Journal of Archaeological Science (1998) 25, 331-347 Bit Wear, Horseback Riding and the Botai Site in Kazakstan Dorcas Brown and David Anthony Anthropology Department, Hartwick College, Oneonta, NY 13820, U.S.A. (.Received 18 February 1997, revised manuscript accepted 22 May 1997) Bit wear is the damage that occurs on the occlusal surfaces of the second premolar teeth when a horse chews the bit. Bit wear can provide evidence of horseback riding and/or horse draught from teeth found in archaeological sites. This article revises our earlier definition of bit wear and describes wear on an expanded study collection of modern bitted and feral horses (72 individuals). An experiment is described in which previously unbitted horses were ridden with organic bits to simulate the wear made by pre-metallic bits. Because the evaluation of bit wear depends partially on the age of the horse, a crown-height/age curve for horse lower second premolars is presented. Finally, horse premolars from the archaeological site of Botai in Kazakstan are evaluated using the new definition of bit wear and are found to provide evidence for horseback riding in northern Kazakstan between 3400 and 2700 вс. © 1998 Academic Press Limited Keywords: TRANSPORT, MICRO-WEAR, HORSEBACK RIDING, DENTAL CROWN HEIGHTS, ARCHAEOZOOLOGY, KAZAKSTAN. Introduction bit unobtrusively. Their bad habits, however well concealed from their riders and owners, are clearly The study of bit wear on horse teeth is important revealed by a microscopic examination of their teeth. principally because it can provide a clear In 1989 and 1991 Anthony & Brown presented a answer to a crucial question: were the horses definition of bit wear and described its occurrence on whose teeth are recovered from archaeological sites equid P2s recovered from two archaeological sites: used as transport animals? Bit wear is the damage that Malyan in Iran, dated to 2000-1900 вс (Anthony & occurs on the occlusal (chewing) surfaces of the second Brown, 1989); and Dereivka in Ukraine, dated to premolar teeth (P2s), particularly the lower second 4200-3700 вс (Anthony & Brown, 1991; Anthony, premolars (P2s), when a horse chews the bit. Use of a Telegin & Brown, 1991; Anthony, 1994). The bit wear bit indicates control from behind the horse; therefore, a at Malyan is the earliest unambiguous evidence for bitted horse must have been either ridden or driven. the use of the bit in the Near East, although copper Bokonyi (1972) was the first to suggest bit wear as stains on donkey teeth from Akkadian Tell Brak a likely cause for the anomalous bevelling that he (Clutton-Brock & Davies, 1993) might suggest earlier recognized on the mesial corner of some Iron Age bitting. The bit wear at Dereivka is the earliest evidence horse P2s. Clutton-Brock (1974) observed similar wear for the use of the horse as a transport animal anywhere on the P2s of a horse mandible excavated from the in the world. Egyptian fortress of Buhen, dated c. 1675 вс. Azzaroli Our definition of bit wear was based on a study of (1980: 287) noted the same type of wear on the teeth of natural (occlusal and dietary) wear on the lower Iron Age Venetic and Etruscan horses. However, none premolar teeth of 20 modern feral horses, and of wear of these authorities could be certain that a bit had from metal bits on the teeth of 10 modern domestic caused the wear that they noted. Moreover, like most horses. Recently we have gathered information from writers on horsemanship then and now, they thought 42 additional modern domestic horses, bringing the that bit wear could occur only if a horse was bridled domestic sample size up to 52 individuals. In addition, with a badly fitted bit; otherwise the bit should not we have conducted riding experiments on previously come into contact with the teeth. Bit wear was unbitted horses with bits made of rope, leather, and therefore seen as a symptom of poor management bone in an effort to define the wear made by organic that should occur rarely among bitted horses. As we bits. The data from this larger and more diverse demonstrate below, bit wear is, on the contrary, collection of horse teeth has persuaded us to revise the quite common: over 90% of the frequently bitted definition of bit wear. In this paper we describe modern horses that we have examined exhibited bit how bit wear has been redefined; we then apply wear. Even well-bridled horses apparently chew the the revised definition to horse teeth recovered from the Correspondence to: D. Anthony. E-mail: [email protected]. archaeological site of Botai in Kazakstan. 