"Original kinds have been stable" is a tenet of "scientific" creationism, and scientific evidence can be given to support this tenet. At least, this is the claim of creation "scientists." Evidence usually cited includes probability, thermodynamics, the "impossibility" of beneficial mutations, and the fossil record, all of which are intended to show that evolution from "kind" to "kind" could not have occurred. The intent of this article is to show that the concept of a "basic created kind" is without meaning, especially when applied to fossil animals, and to demonstrate that the fossil record shows all perissodactyls are interrelated, therefore must all be of the same "kind" based on the only logical criterion for assigning fossils to "kinds."
Basic Created Kinds
Creationists conceive of a "basic created kind" as an organism which when created possessed considerable genetic potential for variation. This is commonly cited as "creative forethought" to allow these "kinds" to adapt, within limits, to changing environments (Morris, 1974; Hiebert, 1979). These "basic created kinds" have varied within limits thus accounting for the diversity of modern life forms. Common examples are a basic dog "kind" that gave rise to all varieties of dogs, from jackals to coyotes, a basic finch "kind" to account for Darwin's finches, and a basic horse "kind" that varied to give rise to all modern horses and many, or perhaps all, fossil horses. So variation, or microevolution, is allowed, but creationists emphatically deny that one "kind" could give rise to another "kind" (macroevolution).
The estimates of "basic created kinds" vary enormously. Jones (1973, p. 104) equates "kinds" roughly with the family and concludes (p. 107) that, "The number of animals under Noah's care probably did not exceed 2,000. . . ." These 1000 "kinds" (actually Jones p. 105 argues for 700 kinds) include mostly reptiles, birds and mammals. At the other extreme is Hiebert's (1979, p. 16) conclusion that ". . . species correspond roughly to original created kinds in Genesis chapter one." He does amend this statement by saying biological species ". . . do not always correspond to original kinds."
It is difficult to see how creationists could take either author's concept of a "kind" very seriously, but at least one (Moore, 1983) sees some value in Jones' ideas. Using Jones' concept, it would seem that goats, sheep, musk ox, bison, wildebeests and gazelles all were derived from an ancestral bovid "kind." And of course the okapi and giraffe also must have been derived from a single "kind."
Hiebert simply overloads the ark, even if aquatic kinds are omitted. In addition, Hiebert's formalization of what a "kind" is must surely be too restrictive for most creationists. For example, he insists (p. 114) that new species cannot arise because the chromosome number of each is "rigidly fixed." On p. 113 he argues that more complex animals should show an increasing number of chromosomes if evolution is true, and since there is no such correlation, each organism possessing a different number of chromosomes represents a separately created "kind." The problem is that the basic horse kind of most creationists now has no meaning since the chromosomes vary from 32 in Hartman's zebra, to 46 in Grevy's zebra, to 56 in the onager, to 66 in Przewalski's horse (Gould, 1983, p. 362). Perhaps Hiebert thinks each is a "basic created kind" (he seems to contradict this on p. 60), but it is doubtful that other creationists would agree.
Between Jones' 700 "kinds" and Hiebert's unspecified but undoubtedly large number of "kinds" are somewhat more moderate estimates. Whitcomb and Morris (1961) and LaHaye and Morris (1976) argue for 35,000 and 50,000 animals on the ark respectively. The former estimate seems to be taken from Mayr's list of 17,600 species of mammals, birds, reptiles and amphibians, although Whitcomb and Morris (p. 69) say: ". . . but undoubtedly the number of original 'kinds' was less than this."
Aside from their vague estimates of the number of "kinds," creationists have also left their definition of a "basic created kind" rather vague, a point noted by Judge Overton in the Arkansas decision. This vagueness is probably intentional since any definitive statement invariably leads to problems, as in the case of Hiebert and his criterion of chromosome numbers. Indeed, creationist criteria for "kinds" seem to be variable, depending on the needs of the moment (Awbrey, 1981). The most commonly cited criterion is infertility, but morphology is included when infertility fails (Wysong, 1976, p. 60). Moore (1983 , p. 203) includes both:
A kind is a distinct group of interbreeding organisms found in a particular geographic area which are (sic) genetically isolated from other recognizably different organisms.
Since infertility sometimes fails (see Kitcher, 1982, p. 151-155) "recognizably different" (morphology) is used. But what does "recognizably different" mean? All would agree that whales and sparrows are "recognizably different" and "genetically isolated." However, most would also probably agree that goats and sheep are "recognizably different," but they can hybridize. Two species of zebras, Equus burchelli and Equus grevyi, have overlapping ranges but are not known to hybridize (Keast, 1965). And some species of fruit flies are "genetically isolated" yet not "recognizably different." In the final analysis, the concept of a "basic created kind" becomes meaningless; "kinds" are simply whatever creationists want them to be.
But, if these criteria of "genetic isolation" and "recognizably different" are objectively applied, what is the result? Hyenas are certainly dog-like in appearance, have a social structure and habits similar to those of some dogs, but they do not hybridize, so they are "genetically isolated." Nevertheless, if hyenas were derived from an ancestral dog "kind" one would still expect to see some indication of this genetic relationship. In fact, hyenas are more closely related to the viverrids (genet cats, civits and mongooses) and cats: "This has been established from recent studies of chromosome patterns . . ., and especially from fossil evidence. . . ." (Kruuk, 1972, p. 269). Gregory and Hellman (1939, p. 331), Romer (1966, p. 233) and Colbert (1980, p. 345) all note that the late Miocene-early Pliocene genus Ictitherium is transitional between viverrids and hyenas.
Gish, in Evolution: The Fossils Say No! (1978), gives his views on "basic created kinds," but his discussion is even less informative than those of other creationist writers. Humans are of course a "kind" (p. 32), and gibbons, chimpanzees, and gorillas are also each "kinds" (p. 35). But on p. 47 he lists apes as a "major kind," and also dinosaurs. Based on this statement alone, it would seem that a "major ape kind" gave rise to all other apes. However, Gish's previous discussion (p. 35) muddies this point, for although he states that each ape is a "kind," he further discusses "kinds" within "kinds," whatever that means. So, Gish's concept of a "basic created kind" is confusing at best, yet it appears in a book advertised for use in public schools.
