Sense of Vision

For most humans, vision is the primary sense, and we are surprised to learn that it is not necessarily that important to other species. It can also be difficult to imagine how animals with visual capabilities different from our own see the world around them. For dogs, the sense of vision is based on what might be needed for a hunter. The development of the canine visual system is incomplete at birth, as evidenced by the sealed eyelids of a neonate. The palpebral reflexes are present at birth, but most of the other protective responses do not appear until later (Table 2-1). Even after the eyes open, development continues. Although we typically think of the palpebral reflex as the only one for eye protection, dogs also have a vibrisso-palpebral reflex, so that they blink if the vibrissae are touched.¹²⁰

Like most predators, the dog has a visual field with a relatively large amount of binocular vision and a large blind spot behind the head (Fig. 2-1). The dog’s binocular field is 60 to 116 degrees versus 140 to 160 degrees for humans.^290,366 Each eye has an additional monocular field of 86 to 90 degrees, resulting in a total visual field of 240 to 290 degrees.^* In humans, the visual field is approximately 180 degrees.^366,457 The extreme variation in head shapes among the breeds allows for tremendous variation in the extent of the blind area, from 70 to 120 degrees. Contrast the eye placement of a Borzoi with that of a Pekingese (see Fig. 2-1). Head shape also affects eye placement somewhat. In general, the eyes are placed at the front of the head with a slight angle of divergence varying from 15 to 52 degrees, depending on the head shape.^†

Figure 2-1 The visual field of the dog.

Visual deprivation during the first 5 weeks of life does not affect myelinization of the visual cortex, but it does result in significant structural, biochemical, and electrophysiologic changes.^130,132 If a puppy has one eyelid fixed so that it never sees from that eye, later tests show that normal vision will not occur when the lid is unsealed.³⁹⁷ Suppose a puppy’s right eyelid is sealed past the normal time after birth. The pup’s normal left eye would have a visual field from approximately 120 degrees to the left to 30 degrees to the right of the midline, or the normal 150-degree range. The deprived right eye, however, would see only the 90 degrees farthest to the right and would not see in the forward binocular area.³⁹⁷ Thus, there would be a full range of vision, but it would be monocular.

The development of cells within the brain also is affected by early visual deprivation. If the right eye has been covered, the cells of the right geniculate dealing with monocular vision for that eye will be normal, but lamina A cells in the right geniculate will be 40% larger than normal, and those in the lamina A₁ layer will be smaller than normal.³⁹⁷ Cells in the left geniculate will be normal for the monocular area, smaller in the left lamina A layer, and 17% larger than normal in the left lamina A₁ layer.³⁹⁷ Compensatory hypertrophy in the auditory and somatosensory cortex can also occur.¹³⁰

Internal anatomic features indicate several differences between canine and human vision. The amount of fiber decussation in the optic nerves is an indication of the relative amount of binocular vision. In dogs, one fourth of the fibers decussate at the optic chiasm. In humans, one half do.^108,457 In dogs, the lens of the eye is large relative to the globe of the eye. In humans, the lens-to-globe ratio is 1:18, but in the dog it is 1:10.2.⁴¹¹ The retinal magnification factor, a function of axial length, is 0.19 mm per degree, for an axial length of 19 mm.¹⁹⁶

Dogs may not focus as well as humans. At least five factors suggest that conclusion. First, both the fovea centralis and macula lutea are areas that enhance visual acuity in the human eye. The canine retina lacks the macula lutea.^206,366,411 Second, dogs tend to be hypermetropic (farsighted), with images focusing behind the retina at 0.25 to 1.5 diopters (D).^290,411 If 20/40 vision means a person sees at 20 feet what a person with normal sight would see at 40 feet, it is estimated that the visual acuity for dogs is 20/50 to 20/100.²⁹⁰ Functionally, this farsightedness is insignificant for canine hunters, because dogs with a lens size of 0.75 to 1.5 D have no diminution of their ability to find and chase rabbits.⁴¹¹ Motion is probably more important to canine vision than detailed images. Of interest is a breed predisposition for myopia (nearsightedness). German Shepherds have a 53% incidence of myopia, and Rottweilers have a 54% incidence.²⁹⁰ Astigmatism is rare in dogs, occurring in only 4.2%.²⁹⁰

