DISCUSSION AND CONCLUSIONS

Analysis of Individual Variation

Considerable individual variation occurs in both the muscles and the nerves of the leg of the three species studied. The amount of variation reported by a worker depends in large part on the degree of variation that he considers significant.

Individual variation in the muscles and in the nerves will be discussed separately; that of the muscles (excluding innervation) will be considered first.

Muscles

Considering the number, rather than degree, of variations, the most variable muscles are: Mm. flexor digitorum longus, obturator, caudofemoralis, and extensor hallucis longus. The first-mentioned muscle exhibits 14 different variations in the specimens studied. Mm. vastus lateralis, flexor perforans et perforatus digiti II, and piriformis also showed a considerable number of variations. The following muscles did not exhibit any variations considered significant in this study: Mm. vastus medialis, femoritibialis internus, flexor perforatus digiti III, extensor brevis digiti III, and abductor digiti IV.

Muscles showing a great degree of individual variation included the following: M. extensor proprius digiti III was present in two legs of Pedioecetes but absent in the other legs studied. A fleshy muscle slip connected M. caudofemoralis pars caudifemoralis with the tendinous raphe between Mm. flexor cruris lateralis and femorocruralis in two legs, whereas in others this connection was tendinous or even absent altogether. M. caudofemoralis pars caudifemoralis had a tendinous area within the belly in only three legs. A vinculum connected the insertional tendons of Mm. flexor perforans et perforatus digiti II and flexor perforatus digiti II in only one leg. The fleshy belly of M. iliotrochantericus medius was completely split into two parts in one leg. M. flexor cruris lateralis had an accessory slip arising from the caudal musculature in one leg.

Certain individual variations reported in the accounts of the muscles formed a graduated series, as far as degree is concerned, from the typical to the extreme condition. Therefore it was difficult or impossible in some cases to state whether or not certain specimens exhibited such a variation. Elimination of the doubtful instances of variation leaves a total of 50 different variations (excluding variations between species) which can be attributed to a definite number of specimens. The remainder of the discussion of individual variation in the muscles concerns these 50 variations. See table 3.

The typical condition of any structure is considered to be the condition of that structure in the majority of the legs studied. Some conditions considered as typical in the present study might not be so considered if a larger number of specimens had been studied. If exactly half of the legs of one species shows a particular condition of a structure, the condition typical for this species is considered (for purposes of the following discussion) to be that found in the majority of the legs of the other species.

In all instances except two (of 50) the typical condition of the muscles in T. pallidicinctus was also the typical condition in T. cupido. The majority of the legs in T. cupido had an additional dorsal slip on the tendon of M. flexor digitorum longus in digits II and III. In all instances except seven the typical condition in T. pallidicinctus was also the typical condition in Pedioecetes. In these seven instances a variation in the former was the typical condition in the latter. These were: an additional dorsal slip on the tendon of M. flexor digitorum longus in each of three digits, a vinculum between the latter and M. flexor perforatus digiti IV, a partly fleshy insertion of M. flexor cruris medialis, an unossified lateral branch of the insertional tendon of M. extensor digitorum longus, and an independent insertion of the distalmost fibers of the distal head of M. extensor hallucis longus. For all characters except the number of the dorsal slips on the tendon of M. flexor digitorum longus in digits II and III, the typical condition in T. pallidicinctus was also the typical condition for all species considered together. To facilitate comparison, in the following discussion all of the above-mentioned characters are considered in all species as variants from the typical condition.

Certain legs showed a greater number of variations from the typical condition than did others. The majority of legs showed from four to seven variations in the muscles of the leg. The extremes were P.p. 1L, which showed 11, and T.c.p. 2L, which exhibited only one variation.

Twenty-three of the 50 variations were found in only one leg (out of 23). It would be expected that if additional specimens were studied, more kinds of variations would be found. Nine variations were found in only two legs, five in three legs, five in four legs, and four in five legs. One variation was found in nine legs, one in ten legs, and two in 12 legs; the last four variations were in the number of dorsal slips of the insertional tendon of M. flexor digitorum longus in digits II, III, and IV and in the ossification of the insertional tendon of M. extensor digitorum longus.