Gish is, however, clear and partly correct on one point. Some animals, although not all that he lists, do appear in the fossil record with all the characteristics of a "kind." Bats are full-fledged bats when they appear in the Eocene, and, so far at least, no intermediates between them and their insectivore ancestors have been found. Creationists are quite fond of this point and use it often. They should derive only slight comfort from this fact, however, since there are numerous well documented examples of evolutionary relationships that go beyond what is accepted as variation within "kinds."
Obviously infertility cannot be the criterion to determine fossil "kinds." Morphology will have to be used, so the question becomes, will creationists apply this criterion objectively and consistently, or will they simply establish arbitrary "kinds" as the need arises? Most likely the latter, because if morphology is applied objectively and consistently, animals as "recognizably different" as hyenas and civits would end up as members of the same "basic created kind."
Perissodactyls
The living perissodactyls are grouped into three familiesEquidae, Rhinocerotidae, and Tapiradae, all of which are "reproductively isolated" and "recognizably different." Nevertheless, they are united by several shared characteristics. For example, the cusp pattern of the cheek teeth is similar, a condition referred to as lophodont. The dentition does vary in crown height, being highcrowned (hypsodont) in grazers (equids and one rhinoceros), and low-crowned (brachyodont) in browzers (the tapirs and other rhinoceroses). The digits in the hind foot are reduced to three or one, and the front foot has one, three or four digits. But in all, the plane of symmetry of the foot passes through the third toe. In addition, the calcanium and astragalus, although present in the ankle of all mammals, are uniquely perissodactyl.
There is considerable variation in living and fossil perissodactyls, but most such variation is related to specializations in diet as reflected in the dentition, and skeletal modifications related to locomotion and size. For example, limb element reduction, an adaptation for running, is extreme in horses, while the heavy rhinoceroses and extinct titanotheres have those skeletal modifications related to large size. It seems unlikely that creationists would consider horses, rhinoceroses and tapers to represent variation within a single "created kind," but if perissodactyls are considered in detail, some questions arise.
It seems odd that the Creator saw fit to give zebras perissodactyl teeth and ankles and one toe, while giving Thompson's gazelle and the wildebeest artiodactyl teeth and ankles and two toes. After all, these animals live side by side, eat the same grass, and flee from predators. There are behavioral differences which reduce interspecific competition (Bell, 1971), but overall they seem to have been "designed" to do the same thing. Did the Creator have two plans for plains-dwelling, grazing, running animals? If so, it seems that one plan was inferior to the other since perissodactyls of this type were formerly much more abundant and varied, but now constitute only a small part of the mammalian fauna. Indeed, all perissodactyls have declined in abundance and diversity. But we are told that "design" is a strong argument for creation, and that creation was perfect and complete (Morris, 1974). Of course the entire earth, and all upon it, is in a state of decline (the principle of disintegration according to Morris, 1974), but why should this affect perissodactyls and not artiodactyls?
At the family level, modern perissodactyls are quite different one from the other, but does this hold up if each family is traced back in the fossil record? It should according to Gish (1978, p. 47) who claims:
We would thus expect to find the fossilized remains, for example, of cats, dogs, bears, elephants, cows, horses, bats, dinosaurs, crocodiles, monkeys, apes, and men without evidence of common ancestors. Each major kind at its earliest appearance in the record would possess, fully developed, all the characteristics that are used to define that particular kind.
This prediction is simply not borne out by the fossil record. Creationists will no doubt disagree and gleefully point out that bats and rodents appear abruptly with no evidence of ancestral forms. However, creation "science" is all or nothing, either nothing evolved or everything evolved. With this in mind, let us look at the fossil record of the perissodactyls-horses, tapirs, rhinoceroses, and the extinct titanotheres and chalicotheres.
Horses
The following account is concerned with those equid genera and evolutionary trends that led from Hyracotherium to Equus. This is not to minimize the fact that horse evolution was actually a complex of diverging lineages, at least after the appearance of Miohippus in the late Oligocene. These other lineages are important, interesting, and well documented by fossils, but are peripheral to the main argument advanced in this article. The following account briefly reviews the overall trends in horse evolution, presents a brief description of the "main line" genera, and concludes by addressing those criticisms voiced by creationists.
Some major trends in horse evolution were: 1) increase in size; 2) lengthening of the legs and feet; 3) reduction of lateral toes; 4) molarization of premolars; 5) development of high-crowned, cement-covered cheek teeth; 6) increasing complexity of the enamel pattern of the cheek teeth; and 7) changes in skull proportions to accommodate high-crowned cheek teeth. These trends were not uniform, nor did they all occur simultaneously. For example, horses actually decreased in size slightly in the Eocene, but the "main line" genera increased in size thereafter. Molarization of the premolars largely preceded the reduction of lateral toes. Some trends, however, were interrelated and proceeded at more-or-less the same rate. The change in skull proportions, for example, occurred along with the development of high-crowned cheek teeth.
Horse evolution begins with Hyracotherium, first described by Owen (1841) based on specimens from the London Clay. Early North American discoveries were referred to as Eohippus, but it eventually became clear that both Hyracotherium and Eohippus were similar enough to be included in the same genus, so the earlier name applies. Owen did not realize that Hyracotherium was related to modem horses, and, in fact, he compared it with some other animals. This is a point exploited by creationists and will be discussed later.
Horse evolution was largely a North American phenomenon. There were Old World Eocene genera, and the European paleotheres diverged from the ancestral equid stock but died out in the Oligocene. Some later Cenozoic genera, especially those in the Miocene and Pliocene, migrated from North America to the Old World and to South America, but Hyracotherium to Equus evolution was a North American event. The "main line" horse genera are briefly described in the following paragraphs. See Figure 1.
Hyracotherium (Eohippus):
This little animal varied from about ten to twenty inches at the shoulder (Simpson, 1951), and served as the ancestral stock for all later horses. Hyracotherium is reported from late Paleocene age sediments (Morris, 1968; MacFadden, 1982), but Savage and Russell (1983) note that these specimens are probably early Eocene. In any case, early Eocene specimens are common. The forefoot had four fully functional toes; the fifth toe was smallest, and no vestige of the first metacarpal remained. The hind foot had three toes. In each jaw there were four premolars and three molars, all of which were low-crowned.
