Did Ctenosauriscus Produce Chirotherium Tracks?

Probably, the somewhat mysterious Ctenosauriscus koeneni was at least one of the producers of many Chirotherium tracks.

These fossil tracks are well known from the Triassic period of Europe, in particular from Germany. Two different types occur, a quadrupedal track consisting of marks of small fore-feet and of big hind-feet and a bipedal one consisting of imprints of the hind-feet only. The animal must have had rather long hindlimbs, since it walked on a straight line.



Fig.1. Chirotherium footprints, produced by imprints on a moist ground. Two different types occur, a quadrupedal track consisting of marks of small fore-feet and of big hind-feet and a bipedal one consisting of imprints of the hind-feet only.

It is not entirely clear whether these tracks were produced by a single species or by several.

For a long time Ctenosauriscus was known only by parts of a vertebral column that had been found in 1871 near Göttingen, Germany. Meanwhile further skeletal remainders are known from Waldhaus, Southern Germany, which can be attributed to Ctenosauriscus (Ebel et al. 1998).

Former ideas concerning the producers of Chirotherium tracks

Several workers have speculated which animal might have produced the Chirotherium tracks. Many candidates were taken into consideration, for example by the well-known fossil hunter Richard Owen. But for more than 50 years the efforts could not yield a serious success. W. Soergel (1925) arrived at the conclusion that the tracks could be attributed to a thecodontian. B. Krebs (1965) proposed a reptile named Ticinosuchus which had been found in marine sediments. However, certainly this animal had a quadrupedal gait and, therefore, cannot be made responsible for bipedal tracks. As can be seen from the tracks the hind-limbs must have been considerably stronger than the forelimbs. The tracks suggest that the producing animal walked on four feet using a slow pace, yet the faster gait was practised on the long hindlimbs only.



Fig.2. Ticinosuchus, after R. Carroll (1993), regarded as a track producer by B. Krebs (1965). It had a quadrupedal gait and was no mere land dweller.


The different size of fore- and hindlimbs has never been taken into account. But this size hints at the forces that are transmitted to the ground. It is clear that the stress of the feet of an animal is proportunately distributed to all feet, that is the force per foot area respectively the pressure acting on the ground is the same. Therefore, in the Chirotherium trackmaker the highest weight portion must have been transmitted by the hindfeet. An obligatory quadrupedal animal such as Ticinosuchus is not suited. Unfavourably stressed feet do occur only in women’s high-heeled shoes producing an enormous pressure on the ground which damages parquet floors.

Ctenosauriscus was a likely creator of many Chirotherium tracks

Although for a long time no definite remainders of an animal were known that might belong to Chirotherium tracks nevertheless appropriate informations were available, but palaeontologists have enormous problems in interpreting the former functions of fossil vertebrate remainders. They are familiar with the anatomy, the correct position of all bones in a skeleton, but they do not know why a bone looks as it looks. They are unable to recognize which function should have been connected with a particular skeletal shape. This is a handicap that cannot be overcome by self-teaching, it has to be learned by long-lasting practical experience in applying mechanics, no chance for fossil hunters.

Ctenosauriscus was known since 1871. The remainders consist of parts of a vertebral column with spectacularly elongated spinous processes. Since these processes are reminiscent of a comb it was called comb lizard. For many decades these isolated bones were unattractive to palaeontologists, at least until 1965 when B. Krebs produced an excellent reconstruction. As a matter of fact, these skeletal remainders are highly interesting, since they allow statements as to the mechanics of early land-dwelling vertebrates. Obviously, this animal was extremely specialized and well adapted to its unusual environment on land.