Five of the variations were found only in specimens in which only one leg was dissected. Considering only those eight specimens in which both legs were dissected, five of the 45 variations were found in both legs of each specimen exhibiting the variation; 28 variations were found in only one leg of each specimen exhibiting the variation; 12 variations were found in both legs of some specimens but in only one leg of other specimens. Of the six muscle features showing the greatest degree of individual variation (described previously), only two (both pertaining to M. caudofemoralis) were found in both legs of the specimens exhibiting the variation.

For one leg (the one showing the most variations) of each specimen of which both legs were studied, the number of variations that this leg had in common with every other leg (of all species) was determined. Then the number of variations in common between the two legs of one individual was compared with the number of variations in common between one leg of this individual and each leg of every other individual. See table 4. One leg of six of the eight specimens showed at least as many variations in common with a leg of another individual as with the other leg of the same individual. The two exceptions were T.p. 2R and T.c.a. 1R. Thus for most specimens there was as much variation in the muscles between the right and left legs of one individual as there was between individuals.

Of the 50 muscle variations seven were found only in T. pallidicinctus (eight legs), 16 were found only in T. cupido (nine legs), and ten were found only in Pedioecetes (six legs). Two were found in both species of Tympanuchus (but not in Pedioecetes). Fifteen were found in both Tympanuchus and Pedioecetes; of these, five were found in all three species studied, eight were shared by T. pallidicinctus and Pedioecetes, and two occurred in T. cupido and Pedioecetes.

Nerves

The lumbosacral plexus, femoral nerve, sciatic nerve, and tibial nerve all showed numerous individual variations. The peroneal nerve, however, was relatively constant. Variations in the obturator nerve were considered to be insignificant. See table 5.

In all instances except one (of 40) the typical condition in T. pallidicinctus was also the typical condition in T. cupido. In most of the legs of the latter the nerve to M. flexor cruris lateralis did not perforate M. caudofemoralis. In all instances except four the typical condition in T. pallidicinctus was also the typical condition in Pedioecetes. These exceptions were: prefixation of the lumbosacral plexus, six roots of the sciatic nerve, femoral nerve formed mainly from S2 to S4 and two twigs to M. flexor ischiofemoralis. In all instances the typical condition in T. pallidicinctus was also the typical condition for all species considered together.

Certain legs showed a greater number of variations from the typical condition of the nerves than did others. The greatest number of variations was shown by P.p. 3L, which had 12. T.p. 1R and T.c.p. 1L both showed only one.

All six variations in the lumbosacral plexus were found on both sides of each specimen exhibiting the variation. In marked contrast to the other nerves, there was no significant variation in the lumbosacral plexus between the right and left sides of one individual. (This might not always be true, however, if a larger number of specimens were studied.) Of the variations in the lumbosacral plexus, one was found in only one specimen (of 15), one was found in three specimens, one in four specimens, two in six specimens, and one in seven specimens. Of the 34 variations found in the other nerves, 14 were found in only one leg (of 23), six occurred in two legs, four in three legs, three in four legs, three in five legs, two in six legs, one in seven legs, and one in nine legs.

Four of the variations were found only in specimens in which only one leg was dissected. Considering only those eight specimens in which both legs were dissected, and excluding the lumbosacral plexus, ten of the 30 variations were found in both legs of each specimen exhibiting the variation; 16 variations were found in only one leg of each specimen exhibiting the variation; four variations were found in both legs of some specimens but in only one leg of other specimens.

The number of variations in common between the two legs of one individual was compared with the number between individuals in the same manner as for the muscles; the lumbosacral plexus was excluded from consideration. See table 6. One leg of six of the eight specimens showed at least as many variations in common with a leg of another individual as with the other leg of the same individual. The two exceptions were T.p. 2L and T.p. 3R. Thus for most specimens there was as much variation in the nerves other than the lumbosacral plexus between the right and left legs of one individual as there was between individuals.