Orohippus:
The differences between middle Eocene Orohippus and Hyracotherium are slight.
Orohippus and Hyracotherium are very similar to each other in almost all known anatomical characters (Kitts, 1957, p. 1).
The fact that advanced species of Hyracotherium and primitive species of Orohippus resemble one another so closely clearly indicates that Hyracotherium was the immediate ancestral form
. . . (Kitts, 1957, p. 32).
The main difference between these two genera is that the third and fourth upper premolars of Orohippus have four cusps of roughly equal size, and the heel of the third lower molar is relatively shorter than in Hyracotherium (MacFadden, 1976). That is, the upper premolars of Orohippus were more molariform than in Hyracotherium.
Epihippus:
Epihippus appeared in the late Eocene, and differs little from Orohippus. The upper and lower third and fourth premolars were molariform, and the first lower premolar was single-rooted rather than double-rooted as in Orohippus and Hyrocotherium. According to MacFadden (1976, p. 11): "This degree of premolar molarization is nearly approximated in some advanced specimens of Orohippus."
Mesohippus-Miohippus:
Mesohippus was the early Oligocene descendant of Epihippus. It differs from its ancestor in the reduction of metacarpals to three (II, III, and IV), but a small vestige of a fourth remained. The most notable difference was in the second upper premolar which was molariform and thus advanced over the stage in Epihippus. Various species of Mesohippus vary in size but average about 24 inches at the shoulder.
Mesohippus died out in the middle Oligocene, but not before giving rise to Miohippus, In general, species of Miohippus were larger than Mesohippus, and differed from their ancestor in some other details. For example, a small infold of enamel of the upper molars, the crochet, appeared as an occasional variation, but became a constant feature in many later horses. Also in Miohippus, and all later horses, the cannon bone (third metatarsal) was in contact with the ectocuneiform and cuboid, while in earlier forms it only contacted the ectocuneiform. Miohippus is not known after the early Miocene.
Parahippus:
Parahippus, an early to middle Miocene genus, intergrades with Miohippus on the one hand, and with its descendant, Merychippus, on the other. Molarization of premolars was already completed in Mesohippus, but that genus showed the first indication of lengthening of the limbs and feet, a trend also seen in Parahippus. The crochet, seen as a variant in some specimens of Miohippus, was consistently present on the upper molars of Parahippus. The trend toward high-crowned teeth and the addition of cement to the cheek teeth are both first seen in Parahippus.
Cement first appears as a mere film on the teeth of some members but not others in single populations. Gradually it comes to characterize whole populations and, still varying, it increases in average thickness until it reaches an evident optimum about which it fluctuates without further secular change down to recent Equus (Simpson, 1953, p. 106-107).
Figure 1: Evolution of the horse family (greatly simplified), showing the transition from Hyracotheriumto the modern horse, Equus, and the evolution of the forelimb.
Merychippus:
Although later species of Parahippus are difficult to distinguish from Merychippus, a middle and late Miocene genus, the latter is generally considered to be the first grazing horse. All of the molariform cheek teeth were high-crowned (hypsodont), richly covered with cement, and the enamel pattern on the chewing surfaces was more complex, Merychippus stood about 40 inches at the shoulder, the size of some modern ponies. The feet were still functionally three-toed, but the side toes were further reduced compared with earlier genera.
Pliohippus-Dinohippus-Equus:
The modern stage of evolution is closely approached with the appearance of Pliohippus in the Miocene. Pliohippus shows further progression in those trends established much earlier, For example, Pliohippus was functionally one-toed, although some early species still possessed minute side toes not seen in later species. The molariform cheek teeth were much like those of Merychippus except they were higher-crowned, had more cement, and a more complex enamel pattern. The cheek teeth were, however, markedly curved unlike those of Equus.
Dinohippus was the later Miocene descendant of Pliohippus, and in the Pliocene gave rise to Equus. Both genera are very similar to one another. Dinohippus was restricted to North America, but Equus migrated to the Old World where it survives, although in considerably reduced numbers.
To summarize, size increase was rather constant in the "main line" genera from Mesohippus to Equus. Molarization of the premolars was completed when Mesohippus appeared, but hypsodonty and cement are first seen in their incipient stages in Parahippus. Changes in skull proportions occurred more-or-less with the continued development of increasingly high-crowned cheek teeth. Reduction of toes from four to three in the forefoot probably occurred within Mesohippus, and further reduction of toes to one, in both forefeet and hind feet, occurred in Pliohippus. Limb elongation is first seen in Mesohippus with later genera showing a continuation of this trend.
Surprisingly, creationists have written very little on horse evolution. Wysong (1976), Gish (1978), and Hiebert (1979) each devote little more than a page to the topic. Cousins (1971) presents the most complete coverage, but his paper is mostly a report on the work of Nilsson in 1954 who derived some of his data from Abel, a 1929 source; the latter work claimed by Cousins/ Nilsson (?) ". . . to be representative of the present position of relevant research." This is quite a remarkable statement in view of the hundreds of papers on horse evolution which have appeared since 1929. Nonetheless, it is true that only one general summary work has appeared in recent decades (Horses by G. G. Simpson, 1951), and, as Woodburne and MacFadden (1982) point out, the early workers had the overall story essentially correct.
Cousins (1971) concentrates on two transitions in the horse lineage, Epihippus to Mesohippus and Parahippus to Merychippus. According to Cousins (p. 106):
Epihippus is the last of the old horses, while Mesohippus is the first of the new horses. Between these we have a very considerable jump. For the first were small animals, only as big as foxes, with four-toed forefeet; only with the latter did the large, three-toed type first occur.
It seems that size and the number of toes in the forefront is the evidence for this "considerable jump." Kofahl (1977, p. 66) has almost certainly paraphrased and elaborated a bit on Cousins' work in his discussion of Hyracotherium, Orohippus, and Epihippus:
. . . the average size of these creatures, sometimes called 'old horses', decreases along the series, which is contradictory to the normal evolutionary rule, and they were all no larger than a fox.