The third reason dogs may not focus as well as humans is related to their enhanced ability to see at night. The tapetum reflects light that enters, causing it to be bounced around and magnifying its effects. This gives the dog a light collection factor of 5.6 × 10⁵ compared with a high in other night-hunting species of 1 × 10⁶.¹⁹⁶ The minimum threshold for light is approximately four to five times lower for dogs than for humans.²⁹⁰ As a result of the reflected light, the dog can see better in dark environments but the trade-off is a fuzzier retinal image.

Limited accommodation is a fourth reason dogs do not focus well. Dogs have an accommodative range of 2 to 3 D and are unable to accurately focus on objects closer than 33 to 50 cm.²⁹⁰ A fifth reason that dogs have reduced focusing ability compared with humans is that behaviorally they have less need for it. Their activities do not require fine discrimination of static forms, but rather are more dependent on the motions, sounds, and smells of potential prey.¹⁴⁸ The image seen by the generic dog is like a general form without a lot of details. This image has been equated to what a person sees near sunset,³⁵⁵ having less form and pattern than a view in full sun.¹⁴⁸

An estimate of canine visual acuity is 1.3 to 2 times that reported for cats and 0.2 to 0.4 times that of primates tested under similar conditions.³²¹ Also supporting the concept of reduced visual acuity are the canine histologic features of unadjusted central ganglion cell densities and lower numbers of optic nerve fibers (Table 2-2).³²¹

TABLE 2-2 Comparison of visual acuity in dogs, cats, and primates

In repeating patterns, like the alternating light and dark bars of a grate, the visual-evoked cortical potentials (VECP) and pattern-evoked retinal potentials (PERR) help evaluate how thick the lines must be to be perceived as distinct when viewed from different distances and at different intensities of illumination. These are measures of fundamental spatial frequency.³²¹ The mean VECP for dogs is 12.5 cycles per degree. The mean PERR is 11.61 cycles per degree.³²¹

Most dogs do not watch television, probably for other reasons than the quality of the shows available. Although certain sounds tend to draw a dog’s attention, the poor form recognition of their vision may make the picture insignificant. It has also been suggested that the scanning rate of the gun in the picture tube, which updates the screen 60 times per second, is too slow to keep a full picture for a dog, which sees only a flicker.²⁹⁰ In support of this supposition, dogs can discriminate individual flickers of light at rates much higher than humans can.⁷² Whereas humans can usually detect 50 to 60 Hz, dogs can detect 80+ Hz.²⁹⁰

The histologic makeup of the retina indicates some of the differences between canine and human vision. In addition to the presence of a tapetum, as already mentioned, dogs have a different distribution of rods and cones. Even though the central 25-degree area of each retina has the densest population of cones,²²⁹ in the dog’s eye the area is still predominantly rods.²⁹⁰ In humans, the area has significantly more cones.²⁹⁰ The rod photopigment, rhodopsin, in dogs has a peak light sensitivity to wavelengths of 506 to 510 nm, as is typical for night-adapted species.²⁹⁰ The cone sensitivity for humans is around 496 nm, and this pigment regenerates much faster in humans after being exposed to bright light.²⁹⁰

It is generally believed that dogs have some ability to see color, but whether dogs actually use color in their daily activities has been the subject of much debate. It has been argued that the mere presence of cones in the retina signify color vision. This may or may not be true,⁵ but other research provides more information.

Within the eye of a dog, cones have two classes of photopigments. One peaks at 429 to 435 nm (violet) and the other at 555 nm (yellow-green).^206,290,314 The middle range for humans, 475 to 485 nm (blue-green), may appear to dogs as gray.²⁹⁰ This characteristic compares favorably to the dichromatic characteristic in animals, giving them a yellow, bluish, black, white, and gray discrimination. Humans have a total range of violet-to-red (350 to 700 nm) spectrum of visible light.