Of the 40 nerve variations (including the lumbosacral plexus) 11 were found only in T. pallidicinctus, seven were found only in T. cupido, and seven were found only in Pedioecetes. Four were found in both species of Tympanuchus (but not in Pedioecetes). Eleven were found in both Tympanuchus and Pedioecetes; of these, four were found in all three species, three were shared by T. pallidicinctus and Pedioecetes and four occurred in T. cupido and Pedioecetes.

The average number of variations per leg in both muscles and nerves was 11 in T. pallidicinctus, nine in T. cupido, and 16 in Pedioecetes. The high number in the last is in part the result of these being variations from the typical condition of T. pallidicinctus (rather than from Pedioecetes).

Analysis of Variation Between Species

No constant differences in the muscles or nerves was found between T. cupido pinnatus and T. cupido attwateri. Only one constant difference was found between T. cupido and T. pallidicinctus: a thicker fleshy origin of M. extensor iliotibialis lateralis in T. cupido (associated with a thicker edge of the lateral iliac process).

Although no constant differences in the nerves were found between Pedioecetes and Tympanuchus (both species), 17 constant differences in the muscles were found between these two genera. Seven of these differences pertain to features of a single muscle—M. flexor cruris medialis. Compared with the condition in Tympanuchus, M. flexor cruris medialis in Pedioecetes has a wider origin, a partly fleshy (instead of entirely tendinous) origin, a more pronounced curvature of the line of origin, a wider insertion, an insertion posterior (rather than anterior) to the medial collateral ligament, an insertion that attaches in part to the articular capsule, and a shorter tendon of insertion (resulting in the fusion of the common insertional tendon of Mm. flexor cruris lateralis and femorocruralis with the fleshy belly rather than with the insertional tendon). Other differences include the following. A more extensive posteroproximal aponeurosis of M. extensor iliotibialis lateralis in Pedioecetes (resulting in a narrower fleshy origin); a more nearly straight line of origin of this muscle (associated with a less pronounced lateral iliac process); a thinner fleshy origin of this muscle (associated with a thinner edge of the lateral iliac process); a wider M. flexor cruris lateralis that is fleshy up to the origin from the vertebrae; a wider fleshy origin of M. iliacus; the origin of M. caudofemoralis pars iliofemoralis not reaching the ventral edge of the ischium; a narrower origin of M. adductor superficialis; a wider M. femorocruralis; and a shorter belly of M. extensor digitorum longus. Some additional differences between these two genera, which are slight in degree, are given in the accounts of the muscles. If additional specimens were studied, some of the differences listed above possibly would prove to be subject to individual variation and so could not properly be listed as constant differences between the two genera.

The picture of the differences between Tympanuchus and Pedioecetes that the present study presents is radically different from that presented by the study of Hudson, et al. (1959). These authors reported the following differences between these two genera. (I am using my terminology.) The origin of M. piriformis is narrower in Pedioecetes and is more posteriorly situated; the belly of M. extensor iliotibialis anticus is broader in Pedioecetes; the belly of M. tibialis anticus is longer; the belly of M. peroneus brevis is shorter; the insertional tendon of the anterolateral head of M. flexor perforatus digiti III is shorter; the belly of M. flexor digitorum longus is shorter; only two (rather than three) of the branches of M. extensor digitorum longus on the tarsometatarsus are ossified; the posterior metatarsal crest is shorter; M. flexor perforans et perforatus digiti II has two heads in Pedioecetes but only one in Tympanuchus; the roof over the hypotarsal canal enclosing the tendon of M. flexor digitorum longus is bony in Pedioecetes but fibrous in Tympanuchus; M. flexor cruris lateralis is wider in Pedioecetes; and the origin of M. femorocruralis is wider. I paid particular attention in my study to these 13 features given by Hudson, et al.; of these the only differences that I found to be constant were the last two. The apparent reason for this great discrepancy is the small number of legs of Tympanuchus studied by Hudson, et al. They studied eight legs of Pedioecetes but only two legs of Tympanuchus. This emphasizes the danger of making comparisons based on a very small number of specimens (a criticism which may prove to apply to the present study as well). The reason why Hudson, et al. did not report most of the differences found by me is not so apparent. Either the specimens studied by the former workers showed a greater variation in these characters than did my specimens or else those workers overlooked the differences. Probably both factors are involved. It remains to be determined how many specimens need to be studied in order to obtain a fairly accurate picture of variation.