Between Epihippus and Mesohippus, the next genus in the horse series, there is a considerable gap.
In addition to these statements simply being wrong, Kofahl has cited none other than Simpson (1951) as his source. Simpson said no such thing, and in fact on p. 117 stated:
The larger species of eohippus were not particularly tiny animals: they were about half the size of a Shetland pony.
It is true that later Eocene horses, Orohippus and Epihippus, were somewhat smaller than Hyracotherium, but not by much. Species of Hyracotherium varied from 10 to 20 inches at the shoulder, and the mounted skeleton of Orohippus in the Peabody Museum of Natural History measures 13 or 14 inches at the shoulder. "The typical height [of Mesohippus] was probably about 6 hands (24 inches) although species considerably smaller and larger are known" (Simpson, 1951, p. 124). Lull (1931, p. 18) described Mesohippus bairdi as about 18 inches in height, but noted that Mesohippus intermedius was much larger. Scott and Jepsen (1941, p. 911) described Mesohippus as about the size of a greyhound.
Cousins' "very considerable jump" in size is a creationist myth. But can the three-toed forefoot of Mesohippus substantiate Cousins' claim? Paleontologists have for a long time maintained that Mesohippus was the first threetoed horse. For example, Scott (1891, p. 324-325) states:
The metacarpus . . . consists of three functional members, the second, third, and fourth, and one rudimentary, the fifth.
The fifth metacarpal is represented by a rudiment which carries no phalanges. The head is as large as in No. IV, but the shaft is very slender and tapers rapidly to a point.
Taken with the other similarities between Mesohippus and Epihippus, it would seem the rudimentary fifth metacarpal would be compelling evidence for a close relationship. Creationists of course demand more.
It is not widely known, except among those who study mammalian evolution, but Mesohippus with four metacarpals and only three toes has been known at least since 1975. According to MacFadden (1976, p. 12) who cites a personal communication from Emry (1975):
Recently . . . Mesohippus has been found with a fourth metacarpal (V) that is nearly as long as metacarpal IV, which articulates with three phalanges.
In his response to my inquiry for additional data on four-toed Mesohippus, Dr. Emry of the National Museum of Natural History kindly gave permission to use his data. He pointed out that forefeet of middle Oligocene Mesohippus are well known, and as far as he is aware all are three-toed. But there are specimens representing at least two individuals from the early Oligocene of Wyoming which have four metacarpals as noted by MacFadden above. Dr. Emry further noted that the Wyoming specimens have dentition much like that of Mesohippus hypostylus, which means that the second premolars are molariform.
It seems that the main difference between Epihippus and Mesohippus is the molarization of the second premolars. The number of toes will no longer help creationists since: "The present specimens suggest that the reduction of the most lateral digit took place within Mesohippus rather than between Epihippus and Mesohippus" (Emry, 1984, personal communication). Some specimens of Mesohippus have forefeet not quite like those of Epihippus, three toes but four metacarpals, which is not characteristic of typical Mesohippus. These same specimens do have typical Mesohippus dentition, however. In addition, some Mesohippus species were not particularly larger than their ancestors. Even before Emry's data were available, a compelling case for an Epihippus-Mesohippus relationship could be made. In view of the data now available, continued use of Cousins' argument would be meaningless.
Cousins (1971) apparently thinks that Mesohippus and Parahippus are of the same kind, but draws a clear distinction between Parahippus and Merychippus. On p. 107 he says:
With Merychippus and Hipparion there is a rich group of Equus-like forms which are all separated from the former 'brachydontal' groups by a gaping evolutionary gap.
Cousins tries to make a case for this "gaping evolutionary gap" by concentrating on toes and teeth. As for toes he claims: "One-toedness dominated, although quite clear rudiments of two side-toes may occur." He claims that Merychippus had hypsodont, cement covered teeth, which is correct, but that horses of this type appeared suddenly with no indication of ancestors. Cousins' choice of toes and teeth was a poor one, at least in support of the creationist cause, since these two evolutionary trends did not proceed at the same rate. Cousins is simply wrong with respect to toes; Merychippus was three-toed as was Parahippus, but Pliohippus was one-toed. Typical Merychippus had teeth more similar to Pliohippus than Parahippus, however. Cousins' "gaping evolutionary gap" may exist in his mind, but the fossils tell a completely different story.
What do authorities on horse evolution have to say about Parahippus-Merychippus relationships? Stirton (1940, p. 178) says:
Differences between the most primitive species of Merychippus and the most advanced species of Parahippus are hardly distinguishable.
Forsten (1975, p. 395) in referring to fossils from the Burkeville fauna of the Texas Gulf Coastal Plain noted that: "Merychippus gunteri . . . resembles Parahippus leonesis in many characteristics." Woodbume and Robinson (1977) point out that one author (White, 1942) identified Merychippus gunteri in the Thomas Farm fauna of Florida, but a second author (Bader, 1956) regarded these same specimens as Parahippus leonesis. Simpson (1953, p. 104) has made one of the strongest statements in defense of this transition:
. . . he [Dietrich, 1949] says of Parahippus and Merychippus that no intermediate form bridges the gap between the two, 'no gradual transition can be established.' The statement is . . . false. There are unified samples, surely representing local populations, perfectly intermediate between Parahippus and Merychippus and so varying in the 'diagnostic' characters that assignment of individuals in a single population could be made to both genera and assignment of the population to one or the other is completely arbitrary.
The most significant evolutionary event seen in Merychippus was the development of hypsodont (high-crowned) cheek teeth which were covered by cement. Most paleontologists interpret this change in dentition as an adaptation for grazing, but it was in Parahippus ". . . that the inception of hypsodonty took place" (Stirton, 1940, p. 177).
From Merychippus to Equus not much need be said. Creationists no doubt realize the futility of trying to draw distinctions between these genera and the intermediates Pliohippus and Dinohippus. Merychippus was functionally three-toed while the others were one-toed, and the latter genera had higher-crowned cheek teeth, but all are recognizably horses.