Among studies to determine if dogs can distinguish one color from another, 10 of 17 concluded that dogs had this ability.^205,290 Some of these studies strongly indicated that form and brightness were more significant than color.^5,457 In a few studies, care was taken to negate the intensity of reflected light as an inadvertent cue. Results, even from these studies, also suggest some color vision is present. Two behavioral reasons lend support to the conclusion that color vision is used at least minimally by dogs.³¹⁴ Dogs can be trained in color-discrimination tests. This quality is in contrast to many other dichromatic species. In addition, once trained in color discrimination, a research dog will tend to ignore obvious brightness cues in favor of color choices. This tendency, too, distinguishes the dog from other dichromatic species.

The detection of motion is extremely useful to a canine hunter. Experimentally, dogs could recognize moving objects at 810 to 900 m and stationary ones at 585 m.^290,457 Determining clinically how much vision an individual dog has is difficult. Tests currently available include the menace response, an ophthalmologic exam, an electroretinogram (ERG), a pattern electroretinogram (PERG), a visual-evoked response (VER), preferred looking, and the optokinetic nystagmus (OKN) technique.^305,323

Preferred-looking tests present the dog with a preference test between figures with irregular contours. The OKN reflex is based on the natural response in which the eyes and head move to follow the direction of motion of a visual stimulus. A special unit can present specific visual target grids of various sizes, and responses can be monitored to determine when a target’s size is too small to be perceived. The normal range for the OKN response in dogs is a 5.04’ to 5.75’ arc,³⁰⁵ and normal dogs perform better on visual acuity tests at a distance of 30 cm.¹¹⁰ In normal dogs, the PERG visual resolution limit is 6.9 ± 2.6 minutes of arc/phase in the center 15 degrees of the retina and 11.8 ± 2.3 minutes of arc/phase peripheral to that.³²³

The optic nerves carry retinal information to the brain. The mean number of nerve fibers in each optic nerve of a dog is 153,712.^13,14,321 At the intersection of the right and left nerves, there is decussation of approximately one fourth of the fibers.^108,457 Within the brain, the occipital lobes are responsible for vision. Bilateral ablation causes object blindness but not the loss of the ability to discriminate light intensities.³⁶⁶

The tapetum, something so visible in adult dogs, does not develop until after birth.³⁶² It is common to notice its absence during ophthalmologic exams of 6- to 8-week-old puppies presented for vaccinations, but to see it present by 12 weeks. Eye color, too, changes from the blue in most young puppies to its adult color during the first several weeks of life. The final color has not been found to correlate with reactivity to environmental stimuli.²⁴³

Sense of Audition

The auditory system, like the visual system, is incomplete at birth and continues to undergo maturation in the postnatal puppy (Fig. 2-2). The ear canals begin to open around 12 to 14 days of age. Although this opening seems to occur in a few days, the canal actually continues to widen for approximately 5 weeks.⁴⁶ Associated responses, such as auditory startle and auditory-evoked cortical potentials, also appear about the time the ears first open (see Table 2-1). Auditory-evoked responses develop rapidly and expand toward the high frequencies during the first 3 weeks and are relatively mature by the fourth or fifth week.^129,363

Figure 2-2 Ear canals and eyes are physically sealed in newborn puppies.

The ability to hear a particular tone is actually dependent on two factors: the frequency of the tone and its intensity (loudness). In theory, then, very low or high frequencies could be heard if the volume were loud enough. Frequency plotted against intensity would produce a U-shaped curve. Practically, however, human hearing usually ranges from a low of 13 to 20 Hz (cycles per second) to a high of 16,000 to 20,000 Hz-—about seven octaves.^* Audiologists say that humans hear best between 1000 Hz and 4000 Hz, and the average person can distinguish 2000 pitches. The average conversational voice has a pitch of approximately 120 Hz for the male human and 250 Hz for the female, with the typical range varying from 300 to 4000 Hz.²² The intensity for auditory testing is approximately 20 decibels (db), about the loudness of a whisper.

