frequency: duration = 1:3.75). Spinal crest sniffing exhibited a characteristic pattern: the nose (probably guided by the contour of the underlying vertebrae) was rubbed anteriorly within the confines of the black hair of the crest. The vast majority of the time the companion's head was the head-darting target; the neck, fore- legs, and tail were struck at less frequently. Other body areas were infrequent targets (Figure 36). BODY TARGETS IN Civettictis.?An interesting comparison with Genetta is presented by this species. While the head, neck, and sides were the most commonly sniffed anterior regions, in the pair of Civettictis the majority of sniffing was directed to posterior body targets. The female directed 60% of all nasal contacts to the male's hindquarters, while the male addressed 65% to this area of the female's body (Figure 37). In a number of instances it was possible to deter- mine the precise orientation of sniffs to specific body FIGURE 37.?Body targets of head-darting and sniffing in the female and male of Civettictis (each category exclusive of others): A, general body targets of head-darting and sniffing in the male and female; B, specific targets of head-darting and sniffing on the head of the male and female; c, specific targets of sniffing on the anogenital region of the male and female (female not shown here to conserve space; because body targets were not always noted specifically during an encounter, general targets such as head and anogenital region have been duplicated in B and c, e.g., the undertail category of c was in- cluded as part of the anogenital category in A; numerals ac- companying each segment of the graph represent the number of observed instances from which the percentages are cal- culated). NUMBER 239 73 4 0 2 0 6 0 Z 2 0ilua. OL 20 20 ? n . 178I . 21 n1 7 usiW f i 35 | 3 J 0 . SNIFFING HEAD-DARTING m 4 1 1 3 0 40 * 22 n 13 ? 6 [~I ? . 1 n I ji. ? 1 .1 HI 9 cf 1 n 17 58 1 1 n i ? 0 : 18 14 4 - 15 head forefoot breast chest midback crest tailbase tail hmdfoot .3. A J . HEAD-DARTING rf 37 ^ ? j . l 3 A. ? I SNIFFING 9 15 ii mouth eye cheek ear ear ear ear mask head nose ? forehead ? in front ? behind belc B 74 SMITHSONIAN CONTRIBUTIONS TO ZOOLOGY parts (Figure 37B,C). Based on a limited number of observations, the nose, cheek, and ear received the majority of nasal contacts to the head. Sniffing tar- gets of the hindquarters differed only slightly be- tween the two subjects. Targets in this area on the female can be ranked in decreasing frequency from scent-gland, under-tail, to anus. On the male's body, the order of the first two targets was the same, the scrotum was the third, and the anus was sniffed least commonly. As in Genetta the companion's head was the most common target of head-darting, but in addi- tion the neck, foreleg, and hindleg were also hit. While these areas were also struck in the genets, the relative proportion of darts to them was less, par- ticularly in the case of the hindleg (Figure 36). There is an obvious difference in targets between animals. The male darted primarily at the neck, foreleg, and hindleg (totaling 80% of 21 darts), and the head and side were infrequently struck. The female directed over 65% of 262 darts to the male's head and struck at the neck less than 15% of the time. The neck, hindleg, and foreleg were the next most common targets (Figure 37A). Thirty percent of the male's darts directed to the head struck the ear region, while the mask and cheek were less common targets. In comparison with Genetta several quantitative differences can be cited. Whereas the genets sniffed the tail more often than the anogenital region (71% of all sniffs to both areas), the civets exhibited only a slight bias in favor of the tail (54% and 63% in the male and female). In Genetta 33% of all sniffs to the tail were directed to the base while in the civets 54% of all such sniffs were aimed at the base (male = 48%, female = 65%). Thus in Civettictis the proximal portion of the tail and the anogenital region seem to be more preferred or accessible targets for sniffing. From the standpoint of hair patterning these areas are quite distinctive (Figure 37c). The scent gland is a visually and physically prominent attribute comparable in its alternating pattern of white and black to the sides of the neck. The tail of Civettictis differs from Genetta in that most of the white hairs are concentrated on the sides and ventral surface of the base. It is also significant that the proportionately shorter tail of the African civet provides a relatively smaller sniffing area than in the genets. Interaction through Sniffing and Head-darting: the Sniff-Dart Axis As sniffing and head-darting are prominent and causally linked elements of interaction, it is impor- tant to consider some aspects of their occurrence. The following analysis is based on the assumption that head-darting occurs predominantly as a re- sponse to a proximate animal's sniffing or attempt- ing to sniff its companion's body. ROLES OF "SNIFFERS" AND "DARTERS".?Table 12 reveals that if a series of encounters is considered the animal with the highest sniffing score may or may not have the highest darting score. The mean- ing of these summated values, though, is difficult to interpret because the relationship between two animals can change during a series of encounters. A number of patterns emerged, however, when head-darting frequencies were compared for all single encounters in which a given animal was the sniffer. Three of the six animals exhibited statis- tically significant differences (Appendix 4). Animal K, for example, darted significantly less frequently than his companions when he dominated sniffing (P <0.005). Animal C also darted less during en- counters with F (P <.O5). There was no significant difference in her darting frequency, however, when these encounters are combined with the other ones in which she was the predominant sniffer. Animal O, on the other hand, darted more frequently in encounters even though he was the dominant sniffer (P = 0.025). The remaining three animals displayed no consistent tendency to dart more or less than their companions. Thus, while there is usually a disparity in any given encounter in the relative frequencies of sniffing and darting, some sniffers predictably darted either more (O) or less (K, C) than their companions. In I, F, and Mo the level of darting bore no consistent relationship to frequency of sniffing. Similar results were found in the two African civets. The male, who domi- nated sniffing in eight of nine encounters, darted significantly less than the female (P<0.01, Appendix 5). Appendix 6 records the darting scores of both animals in 21 encounters in which only one genet sniffed its companion. In 15 out of 21 cases the sniffer received darts but did not dart itself. In the remaining 6 encounters, both sniffer and nonsniffer darted and the nonsniffer predominated. There ap- NUMBER 239 75 pears to be a tendency for the sniffer to refrain from darting, but this is not a statistically significant pattern (P >0.05). On the other hand, of 68 en- counters in which both animals sniffed, on only 10 occasions (14.7%) did the less sniffed animal refrain from darting. When the less sniffed ani- mal does dart, the level of darting varies indi- vidually both within and between encounters. These findings tentatively suggest that the sniffer refrains from darting as long as the companion confines its activity to darting but does not smell back. INDIVIDUAL VARIATION OF TARGET SELECTION.? The overall distribution of sniffs and head-darts to body targets may not be