Fig. 3. Fossil remainders of the vertebral column of Ctenosauriscus, re-assembled from two parts on a slab, after B. Krebs (1965)



I got into touch with Ctenosauriscus in 1995 on occasion of a presentation by the well-known vertebrate expert Dr. Rupert Wild in Stuttgart who reported on the new finds of Chirotherium tracks and remainders of rauisuchians from Waldhaus. The new finds included a left Ilium, the shape of which allowed to place this fossil within the bipedally running poposaurid rauisuchians. Furthermore, dorsal vertebrae similar to those of Ctenosauriscus were found. They are characterized by extremely elongated, slightly curved, laterally strongly flattened spinous processes with a very slender cross-section. Dr. Wild supposed that all remainders were part of a single skeleton corresponding to Ctenosauriscus koeneni (v. Huene), which therefore was a rauisuchian capable of running bipedally.



Fig.4. Reconstruction of the vertebral column of Ctenosauriscus koeneni after B. Krebs (1965)                              


Although B. Krebs (1965) had presented a fine reconstruction using the preserved imprints on the slab (figs 3 and 4) he was unable fo find a plausible function of the extremely elongated spinous processes. He presumed a function in connection with temperature regulation as it had been proposed for other fossil vertebrates with elongated spinous processes such as Dimetrodon from the Permian period.

These vertebral remainders appeared mysterious to Dr. Wild, too, and he asked a question quite unusual for a palaeontologist, namely whether somebody had an idea concerning the meaning of such long spinous processes. I answered spontaneously  I had the impression that they served the suspension of muscles for forces with large vertical components. A later thorough investigation revealed a very interesting regularity to me indicating a mechanical function (see fig. 11). Obviously there was a problem of shape and function, and the shape followed from the function. The new function determined a change of the shape of the vertebral column.

 Two questions arose from the skeletal remainders that I intended to investigate, namely

     -   why are the spinous processes spectacularly elongated ?

     -   why do the directions of the spinous processes at their tips point to a common intersection ? 

The cause of  the elongation of the spinous processes in Ctenosauriscus

The spinous processes of the vertebrae in Ctenosauriscus are conspicuously elongated and in part considerably curved. They are laterally flattened, and the cross-section is rectangular. These features hint at an adaptively intended enlargement of the resistance moment of the vertebral column against deformation respectively against bending. To make clear this context imagine a ruler with rectangular cross-section. The flat side can easily be bended, but there is an enormous resistance against bending the high side. For this reason it is most probable that a (bio-) mechanical function has caused the development of this extremely shaped vertebral column. In addition, the change of shape is connected with the animal’s motility. In comparison to a water-dwelling vertebrate the weight of which is compensated by its buoyancy a land-dwelling vertebrate has to bear greater weight forces, and correspondingly stronger muscles must be present that have to lead the weight to the ground. As a skilled engineer I know how a design has to be shaped for a special function. On the other hand, a probable function can be derived from a conspicuously shaped bone or in this case from several. Nature finds in any case an excellent solution by mutation and natural selection as an engineer could not do it better. It is a kind of try-and-error method which finally leads to an optimum result by selection and survival of the most suitable individuals. However, the pre-conditions must be fulfilled so that a modification is basically possible. ”An elefant will never learn to fly”. The pre-conditions for a modification were obviously favourable in the ancestors of Ctenosauriscus. Presumably, their spinous processes had a normal length as outlined in fig.5



Fig. 5. Modification of spinous processes in a primordially unspecialized vertebral column of an ancestor with normal spinous processes resulting in the elongated and curved spinous processes of Ctenosauriscus koeneni (v. Huene)


Apparently, hitherto the spinous processes of tetrapods had never been submitted to a really thorough investigation as to their biomechanical function. This is all the more surprising as the spinous processes exhibit great variability concerning shape, structure, length, and direction. Maybe, just this variability has prevented the recognition of biomechanical regularities.

Biomechanics of a spinous process as an analogy with a jibcrane 

The shape of the spinous processes in Ctenosauriscus koeneni (v. Huene) makes clear that the vertical loads on the vertebral column of this rauisuchian species would have considerably grown in comparison to its ancestors. Two possible arguments for a change in lifestyle can be advanced, either the transition from an amphibious lifestyle to an exclusive life on land and additionally the transition from a quadrupedal to a facultative bipedal gait on land. Both modifications must lead to a substantial load increase of the vertebral column by essentially vertical loads. The weight must be transmitted to the ground through the legs, and this change requires skeletal modifications facilitating the load transmission.