331 0305^403/98/040331 + 17 $25.00/0/as970242 © 1998 Academic Press Limited 332 D. Brown and D. Anthony Figure 1. A properly placed bit. The Modern Comparative Sample with polyvinyl siloxane, a high-resolution moulding compound. Casts were made from the moulds with We have examined 97 permanent and deciduous Tapox. an epoxy resin. Moulds were easily made on mandibular P2s from 52 modern domestic horses of 10 dry teeth from autopsy laboratories and museums. For breeds. We were able to obtain detailed riding and the live horses at the Cobleskill campus of the State bitting histories (riding frequency, date of last riding University of New York, the making of moulds episode, type of bit) for 17 of these horses. We assume required a protocol that overcame the problems raised that most of the other 35 domestic horses were bitted by head motion, tongue action, and saliva. We worked at least occasionally, because most of them were with lightly sedated horses that were still able to stand. identified in autopsy or institutional reports as having Upper premolars were not studied partly because it been used in ways that require a bit (racing, harness proved impossible to make casts of upper teeth on racing, draught, hunting, dressage, jumping, etc.). live animals unless they were fully sedated and lying Our collection of modern feral horse teeth consists of down, which would have been dangerous to their 37 permanent and deciduous P2s from 20 modern feral health. horses. Two populations are represented: four indi viduals were Assateague Island ponies from the Atlantic barrier islands of the Virginia and Maryland The Definition of Bit Wear coast, and 16 were mustangs from the Nevada Plains. These horses were never ridden or bitted. In addition, A bit normally rests on a horse’s tongue and gums in we examined the P2s of four live domesticated horses the diastema between the molars and the incisors, a that had never been bitted; these animals were used location called the “bars” of the mouth (Figure 1). in a riding experiment with organic bits, described When the reins are pulled the bit presses into this below. When first examined, their teeth did not differ sensitive soft tissue on the mandible, causing the significantly from those of the feral horses. horse to turn its head or tuck its chin to alleviate Most of the P2s examined in this project were casts, pressure. Horses can use their tongues to lift the bit off or replicas. Moulds of the original P2s were made their gums and push it back into the grip of their Bit Wear Damage 333 step-fracture”. Second, fracture features often cluster in the centre of the enamel ridges and propagate longitudinally, creating microscopic linear trenches up to 1 mm deep that run down the centre of the occlusal enamel in a worn area. We have called these features “centre-origin spalls”. Centre-origin spalls can weaken the enamel sufficiently so that a large chip may break off laterally. Together, these types of microscopic wear (abraded step fractures and, centre-origin spalls) are referred to as “a” wear in this paper. The “a” wear caused by a horse chewing the bit generally will not be confused with taphonomic, or post-mortem fractures and abrasions on the tooth surface caused either by nature or by clumsy curators. Taphonomic features are distinctive in three ways: they are not focused exclusively on the paraconid cusp of the P2, the step-fractures are not abraded and there Figure 2. Plan of occlusal surface of right side P2. are no centre-origin spalls. “A” wear does occur sporadically on feral horse teeth, which is potentially a second premolars (P2s), where it can no longer cause more serious problem. In the wild it is probably caused discomfort. The position of the bit between the pre by irregular occlusal contact or by biting on stones. molar teeth'was documented by Clayton & Lee (1984; The “a” wear that we observed on domesticated horse Clayton, 1985) in fluoroscopic X-ray photographs that teeth, however, was much more extensive than that showed the action of a bit in living horses’ mouths. observed on the teeth of feral horses. They established that both bar bits and jointed or Quantitative evidence supporting this conclusion “broken” bits can be gripped in the P2 teeth. came from scanning electron micrographs of the It is difficult for a horse to push a bit back into its occlusal surface of the paraconid cusps of six feral P2s teeth because the corners of the lips are positioned and 10 domestic P2s. These micrographs were enlarged forward of the premolars. The bit must be gripped very and placed on a digitizing pad. The occlusal enamel firmly to prevent the cheek corners from pulling it was outlined and its area measured. “A” wear features forward onto the bars of the mouth, and the bit can be were outlined and their total area was measured in held only between the tips of the teeth. Very powerful the same manner. Among the feral horses, “a” wear pressures are loaded onto a small occlusal space on the covered an average of 6-7% of the occlusal enamel on P2s. Bit wear occurs on the occlusal surfaces of the P2s the paraconid cusp, while for the domestic horses, “a” because the bit slips across these surfaces as the horse wear covered an average of 66-7% of the occlusal grasps and releases the bit. In most cases, bit wear is enamel on the paraconid cusp—it was 10 times more confined to the first or paraconid cusp of the P2. extensive (Table 1). For feral horses, the maximum Occasionally it extends as far back as the second extent of “a” wear observed on the paraconid cusp was (metaconid) or even third (metastylid) cusps (Figure 2). 