indicative of the specific distribution exhibited by a given animal. Also, the sniffing and head-darting orientation of any one animal may or may not vary with different com- panions. To test the hypothesis that target distri- bution was independent of introduction of differ- ent companions, sniffing and head-darting targets were grouped into four arbitrary categories: (1) head and neck, (2) trunk (including shoulders, chest, spinal crest, midback, rump, anogenital region, and belly), (3) legs (fore- and hindlegs and feet), and (4) tail (including tailbase); frequency tabulations were ranked for these four body regions on each companion that was encountered by a given animal; and the results were tested with the Friedman two-way analysis of variance (Appendices 7 and 8). Though the head and neck were the predominant targets, the absence of scores for the other regions resulted in rank ties that produced insignificant variance. The null hypothesis for sniffing was re- jected at a probability of <0.017 for four of the six animals (K, C, I, and O). For head-darting only animals C and I deviated from a random variance (P< 0.054). These results indicate that though most animals darted and sniffed more often at certain regions than others, the relative frequencies to these regions were not necessarily consistent with different companions. Thus, while almost all animals exhib- ited disproportionate contact distributions to their companion's body regions only two of the six ani- mals were inconsistent in the way they sniffed their companions, and only two were consistent in head- darting at their companions. SNIFF-DART RATIO AND TARGET SELECTION.? Tables 17 and 18 summarize the relationship of head-darting to sniffing and sniffing targets in Genetta and Civettictis. The frequencies of these two patterns were converted to rates to facilitate comparison. Though the rates varied between en- counters, the overall sniffing rates of the civets were comparable to the higher rates shown by the genets. The overall darting rate in the female civet was two times greater than the highest rate dis- played by a genet; the male displayed a darting rate comparable to the lower rates in the genets. On the average, then, the civets sniffed and darted at relatively high rates. It is also evident that the variation in sniff-dart ratios is not correlated with any fluctuation in overall rates of sniffing and dart- ing during an encounter or series of encounters. It can be hypothesized that the distribution of sniffs to the body parts is partially determined by the relative proportion of sniffs that are responded to with darts (i.e., the sniff-dart ratio). As head- darts usually cause a sniffing animal to desist (at least momentarily), the predominance of tail and hindbody sniffing appeared to be a strategy permit- ting maximum olfactory sampling at a minimum risk of sustaining a dart. To test this hypothesis, the body was divided into three sections that maximally reflected the longi- tudinal polarity of sniffing distribution (Tables 17 and 18) and the frequency distribution of sniffs to these three body parts was tested for low and high sniff:dart ratio encounters with the Friedman two- way analysis of variance (Appendices 9 and 10). In genets exhibiting sniff-dart ratios < 1:0.5, there was a highly significant variance of the sniff distribution to the three regions (P<0.001). When the same genets had sniff:dart ratios > 1:0.33, the variance between body regions did not differ from random. Thus, though both groups showed an overall in- creasing gradient of sniffs from fore- to hindbody, the pattern was consistent only in encounter series characterized by low sniff:dart ratios. These results are at odds with the prediction that high levels of darting restricted sniffing to the hindquarters. On the contrary, low sniff:dart ratios do not predictably indicate a sniffing bias to the hindbody, and genets strongly inclined to sniff the fore- and midbody did so in spite of the frequent darts they received. It appears that although head-darting is often elicited by sniffing, it does not predictably constrain the companion's sniffing to the rear end. In the African civets, sniffing variance between 76 SMITHSONIAN CONTRIBUTIONS TO ZOOLOGY TABLE 17.?Relationship of sniffing to head-darting frequency and to sniffing body targets (percentage in parentheses) in Genetta tigrina 1 2 3 4 5 6 7 8 9 10 11 Animal Pair F 0 C K Mo K I K C I C Mo C F I 0 C 0 o"- ? o" ? ? 9-. rf.. d?'.'. d'.\ 2.. 2.. 2.. 2.. 2.. 2 - 9 ?? I ?.. Mo 2 ?? I F 9 ?? Sniff Total 401 444 233 161 23G 218 160 92 58 60 58 64 79 1 0 43 5 38 8 30 2 1 per 10 min. 10.4 11.5 6.6 4.6 7.7 7.1 5.0 2.9 2.6 2.6 1.4 1.5 2.4 0.3 1.2 0.2 1.1 0.2 0.7 <0.1 <0.1 Head-dart Total 31 62 170 42 16 8 24 8 49 4 17 21 0 57 9 30 33 4 4 7 12 2 per 10 min. 0.8 1.6 4.8 1.2 0.5 0.3 0.7 0.3 2.2 0.2 0.4 0.5 1.7 0.3 0.9 0.9 0.1 0.9 0.2 0.3 Sniff:Dart ratio! 1:0.15 1:0.07 1:0.18 1:1.10 1:0.03 1:0.07 1:0.05 1:0.26 1:0.07 1:0.82 1:0.36 1:0.26 1:0.72 1:0.21 1:0.80 1:0.86 1:0.87 1:0.13 1:2.00 1:6.00 Sniffing Forebody 129 65 44 24 54 47 30 21 15 18 15 25 6 0 0 3 2 11 6 5 1 0 (32.3) (14.6) (18.8) (15.0) (22.7) (21.6) (18.9) (22.8) (26.3) (30.0) (25.9) (39.1) (7.6) (6^9) (28.9) (75.0) (16.6) -- Body Targets2 Midbody 149 137 58 50 52 44 60 33 12 19 9 6 9 1 0 3 3 2 2 5 0 0 (37.3) (30.8) (24.9) (31.3) (21.8) (20.3) (65.2) (35.8) (21.1) (31.6) (15.5) (9.4) (11.4) (6^9) (5~3) (15.0) (16.6) -- Hindbody 122 242 131 86 132 126 68 38 30 23 34 33 64 0 0 37 0 25 0 20 1 1 (30.5) (54.5) (56.2) (53.7) (55.5) (58.1) (73.9) (41.3) (52.6) (38.3) (58.6) (51.6) (81.8) (86.0) (65.8) (66.6) -- Total duration of encounters per pair (hr :min:sec) 6:24:06 5:52:57 5:08:39 5:18:24 3:57:25 6:41:20 5:29:40 5:54:02 5:34:16 6:45:39 5:52:44 Ratio of the number of companion-directed sniffs to the number of head-darts received from the companion. 2 Forebody = head and neck; midbody = forelimbs, chest, shoulder, side, back, spinal crest, and bel ly; hindbody t a i l , rump, hindlimbs, anogenital region. Percentages calculated only when sniff ing frequency ^30. the three body regions was statistically significant in situations of high and low sniff:dart ratios (P<.05), but a distinct difference in rank order was exhibited. Though the hindbody was usually the most frequently sniffed region in both groups (Appendix 10), the forebody ranked second highest in high ratio encounters and lowest in low ratio encounters. So while there was no difference in variance between the two conditions, the forebody was a high frequency target more often when sniff-dart ratios were high. In view of the limita- tions of the small number of animals, the data do not lend themselves to further interpretation. DISCUSSION Vocalizations among Viverrids Table 11 reveals that there is relatively little variation between species in the number of calls and their distribution within basic call groups. The most widespread calls, namely hissing, whining, screaming, and growling are known to occur in several other mammalian orders. Whining, a char- acteristic call of infants given under a broad variety of circumstances, is probably homologous to the tremelo groan of Fossa and the bleating call in NUMBER 239 77 TABLE 18.?Relationship of sniffing to head-darting frequency and to sniffing body targets (percentage in parentheses) in Civettictis civetta Encounter and animal 1 2 3 4 5 6 7 8 9 _ 30. Entries in columns 2 and 4 are means. Nandinia. All are emitted repeatedly as responses to temperature changes, hunger, and other sources of discomfort or pain. Nandinia differs most from the other species in