Fig. 6. Acting forces at the location (K) of a jibcrane respectively at the tip of a spinous process, represented by arrows. The length of an arrow is a scale of a force compared with the load G1, acting at (K).

As can be seen, tensile force and compressive force are considerably lower in case of an elongated spinous process. In case of equilibrium the forces form a closed triangle


Lateral muscles in the water-dwelling and swimming ancestors that had been important for locomotion are no longer needed after the transition of tetrapods to the dry land. Now the muscular system must develop muscle fibres which can transmit vertical loads. To be able to carry the body on its legs the vertebral column must be modified so that it can transmit weight forces through the legs to the ground. For this purpose vertical hard parts must be present which can serve the suspension of muscles and tendons. This task is fulfilled by the dorsal spinous processes which represent the highest hard parts in a horizontally aligned vertebrate. Furthermore, the pull in the muscle fibres shall be as low as possible. This can be achieved most effectively by long spinous processes. A technical analogy can illustrate this context. Although there is no technical construction which completely simulates the function of a vertebral column nevertheless the pylon of a jibcrane is well suited to demonstrate the manner of action. The oblique ropes of the jibcrane transmit tensile forces. The size of these forces is dependent on the angle under which the ropes are suspended. The horizontal parts transmit compressive forces. As can be seen in fig. 6, the tensile force can be diminished by an elongation of the pylon, with the external force of the weight remaining unchanged, of course. It is no witchcraft, just simple mechanics.

The principle of suspending large loads on long pylons is often and very successfully applied. A wonderful example is presented by a recently opened new bridge over the Tarn valley in Southern France. The ropes carry the bridge.



Fig. 7. Oblique rope bridge over the Tarn valley in Southern France as an analogy with long dorsal spinous processes on which muscles are suspended. The ropes carry the bridge. See also fig. 6.



Thus, the spinous processes represent a very important element of the transmission of forces to the ground, and their elongation is connected with considerable advantages by reduction of the occurring loads in the muscles. Apparently, there was a strong stimulus in Ctenosaurisus respectively in its direct precursors to reduce skeletal loads by elongation of the dorsal spinous processes. Most probably this occurred during the transition to the life on the dry land. First of all the problem of modification of the vertebral column had to be resolved. A semi-aquatic lifestyle would have offered the best pre-conditions. The transition could proceed slowly and without time pressure, however very fast on a geological scale.

Accordingly, the reason of the elongation of the dorsal spinous processes is found, it was the transition to life on land         

The transition from quadrupedalism to bipedalism effected an additional stress of the vertebral column and made further modifications necessary. However, after the initiation of modifications for the life on land the additional loads were no entirely new problem. Presumably, special modifications of the brain functions may have been required, since bipedal locomotion needs an improved ability to balance the body.

The length of the spinous processes and their distribution along the vertebral column offer an excellent criterion to judge the ability of a fossil vertebrate to live on the dry land and its ability to perform the bipedal or quadrupedal gait. Ctenosauriscus koeneni (v. Huene) can unequivocally be recognized as a terrestrial animal by its long dorsal spinous processes, since these can be interpreted as an adaptation to relatively large loads that have to be transmitted to the ground.                                                                                 

However, the question arises why there is such an enormous elongation of the spinous processes in Ctenosauriscus if other vertebrates with considerably shorter spinous processes were capable of a bipedal locomotion on land. Many dinosaurs are usually depicted running or walking bipedally on the land. In my opinion, it is very questionable whether such an idea is based on reality.

Alignment of the spinous processes in the shoulder area of recent ungulate mammals, as an analogy

A second problem as to the shape of the vertebral column in Ctenosauriscus consists in the strong curvature of the spinous processes and how to explain it. It is connected with the elongation of the spinous processes, but it follows from a mechanical requirement, that is from an necessary alignment of the spinous processes. A comparison with recent ungulates can help to understand the problem.