30%, and this was observed on only one of the six teeth Bit wear is indicated by both microscopic and examined. In contrast, eight of the 10 domestic teeth macroscopic traits. exhibited “a” wear over more than 58% of the occlusal enamel of the first cusp. Finally, most of the “a” wear in feral horses occurred on the buccal or cheek side of Microscopic bit wear—“a” wear the paraconid cusp. This trait was observed on many of The occlusal surface of a horse premolar consists of the other feral P2s in our collection. The cheek side enamel ridges that rise above a softer core of dentine. of the lower tooth row naturally bears most of the The enamel ridges, which bear the brunt of all wear, damage associated with occlusion, and this buccal-side form a series of cusps. The microscopic aspect of bit contact probably accounts for much of the “a” wear wear consists of abraded fractures on the occlusal on mandibular feral horse teeth. Only one feral horse enamel of the paraconid cusp. Under magnification P2 displayed “a” wear on the lingual side. Among (12-15 x is sufficient) microscopic wear has two bitted horses, “a” wear occurred on both the lingual characteristics (Figure 3(a) & (b)). First, bit wear and buccal sides. causes step-like fractures that reflect the underlying It is the location and intensity of “a” wear that is linear alignment of the crystal rods composing the diagnostic of bit wear. If “a” wear is observed on 50% enamel. The tops of the rods, normally worn to an even or more of the occlusal enamel of the paraconid cusp, plane by occlusion with-the opposing tooth, are broken including the lingual side of the cusp, and similar unevenly by the bit. The sharp edges of the broken rods abrasion does not occur, or occurs only sporadically, are then abraded by the action of the bit and by dietary on other parts of the tooth, bit wear is indicated. This wear, producing a pattern we have called an “abraded pattern of wear is called “significant ‘a’ wear”. 334 D. Brown and D. Anthony Figure 3. (a) Bit worn P2 of 10-year-old race horse, (b) Unworn P2 of mature feral horse. Macroscopic bit wear—bevelling bit slipping back and forth over the mesial edge, or The macroscopic aspect of bit wear is a bevelled wear prow, of the P2. The worn area can take the shape of a facet on the mesial (anterior) corner of the occlusal real facet, with a well-defined, angular bevel, or it can surface of the P2. The facet appears to be caused by the have a gradual slope, without a sharp change in angle. Bit Wear Damage 335 Table 1. Digitized analysis of “a" wear on P2 teeth Table 2. Bevel measurement statistics for feral and domestic horses Specimen number “A”’ wear (%) Sample size of horses over 3 years old Average bevel Range Median bevel SD Feral horses 4R 0 7R 2 Feral horses 17L 3 25 teeth 18L 0 13 horses 28L 5 2 populations 0-78 0-2 0-5 0-66 28R 30 Average “a” wear on feral horse teeth = 6-67% Domestic horses Median “a” wear is between 2% and 3%. (includes known frequently bitted horses) 73 teeth Domestic horses 47 horses 12R 80 13R 86 10 breeds 3-11 0-10 2-5 1-93 14R 69 45R 85 Ail measurements are in mm. 45L 58 50R 95 57R 8 are then used to measure the distance between the 61R 4 63R 100 bottom of the ruler and the point on the tooth at which 66L 82 the occlusal surface meets the mesial edge. This Average “a” wear on domestic horse teeth = 66-70% measurement is rounded to the nearest 0-5jnm. Median “a” wear is between 80% and 82%. In our study collection, bevel measurements for mature domestic horses (mean 3-11 mm) were consist ently and significantly larger than those for mature The depth of the wear facet is measured by placing a feral horses (mean 0-78 mm), and bevel measurements ruler across the top of the 2nd and 3rd lingual cusps so for mature horses that were known to have been bitted that the ruler forms a right angle with the lingual side almost daily (mean 3-6 mm) were larger than those for of the tooth (Figure 4). The ruler serves as a plane, other domestic horses (Table 2 and Figures 5 & 6). The oriented in one dimension by the crests of the 2nd and mean bevel measurement for mature domesticated 3rd lingual cusps and in the other by a perpendicular horses (З-11 mm) exceeded the maximum measurement angle with the lingual surface of the tooth. Calipers observed for mature feral horses (2 mm). (Immature Figure 4. Method of measuring bevel. 