The auditory range of dogs is somewhat greater than that of humans. The region of maximum peripheral sensitivity, or best hearing, is 200 to 15,000 Hz.³⁵⁷ At the low frequencies of 20 to 250 Hz, dogs and humans hear with the same acuity.^104,134 Above 250 Hz, the dog has a lower intensity threshold to response,^134,252 and its best sensitivity with lowest intensity is at approximately 8000 Hz.^177,178 The upper limit of the canine audible frequency range varies considerably by researcher, from 26,000 Hz to between 44,000 Hz and 100,000 Hz.^* At an intensity of 60 db, dogs can hear as high as 41,000 to 47,000 Hz.^22,177

Behaviorally, frequencies above 35,000 Hz are probably insignificant because of the intensity necessary.¹⁰⁸ Based on possible upper limits of dog auditory abilities, ultrasonic devices may not actually be ultrasonic. For example, the ultrasonic flea collars use pulses of 30,000 to 50,000 Hz, and 73% of dogs show a response to these devices, including distress responses.³⁷⁵ As would be expected from this test and others using pulses ranging from 14,000 to 45,000 Hz at up to 75 db, no device tested so far can be detected by nor repel all dogs.⁴¹ Tests have shown that dogs are more likely to show increased motor activity and a “come” response to short-duration notes of high frequency than to longer notes of lower frequency. The latter work better for “sit” and “stay” responses.²⁷⁵

The general cortical areas for audition are similar in most species and are best studied in cats.³³ Frequencies of 100 to 400 Hz concentrate on the anterior ectosylvian gyrus, and those of 8000 to 16,000 Hz go to the intermediate and posterior ectosylvian gyrus, over the sylvian gyrus and middle ectosylvian sulcus.⁴³⁹ Each ear is represented by a similar but not identical pattern on the right and left cerebral cortices, but the pattern is more strongly present on the contralateral side.^439,440

Partial hearing loss is difficult to determine, and even complete deafness can go unnoticed for a long time. Preyer’s reflex, the pricking of the ears in response to a sound, is the basis of a behavioral test for hearing, but this reflex is very inconsistent in dogs at frequencies below 8000 Hz.³⁰⁵ Behavioral responses to pure-tone stimuli also provide a good estimation of the dog’s hearing ability.⁴⁰³ The most common test for hearing loss is the brainstem auditory-evoked reflex (BAER). The BAER does not develop until after 2 weeks of age and then matures gradually until puppies are 4 to 7 weeks oId.^46,220 Deafness, whether congenital or drug induced, does not produce brainstem-evoked potentials.^{220,226,227,299} Incomplete hearing losses result in intermediate responses.⁴²¹

Mobility of the pinna allows a dog to place the source of a sound to within approximately 4 degrees of its location.¹⁵⁸ Even though there are differences among individual dogs in interaural distance variations and the size of the tympanic membranes, neither type of difference affects hearing ability.^175,177

Studies also have shown that dogs can categorize different sounds. Based on their reactions, they seem to classify them as natural dog sounds and as nondog sounds.¹⁷⁸

Sense of Gustation

The sense of taste in dogs is apparently very similar to that in humans, although palatability is very different. Dogs are capable of responding to substances that taste like acid, bitter, salt, and sweet to humans when tested by nerve responses.^108,456 It has been argued that carnivores do not need to taste sweet, and that the sweet response is minimal in carnivores because they do not eat fruit.¹⁰⁸

Whereas various papillae of the tongue develop at different ages in the fetal puppy, taste buds are first seen on the forty-seventh fetal day.¹¹⁵ Neonates have the ability to show taste preferences; the neonate’s nerve responses to six different sugars are identical to those of an adult.^115,116 The gustatory system continues to mature somewhat after birth, and the olfactory contribution to taste is learned with various experiences.^26,115 Studies of dogs trained to validate specific flavors show their selections are made primarily by the sense of smell instead of taste.¹⁹³

Clinical tests to evaluate the presence or absence of taste are rare. It is important to distinguish between taste and taste preferences.³¹ Palatability tests really evaluate preference and olfaction more than the degree of gustatory function. One test, behavioral gustometry, has been proposed as a clinical test that might have applicability in veterinary medicine.³⁰⁵ The test takes advantage of a dog’s innate lick response when presented with a novel flavor, and this same response occurs to intravenous solutions. In rats, it is even possible to produce taste aversions with intravenous solutions.³⁰⁵ Appropriate licking responses to different concentrations of various sweet and bitter substances have been shown to occur in dogs. Concern about the potential for adversive osmolarity and pH effects has meant salt and sour solutions have not been tried.³⁰⁵ In dogs, the normal range of behavioral gustometry response to sucrose is between –15 and –7 molar concentration in Ringer’s solution.³⁰⁵