Comparison with Other Studies of Innervation

I accept the following concept of muscle-nerve relationship. All muscles of the pelvic limb of birds have developed phylogenetically from either the dorsal extensor muscle mass or the ventral flexor muscle mass. The former was (at least originally) supplied by only the femoral and peroneal nerves ("dorsal" nerves), the latter by only the obturator and tibial nerves ("ventral" nerves). The best guide for determining which muscles are phylogenetically dorsal and which are ventral seems to be their embryogeny (as shown in the studies of Romer, 1927, and Wortham, 1948). In the phylogenetic changes undergone by the muscles under consideration, the innervation may have changed in some instances, although this is less likely to occur than changes in the attachment or function of the muscles. If a change in innervation has occurred, it would be more likely to be a change from one dorsal nerve to the other or from one ventral nerve to the other rather than from a dorsal nerve to a ventral one or vice versa.

Thus, in my opinion, a report of a dorsal muscle supplied by a ventral nerve, or vice versa, should be viewed with suspicion until it is verified. I suspect that many previous workers have ignored this concept of muscle-nerve relationship, or else do not accept it, since they report, without comment, dorsal muscles (as determined embryologically) innervated by ventral nerves, or vice versa. Owing to the intimate association between the proximal parts of the tibial and peroneal nerves, the true relationship may be difficult to determine. I suspect that this relationship has been misinterpreted by a number of workers. I found in Tympanuchus and Pedioecetes a branch of the tibial nerve that is closely associated with, and distributed with, the peroneal nerve and has been mistakenly considered a part of the peroneal nerve by some workers. In the study here reported on, I have found no definite exceptions to the expected innervation. The only possible exception is an extra branch, which could not be traced to its origin, supplying M. extensor iliofibularis in one leg. Thus my study of innervation agrees with the embryological determination of the (phylogenetic) dorsal and ventral muscles and lends strong support to the above-stated concept of muscle-nerve relationship.

I have compared my findings on the nerves with those of other workers, who have studied the nerves with a varying degree of thoroughness. The important differences in innervation between these studies and the present one are discussed below.

In neither of Gadow's works did he distinguish tibial and peroneal components in the thigh. In his later work (1891), covering a wide variety of birds, he found that M. piriformis sometimes has a femoral innervation in addition to the constant sciatic one and that M. gluteus profundus may or may not have a sciatic supply in addition to the femoral one. A comparison of Gadow's terminology of the sciatic nerve branches in the shank and foot (in both works) with mine shows that his branch I represents my peroneal nerve plus my paraperoneal branch of the tibial nerve (Ic); his branch II represents my medial division of the tibial nerve; and his branch III represents my posterior (IIIa) and lateral (IIIb) divisions of the tibial nerve.

Gadow's study (1880) on the ratites included Struthio, Rhea, and Casuarius. Only in Casuarius did Gadow find a branch (IIe) of the sciatic nerve supplying Mm. lumbricalis, adductor digiti II, and abductor digiti II. The two former muscles are typically supplied (as in Rhea) by the paraperoneal branch of the tibial nerve; Gadow's branch IIe presumably represents a segregated branch of this nerve. More surprising is his finding that M. abductor digiti II is innervated in Casuarius by both the deep peroneal nerve and branch IIe and in Rhea by branch Ic (paraperoneal branch of tibial nerve). The deep peroneal innervation is typical. Also unexpected is his finding that the posterior division of the femoral nerve gives minute twigs into M. gastrocnemius pars interna in Struthio and Casuarius. Since the other terminal branches of this nerve in these birds are nonmuscular, since this muscle is chiefly supplied by other nerves, and since the innervation from the femoral nerve is apparently atypical for most birds, the possibility should be considered that the femoral twigs are sensory rather than motor.