retaining the call into adult- hood, when it is produced in response to tactile stimulation and apparently mild pain. Low intensity whines are occasionally produced by adult genets and African civets, but screaming is the call that is usually emitted. Unlike whining, screaming is a response to specific painful stimuli that often arise during interaction with other animals. It appears that during development whining becomes decreasingly responsive to general discomfort stimuli and is almost always exhibited in an intense form (as screaming) to painful stimuli. From the standpoint of avoidance context, hissing and snorting are essentially identical calls. In all species they are often associated with head-darting, but in Paguma this connection appears to be obli- gatory. Humming seems to be a specialized deriva- tive of growling. While long duration is shared by both calls, humming is frequently modulated and the species that display it do not growl. Though humming was usually not produced under circum- stances conducive to growling, when disturbed from rest the Paguma often hummed after a series of 78 snorts. In Nandinia, animals mimiced the gradual frequency oscillations of one another, but the func- tion of the call is obscure. Hooting and neighing are unique to the two palm civets (Nandinia and Paguma) and no repe- titive sound having equivalent intensity has been heard in any other species. Eisenberg (pers. comm.) heard a whinnying call in wild Ceylonese Para- doxurus that approximates my description of neighing in Paguma. All three of these species are solitary arboreal omnivores with a strong predis- position for fruit. It is feasible that these calls function in spacing, for their brief, loud, and repetitive sound properties make them easily local- ized over relatively long distances (Whitfield, 1971). Though these three species of palm civets also have perineal scent glands, an examination of animals and museum specimens indicated low sec- retory rates of relatively weak scents. If it is assumed that the paradoxurines evolved from semi- arboreal ancestors in which scent markings served a spacing function, it is conceivable that high intensity calls have assumed at least some presump- tive scent functions. The advantage of sound com- munication in a "volume environment" is that it can be broadcast over long distances, its transmis- sion is dependent on and enhanced by air current, and it is doubtful that all parts of civets' living volume are equally accessible for scent sampling. It is curious that none of the terrestrial foraging civets (Genetta, Civettictis, Fossa) have repetitive high intensity calls in their repertoires. These calls occur predominantly in small arboreal mammals (galagos, tree hyraxes, monkeys), and terrestrial species that are large (felids), or medium-sized and social (canids). This pattern of occurrence suggests that such a call might make a small, ground-living species more vulnerable to predation. In terrestrial species scent would seem to be a more effective means of distance communication because (1) the probability of a strange animal entering the active space of a scent mark would be greater in a two-dimensional than three dimensional environment, and (2) scent marks do not betray an animal's immediate whereabouts. Coughing vocalizations were exhibited only by the terrestrial foraging species. While the physical properties of this call are probably quite similar between species, the temporal patterning is distinc- tive. Coughs in Genetta and Fossa are emitted SMITHSONIAN CONTRIBUTIONS TO ZOOLOGY singly or in volleys, and the interval distribution is dispersed over a wide range of time ( < l - > 6 0 sec). In Civettictis and probably Viverricula, individual coughs are almost always temporally clustered into bouts composed of 5-6 elements, and these volleys are usually separated from others by relatively long periods of time (>60 sec). In Civettictis, Viverricula, and Genetta, coughing is probably homologous. In all three species it clearly results from respiratory hesitation often associated with sniffing, and stance and locomotory transitions. In the adult African civets it was infre- quently exchanged between animals. In mother- infant associations, however, it clearly functions as a contact call: "the call is most frequently heard from kittens and is given by an individual who finds himself alone. The litter-mate at once responds by repeating the call and moving to the caller . . . the response of young kittens is completely auto- matic and invariable, and I never heard a call that failed to evoke a response" (R. F. Ewer, pers. comm.). The contingency analysis of this study revealed that the callers were approached or sniffed 12%-20% of the times they coughed. Although responsiveness to the call does not disappear, Ewer noted that it gradually becomes more and more conditional upon other factors in the situation (Ewer and Wemmer, 1974). The presumptive stimuli evoking this call are always present, but animals do not cough continu- ously. Certain conditions appear to lower the thresh- old for calling. Confinement in an unfamiliar area and separation from other animals or the keeper are common conditions that elicit this call. When an animal is sufficiently aroused, the coughing may be spontaneous and bear no clear relationship to ongoing activities. Prior to copulation, for example, male genets coughed repeatedly while following the female. Coughing in volleys probably arises from a ten- dency for the glottis to close immediately after a single cough and then snap open again. In Genetta coughs are usually linked with expiration one for one, but doublets and triplets also occur during single expirations. The expulsion of volleys of coughs during expiration in Civettictis could have arisen from the doublet and triplet condition seen in Genetta. Temporally dispersed coughing then probably represents the primitive condition. Coughing in Fossa resembles that of Genetta, NUMBER 239 79 except that it can be evoked by human imitation (Wemmer, 1971). It is probable that the call arose independently in this species. Curiously, like Civet- tictis, Fossa is strictly a ground-living species. Nonvocal Interaction in Genetta and Civettictis A large number of behavior patterns are shared by both Genetta and Civettictis; a small number are different. The differences are best elucidated by a consideration of timing and spacing. In Genetta back-arching, crouching, and tail-flaring are asso- ciated with mutual or unidirectional visual fixation; they occur in the companion's proximity, but out of his reach. Back-arching and crouching are char- acterized by the tonic retention of a lateral or frontal orientation to the other animal. During minimal contact, sniffing may be unidirectional or reciprocal; sniffing and head-darting are frequently distinct roles, and mutual head-darting is rare. Interaction is terminated by the departure of one or both animals. In Civettictis, back-arching and tail-flaring are absent, and crouching is infrequent. Visual fixation thus is not specified by tonic postural variants. Minimal contact consists of mutual sniffing, and sniffing and head-darting; unlike Genetta, recipro- cal head-darting (sparring) is also common. The recipient of darting frequently assumes (1) a lateral or tangential orientation to the companion, with the face deflected away, and/or (2) a reclining posi- tion on the ground. After the delivery of additional head-darts, the episode is terminated by the darting animal's departure. Lateral or tangential body orientation to the companion is a common