Sense of Olfaction

Dogs have long been recognized as having an outstanding sense of smell. Their ability to use this sense has been tapped for tracking, drug detection, locating accelerants in suspected arson, and even finding people buried under avalanches, water, or building rubble. Understanding the extremes of canine smell has been a difficult task for scientists because limited human olfactory capabilities make experimental design difficult. For example, one study attempted to determine whether dogs could be trained to be living smoke detectors. Behavior was shaped so that a dog would push a lever if smoke was detected at the test site and not press it if smoke was not smelled. The dogs were then taught to check a specific location in each of three rooms. A problem occurred in that when the dogs smelled smoke in either the first or second room, they responded with a positive response in all rooms thereafter.²²⁵ It was ultimately decided that those results would be considered acceptable, because in a real-life situation, if one room was on fire it would not matter whether the others were.

To appreciate the olfactory sense, it is useful to make anatomic comparisons between dogs and humans. The total surface area of olfactory epithelium is 2.0 to 11.5 cm² in humans and 75 to 150 cm² in dogs.^{53,134,148,422,423} Humans have 5 to 20 × 10⁶ cells in the olfactory bulb, and dogs have 2.8 × 10⁸ cells.^53,134,148 Differences in the structure of the olfactory bulbs in dogs and cats as compared with humans include more surface area (the cat has 13.9 cm²),⁴²³ a cribriform plate that wraps around the bulbs instead of being flat, and olfactory nerve fibers entering from the sides, above, and below instead of just the front.³⁶⁸ These types of differences suggest an early specialization toward a sophisticated sense of smell.³⁶⁸ From the olfactory bulbs, impulses travel through the olfactory tracts, olfactory tubercle, pyriform area, and hippocampus to the amygdala.¹⁸⁷

The importance of early stimulation by various odors is not well studied in newborn puppies, although it has been shown that they are capable of smelling specific odors within the first 15 minutes after birth.⁴⁶³ Studies in rats have shown that surgical closure of one of the external nares for the first month after birth results in a 25% reduction in size of the ipsilateral olfactory bulb.¹⁴²

The ability of dogs to detect dilutions of various odors has been researched. In interpreting the results of any such study, however, it is important to realize that a dog can respond to the primary odor via cranial nerve I or through the chemoreceptors of the ethmoid and palatine branches of cranial nerve IV.³⁰⁴ Odors that have been shown to stimulate only the olfactory nerve include cloves, lavender, anise, asafetida, benzol, and xylol.⁶ The odor of lavender has been successfully used as aromatherapy to reduce travel-induced excitement.⁴⁶¹ Lavender and chamomile in animal shelters result in apparently calmer behaviors.¹⁵⁹ The odors that trigger responses of both the olfactory and trigeminal nerves include camphor, eucalyptus, pyridin, butyric acid, phenol, ether, and chloroform.⁶ Amylacetate works only on the olfactory nerve at low concentrations, but it will trigger the trigeminal as well at higher levels.²³¹

Whereas humans can detect odor concentrations at 10^−4.5 molar (M) to 10^−5.0 M, and most animals can detect concentrations of 10⁻⁶ M to 10⁻⁹ M, dogs can detect some concentrations at 10⁻¹⁷ M.^162,231,353 Dogs are at least 2 to 4 log₁₀ units more sensitive to odors than humans.^231,301 Odor-detection studies have covered a wide variety of chemicals, both in solutions and as vapors (Table 2-3). As might be expected, the level of detectability is directly related to molecular size, at least for straight-chain fatty acids.¹⁵

TABLE 2-3 Molar concentrations of chemicals detectable by canine

Data from references 15, 134, 148, 231, 300, 301, 305, 353, 422.