336 D. Brown and D. Anthony suggests that either significant ‘a” wear or a significant bevel measurement on a horse older than 3 years is diagnostic of bit wear. Bit Wear on Frequently Ridden Horses We were able to collect detailed histories about bitting practices and frequency of usage for 17 horses. Of these, 13 animals were bitted and ridden or driven regularly, two to five times per week, up to the time of death or the day the tooth mould was made. We examined 24 P2s from these 13 horses. All 24 teeth were expected to exhibit bit wear—a significant bevel, significant “a” wear, or both. Table 3 shows the incidence of bit wear on these 13 animals. Only one horse had no wear on either the right or left side P2. This 13-year-old quarterhorse mare was bitted and ridden several times a week at a riding school. She simply did not chew the bit. Three horses exhibited bit wear on only one side. Altogether, five teeth (21%) derived from four horses exhibited no bit wear. Bit wear did occur on 79% of all the P2s, and on at least one side in 92% (12 of 13) of the frequently bitted horses. Archaeologists should note that an Figure 5. Box plots of bevel measurements—note that modern domestic horses are broken into two groups, those of unknown analysis based on individual teeth will undercount the bitting history and those known to have been bitted frequently. Only number of horses that have been bitted, because wear horses over 3 years are included. occurs on only one side in some horses. Nevertheless, bit wear is common and expected on horses that are ridden or driven almost every day. horses, 3 years or younger in age, are discussed separately below.) We have established the threshold for a bit-induced bevel at 3 mm, which is more than The Longevity of Bit Wear three standard deviations greater than the mean bevel measurement for mature feral horses. We refer to this Horses have high-crowned teeth that erupt continu as a “significant bevel”. ously as they are worn away by occlusal and dietary wear. Depending on diet and environmental factors, it may take 20-40 years for horses’ cheek teeth to wear down to the roots. Feral horses in the wild rarely live The new definition longer than 20-25 years; domestic horses can survive Previously we defined bit wear as the combination of more than 40 years. Newly erupted permanent P2s in a significant bevel measurement and significant “a” our sample (P2s of 3-year-old horses) had a maximum wear. However, after observing more domestic horse crown height (a measurement of length) of 63 mm. teeth, and gathering more detailed information about Assuming a P2 life-span of 20 years (which means a many of them, we have found this definition to be horse life-span of 23 years), the average rate of occlusal excessively conservative. Analysis of our new collection wear should be 3T5 mm per year. Other published Table 3. Bit wear on 13 horses known to have been bitted frequently No. of teeth % of 24 No. of horses % of 13 No wear 5 21 1 8 Significant “a” wear 12 50 7 54 Significant bevel 17 71 11 85 Significant bevel but not significant “a” wear 7 29 5 38 Significant “a” and bevel 10 42 6 46 (old definition) Significant “a” or bevel 19 79 12* 92 (new definition) ‘Only nine horses (69%) had some kind of bit wear on both P2 teeth. Bit Wear Damage 337 Figure 6. Bevel measurements of domestic, feral and frequently bitted horse teeth plotted against age. □, Domestic horse teeth (excluding known frequently bitted teeth) (D), 65 teeth, 34 horses. A, feral horse teeth (F), 33 teeth, 20 horses. O, known frequently bitted teeth (KJ, 24 teeth, 13 horses. 2D, F or K, multiple teeth on this point. Area inside arrows excludes feral horse teeth. estimates of mean annual cheektooth wear rates for Bevelled wear facets are larger features, so they last grazing equids are slightly higher, between 3-3 and longer than “a” wear. We have recorded bevel 4-2 mm per year (Hulbert, 1982). Estimates of maxi measurements of up to 10 mm on bitted horses. A mum crown height also vary (Klein & Cruz-Uribe, bevel facet 10 mm deep might be preserved for more 1984: 54). If occlusal and dietary wear remove 3-15 mm than 3 years after the horse was last bitted. In addition, of occlusal surface per year, most “a” wear features by our definition, “a” wear must be observed on more should disappear within 4 months. If the rate of wear is than 50% of the occlusal enamel of the first cusp if it is higher, “a” wear will disappear more rapidly. to be categorized as “significant”. If portions of the “A” wear does not survive very long. Occlusal and paraconid cusp are destroyed by taphonomic processes dietary wear probably remove 3-4 mm of occlusal the tooth might fall short of the minimum surface surface per year on horse P2s, although the rate of required to confirm significant “a” wear. A bevel wear varies with age and diet. Most “a” wear abrasions measurement, on the other hand, can be taken as long are less than 1 mm deep, so should disappear within as the critical part of the paraconid cusp is not chipped 1-4 months if the horse is not bitted during that off. A significant bevel can survive the effects of period. occlusal wear longer, and can be Observed even 338 D. Brown and D. Anthony when the tooth is partially damaged by taphonomic processes. Bevel Measurements and Crown Heights Are bevel measurements alone a reliable indicator of bit wear? The