In all quadrupedal mammals during the fast locomotion the weight is transmitted to the ground essentially through the fore-feet, while the hind-feet are mainly used for propulsion. This fact becomes particulary evident in mammals with a heavy head and thus large vertical loads, compared with the otherwise delicate skeleton. As an example I have used the Pleistocene elk from SMN Stuttgart, an ungulate with a special adaptation to hard soils during the ice age. This skeleton shows a considerable elongation of the spinous processes of the pectoral vertebrae, with the remaining cervical and dorsal processes being unspectacular. Following the direction of the spinous processes of the



Fig. 8. Directions of resulting weight and inertial forces in a Pleistocene elk during a step. The forces are transmitted through vertebral column, shoulder joint and fore-feet to the ground. The elk is exhibited in the SMN Stuttgart   


shoulder area we find that they are exactly aligned towards the shoulder joint. This alignment is absolutely necessary, since through a joint forces can be transmitted only, but never moments, and for balance reasons the line of action of the transmitted force must at any point of the motion pass as exactly as possible through the joint and the foot touching the ground. During a step all spinous processes are stressed one after another for a short moment.

Important for the comparison with Ctenosauriscus is the alignment of the spinous processes towards a joint.

External and internal forces, oh my God, what is that ?

Possibly, readers who are not familiar with the subject of mechanics might be puzzled by the fact that the action lines of forces in fig.8 do not always follow the hard parts of the skeleton. This arrangement of skeletal parts is by no means a matter of chance, on the contrary the result of an adaptation. It is very advantageous for a jumping or galloping vertebrate if the forces are not transmitted to the ground through a bony straight line, because the high dynamic forces on arrival on the ground are reduced by muscles and tendons in similar way as M. Bennett & G. Taylor (1995) have demonstrated in kangaroos. Every motor-cyclist enjoys the elasticity of the front fork. The elasticity means comfort, it diminishes the dynamic loads.

All four-legged mammals make clear that during locomotion at full speed considerable forces are transmitted through the fore-feet on arrival on the ground. While during walk or trot the fore-feet make single steps, this behaviour changes abruptly in case of a further speed increase. During galloping the forces have become so strong that the fore-feet must touch the ground simultaneously in order to be able to absorb the forces safely.

But it must be emphasized that it is irrelevant to balance considerations how in detail the forces between two definitely known points pass through a system, in this case through the body. There is a great difference between external and internal forces. External forces are those which lead to a movement of the elk or keep it in balance. External forces are weight and inertial forces resulting from the accelerated mass of the animal, as well as bearing forces and frictional forces on the ground. In case of a balanced situation of the external forces the system is at rest or continuously moving without an acceleration. 



Fig. 9.  Illustration of the difference between external and internal forces                      

a: The balance condition of a load suspended on a hook can easily be determined without knowledge of the internal forces passing through the hook. The shape of the hook is likewise unimportant. The forces pass through the air.

b: Alignment of the hook if a joint is inserted at the location x. In this case the hook assumes an alignment with the line of action passing through the joint.

           G = weight, Z = rope force            

Internal forces lead the external forces through the body, stressing bones, muscles and tendons. Mostly, it is very difficult to determine the internal forces, but in some cases the course of an internal force can easily be evaluated, namely if its line of action is known, for example the stress of a tendon. In general, for balance considerations internal forces can be neglected, their consideration is unimportant. Fig. 9 shall illustrate this context. Fig.9a shows a load suspended on a hook. It is well known that this system is in balance if the load’s centre of gravity has an exactly vertical position below the suspension point. The rope force equals the load’s weight. But for the balance condition it is entirely unimportant how the force is guided through the hook, which shape the hook has and which forces and moments are active in it. The only important pre-condition is that the hook is capable of sustaining the load without suffering damage.