Humans have harnessed the olfactory capabilities of dogs to extend our own abilities. We typically think of the Bloodhound and German Shepherd breeds as the scent-tracking and drug-detection dogs, but other breeds are used by federal and state agencies, as well as by private firms. A military working-dog study found small breeds could also be successfully trained for odor detection of drugs and explosives and could be useful where limited space was a problem.⁸¹ Recent studies are focusing on the canine olfactory capability to detect various diseases in humans, including bladder, lung, melanoma, and breast cancers.^276,360 Also, with the advent of gene research, the relationship between success in detector trained dogs and their genetic makeup is under study.³²⁴

The ability of dogs to identify and track people is certainly the best known use of their sense of smell. Dogs have been used to identify criminals in scent lineups with approximately 75% accuracy.³⁸⁹ Trained dogs can easily distinguish among the odors of different individuals, unless the people are identical twins.^215,414 Even identical twins can be differentiated if the scents of both are available for comparison.²¹⁴ Interestingly, clothing from human infants was indistinguishable to the dogs in one study if the infants were on the same breast-fed milk diet.⁴¹⁴ Dogs that are cross-trained to find either live people or cadavers do not do as well at finding live people compared with dogs trained to find the living only when both live and cadaver scents are present.²⁵³

Studies using a chromatogram differentiated over 200 peaks in four fractions for human body odor. Dogs apparently use scents in the second and third fractions.⁴¹⁴ Although individual aptitude and day-to-day performance can vary, the body location for odor source usually does not make a difference in matching success for dogs, even though such odors may be very different to a human nose.^{48,49,215,396}

Actual tracking success by experienced dogs can be as high as 93.3% under general conditions and 100% when the dogs are given the scent with the same article of clothing actually used to lay the track.⁴³² Tracking behavior occurs in three phases.⁴²⁹ The initial searching phase is when the dog is seeking the particular scent. Once located, the track is investigated to determine direction—the deciding phase. Here the dog moves about half as fast as before and sniffs for a much longer period. Sniffing greatly increases the contact of odor molecules with olfactory mucosa. Up to 10,000 times greater concentration occurs at the mucosa than in the air.³⁰⁴ Nasal flow rate is an important factor in determining the response to odors.⁴³³ During the tracking phase, direction of travel is apparently determined by comparing odor strengths from consecutive footprints and following those that have increasing intensity.⁴²⁰ Male dogs and younger dogs tend to be more accurate.⁴⁶²

Under field conditions, dogs may use more than just the human odor cue. Smells from crushed vegetation and the compression of earth may provide additional information.¹³⁴ Optimal tracking time occurs when the ground temperature is a little higher than the air temperature and there is a moderate amount of moisture, making early evening and early morning favorite hunting times for many keen-scenting carnivores.^108,187 Although little correlation was shown between climate changes and performance scores in studies, there is experimental evidence that certain outdoor conditions can result in the loss of odor on a glass object in 3 weeks.²²⁴ This loss of odor did not occur when the glass object was kept indoors. It has also been found that when dogs overheat, they may not work as well.¹⁵² The need to pant diverts air from the olfactory epithelium.

In addition to tracking, dogs have been trained to detect many types of odors. Dogs have the ability to sniff out truffles because of the odor of the δ-16 sex hormone steroid present in them, as well as in other plants like celery and parsnips.²⁰ They are also capable of differentiating between the feces from different species, such as various bears, wolves, coyotes, and foxes.⁴¹⁰ Their success in this search increases after castration.²⁰ Dogs are used daily in airports, in schools, and on border check stations throughout the United States to sniff out drugs and explosives. Efforts to hide or mask odors from a well-trained dog are rarely successful, even with multiple layers of insulation, different light intensities, or strong, unpleasant smells.^82,153,215

Dogs have also been successfully trained to identify estrous cows by scents in milk samples, vaginal mucus, and other body fluids. The dogs were able to detect pre-estrous cows within 1 to 2 days of estrus in 64% of the milk-sample trials.¹⁷³ They were 83% correct during estrus but responded to only 8% of cows in the luteal phase.¹⁷³ Accuracy in detecting estrus from swabs of the vulva and vestibule was 97%, from deep vaginal fluid 86%, and from voided urine 86%.²²³