Cousins (1971, p. 108) in his conclusion seems unable to clearly state what his argument for creation is. For example, he criticizes a study done by Stecher (1968) in which that author drew evolutionary conclusions based on variability in the chromosome count of modern equids and the variability in their spinal columns. Cousins disagrees with Stecher's evolutionary conclusions:
It suggests, to my mind, nothing of the kind; it shows conclusively that the spines and chromosome counts are different in different animals and absolutely no evolutionary argument can legitimately be imported into his researches.
The key phrase here is "different in different animals." Yet Cousins' final sentence reads:
The horse family is unique and separate and the evidence can, without any weighting, be fitted to the case for special creation.
Cousins has argued that Eocene horses are different from Mesohippus, and that Mesohippus-Parahippus are different from later horses. If they are really as different as he claims, they cannot all represent variation within a single "kind," so there must be three "kinds," a point with which Kofahl (1977, p. 66) seems to agree. Cousins then argues that different chromosomes and vertebral spines occur in different animals. Is each then a different "kind"? This isn't clear. But, if Kofahl's view is accepted, some horses must have evolved, even if we consider it to be only microevolution. In this case, however, Stecher was correct in the beginning. So why does Cousins argue against him? And, worse yet, why does Cousins finally argue that the whole horse family was specially createdas though we were back to thinking of it as a single kind? Consistency such as this is the hallmark of pseudoscience.
Hiebert (1979, p. 60-61) concentrates on problems of size and ribs, and makes no mention of other features showing the relationships among fossil and living horses. He notes that Eohippus (Hyracotherium) has 18 pairs of ribs, Orohippus has 15 pairs, Pliohippus has up to 19 pairs, and Equus scotti has 18 pairs, and concludes (p. 61): "The rib count denies any continuous evolution here." Perhaps Hiebert is unaware, but in mammals, ribs are found on the thoracic vertebrae, and the number of thoracic vertebrae and hence the number of ribs varies (Romer, 1962). The rib count is usually consistent in a species, but even here there is some variation, for example, in some individual humans (Crouch, 1965). And among the equids, Epstein (1971, p. 422) reports that modern horses may have 17, 18 or 19 pairs of ribs.
Creationists are particularly fond of small modern horses, dwarfed Argentine horses for example, and try to make a case for their similarity to some of the smaller fossil horses (Wysong, 1976, p. 304; Hiebert, 1979, p. 61). The dwarfed Argentine horse is similar only in size to Eocene and perhaps early Oligocene horses, genera which Cousins (1971) has claimed are quite different from later equids both in size and morphology. Moore and Slusher (1974, p. 420) think that "poor feed" may account for some small fossil horses, and give an example of small modern horses discovered in 1942 which reportedly were small for this reason. It seems, however, that all known specimens of Eocene and early Oligocene horses were small. Surely if "poor feed" were the cause, there must have been some that enjoyed an adequate diet and "normal size." And claiming that these small horses were simply size variants of other larger genera will not work. "Poor feed" may account for smaller size, but it will not change molars into premolars, nor will it add toes to the feet.
Paleontologists have long been aware that there was size variation in fossil equids. Archaeohippus (Miocene) and Calippus (Miocene-Pliocene) do in fact show a decrease in size compared to their contemporaries. In the "main line" leading from Hyracotherium to Equus, however, there was a general increase in size, a point exploited by Hiebert (1979, p. 61):
Once horse fossils were found in a variety of sizes, it took little ingenuity to line them up from smallest to largest and to insist that the evolution of the horse has been proven.
Obviously Hiebert is charging paleontologists with outright deception. But it has never been claimed that size increase through time was uniform and continuous in all genera, nor is size increase the only evidence for horse evolution.
Moore and Slusher (1974) and Kofahl (1977) claim that horse fossils have a scattered distribution making them useless for evolutionary studies.
. . . the fossils of these horses are found widely scattered in Europe and North America. There is no place where they occur in rock layers, one above another (Moore and Slusher, 1974, p. 420).
This quote contains two statements, both of which are only partly correct. As for the scattered distribution, only two of the "main line" genera, Hyrocotherium and Equus, are known from both Europe and North America. All others on the "main line" are uniquely North American, and all of the relevant genera have been recovered from sediments in the western United States. In fact, the relevant genera are known from Utah, Wyoming, Nebraska, and South Dakota, although the geographic distribution of some was much greater than this. While it is true that equids lived in Europe and Asia, it seems that the scattered distribution is true only in the broad sense; the "main line" genera occur in a considerably more restricted area. However, this is still a rather large area, and that brings us to the second part of the above quote: "There is no place where they occur in rock layers, one above the other."
A full response to this claim would be rather lengthy, because what is really being questioned here is mammalian biostratigraphy. This is an area in which "stage of evolution" actually has been the basis for relative age determinations and correlations, Fortunately this issue has been dealt with in some detail (see Schafersman, 1983, p. 238-241), so it will not be repeated here. Suffice it to say that the relative sequence of continental mammal-fossil-bearing strata has been independently verified by radiometric dating. And, while it is true that the entire horse lineage is not represented by fossils in a single area of superposed beds, there are many places where at least parts of the sequence have been found. The fact is that in those superposed strata containing a part of the horse lineage the sequence is consistent.
Cousins (1971) and Wysong (1976) compare Hyrocotherium with the modern hyrax:
Hyrax, like Hyracotherium, is a small animal, about the size of a rabbit or fox. Like these, Hyrax has four toes on the fore-limbs and three on the hind limbs, a quite striking similarity. The back teeth of the two genera exhibit many similarities and resemble those of the rhinoceri more than those of horses (Cousins, 1971, p. 106).
Cousins notes that Hyracotherium is not much like the modern horse, which is true, but then Archaeopteryx does not resemble any modern bird, except for having feathers, yet creationists consistently claim it was a bird. Cousins also tries to capitalize on the fact that Owen (1841) derived the name Hyracotherium to suggest the similarity to hyrax. There are some problems with Cousins' arguments. For one thing, Owen did not have other fossil equids with which to compare. Also, Owen compared Hyracotherium with pigs and rodents. Apparently, he felt the similarities with hyrax were greater, hence the name he chose. Superficial similarities can, however, be misleading and Owen was not the only one to make such a mistake. For example, Colbert (1980, p. 423) notes that: "Among ancient peoples, and even among the earlier modern naturalists, these animals [hyraxes] were thought to be rabbits of some sort . . ." In addition, hyrax is also called a coney as are some lagomorphs.