It is extraordinarily advantageous that for balance considerations of the complete system (in this case an animal) the internal forces do not need be regarded; they are generally unimportant. The balance condition is established by the balanced external forces only.                                                                                                                                                                                        

A moment cannot be transmitted through a joint

Moreover, it may not easily be understandable from our own experience that moments cannot be transmitted through joints, although apparently we can carry a load on our arms and lead a moment (force times lever arm) through the elbow joint (fig.10). However, this impression is erroneous. The load is transmitted through muscles and tendons having a certain distance from the elbow joint. It is easy to carry a small load on your arms, but if you have to carry a cask of beer you have to use your arms stretched, that is the force must pass through the joint.



                            Fig. 10. Moments (force * lever arm) cannot be transmitted through a joint


I mention these facts in detail because our publication (Ebel et al. 1998) has been attacked by  several authors (H. Pfretzschner 1999 and R. Butler et al. 2011), instead of simply making an attempt to understand.my argumentation. These authors have demonstrated by their statements that they do not have the slightest idea of mechanics. It’s the physics, stupid. They have erroneously argued that the action line of forces could not simply pass through the air to the ground. Obviously, they have never heard about a difference between external and internal forces. Anticipating certain difficulties of understanding I had explained the facts at some length, however they were entirely unable to understand the problems (Fig. 9). Moreover, they have denied the correctness of our results. To reproach experienced engineers with a missing knowledge of one of their most important tools, mechanics, means a tall order. Incompetent laymen in this field should not claim to be qualified, it is ridiculous. The arrogance of ignorance is hardly tolerable. In addition, statements of this kind are indicative of the doubtful quality of such palaeontogical publications and of the simplicity of their authors. Biomechanics requires a profound familiarity with mechanical laws. Critics should keep this in mind.                   

The cause of the curvature of the spinous processes in Ctenosauriscus

In the vertebral of Ctenosauriscus koeneni (v. Huene), reconstructed by B. Krebs (1965), an obviously intended alignment of the spinous processes can also be recognized (fig.11). The spinous processes are not only elongated, but in addition they are curved. Furthermore, the spinous processes show another development, namely they are extremely flat, forming a rectangular cross-section with respect to the longitudinal plain. By this feature they are very stiff and can hardly be bended. The whole dorsal vertebral column in connection with a surely strong musculature around the “sail” resembles a strong plate. Certainly, it is not a matter of chance that this part of the fossil skeleton remained in contact and was the only remainder of the first find. Since there is probably a relationship between the shape of the spinous processes and a biomechanical requirement, the modified spinous processes would result from the transmission of forces during locomtion, comparable to the shoulder joint of the previously regarded elk. The upper tips of the spinous processes are of prime importance because here the compressive forces through the connections with muscles and tendons are transmitted to the ground, as through the pylon in fig.6. Contrary to the situation in fig. 6 in this case the force consists not only of weight forces, but additionally from inertial components resulting from the locomotion. Only at rest there is no additional force besides the weight.



Fig. 11. Ctenosaurisus koeneni (v. Huene) after B. Krebs (1965) with the directions of the compressive force during a step. It is surprising and certainly not a matter of chance that all directions meet in a common intersection, with the exception of the most forward and backward spinous processes.


A straight line through the tips of spinous processes with these directions reveals the surprising result that all action lines of the force meet in a common intersection. Certainly, this is not the result of an accidental development. Most probably this intersection corresponds to the position of a joint. However, it cannot be be the former position of the hip joint, rather it is the position of the knee joint. Since all strongly elongated spinous processes show an alignment towards this point of intersection the conclusion appears justified that during locomotion the main portion of external forces was transmitted to the ground through the hindlimbs only.

Most probably, Ctenosauriscus used a predominantly bipedal gait, and this fact is the cause of the strong curvature of the dorsal spinous processes.

There are no indications of an increased load on the foremost shorter spinous processes. Of course, the use of the fore-feet for locomotion was not precluded by this fact, but the animal had not to load them permanently and could carry its weight completely on its hindlimbs walking bipedally. Presumably, because of the different length of forelimbs and hindlimbs the fore-feet were loaded to a small extent only even during quadrupedal locomotion, and this gait allowed a low speed only. A comparison with otherwise very different kangaroos can serve as an example which use the hindlimbs only during fast locomotion.