Attempts to alter the olfactory capabilities of dogs have centered on various types of drugs. Steroids appear to elevate the threshold values, at least to eugenol and benzaldehyde, as measured by an electroencephalogram (EEG) and by innate behavior responses.¹⁰⁹ Amphetamines, alone or with caffeine or bromine, have been shown to result in a sharpened sense of smell in trained police dogs, an increase in intensity during searching activity, and improved tracking and selection skills.²³² Prolonged use, however, has the opposite effect.²³² A 3.5% zinc sulfate solution infused into the nasal cavity will result in a peripheral anosmia lasting at least 6 weeks.^193,194,309 (The use of an inflatable cuff on a tracheostomy tube prevents the animal from inhaling air over olfactory mucosa and can be used as another method to produce reversible anosmia.)¹⁹⁴

All mammals have a vomeronasal organ as part of a second chemoreceptive olfactory system.⁴²² Anatomically, each palatine duct connects the incisive papilla in the mouth to the ipsilateral vomeronasal organ at the base of the nasal septum. Cilia are present on the vomeronasal receptor cells in dogs, although not in other animals.¹⁶³ The vomeronasal nerve consists of bipolar neurons that leave the organ to penetrate the cribriform plate and terminate in the glomerular layer of the accessory olfactory bulb.^162,163 Because the palatine ducts are narrow, they must be actively opened to allow substances to enter and be moved by the cilia up the 50- to 60-mm duct to the vomeronasal organ.²

Flehman behavior, licking, and tongue flicking will open the ducts to allow nonvolatile, high-molecular-weight substances and pheromones access to the vomeronasal organ.^163,304 Many of the individual neurons are very selective for either male or female urine, and the firing rates increase dramatically in the vomeronasal nerve.¹⁸⁸ Unlike classic olfactory neurons, those of the vomeronasal organ do not adapt to prolonged stimulus exposure.¹⁸⁸ The result mediates endocrinologic and behavioral responses to sex-related odors. Flehman behavior, although obvious in horses and cattle, is inconspicuous in dogs because the philtrum and pendulous lips limit the mobility of the upper lip. Instead, dogs tend to hold their mouths open with their heads slightly extended. Dogs can also be seen using a slow licking motion to introduce substances like horse or cow urine into the mouth to taste or smell. In dogs, there is also a connection between the palatine duct and the ipsilateral nasal cavity, so appropriate odors can come that way too.²

Dogs use olfaction to gather part of their information about individuals they encounter. The canine-specific greeting of sniffing the nose of the least familiar dogs and the flank or perineal area of those they know are examples of this. Similarly, dogs sniff familiar adult persons with more interest in the thighs and perineum and less interest in the thorax and arms.¹¹⁷ The interest in body odors of unfamiliar children is directed toward the anogenital area, just the opposite of the interest in adult humans.²⁹¹ Duration of sniffing familiar and unfamiliar people is the same, however.¹¹⁷

Olfactory problems are common complaints in veterinary practice, so it is helpful to be able to assess function. About 85% of owners of hunting dogs will complain about an olfactory problem in their dog at some time.^187,305 Several conditions can produce temporary or partial anosmia in any dog, including canine distemper, canine parainfluenza, some seizures, hyperadrenocorticism, diabetes mellitus, hypothyroidism, nasal tumors, head trauma, modified live canine distemper vaccinations, and car exhaust.^* There is unpublished data suggesting there is some reduction in a dog’s sense of smell around 10 years of age, but there are also many individual exceptions.³⁰⁶ Clinical tests to evaluate a dog’s ability to smell include electroolfactography (also known as EEG olfactometry, electroolfactogram, or behavioral olfactometry). Behavioral responses of avoidance, approach, sniffing, and licking are used to determine when and if the dog detects gradually increasing concentrations of a particular odor.^304,305 Although normal threshold values have been determined for eugenol, benzaldehyde, amylacetate, cyclohexanone, and cocaine HCl, only eugenol does not affect the trigeminal nerve as well as the olfactory and would therefore be useful in such a test.^304,309