In view of these facts, and the fact that hyrax incisors are rodent-like, are we to assume that the hyrax, rabbits and rodents are "amazingly similar"? One could certainly make as good a case for this as Cousins has for the Hyracotherium-hyrax similarities. Perhaps all Eocene horses, modern rabbits, hyraxes and rodents (and maybe pigs too) were all derived from a single "basic kind." All of these resemblances are, however, rather superficial. In fact, the differences among these animals are much greater than the differences between Epihippus and Mesohippus or between Parahippus and Merychippus.
As evidence that the fossil record is supposedly more in accord with creationism, Gish (1978, p. 157) quotes Goldschmidt (1952, p. 97): "Moreover, within the slowly evolving series, like the famous horse series, the decisive steps are abrupt without transition." Wysong (1976, p. 301) makes a similar statement: "There are no gradations from one link to another. All suggested links appear suddenly in the fossil record."
However, these same fossils, at least some of them, are sometimes cited by other creationists as evidence for variation within a "basic created kind" (Moore and Slusher, 1974, p. 420). To sum up creationist opinion on this, then, it seems that all the following things must be true: (1) horses show variation within a "basic created kind," (2) all of the thousands of fossil horses were alive at the same time, (3) all were buried in deposits of a single flood, (4) but only distinctive types without intermediate variants were preserved. "Gaps" between these distinctive types are used as evidence against evolution, yet the same distinctive types show variation with a "basic created kind." Clearly, the creationist position on horse evolution is self contradictory.
Furthermore, contrary to creationist critiques, the evolutionary relationships among horses are not based upon conjecture, supposition, or a fragmentary or incomplete fossil record. Thousands of fossils, some of which are exactly what creationists have demanded (intermediates between intermediates), have provided the evidence for these relationships. So why don't creationists simply accept this evidence? They could still claim that Hyracotherium to Equus only shows variation within a "basic created kind." They will not likely do this, however, since Hyracotherium is just too different from Equus. In fact, Cousins particularly has argued that Hyracotherium is very unhorselike, and more similar to the modern hyrax. No doubt creationists will simply continue to rely on Cousins' work, continue to selectively quote evolutionary biologists, and/or largely ignore horses and concentrate on bats and rats.
Tapirs
Living tapirs are represented by a single genus and four species. Like other perissodactyls, they were formerly more varied and abundant, but are now geographically restricted occurring only in the New World tropics and the Malayan area. Tapirs are large animals measuring up to eight feet long and weighing as much as 700 pounds.
Since tapirs appear to have always been forest dwellers, their fossil record is not as good as that of most other perissodactyls. Nevertheless, fossils are abundant enough to document their ancestry with a fair degree of accuracy. See Figure 2.
Tapirs are closest to the ancestral perissodactyl condition since they have changed far less than members of the other groups. Indeed, some authors (Scott and Jepsen, 1941) consider them living fossils because so little change has occurred, especially since the Oligocene. All tapirs, living or fossil, have four-toed forefeet and three-toed hind feet, and have low-crowned teeth. However, modern forms have only two premolars in each jaw. The most notable evolutionary trends were the development of a short proboscis and an increase in size with the skeleton becoming stouter.
The earliest tapiroids, Homogalax (Family Isectolophidae) and Heptodon (Family Helaletidae), are found in strata of early Eocene age. Neither is particularly different from the earliest of the horse series, Hyracotherium, but Homogalax shows a greater similarity than does Heptodon. Both genera were small, a little larger than Hyracotherium, and both were also similar to a third middle Eocene tapiroid genus, Hyrachyus (Family Helaletidae).
Hyrachyus differs from Heptodon only in being slightly larger and in having slightly higher crested teeth and no third molar hypoconulid . . . (Radinsky, 1968, p. 317).
This is an important point since most authorities think Hyrachyus was ancestral to "one group of primitive rhinocerotoids" (Radinsky, 1968, p. 317). In fact, Hyrachyus has been classified as a rhinoceros by some, but the consensus now seems to be that it is a tapiroid.
True tapirs (Family Tapiridae) are found in the early Oligocene, being derived from ancestors like Heptodon. Protapirus has a skull length of about one foot, not quite one-half as large as the skull of the largest modern tapirs, but the "proportions of the limb bones are decidedly more slender" (Scott and Jepsen, 1941, p. 749). Scott and Jepsen also noted that while the dentition is tapir-like, it is not as specialized as in modern tapirs. Protapirus may have been developing an incipient proboscis as evidenced by some retraction of the nasal bones.
Miotapirus of the Miocene is the direct descendant of Protapirus, and is ancestral to Tapirus, the modern tapir. It was somewhat smaller than Tapirus, but had strongly retracted nasals indicating the presence of a proboscis.
Rhinoceroses
The Family Rhinocerotidae contains four living genera and five species, all being confined to Africa or southeast Asia. These are large animals weighing up to 3600 kg (7937 lbs.) (Kingdon, 1979). Among the fossil rhinoceroses is found the largest known land mammal: Baluchitherium (or Indricotherium), an Oligocene-Miocene rhinoceros of Asia, probably stood 16 to 18 feet at the shoulder. Like all perissodactyls, rhinoceroses were formerly more varied, abundant, and more widespread geographically, especially in the Oligocene and Miocene, but now seem to be headed for extinction (see Martin, 1984).
Modern rhinoceroses are all recognizably rhinoceroses, but they do show a great amount of variability. For example, three genera, Diceros, Ceratotherium, and Dicerorhinus, have no upper or lower tusks, but do have nasal and frontal horns. In contrast, Rhinoceros has both upper and lower tusks, but has only a nasal horn (Matthew, 1931). In addition, the Asian forms have folds in the skin giving them an armor-plated appearance not seen in African forms. Rhinoceroses also vary considerably in size from the relatively small Sumatran rhino (Dicerorhinus sumatrensis) to the large white rhino (Ceratotherium simum). There is also variation in the crown height of the cheek teeth.