component of contact interaction in canids and has been referred to as the T-position (Golani and Mendelssohn, 1971) or scruff-shoulder orientation (Fox, 1971). In the golden jackal, for example, the lateral animal is the contact recipient; the companion may sniff, paw at, or bite him, or may rest his head or forelegs on the lateral animal's back. The T-position terminates with circling or genital licking (Golani and Mendelssohn, 1971). In the wolf, dog, and coyote, a reclining position may be assumed if the lateral animal is pushed (Fox, 1971). Lateral recumbency (passive submission: Schenkel, 1967) is also a reac- tion to being sniffed in the genital region. Ac- cording to Schenkel (1967:324) the most obvious characteristic of the recumbent animal is "the readi- ness to actively enter into contact with the superior" animal; the reclining animal may whine, make licking movements, push with its nose, and wag the tail. In reclining civets, only the head and neck are moved in response to the proximity and head- darts of the standing animal; motion only appears to provoke further darting. Nevertheless, while types of contact differ, Civettictis is basically similar to the canids in that contact is received during lateral orientation and recumbency is a response to contact. While there are individual differences in the orientation of sniffs and darts, in both Genetta and Civettictis there is a general trend for the forebody to be the darting target, and the fore- and hindbody to be sniffing targets. Though other workers have not quantitively documented observations of ori- ented contact, it has been suggested that certain body markings serve as contact targets (Fox, 1969; Kleiman, 1967). The body targets of sniffing and darting are the same in these two species, but the hair patterning differs. There is thus an indication that white body markings may serve several different functions. Pre- sumptive general functions can be deduced by evaluating (1) whether the marking is always promi- nent or can be hidden, (2) whether the size, propor- tions or shape of the design is strongly influenced by body configuration, and (3) whether the marking is characteristically oriented to companions or oriented to by companions. The contrasting mark- ings on the bodies of Genetta and Civettictis do not exhibit these characters equally. In the African civet, the white spinal stripe is visible only during crest erection and the white perineal gland is often concealed by the tail; only the gland receives oriented contact. Neither marking's configuration is altered significantly when viewed from different positions. The design properties of the eye spots and tail of Genetta and the face and neck markings of Civet- tictis, however, alter considerably during positional or postural changes in relation to the viewer. Only in Civettictis, though, is there a correspondence between the location of contrast patterns and oriented contact. On the basis of these different properties, contrast markings may (1) enhance visibility of the body or its movements in animals that are nearby but out of contact range, (2) specify 80 SMITHSONIAN CONTRIBUTIONS TO ZOOLOGY postural configurations of the body or the part of the body carrying the mark, and (3) specify targets for companion-oriented contact. The following outline summarizes some of the possible changes associated with the transition from a genet-like grade of behavioral organization to that exhibited by the African civet. It is based on the assumption that Genetta is a relatively primitive viverrine and that Civettictis evolved from an ancestor resembling Genetta. Items are numbered for convenience, but the separation does not imply a lack of causal relationship between them. 1. Relative shortening of tail length and degen- eration of its highly patterned design; loss of tail- flaring capacity. 2. Loss of ability to maintain protracted arched- back stances; compensatory development of the spinal crest but retention of the tendency to maintain a lateral body orientation to other ani- mals during maximal crest erection. 3. Origin of lateral body orientation with neck deflection, and its incorporation with reclining body positions as intermediary components in the sniff-dart axis; concomitant elaboration of highly patterned head and neck markings from a basic genet-like design. 4. Enlargement of the perineal gland and elabo- ration of its coloration; derivation of the normal quadrupedal stance from a squatting or reversed quadrupedal upright; specialization of Flehmen. The genet and African civet are more similar behaviorally to one another than to other species of carnivores. While lateral orientation and reclin- ing are also components of contact interaction in canids, it is probably a simplification to view Civettictis as a more social species. There is no evidence suggesting it is anything but solitary. Per- haps the exigencies of a totally terrestial existence select for an interaction design that permits the establishing of social roles in minimal time. It is conceivable that the assumption of oriented stances in the African civet facilitates transition from sniffing and darting to maximal contact interaction. Addendum Shortly after this manuscript was revised for publication two important papers appeared dealing with the natural history and behavior of the small-spotted genet, Genetta genetta, in Europe. They are B. Gangloff and P. Ropartz, Le Repertoire Comportemental de la Genetta, Genetta genetta (Linn6), La Terre et la Vie (1972) 26(4):489-560; and M. Delibes, Sobre Alimentacion y Biologia de la Geneta (Genetta genetta L.) en Espana, Donana, Acta Vertebrata (1974) 1:143-199. Gangloff and Ropartz' study is a lengthy description of the social, nonsocial, and predatory behavior of captive animals. It is evident that the social behavior of this species is qualitatively similar or identical to that of Genetta tigrina; however, quantitative data on social interaction are not presented on which to base further comparisons. Delibes' analysis of feeding habits reveals the pre- ponderance of small vertebrates and insects in this species' diet. This is the only large contribution available on this aspect of genet ecology. Appendices ENCOUNTER SCHEDULES AND STATISTICAL PROCEDURES Appendix 1: Encounter number Date 1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 1 2 3 4 5 6 7 8 9 10 1 2 3 4 6Sep69 8 Sep69 10 Sep 69 12 Sep 69 14 Sep 69 16 Sep 69 18 Sep 69 20 Sep 69 22 Sep 69 24 Sep 69 Encounter Schedules for Genetta First animal introduced Duration in cage prior to encounter (hrs:min) MALE-MALE ENCOUNTERS: F AND O F O F O F O F O F O 3:30 3:50 3:50 4:10 3:45 3:25 4:00 4:15 4:10 4:20 Second animal introduced 1900 1940 1915 1910 1645 1630 2010 1935 1945 1930 FEMALE-FEMALE ENCOUNTERS: C AND I (familiar) 20 Jun 69 9 Jul 69 17 Jul 69 25 Jul 69 31 Jul 69 7 Aug 69 15 Aug 69 I I I C c I I 4:15 ? 