answer to this question depends on the age of the horse. For horses with permanent teeth that have come fully into wear, our studies suggest that the answer is yes. To support this statement, we should first look at our statistics on bevel measurements, and then consider methods for eliminating young horses, 3 years of age or younger, from bit wear studies. Among the 13 frequently bitted horses, significant bevels were observed without significant “a” wear on seven of 24 teeth (29%) and in five of 13 horses (39%). These horses were bitted two to five times per week and clearly chewed their metal bits, yet did not show Figure 7. Photograph showing deciduous teeth on top of developing significant “a” wear. Bit wear apparently is not pro permanent P2 teeth as they grew in the jaw of a 21-month-old horse. duced by a continuous process of bit-chewing that White arrows point to the mesial ends..of right and left deciduous occurs each time the horse is bitted; rather it can be teeth. This view shows the lingual sides. The natural bevel at the caused by episodic bit-chewing events. The “a” wear mesial end of the developing permanent teeth can be clearly seen. Photo by Kim Ilowit. from these events is worn away relatively quickly, but the bevelled wear facets survive longer. For these reasons, significant bevel measurements will be seen rounded, conforming to the underside of the deciduous without significant “a” wear even among horses that tooth that caps it (Figure 7). After the deciduous P2 is are bitted frequently. shed the occlusal surface of the new permanent tooth is Bevel measurements alone clearly separate mature gradually worn flat by contact with the opposing tooth. bitted horses from mature feral horses. Among the 13 The low, rounded prow is one of the last parts of the frequently bitted horses, the average bevel measure new P2 to come into wear. Before it is completely ment was 3-6 mm, the median was 4-0 mm, and the flattened, the mesial corner often exhibits a bevel of range was 1-7 mm (Table 2 and Figures 5 & 6). Among more than 2 mm. We have observed this temporary feral (e.g. unbitted) horses older than 3 years there phenomenon on two live domesticated horses, never were no bevels greater than 2 mm. ridden, whose P2 bevels were 3-0 mm at age 2-5 years, but declined to 1-5 mm or less 4 and 5 months later. A “significant bevel” was present at age 2-5 years, but Problems with bevel measurements on young horses disappeared naturally within months. Clearly, the teeth One feral horse in our study collection had bevels on of horses 3 years old and younger should be eliminated both P2s of 3 mm. This animal was less than 3 years from analyses of bevel measurements. old. Unusually high bevel measurements occasionally occur among never-bitted horses aged 3 years or younger. The causes of the anomalous bevels lie in the Identifying the teeth of З-year-olds: crown heights and process of P2 eruption and early wear. Permanent P2s age erupt at approximately 2-5 years in horses. Both In order to eliminate З-year-olds from bit wear studies, deciduous teeth and newly erupted permanent teeth one must be able to identify their teeth in archaeo can exhibit irregular occlusal surfaces. logical contexts, where teeth are often found isolated Deciduous teeth are shaped differently from from the jaw. Specialists disagree about the best permanent teeth and frequently are worn irregularly method for determining the ages of horses based only because they are not shed simultaneously. An upper on isolated horse teeth (Spinage, 1972; Hulbert, 1982; might fall out while a deciduous lower tooth is still Klein, 1982; Levine, 1982; Burke, 1993; O’Brien, 1995). in use. The wear in this case becomes uneven until We chose to analyse cheektooth lengths, or crown the new upper tooth descends to occlude evenly heights (which decrease as the horse gets older). Levine with the lower deciduous tooth. If the deciduous has previously published mathematical curves which tooth is loose, malocclusion might cause further she used to determine horses’ ages from crown heights irregularities. in archaeological assemblages (Levine, 1982, 1990). Newly erupted permanent P2s also are irregular and These curves were based largely on indirect measure can exhibit anomalously large bevel measurements. ments (from X-rays) taken from a relatively small This occurs because, prior to eruption, the unworn sample of modern horses of estimated age. We occlusal surface of the permanent tooth is somewhat re-analysed the correlation between crown height and Bit Wear Damage 339 Figure 8. P2 crown heights plotted against age. O, Known age (55); A, feral (24); DK, deciduous known age (11); DF, deciduous feral (9); RF, roots feral (7) (i.e. permanent teeth with roots incompletely formed). age based on our collection (Figure 8). For the The range of crown heights at each age is very broad. crown height measurement we used 62 permanent and In addition, we found that there can be as much as deciduous