Figure 2: Evolution of the rhinoceros and tapir families (greatly simplified), showing the early Perissodactyl common ancestor.
Fossil rhinoceroses are placed in three families. One, the
Amynodontidae, was restricted to the Eocene and Oligocene and became extinct.
Amynodonts were probably derived from a tapiroid stock (Figure 2), but
their ancestry is not so well documented as compared to the other two families. The
other
families are the Hyracodontidae, or running rhinoceroses, and the
Rhinocerotidae, or true rhinos. The former appeared first and was ancestral
to the latter.
Rhinoceros evolution is more complex than that of other perissodactyls. Rhinoceroses were numerous and varied earlier in the Cenozoic, and several lineages show evidence of parallel evolution. In any case, the details would require considerable space, so I have chosen instead to concentrate on the earliest rhinoceroses and those descendants that led to animals that are undeniably rhinoceroses. Accordingly, the discussion will be restricted mostly to Eocene and Oligocene forms.
As noted earlier, the Eocene genus Hyrachyus has in the past been classified as a rhinoceros, but most experts now agree that it belongs in the extinct tapiroid family Helaletidae. Hyrachyus was quite similar to other early tapiroids, especially Homogalax and Heptodon, which were in turn quite similar to Hyracotherium. There seems to be little doubt that the earliest rhinocerotoids, Triplopus (Family Hyracodontidae), were derived from Hyrachyus or a Hyrachyus-like tapiroid ancestor:
Characteristic rhinocerotoid dental features are approached in some variants of a late middle Eocene species of Hyrachyus, which is overlapped in dental morphology by primitive variants of an early late Eocene species of Triplopus, a hyracodontid rhinocerotoid: thus it appears that at least one line of hyracodontid rhinocerotoids evolved from Hyrachyus (Radinsky, 1967, p. 12).
Hyrachyus had four toes in the front foot and three in the hind foot, while Triplopus had three toes in all feet. Some hyracodontids were fairly large animals measuring up to five feet in length and two and one-half feet at the shoulder. Forstercooperia, for example, had a skull about seventeen inches long (Lucas et al., 1981, p. 834), but Triplopus was considerably smaller.
Triplopus was ancestral to other hyracodontids in North America, but more importantly, in Asia it gave rise to the earliest member of the Rhinocerotidae, Prohyracodon, from which it differed very little. Prohyracodon, from the late Eocene, and the related Oligocene genera Caenopus, Trigonias and Subhyracodon represent the central stock of true rhinoceroses. These were large animals, Caenopus was up to eight feet long, and all were rhinoceroses in every sense of the word except for being hornless.
Horned rhinoceros, es appeared in the latest Oligocene and Miocene. One of the earliest was Dicerorhinus, the genus to which the modern Sumatran rhino belongs. Rhinoceros, which includes the living Javan and Indian rhinos, is known as far back as the middle Miocene. Kingdon (1979) reports that Paradiceros mukiri from the Miocene of Africa may be ancestral to the modern African species. He further notes that both modern African genera, Ceratotherium and Diceros, were present in the Pleistocene, the latter in its present form, but Ceratotherium praecox of the late Pliocene still shows resemblances to Diceros, but is probably directly ancestral to Ceratotherium simum, the modern white rhino.
The preceding discussion has been intended only to demonstrate the continuity in rhinoceros evolution. There were certainly many other Cenozoic rhinoceroses, such as the common North American forms Hyracodon, Aphelops and Teleoceras, but these were not ancestral to modern forms and therefore have not been considered. In addition, there were rather bizarre Asian forms, Sinotherium and Elasmotherium, that, unlike other horned rhinoceroses, had a single rather large horn located on the frontal bones of the skull.
Brontotheres (Titanotheres)
Compared with other major perissodactyl groups, the titanotheres were short lived, first appearing in the early Eocene and disappearing at the end of the early Oligocene. Nevertheless, they persisted for about 21 million years, during which time they evolved from small Hyracotherium-sized animals to giants measuring up to eight feet at the shoulder. Numerous genera of titanotheres have been described, and while their interrelationships are complicated, the overall trends in titanothere evolution are quite clear. The dominant trends were the attainment of large size, and the development of large horns on the skull. In contrast to the equids, the cheek teeth remained unprogressive and simple; the molars were low-crowned, and the premolars remained small and became only partly molariform. All titanotheres had four toes in the forefoot and three in the hind foot. There were, however, skeletal modifications related to the large size of the later members of the family, and skull modifications related to the development of horns.
The first titanotheres, Lambdotherium and Eotitanops, appeared in the early Eocene. The former measured about fourteen inches at the shoulder (Osborn, 1929), and had a skull about seven inches long (Gazin, 1952, P1. 10). Except for details of the dentition, Lambdotherium differed little from the earliest horse, Hyracotherium. Eotitanops was about 50% larger than Lambdotherium. In this genus there was established "the basic molar pattern that remained essentially unchanged throughout titanothere evolutionary history" (Radinsky, 1968, p. 314).
Palaeosyops and Manteoceras are typical middle Eocene genera. The former gave rise to a branch of titanotheres which was hornless but had enlarged canines. The latter measured a bit over four feet at the shoulder, and can be considered on the "main line" to the giant early Oligocene forms. Incipient horns first appeared in Manteoceras being represented by "the paired roughening of the nasal bones, to which horns must have been attached" (Scott, 1945, p. 239).
The late Eocene is characterized by genera like Dolichorhinus and Protitanotherium which had rudimentary horns on elongate nasal bones, and which were larger than modern tapirs. With the development of horns came a change in the configuration of the skull: it became saddle shaped. This trend is first seen in these later Eocene genera, but is more pronounced in early Oligocene forms.
Titanotheres of the early Oligocene such as Brontops and Brontotherium, were very large animals. The nasal bones had large rugose horn-like structures which Stanley (1973, p. 456) thinks served "to protect the head and neck region against injury during butting, which was probably chiefly intraspecific in nature." Brontops, Brontotherium, and other large contemporary genera are the last titanotheres known. Their extinction may be accounted for by their unprogressive dentition not being suitable for the harsher vegetation that characterized the early Oligocene and later parts of the Tertiary.