3:50 3:45 2:45 3:40 2:05 2030 1900 2010 1930 1900 2040 2000 FEMALE-FEMALE ENCOUNTERS: C AND M O 18 Aug 69 20 Aug 69 22 Aug 69 24 Aug 69 26 Aug 69 28 Aug 69 30 Aug 69 1 Sep 69 3 Sep 69 5 Sep 69 25 Sep 69 27 Sep 69 28 Sep 69 1 Oct 69 c Mo C Mo C Mo C Mo C Mo 4:30 3:15 1:15 2:30 4:47 2:45 3:12 4:05 3:50 4:50 2130 1945 1630 1845 1855 1930 1930 2000 1910 2020 FEMALE-FEMALE ENCOUNTERS: I AND M O Mo I Mo I 3:20 4:25 3:00 3:00 1950 2050 1905 1900 tigrina Encounter duration (min:sec) 52:53 45:47 31:24 34:20 34:09 33:35 40:24 46:10 31:39 33:45 30:58 33:44 32:56 32:22 32:55 33:11 31:19 34:57 40:50 36:31 33:34 34:45 49:22 46:15 35:55 39:01 50:10 35:25 37:17 37:27 41:01 81 82 SMITHSONIAN CONTRIBUTIONS TO ZOOLOGY Appendix 1?Continued Encounter number 5 6 7 8 9 10 Date 3 Oct 69 5 Oct 69 7 Oct 69 9 Oct 69 11 Oct 69 IS Oct 69 First animal introduced Mo 1 Mo 1 Mo I Duration m cage prior to encounter (hrs:min) 3:35 2:40 3:00 3:20 3:30 3:30 Second animal introduced 1915 1930 1615 1940 1630 1930 Encounter duration (min:sec) 45:19 39:42 45:23 48:38 36:34 38:53 MALE-FEMALE ENCOUNTERS: K AND MO 1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 9 Jun 69 16 Jun 69 23 Jun 69 30 Jun 69 7 Jul 69 14 Jul69 21 Jul 69 28 Jul 69 4 Aug 69 11 Aug 69 Mo Mo K Mo K Mo Mo K Mo K 4:00 3:10 3:30 3:50 3:45 3:00 2:25 6:50 2:15 3:15 2100 2010 2030 1620 1915 1930 1715 1630 1915 1930 MALE-FEMALE ENCOUNTERS: K AND C (familiar) 14 Oct 69 16 Oct 69 18 Oct 69 20 Oct 69 22 Oct 69 24 Oct 69 26 Oct 69 28 Oct 69 30 Oct 69 1 Nov 69 K C K C K C K C K C 3:00 3:40 3:40 4:05 3:05 3:25 3:30 4:00 3:20 4:05 MALE-FEMALE ENCOUNTERS: F AND I 2 Nov 69 4 Nov 69 6 Nov 69 8 Nov 69 10 Nov 69 12 Nov 69 14 Nov 69 16 Nov 69 18 Nov 69 20 Nov 69 F I F I F I F I F I 3:45 2:50 6:20 3:00 3:15 3:22 3:50 3:25 5:05 4:30 MALE-FEMALE ENCOUNTERS: O AND C 21 Nov 69 23 Nov 69 25 Nov 69 27 Nov 69 29 Nov 69 1 Dec 69 O C o c o c 3:00 2:55 2:55 4:10 3:48 3:00 1635 1940 2030 2050 1940 2040 2000 2000 1940 2010 1925 1620 1950 1615 1645 1652 1940 1655 2030 2100 1850 1625 1840 1740 1658 1920 29:54 30:06 31:22 36:16 32:11 33:06 22:32 27:04 32:49 33:19 35:14 35:44 34:04 36:21 36:11 35:31 39:47 34:24 33:44 31:57 39:10 33:32 36:36 31:16 33:06 39:48 34:14 33:14 38:38 33:10 39:32 36:39 34:55 31:39 30:23 33:56 NUMBER 239 7 8 9 10 3 Dec 69 5 Dec 69 7 Dec 69 9 Dec 69 O C O C 3:55 4:00 3:00 4:20 2015 1930 1630 2020 28:11 32:18 35:24 31:19 83 MALE-FEMALE ENCOUNTER: F AND M O 11 Dec 69 F 3:00 MALE-FEMALE ENCOUNTERS: O AND I 1600 11:40 1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 10 30 Dec 69 1 Jan 70 3 Jan 70 5 Jan 70 7 Jan 70 9 Jan 70 11 Jan 70 13 Jan 70 15 Jan 70 17 Jan 70 oi o I o I o I o I 3:30 3:20 3:35 4:05 3:50 4:50 2:40 3:40 3:00 3:10 1930 1905 1935 2040 1950 2050 1940 1940 1900 1610 MALE-FEMALE ENCOUNTERS: F AND C 18 Jan 70 20 Jan 70 22 Jan 70 24 Jan 70 26 Jan 70 28 Jan 70 30 Jan 70 1 Feb 70 3 Feb 70 5 Feb 70 F C F C F C F C F C 3:30 2:20 3:05 3:20 3:15 4:10 3:00 4:25 4:15 4:05 MALE-FEMALE ENCOUNTERS: K AND I 6 Feb 70 8 Feb 70 10 Feb 70 12 Feb 70 16 Feb 70 18 Feb 70 20 Feb 70 22 Feb 70 26 Feb 70 2 Mar 70 K I K I K I K I K I Appendix 2: Encounter 3:00 3:30 4:15 3:20 4:00 3:30 4:00 3:45 3:30 3:00 1915 1845 1930 1650 1845 2010 1845 2020 2015 2000 (familiar) 1945 2000 2015 1640 2000 2000 2000 1845 1630 1900 Schedules for Civettictis 32:14 34:44 52:50 32:54 35:08 33:16 33:01 34:16 32:43 32:56 32:11 32:58 35:40 34:42 29:45 32:27 28:20 33:35 36:03 33:59 32:57 27:21 30:32 29:42 35:20 33:30 33:27 36:27 29:56 29:12 civetta Encounter number 1 2 3 4 (male-female encounters) Date 2 Nov 70 3 Nov 70 4 Nov 70 5 Nov 70 Encounter onset time 1500 1530 1445 1500 Encounter duration (min:sec) 44:00 34:39 53:54 59:24 84 SMITHSONIAN CONTRIBUTIONS TO ZOOLOGY Appendix 2?Continued Encounter number 5 6 7 8 9 Appendix 3: Date 9 Nov 70 10 Nov 70 12 Nov 70 13 Nov 70 16 Nov 70 Statistical Procedures Encounter onset time 1530 1515 1500 1500 1530 for Tail-flaring Encounter duration (min:sec) 36:30 33:15 38:30 48:24 40:25 Frequency in Genetta tigrina Wilcoxon matched-pairs, signed-ranks test; N = number of series encounters in which at least one animal jaw-gaped and both animals tail-flared (see Table 13); variable: tail-flaring fre- quency; significance level (two-tailed): 0.05 Jaw-gaper K Mo I C I F I I 7 5 20 22 2 1 3 2 c K Mo Mo O C F K Companion 13 19 27 23 11 8 1 1 Rank Statistics - 6 - 1 4 - 7 - 1 - 9 - 7 +2 + 1 -4 .0 -8 .0 -5 .5 -1 .5 -7.0 -5.5 +3.0 + 13 N T P = 8 = 43 = >0.05 Appendix 4: Statistical Procedures for Head-darting Frequency in Genetta tigrina (between Sniffers) Wilcoxon matched-pairs, signed-ranks test; N = number of single encounters in which (1) each animal sniffed its companion, and (2) at least one animal head-darted; sniffer = animal that sniffed more frequently than its companion; variable: head-dart frequency; significance level: 0.05 Sniffer Sniffed Rank Statistics K 0 8 0 0 7 3 0 0 2 C 33 15 20 13 3 9 0 0 8 I Mo K I Mo 3 37 9 4 32 6 4 1 5 9 5 5 4 0 1 8 2 4 - 3 - 2 9 - 9 - 4 -25 - 3 - 4 - 1 - 3 +24 + 10 + 15 +9 +3 +8 - 8 - 2 + 4 - 3 - 9 - 7 -53 - 8 - 3 -53 - 1 - 3 + 18 + 14 + 17 + 12.5 +5 + 10.5 -10.5 - 4 +6 N T P = 9 = O = <0.005 NUMBER 239 85 Sniffer Sniffed Rank Statistics c I Mo O F 1 5 0 0 0 0 0 0 0 0 0 0 0 0 0 0 18 0 2 0 1 1 0 3 0 3 0 0 7 0 0 0 0 2 0 1 0 5 10 6 6 2 9 5 0 6 0 0 0 0 0 1 0 1 0 5 2 O F F K F Mo C I C K F I C O 0 4 7 6 9 11 1 14 5 7 6 9 11 1 14 5 0 2 0 4 2 0 3 0 37 0 1 1 0 1 2 1 4 1 1 3 1 6 8 8 2 0 2 1 1 1 2 9 2 4 4 0 1 17 1 14 8 + 1 + 1 - 7 - 6 - 9 - 1 1 - 1 - 1 4 - 5 - 7 - 6 - 9 - 1 1 - 1 - 1 4 - 5 + 18 - 2 +2 - 4 - 1 + 1 - 3 +3 - 3 7 +3 - 1 - 1 + 7 - 1 - 2 - 1 - 4 + 1 - 1 - 2 - 1 - 1 +2 +2 +4 +2 +7 +4 - 1 +5 - 2 - 9 - 2 - 4 - 4 + 1 - 1 - 1 6 - 1 - 9 - 6 +2 +2 - 9 - 8 - 1 2 3 - 1 5 - 2 - 1 6 - 7 - 4 - 3 - 5 - 6 - 1 - 7 - 2 +8.0 -3.5 +3.5 -7.0 - 1 3 + 13 - 5 3 +53 +9.0 +9.0 - 3 3 -3.5 + 11.0 - 3 3 - 7 3 - 3 3 -10.0 + 3 3 -3.5 - 7 3 -2.0 -2.0 +6.0 +6.0 + 103 +6.0 + 14.0 + 103 -2.0 + 13.0 -6.0 + 15.0 -6.0 + 103 + 103 +2 - 2 - 6 - 2 - 5 - 4 N T P N T P N T P N T P N T P N T P = 18 = 54 = >0.05 (nj.) - 7 = O = 0.05 = 9 = 173 = >0.05 (n.$.) = 11 = 233 = >0.05 (nj.) = 15 = - 2 4 * = 0.025 = 6 = 2 = >0.05 (n.s.) ? The negative sign of the T value indicates statistical significance for the opposite hypothesis. 86 SMITHSONIAN CONTRIBUTIONS TO ZOOLOGY Appendix 5: Statistical Procedures for Head-darting Frequency in Civettictis civetta Wilcoxon matched-pairs, signed-ranks test; N = number of encounters; variable: head-dart frequency; significance level: 0.025 Sniffer Sniffed Rank Statistics 4 10 1 3 2 4 0 16 37 38 38 36 27 25 109 0 - 3 3 - 2 8 - 3 7 - 3 3 - 2 5 - 2 1 -109 + 16 -5.5 -4.0 -7.0 -5.5 -3.0 -2.0 -8.0 + 1.0 N T P = 8 = 1 - <0.01 >0.005 Appendix 6: Statistical Procedures for Head-darting Frequency in Genetta tigrina (between Sniffers and Nonsniffers) Wilcoxon matched-pairs, signed-ranks test; N = number of single encounters in which only one animal sniffed its companion, and at least one head-darted; variable: head-dart frequency; sig- nificance level: 0.05 Encounter number 1 8 10 4 8 1 5 6 7 8 1 2 3 6 10 3 4 5 6 8 10 Nonsniffer I I C C o c F I 1 1 1 2 1 2 9 2 4 4 7 6 9 11 5 2 0 2 1 1 1 Sniffer F Mo Mo O C O C O 0 0 0 0 0 0 0 0 5 0 0 0 0 0 0 6 2 9 5 0 6 d + 1 + 1 + 1 +2 + 1 +2 +9 +2 - 1 +4 +7 +6 + 9 + 11 + 5 - 4 - 2 - 7 - 4 + 1 - 5 R' + 2.0 +2.0 +2.0 +7.0 +2.0 +7.0 + 13.5 +7.0 -2.0 +9.0 + 12.0 + 11.0 + 13.5 + 15.0 + 10.0 - 5 - 2 - 6 - 3 + 1 - 4 R" + 3.5 +3.5 + 3.5 +8.5 +3.5 +8.5 + 19.5 +8.5 -3.5 + 12.0 + 17.5 + 16.0 + 19.5 +21.0 + 14.5 -12.0 -8.5 -17.5 -12.0 +3.5 -14.5 Statistics N = T -- P = = 21 = 68.0 = >0.05 Appendix 7: Statistical Procedures for Testing Distribution of Nasal Contacts to Four Body Regions in Genetta tigrina Friedman two-way analysis of variance; conditions = 4 general body targets (I ?? head and neck, II = trunk, HI ? legs, IV = tail); variable: nasal contact frequency; significance level: 0.05 NUMBER 239 87 Animals encountered Individual C I Mo Individual K Mo F O I Individual Mo F C K Individual I C K F Individual F I C Individual I O C Mo K C I Mo O F N 24 21 47 44 14 6 2 15 6 0 18 30 5 25 54 1 65 3 11 0 129 0 1 / Rank 2.0 2.0 2.0 ~6lJ 2.0 3.0 2.0 3.0 3.0 Ts^o 4.0 2.0 3.0 2.0 IsTo 2.0 3.0 2.0 4.0 1L0 1.0 2.0 3.0 To 2.0 3.0 2.0 4.0 ffo" N 57 40 79 52 12 18 3 24 2 0 29 59 7 10 74 0 125 5 1 0 151 1 0 Conditions II Rank 3.0 4.0 3.0 IThO 3.0 2.0 3.0 4.0 4.0 I61T 3.0 2.0 4.0 4.0 