At first sight it may be confusing that all action lines meet in one point, since the knee joint is not fixed and seemingly rotates around the hip joint. However, this impression is misleading. In fact, the knee joint rotates around the foot joint. During a step one foot remains standing on the ground, while the other one is raised, and the knee joint moves forward. Since the knee joint is connected with the tibia the external force must pass through the knee joint, irrespective of the angle between femur and longitudinal axis. Although the action lines from the tips of the spinous processes to the ground seemingly pass through the air nevertheless they pass through the knee joint. Only in this manner the animal can establish balanced conditions. Otherwise it would be impossible to learn the bipedal gait. The animal must be able to foresee the course of forces.

A split-up of the course of locomotion into a series of single pictures shows that at any time the resulting external force is directed from a spinous process through knee and foot to the ground and thereby indicates the direction which ensures balanced conditions (fig.12). For a short time the whole weight of the animal is suspended on a restricted number of spinous processes. In detail the muscular suspension does not consist of two muscles and tendonsonly but presumably of a grid of many, comparable with the oblique rope bridge in fig.7.



Fig.12. During a step the foot on the ground represents the momentary point of rotation for the knee joint. This point stays on the red dotted line while the body moves forward. The resulting external force passes more or less exactly through the animal’s centre of gravity. One after another the spinous processes are stressed.                                                                      

                   Reconstruction of Ctenosauriscus by co-author F.-O. Haderer

This interpretation of the long spinous processes of Ctenosauriscus makes clear that the reason of the elongation did not consist only in the augmentation of the rope angle as in fig.6. Presumably, such an elongation would not have been required, although the whole weight had to be transmitted through one leg. The main reason is not the size of the external force but it is its variable direction which initiated a modification of the vertebral column. The modification follows from the necessity to direct the external load in any case towards the knee joint and thereby to be able to control the complete bipedal course of locomotion. Obviously, the required curvature could be attained only by the elongation of the spinous processes. The advantage of this modification would be a higher attainable speed, however hardly comparable with the maximum speed of ungulates.

Ctenosauriscus was an animal capable of walking bipedally. Basically it was suitable for the production of Chirotherium tracks. This is in harmony with the differing size of fore-feet and hind-feet. Most probably, the fore-feet in Ctenosauriscus were smaller than the hind-feet.

ctenoanCertainly, it is no matter of chance that vertebrates with elongated spinous processes appeared in the upper Middle Buntsandstein. This period  was characterized in Germany by a warm and dry, at times even desert-like climate.

The flat Germanic bassin comprising large parts of Germany and the adjacent European areas was  up with sand and conglomerates. Rivers and freshwater lakes on which all animals were dependent became increasingly flatter and were in danger to dry out. The widespread Chirotherium tracks show that large areas of lakes became dry at least at times. Because of their enormous extension these lakes had essentially to be flat. The struggle for existence forced the reptiles to adapt themselves to these conditions by suitable modifications. In my opinion, these modifications were not primarily initiated by efforts to conquer the dry land, but by the necessity to increase the motility on the dry land. This aim was attained by the elongation of the spinous processes and they are a feature of an adaptation to extreme conditions.



Fig. 13. Remainders of Arizonasaurus show a similar condition of preservation as Ctenosauriscus. In particular, also here the legs are completely missing.


The reconstruction of Ctenosauriscus as a bipedally walking rauisuchian based on the unique shape of its vertebral column with characteristically elongated spinous processes can be transferred to similar forms. I have found in the internet a rauisuchian named Arizonasaurus, maybe published by Berkeley university, that exhibits a corresponding elongation of spinous processes as in Ctenosauriscus. In addition, the skull was preserved, but legs and tail were completely missing. These remainders were used for various reconstructions, unfortunely none of them is reliable. Such reconstructions are based on admirable fantasy, but unfortunately, on a missing insight into the biomechanical background. The biomechanical function of this vertebral column again was not recognized