Chalicotheres
Chalicotheres are the most peculiar perissodactyls. Later types were large animals, about the size of modern horses, and they had a rather horse-like appearance. However, their similarity to horses is rather superficial. The dentition was more like that of the titanotheres, the front limbs were longer than the hind limbs, and there were large claws on the toes. In fact, the feet are so peculiar that in the early fossil finds of these animals the teeth were classified as perissodactyl and the feet as edentate. The use of these clawed feet by a somewhat horse-like animal has been the subject of debate. Some authors (Romer, 1966; Colbert, 1980) suggest that they were used to dig up roots and tubers. Whatever their use, they represent a specialization not seen in any other perissodactyl group. Other than these peculiar feet, however, the Chalicothere skeleton is typically perissodactyl (Peterson, 1907; Romer, 1966).
Chalicotheres are divided into two families, the Eomoropidae and the Chalicotheriidae, the former being ancestral to the latter. Eomoropids were confined mostly to the Eocene, although one genus, Eomoropus, persisted into the early Oligocene. Chalicotherids range from the late Eocene to the early Pleistocene, but were most varied in the middle Miocene when six genera were present. Chalicotheres are much more common in Old World, especially Asian, deposits and it appears that most of their evolution took place there. In fact, only four of the fifteen known genera are found in North America, the last being Moropus and Tylocephalonyx of the middle Miocene.
The earliest eomoropid, Paleomoropus of the early Eocene, is quite similar to Hyracotherium and equid as well as Homogalax (a tapiroid). In fact, Radinsky (1968, p. 308) notes that these three genera "are distinguished from each other by slight differences in molar cusp pattern and in size." Paleomoropus was about the size of a sheep, had four toes in the forefoot and three in the hind foot, but apparently lacked the clawed feet of Chalicotherids.
Eomoropus appeared in the middle Eocene and gave rise to all later chalicotheres. This genus differed from its ancestor, Paleomoropus, in details of the dentition, and from its descendants in being "smaller and more lightly built, with feet unspecialized (digits not sharply flexed)" (Radinsky, 1964, p. 9), and in dental details. The biggest difference between Eomoropus and its descendants is stated by Radinsky (1964, p. 13):
In short, I can find no features in the manus or pes of Eomoropus which suggest in any way the extraordinary modifications which appear in the feet of later chalicotheres.
The earliest and most primitive genus of the Chalicotheriidae, Schizotherium, appeared in the late Eocene. It is the only chalicothere known from the middle and late Oligocene, but was only one of several Miocene genera. Chalicotheres were probably never particularly abundant, but the six middle Miocene genera no doubt represent their greatest diversity. With the appearance of Miocene chalicotheres, such as the North American genus Moropus, not much more occurred. These animals were large, with claws on all the functional toes (three in each foot, although the forefoot retained a large vestige of the fifth metacarpal). The limbs were elongated, and the skull had a long deep face similar to the horses. To be sure, there was some variation in Miocene and later genera, doming of the skull in some, and variations in the anterior dentition (incisors and canines), but for the most part later chalicotheres differed little from typical Miocene forms. Only two genera, Ancylotherium and Nestrotherium, both Old World forms, are known from the Pliocene, and the latter did not become extinct until the Pleistocene.
Summary
Many of the earliest perissodactyls can be differentiated only with great difficulty. For any one lineage there is a sequence of fossils more-or-less continuously linking the earliest forms with their descendants. It may be argued that any one lineage simply shows variation within a "basic created kind." But the earliest members of each lineage are similar enough that, if morphology is the criterion for inclusion in a "kind," they also represent a "basic created kind." Therefore, all perissodactyls must have been derived from a single "basic created kind." See Figure 3.
The Chalicotheres developed a uniquely different kind of foot from the other Perissodactyls. The illustration above shows the front foot of Moropus. Note that the inner toe, rather than the middle toe, was the largest.
In view of the perissodactyl fossil record, it seems that creationists have several options. First, they could simply ignore the evidence, which is not scientific, and continue to concentrate on bats and rats. Second, they could acknowledge that perissodactyls have evolved, but no other group has (which seems unlikely). Third, they could argue that perhaps all perissodactyls were derived from the same "basic created kind," their fossil record thus showing evidence only of microevolution. (This too seems unlikely since "recognizably different" animals, zebras and rhinoceroses, would be of the same "kind.") And fourth, they could arbitrarily divide perissodactyls into several "kinds," as Cousins seemed to do with horses.
Creationists will almost certainly take the last option since they have done so in the past, the mammal-like reptile-mammal transition being a case in point. However, it is quite incredible that anyone could seriously argue that Epihippus and Mesohippus or Parahippus and Merychippus are really that different, especially when one considers the variability allowed in modern "kinds." Likewise, the earliest known equid, tapir, titanothere, and chalicothere are certainly similar enough to be considered members of the same "kind," if the morphological criterion is applied objectively and consistently. The earliest rhinoceroses can also be included in this "kind" since they differ so little from the ancestral tapirs. But since objectivity and consistency would yield negative evidence for creation "science," such a course will not likely be followed. After all, creation "is the basis of all true science" (Morris, 1983), and any evidence to the contrary is irrelevant.
The more that creationists demand of the fossil record, the weaker their case will become. Of course we will never be able to document the pedigree of all organisms, but many can be documented. For example, among the mammals, many artiodactyls, some carnivores, and others can be traced back in the fossil record, demonstrating the interrelationships among many life forms. As these data become more widely known, there will be fewer and fewer "basic created kinds" until it is apparent that all life forms are interrelated. Of course creationists could simply fall back on their Scopes era tactics by claiming that all species, living or fossil, are "basic created kinds." But then Noah's Ark would have had millions of passengers.
Acknowledgements
I thank Philip D. Gingerich for his review of this article, Robert J. Emry for permission to use some of his unpublished data, and James Gillingham for his helpful discussions. Frederick Edwords made several suggestions for improving the article, and Daniel Warren provided the illustrations.