TsTi 3.0 2.0 3.0 2.0 loo 3.0 3.0 1.0 To 2.0 4.0 4.0 2.0 12J0 N 9 4 9 27 4 3 0 6 0 0 5 24 1 2 17 0 69 2 3 0 48 0 0 III Rank 1.0 1.0 1.0 3^0" 1.0 1.0 1.0 1.5 1.0 5T 1.5 2.0 1.0 1.0 5T 1.0 1.0 1.0 2.0 5iO~ 2.0 1.0 2.0 ITo" 2.0 2.0 2.0 2.0 8lf N 70 26 82 97 28 52 0 12 0 1 8 45 17 27 93 0 185 33 23 1 72 0 0 IV Rank 4.0 3.0 4.0 [To 4.0 4.0 4.0 1.5 2.0 T5J0 1.5 4.0 2.0 3.0 ioJ 4.0 4.0 4.0 2.0 liJO 4.0 4.0 4.0 12i0 4.0 1.0 2.0 2.0 ~9J0 Statistics Xr2 = 8.2 P = 0.017 Xr* = 6.6 P = <0.01 Xr1 = 11.7 P = <0.001 XrI = 6.3 P = 0.094 Xr? = 8.4 P = <0.017 X r ? = 1.5 P = 0.754 Appendix 8: Statistical Procedures for Testing the Distribution of Head-darting to Four Body Regions in Genetta tigrina Friedman two-way analysis of variance; conditions = 4 general body targets (I = head and neck, II = trunk, III = legs, IV = tail); variable: head-darting frequency; significance level: 0.05 Animals encountered Individual K C I Mo N 34 2 6 / Rank 4.0 4.0 4.0 Tzo N 1 0 0 Conditions II Rank 1.0 2.0 1.0 To N 2 0 1 III Rank 2.5 2.0 2.5 To N 2 0 1 IV Rank 2.5 2.0 2.5 To Statistics Xr* = 6.6 P = 0.075 88 SMITHSONIAN CONTRIBUTIONS TO ZOOLOGY Appendix 8?Continued Conditions Animals encountered Individual C K Mo F I Individual I Mo O C F Individual Mo I C K Individual O F C I Individual F I O C N 135 16 28 4 3 12 20 7 7 20 6 31 4 12 2 23 36 Rank 4.0 4.0 4.0 4.0 1T6 4.0 4.0 4.0 4.0 I 1:0.36; variable: nasal contact frequency, significance level: 0.05 Animal and companion K I C K Mo Mo C Mo C O O F (I) (*) (K) (Mo) (K) (I) (Mo) (Q (I) (I) (F) (O) N 21 30 44 47 54 5 15 25 15 3 65 129. I Rank 1.0 1.0 1.0 2.0 2.0 1.5 2.0 2.0 2.0 1.5 1.0 2.0 191F Conditions N 33 60 58 44 52 5 9 6 12 3 137 149 II Rank 2.0 2.0 2.0 1.0 1.0 1J 1.0 1.0 1.0 1.5 2.0 3.0 21.0 N 38 68 131 126 132 20 34 33 30 37 242 122 III Rank 3.0 3.0 3.0 3.0 3.0 3.0 3.0 3.0 3.0 3.0 3.0 3.0 34.0 Statistic < 1:0.5 Xr1 = 19.1 P = <0.001 NUMBER 239 K. I I O C C F I ( Q (Mo) (C) (C) (O) (F) W (F) 24 6 18 11 2 6 0 1 1.0 3.0 1.0 2.0 2.0 1.0 1.5 2.5 14.0 50 2 19 2 3 9 0 0 2.0 2.0 2.0 1.0 3.0 2.0 1.5 1.0 TilT 86 0 23 25 0 64 1 1 3.0 1.0 3.0 3.0 1.0 3.0 3.0 2.5 19.5 > 1:0.36 Xr1 = 2.3 P = <0.3 89 Appendix 10: Statistical Procedures for Testing the Distribution of Nasal Contacts to Three Body Regions in Civettictis civetta Friedman two-way analysis of variance; conditions = 3 general body targets (I = forebody, II = midbody, III = hindbody); first test: encounter series in which the sniff-dart ratio was < 1:0.5, second test: > 1:0.33; variable: nasal contact frequency; significance level: 0.05 Encounter and animal 2 2 3 4 4 5 6 6 7 8 9 1 1 3 5 7 $ 9 S $ 9 3 o $ & S s 9 9 9 9 N 1 0 0 0 0 6 2 2 2 0 1 18 19 0 11 3 Rank 1.0 1.0 1.0 1.0 1.5 2.0 2.0 3.0 1.0 1.0 3.0 1.5 2.0 1.5 3.0 2.0 Too N 10 3 7 2 0 5 1 0 11 1 0 18 15 0 1 1 Conditions II Rank 2.0 2.0 2.0 2.0 1.5 1.0 1.0 1.0 2.0 2.0 1.5 nnr 1.5 1.0 1.5 1.0 1.0 To N 34 17 11 6 3 12 11 1 32 15 0 62 29 1 8 17 III Rank 3.0 3.0 3.0 3.0 3.0 3.0 3.0 2.0 3.0 3.0 1.5 3"blT 3.0 3.0 3.0 2.0 3.0 TO" Statistics < 1:0.5 df = 2 Xr1 = 9.8 P = <0.01 > 1:0.33 df = 2 Xr1 = 6.4 P - <0.05 Literature Cited Albignac, R. 1970a. Notes ethologiques sur quelques carnivores mal- gaches: Le Fossa fossa (Schreber). La Terre et la Vie, 24:383-394. 1970b. Notes ethologiques sur quelques carnivores mal- gaches: Le Cryptoprocta ferox (Bennett). La Terre et la Vie, 24:395-402. Allen, J. A. 1924. Carnivora Collected by the American Museum Congo Expedition. Bulletin of the American Museum of Natural History, 47(3):73-281. Anderson, S., and J. K. Jones 1967. Recent Mammals of the World, a Synopsis of Fam- ilies, viii + 453 pages. New York: Ronald Press Company. Andrew, R. J. 1963. The Origin and Evolution of the Calls and Facial Expressions of the Primates. Behaviour, 20:1-109. Ansell. W. F. H. 1960. Mammals of Northern Rhodesia, xxxi + 155 pages, 24 illustrations. Lusaka: Government Printer. Baerends, G. P., and J. M. Baerends-Van Roon 1950. An Introduction to the Study of the Ethology of Cichlid Fishes. Behaviour Supplement, 1: 243 pages. Bartels, E. 1964. On Paradoxurus hermaphroditus javanicus (Hors- field 1824): The Common Palm Civet or Tody Cat in Western Java, Notes on its Food and Feeding Habits, its Ecological Importance for Wood and Rural Bio- topes. Beaufortia, 10:193-201. Bigalke, R. 1968. Evolution of Mammals on Southern Continents, III: The Contemporary Mammal Fauna of Africa. Quart- erly Review of Biology, 43:265-300. Booth, A. H. 1954. The Dahomey Gap and the Mammalian Fauna of the West Africa Forests. Review of Zoology and Botany of Africa, 50:305-314. Bossert, W. H., and E. O. Wilson 1963. The Analysis of Olfactory Communication Among Animals. Journal of Theoretical Biology, 5:443-469. Burger, M. 1959. Eine vergleichende Untersuchung iiber Putzbeweg- ungen bei Lagomorpha und Rodentia. Der Zoo- logische Garten, 23:434-506. Carlsson, A. 1910. Die genetischen Beziehungen der madagassischen Raubtiergattung Galidia. Zoologischen Jahrbuchern, 28:559-602. Carpenter, G. P. 1970. Some Observations on the Rusty Spotted Genet (Genetla rubiginosa zuluensis). Lammergeyer, 11:60- 63. Cobb, S. 1970. A South Turkana Expedition. Oxford, 23:48-52. Cody, M. L. 1969. Convergent Characteristics in Sympatric Species: A Possible Relation to Interspecific Competition and Aggression. Condor, 71:223-239. Coetzee, C. G. In press. Order Carnivora. Part 8 in J. Meester, editor, The Mammals of Africa: An Identification Manual. Washington, D.C.: Smithsonian Institution Press. Colbert, E. H. 1961. Evolution of the Vertebrates. 479 pages. New York: Science Editions, Inc. Copley, H. 1950. Small Mammals of Kenya, viii + 96 pages. London: Longmans, Green. Crawhall, J. C, and S. Segal 1965. Sulphocysteine in the Urine of the Blotched Kenya Genet. Nature, 208:1320-1322. Dane, B., C. Walcott, and W. H. Drury 1959. The Form and Duration of the Display Actions of the Goldeneye (Bucephala clangula). Behaviour, 14: 265-268. Darwin, C. 1872. The Expression of the Emotions in Man and Ani- mals, vi + 374 pages. London: Murray. Dekeyser, P. L. 1956. Le Pare National du Niokolo-Koba, III: Mammiferes. Memoires de I'lnstitut Francois d'Afrique Noire, 48: 35-77. Doran, G. A., and H. Baggett 1971. A Structural and Functional Classification of Mam- malian Tongues. Journal of Mammalogy, 52:427-429. Dorst, J., and P. Dandelot 1970. A Field Guide to the Larger Mammals of Africa. 287 pages. New York: Houghton Mifflin. Ducker, G. 1957. Farb- und Helligkeitssehen und Instinkte bei Viver- riden und Feliden. Zoologische Beitrdge, 3:25-99. 1965. Das Verhalten der Schleichkatzen (Viverridae). Handbuch der Zoologie, 10(20a):l-48. 1971. Gefangenschaftsbeobachtungen an Pardelrollern Nandinia binotata (Reinwardt). Zeitschrift fur Tier- psychologie, 28:77-89. Eibl-Eibesfeldt, I. 1951. Gefangenschaftsbeobachtungen an der persischen Wustenmaus {Meriones persicus Blanford): Ein Beit- rag zur vergleichenden Ethologie der Nager. Zeit- schrift fur Tierpsychologie, 8:400-423. 90 NUMBER 239 91 Eisenberg, J. F. 1963. A Comparative Study of Sandbathing Behavior in Heteromyid Rodents. Behaviour, 22:16-23. 1967. A Comparative Study in Rodent Ethology with Em- phasis on Evolution of Social Behavior, I. Proceed- ings of the United States National Museum, 122: 1-51. Ewer, R. F. 1963. The Behavior of the Meerkat, Suricata suricatta (Schreber). Zeitschrift fur Tierpsychologie, 20:570- 607. 1967. The Behaviour of the African Giant Rat (Cricetomys gambianus Waterhouse). Zeitschrift fur Tierpsy- chologie, 24:6-79. 1968. A Preliminary Survey of the Behaviour in Captivity of the Dasyurid Marsupial, Sminthopsis crass i can data (Gould). Zeitschrift fur Tierpsychologie, 25:319-365. 