P2s from 40 domestic horses of known age, 5 mm difference between the crown heights on the right and 33 permanent and deciduous P2s from 18 feral and left P2s from one horse; i.e. crown heights for two horses. Feral horse ages were estimated on the basis of teeth that have identical histories—age, diet, genetic incisor wear and eruption sequences in articulated traits, and riding histories—can differ by as much as mandibles. All of the teeth were removed from the 5 mm. This greatly complicates attempts to distinguish mandibles and measured directly. Measurements of the teeth of З-year-olds from those of 4, 5, and crown height were taken on the buccal side, from the 6-year-olds on the basis of small differences in crown apex of the junction between the roots, to the midpoint heights. of the occlusal surface. Earlier in the project we found that deciduous Figure 8 demonstrates that it is somewhat problem teeth and very old teeth (which have similar crown atic to determine the age of a horse solely on the basis heights) can be differentiated on the basis of occlusal of a crown height measurement on an individual P2. lengthfwidth ratios (Figure 9). We used the same 340 EL- Brown and D. Anthony horse younger than 4 years, and should be excluded from bit wear studies. This criterion should be sufficient to determine if a loose P2 recovered from archaeological sediments was from an animal sufficiently mature to evaluate for bit wear. The method also will be generally useful for archaeozoologists attempting to analyse the age struc tures of slaughtered horse populations in prehistoric sites. Metal Bit Wear Summary By examining four traits on the P2, we can determine if a horse was bitted, and therefore if it was used for riding and/or draught. These traits are: (1) significant “a” wear, and/or (2) a significant bevel measurement, Figure 9. Occlusal length/width ratios of permanent P2s with crown heights of less than 3 cm. Results of /-test of deciduous vs. to identify bit wear; and (3) a crown height of less than permanent: means of the samples are significantly different. 5 cm, with (4) an occlusal length/width ratio less than 2T2, to identify and eliminate horses of 3 years old and younger. The absence of bit wear on any individual horse does not mean that the horse was not bitted. Twenty-one per cent of the P2s from frequently bitted horses exhibited no bit wear. The absence of bit wear in a very large sample of teeth might suggest that bits were not used on that population. However, the pres ence of a significant bevel or significant “a” wear on the P2 of a mature horse identifies it as a bitted horse. Organic Bit Wear: The Riding Experiment Prior to 1992 our descriptions of bit wear were based entirely on observations of the wear made by modern metal bits. Prehistoric bits made of organic materials—bone, rope, or leather—might have caused different wear signatures. Some archaeozoologists have speculated in conversation that significant bit wear might be produced only by a metal bit. The experiments described below have established that significant bit wear can be caused by organic bits. In 1992 we conducted an experiment to define the wear caused by organic bits. We acquired four horses Figure 10. Occlusal length/width ratios of permanent P2s of horses ages 5/6, 4, and 2/3 years. Results of /-tests comparing means of that had never been bitted or trained. The horses were samples at the 0-05 level: 2/3 vs. 5/6 years, significantly different; 2/3 trained without a bit so that their teeth would remain vs. 4 years, significantly different: 4 vs. 5/6 years, not significantly “clean” until the experiment started. We used a train different. ing technique called “free lunging” in which the rider establishes dominance prior to mounting the animal by controlling its space in a series of one-on-one method to separate З-year-olds from 4—6-year-olds encounters inside a closed pen. After dominance is (Figure 10). The results were less clear, but Student’s established, riding is relatively easy. Similar training /-tests indicate that the mean length-width ratio for techniques, which mimic the behaviours used in nature З-year-olds is significantly different from that of 4, by dominant horses in horse bands, might well have 5, and 6-year-olds. The P2s of З-year-olds can be been attempted by the earliest riders. Once trained, the identified by correlating crown height with a specific experimental horses were bitted and ridden in a ring by length-width ratio. The precise threshold measure two students. When the horses were not being ridden ments might change if we examine larger samples, but they were kept together in an outdoor paddock and for the moment, we propose the following rule: if the were pastured or fed hay rather than prepared feeds to length-width ratio is 2T2 or more, and the crown simulate “primitive” dietary wear on their teeth. The height is 5 cm or more, the tooth is probably from a experiment lasted 8-10 months, depending on the