1973. The Carnivores. 494 pages, 16 plates. Ithaca: Cornell University Press. Ewer, R. F., and C. Wemmer 1974. The Behavior in Captivity of the African Civet Civettictis civetta (Schreber). Zeitschrift fur Tier- psychologie, 34:359-394. Flower, W. H. 1966. An Introduction to the Osteology of Mammalia. 382 pages. Amsterdam: A. Asher. [Reprint of 1855 Lon- don edition.] Fox, M. W. 1969. The Anatomy of Aggression and Its Ritualization in Canids. Behaviour, 35:242-258. 1971. Behaviour of Wolves, Dogs and Related Canids. 214 pages. London: Jonathan Cape. Gaillard, H. 1969. Sur un Trait de Biologie de la Civette (Civettictis civetta (Schreber)). Mammalia, 33:352. Goethe, F. 1952. Vergleichende Studien an unseren Marderartigen- vom "Drohschrei" und anderen Stimmlauten. Pirsch/ Deutscher Jaeger, 4:684-686. 1964. Das Verhalten der Musteliden. Handbuch der Zoo- logie, 10:1-80. Golani, I. 1969. The Golden Jackal. 124 pages. Tel Aviv: The Move- ment Notation Society. Golani, I., and H. Mendelssohn 1971. Sequences and Precopulatory Behavior of the Jackal (Canis aureus Linnaeus). Behaviour, 37:169-192. Gregory, W. K., and M. Hellman 1939. On the Evolution and Major Classification of the Civets (Viverridae) and Allied Fossil and Recent Carnivora: A Phylogenetic Study of the Skull and Dentition. Proceedings of the American Philosophical Society, 81:309-392. Hall-Craggs, E. C. B. 1965. An Analysis of the Jump of the Lesser Galago (Galago senegalensis). Journal of Zoology, 147:20-29. Hediger, H. 1944. Die Bedeutung von Miktion und Defakation bei Wildtieren. Schweizerisches Zeitschrift fiir Psychol- ogie und ihre Anwendung, 3:170-182. Herter, K. 1952. tiber das Verhalten von Iltissen. Zeitschrift fiir Tier- psychologie, 9:56-71. Hollister, N. 1918. East African Mammals in the United States National Museum, Part I: Insectivora, Chiroptera, and Carni- vora. United States National Museum Bulletin, 99: 194 pages. Horwich, R. H. 1972. The Ontogeny of Social Behavior in the Gray Squir- rel. Zeitschrift fiir Tierpsychologie, supplement 8: 103 pages. Howell, A. B. 1965. Speed in Animals, Their Specialization for Running and Leaping. 270 pages. New York: Hafner Publish- ing Company. Huf, K. 1965. Cber das Verhalten des Binturong (Arctictis bin- turong Raffles). Portugaliae Ada Biologica, 9(3/4): 249-304. Kingston, B. H. 1965. The Chemistry and Olfactory Properties of Musk, Civet and Castoreum. Pages 209-214 in part I of Proceedings of the Second International Congress of Endocrinology. Amsterdam. Kleiman, D. 1966. Scent Marking in the Canidae. In Jewell and Loizos. editors. Play, Exploration, and Territory in Mam- mals. Symposium of the Zoological Society of Lon- don, 18:167-177. 1967. Some Aspects of Social Behavior in the Canidae. American Zoologist, 7:365-372. Lawrence, B., and A. Novick 1963. Behavior as a Taxonomic Clue: Relationships of Lissonycteris (Chiroptera). Breviora, 184:16 pages. Le Boeuf, B. J., and R. S. Peterson 1969. Dialects in Elephant Seals. Science, 166:1654-1656. Leyhausen, P. 1956. Das Verhalten der Katzen. Handbuch der Zoologie, 8:1-34. 1960. Verhaltensstudien bei Katzen. Zeitschrift fiir Tier- psychologie, Supplement 2: 120 pages. Lind, H. 1959. The Activation of an Instinct Caused by a "Transi- tional Action." Behaviour, 14:123-135. Lorenz, K. Z. 1953. Comparative Studies on the Behaviour of Anatinae. Avicultural Magazine, 5:1-87. McKinney, F. 1965. The Comfort Movements of Anatidae. Behaviour, 25:120-122. 1966. An Analysis of the Displays of the European Eider Somateria mollissima (Linnaeus) and the Pacific Eider Somateria mollissima v. nigra. Behaviour, supplement 7: 124 pages. Madison, D. M., and R. Schoop 1970. Homing Behavior, Orientation, and Home Range of Salamanders Tagged with Tantalum-182. Science, 168:1484-1487. 92 SMITHSONIAN CONTRIBUTIONS TO ZOOLOGY Mallinson, J. J. C. 1969. Notes on Breeding the African Civet Viverra civetta at Jersey Zoo. International Zoo Yearbook, 9:92-93. Matschie, P. 1902. Cber die individuellen und geographischen Aban- derungen der Ginsterkatzen. Verhandlungen Inter- nationales Zoologen-Kongress, 1902:1128-1144. Mayr, E. 1963. Animal Species and Evolution. 797 pages. Cambridge, Massachusetts: Harvard University Press. 1968. Theory of Biological Classification. Nature, 220:545- 548. Mohr, E. 1952. "Ohrtaschen" und andere tatschenahnliche Bild- ungen am Saugetierohr. Mitteilungen aus dem Ham- burgischen Zoologischen Museum und Institut, 51: 63-68. Montagna, W., and R. Ellis 1959. The Skin of Primates, I: The Skin of the Potto (Perodicticus potto). American Journal of Physical Anthropology, 17:137-161. Morris, D. 1957. "Typical Intensity" and Its Relation to the Problem of Ritualization. Behaviour, 11:1-11. Muybridge, E. 1957. Animals in Motion. 74 pages, 182 plates. New York: Dover Publications. Mykytowycz, R. 1969. The Distribution of Dunghills and the Behaviour of Freeliving Wild Rabbits, Oryctolagus cuniculus (Lin- naeus) on Them. Forma et Functio, 1:333-349. Nuboer, J. F. W. 1959. Optically Aimed Ear Movements in the Rabbit. Experientia, 15:399-400. Pocock, R. I. 1915a. On Some of the External Characters of the Genus Linsang, with Notes upon the Genera Poiana and Eupleres. Annals and Magazine of Natural History, 16:341-351. 1915b. On the Feet and Glands and Other External Char- acters of the Paradoxurine Genera Paradoxurus, Arctictis, Arctogalidia and Nandinia. Proceedings of the Zoological Society of London, 1915:387-412. 1915c. On the Feet and Glands of Other External Characters of the Viverrinae, with the Description of a New Genus. Proceedings of the Zoological Society of Lon- don, 1915 (1):131-146. 1916a. On Some of the External Characters of Cryptoprocta. Annals and Magazine of Natural History, 1:413-425. 1916b. On the External Characters of the Mongooses (Mun- gotidae). Proceedings of the Zoological Society of London, 1916:349-374. Poduschka, W. 1969. Erganzungen zum Wissen iiber Erinaceus e. rou- manicus und kritische Oberlegungen zur bisherigen Literature iiber europaische Igel. Zeitschrift fur Tier psychologic, 26:761-804. Poglayen-Neuwall, I. 1962. Beitriige zu einem Ethogram des Wickelbaren (Potos flavus Schreber). Zeitschrift fur Saugetierkunde, 27: 1-44. 1965. Gefangenschaftsbeobachtungen an Makibiiren (Bas- saricyon Allen, 1876). Zeitschrift fur Saugetierkunde, 30:321-366. Radinsky, L. 1971. An Example of Parallelism in Carnivore Brain Evolution. Evolution, 25:518-522. Rahm, U. 1966. Les mammiferes de la foret equatoriale de l'Est du Congo. Annales du Muse'e Royal de I'Afrique Cen- trale, 149:39-121. Roberts, A. 1951. Mammals of South Africa. 700 pages. Johannesburg: Trustees of "The Mammals of South Africa" Book Fund. Romer, A. S. 1945. Vertebrate Paleontology. 687 pages. Chicago: Univer- sity of Chicago Press. Rowe-Rowe, D. T. 1970. The Development and Behaviour of a Rusty-spotted Genet, Genetta rubiginosa Pucheran. Lammergeyer, 13:29-44. Sanderson, I. T. 1940. The Mammals of the North Cameroons Forest Area. Transactions of the Zoological Society of London, 24:623-725. Schenkel, R. 1967. Submission: Its Features and Functions in the Wolf and Dog. American Zoologist, 7:319-330. Shortridge, G. C. 1934. The Mammals of South West Africa. 2 volumes, frontispiece, plates. London: Win. Heinemann Lim- ited. Siegel, S. 1956. Nonparametric Statistics for the Behavioral Sciences. 312 pages. New York: McGraw-Hill Book Company. Taylor, M. 1969. Note on the Breeding of Two Genera of Viverrids, Genetta spp. and Herpestes sanguineus, in Kenya. East African Wildlife Journal, 7:168-169. Taylor, M. E. 1970. Locomotion in Some East African Viverrids. Journal of Mammalogy, 51:42-51. Thenius, E., and E. Hofer 1960. Stammesgeschichte der Sdugetiere. 322 pages. Frank- furt: Springer Verlag. Thomas, O. 1901. On the More Notable Mammals Obtained by Sir Harry Johnston in the Uganda Protectorate. Pro- ceedings of the Zoological Society of London, 2:85-90. Verheyen, R. 1951. Contribution a I'e'tude ethologique des Mammiferes du Pare National de I'Upemba. 161 pages. Bruxelles: Institut des Pares Nationaux du Congo Beige. Vosseler, J. 1929. Beitrag zur Kenntnis der Fossa {Cryptoprocta ferox Benn.) und ihrer Fortpflanzung. Der Zoologishe Garten, 2:1-9. NUMBER 239 93 Walker, E. P. 1968. Mammals of the World. Volume 2, pages 647-1500. Baltimore: Johns Hopkins University Press. Wemmer, C. 1971. Birth, Development and Behavioural Notes on a Fanaloka (Fossa fossa: Viverridae) at the National Zoological Park. International Zoo Yearbook, 11:113- 115. Wenzel, E. 1972. System der Schleichkatzen (Viverridae). Saugetier- kundliche Mitteilungen, 20(1/2): 110-127. Whitfield, I. C. 1971. Mechanisms of Sound Localization. Nature, 233:95- 97. Winge, H. 1941. Rodentia, Carnivora, Primates. Volume II in The Interrelationships of the Mammalian Genera. Copen- hagen: C. A. Reitzel Forlag. Wurster, D. H., and K. Benirschke 1968. Comparative Cytogenetic Studies in the Order Carni- vora. Chromosoma, 24:336-382. Zannier, F. 1965. Verhaltensuntersuchungen an der Zwergmanguste Helogale undulata rufula im Zoologischen Garten Frankfurt am Main. Zeitschrift fur Tierpsychologie, 22:672-695. Zannier-Tanner, E. 1965. Vergleichende Verhaltensuntersuchung iiber das Hinlegen und Aufstehen bei Huftieren. Zeitschrift fur Tierpsychologie, 22:696-723. REQUIREMENTS FOR SMITHSONIAN SERIES PUBLICATION Manuscripts intended for series publication receive substantive review within their originating Smithsonian museums or offices and are submitted to the Smithsonian Institution Press with approval of the appropriate museum authority on Form SI-36. Requests for special treatment?use of color, foldouts, casebound covers, etc.?require, on the same form, the added approval of designated committees or museum directors. Review of manuscripts and art by the Press for requirements of series format and style, completeness and clarity of copy, and arrangement of all material, as outlined below, will govern, within the judgment of the Press, acceptance or rejection of the manuscripts and art. Copy must be typewritten, double-spaced, on one side of standard white bond paper, with i v i " margins, submitted as ribbon copy (not carbon or xerox), in loose sheets (not stapled or bound), and accompanied by original art. Minimum acceptable length is 30 pages. Front matter (preceding the text) should include: title page with only title and author and no other information, abstract page with author/title/series/etc., following the establish- ed format, table of contents with indents reflecting the heads and structure of the paper. First page of text should carry the title and author at the top of the page and an unnum- bered footnote at the bottom consisting of author's name and professional mailing address. Center heads of whatever level should be typed with initial caps of major words, with extra space above and below the head, but with no other preparation (such as all caps or underline). Run-in paragraph heads should use period/dashes or colons as necessary. Tabulations within text (lists of data, often in parallel columns) can be typed on the text page where they occur, but they should not contain rules or formal, numbered table heads. Formal tables (numbered, with table heads, boxheads, stubs, rules) should be sub- mitted as camera copy, but the author must contact the series section of the Press for edito- rial attention and preparation assistance before final typing of this matter. Taxonomic keys in natural history papers should use the alined-couplet form in the zoology and paleobiology series and the multi-level indent form in the botany series. If cross-referencing is required between key and text, do not include page references within the key, but number the keyed-out taxa with their corresponding heads in the text. Synonymy in the zoology and paleobiology series must use the short form (taxon, author, yearpage), with a full reference at the end of the paper under "Literature Cited." For the botany series, the long form (taxon, author, abbreviated journal or book title, volume, page, year, with no reference in the "Literature Cited") is optional. Footnotes, when few in number, whether annotative or bibliographic, should be typed at the bottom of the text page on which the reference occurs. Extensive notes must appear at the end of the text in a notes section. If bibliographic footnotes are required, use the short form (author/brief title/page) with the full reference in the bibliography. Text-reference system (author/year/page within the text, with the full reference in a "Literature Cited" at the end of the text) must be used in place of bibliographic footnotes in all scientific series and is strongly recommended in the history and technology series: "(Jones, 1910:122)" or ".. . . Jones (1910:122)." Bibliography, depending upon use, is termed "References," "Selected References," or "Literature Cited." Spell out book, journal, and article titles, using initial caps in all major words. For capitalization of titles in foreign languages, follow the national practice of each language. Underline (for italics) book and journal titles. Use the colon-parentheses system for volume/number/page citations: "10(2):5-9." For alinement and arrangement of elements, follow the format of the series for which the manuscript is intended. Legends for illustrations must not be attached to the art nor included within the text but must be submitted at the end of the manuscript?with as many legends typed, double- spaced, to a page as convenient. Illustrations must not be included within the manuscript but must be submitted sepa- rately as original art (not copies). All illustrations (photographs, line drawings, maps, etc.) can be intermixed throughout the printed text. They should be termed Figures and should be numbered consecutively. If several "figures" are treated as components of a single larger figure, they should be designated by lowercase italic letters (underlined in copy) on the illus- tration, in the legend, and in text references: "Figure 9JD." If illustrations are intended to be printed separately on coated stock following the text, they should be termed Plates and any components should be lettered as in figures: "Plate 9b_." Keys to any symbols within an illustration should appear on the art and not in the legend. A few points of style: (1) Do not use periods after such abbreviations as "mm, ft, yds, USNM, NNE, AM, BC." (2) Use hyphens in spelled-out fractions: "two-thirds." (3) Spell out numbers "one" through "nine" in expository text, but use numerals in all other cases if possible. (4) Use the metric system of measurement, where possible, instead of the English system. (5) Use the decimal system, where possible, in place of fractions. (6) Use day/month/year sequence for dates: "9 April 1976." (7) For months in tabular list- ings or data sections, use three-letter abbreviations with no periods: "Jan, Mar, Jun," etc. Arrange and paginate sequentially EVERY sheet of manuscript?including ALL front matter and ALL legends, etc., at the back of the text?in the following order: (1) title page, (2) abstract, (3) table of contents, (4) foreword and/or preface, (5) text, (6) appendixes, (7) notes, (8) glossary, (9) bibliography, (10) index, (11) legends. i