Sudilovskaya (1931), studying Struthio, Rhea, and Dromaeus (Dromiceius), used the same terminology as Gadow except that he designates as branch III Gadow's branch Ic. Sudilovskaya's discussion of the main branches of the sciatic nerve is confusing. He states that in Struthio, branches I, II, and III all pass through the tendinous guide loop for M. extensor iliofibularis; this is hard to believe. As near as I can determine, he has mistakenly given the same designation (branch III) to two separate branches (Gadow's Ic and III). There is no problem, however, in determining to which of these two branches he is referring when he is describing the innervation of a particular muscle, since one supplies only muscles of the shank and the other only intrinsic foot muscles. Sudilovskaya found M. abductor digiti II to be innervated by branch III (Ic of Gadow); thus the innervation of this muscle corresponds to that found in Rhea by Gadow. Although M. adductor digiti II had the expected innervation from branch III (paraperoneal branch of tibial nerve) in Dromaeus, that muscle was found to be supplied by branch II in Rhea. (Gadow, on the other hand, reports a typical innervation for this muscle in Rhea.) Sudilovskaya found M. peroneus brevis to be supplied by the deep peroneal branch (in contrast to the superficial peroneal supply that I found in Tympanuchus and Pedioecetes). He found M. gastrocnemius pars interna to be supplied in Struthio by twigs of the femoral nerve in addition to its typical innervation from branch II of the sciatic nerve; this agrees with Gadow's findings in the same genus. Sudilovskaya reports that M. gastrocnemius pars externa was innervated by branches II and III in Struthio and Rhea and by branches I and III in Dromaeus. (Gadow found only the typical innervation—branch III.)

In the Whooping Crane, Fisher and Goodman (1955) found a peroneal, rather than a femoral, nerve supply for pars postica of M. vastus lateralis. They also report a peroneal nerve supply for M. flexor ischiofemoralis (in contrast to the usual tibial nerve supply) and for M. adductor superficialis (in addition to the usual supply from the obturator nerve). The innervation was not given for the intrinsic foot musculature.

Fisher (1946), studying vultures, reports the following: tibial branches, in addition to the main sciatic branch, supplying M. extensor iliofibularis (typically supplied by the peroneal nerve); an obturator supply, in addition to the usual tibial supply, to M. flexor cruris medialis; a tibial supply, in addition to the typical obturator supply, to M. obturator pars postica; a possible obturator supply, in addition to the typical femoral supply, to M. ambiens; a possible peroneal supply, in addition to the typical tibial supply, to M. flexor digitorum longus; and a peroneal supply to Mm. abductor digiti IV, flexor hallucis brevis, and adductor digiti II (which are typically supplied by the paraperoneal branch of the tibial nerve). Fisher's postfibular branch of the peroneal nerve, which supplies the latter three muscles, apparently represents the paraperoneal branch of the tibial nerve.

Carlsson (1884) did not find a femoral nerve supply for M. gluteus profundus. He found an obturator supply, in addition to the usual sciatic supply, to M. flexor ischiofemoralis in Eudyptes chrysolopha and Mergulus alle but not in the other two forms studied. He reported a peroneal supply, rather than the expected tibial (paraperoneal) supply, to Mm. abductor digiti IV and adductor digiti IV.

DeMan (1873) found a twig of the obturator nerve supplying M. flexor ischiofemoralis, in addition to the typical innervation, in Corvus monedula, but not in the few other forms studied. He did not distinguish tibial and peroneal components in the thigh.

Wilcox (1948), studying a loon, did not find any peroneal supply to M. extensor iliotibialis lateralis or to M. gluteus profundus. He found a femoral, rather than a peroneal, supply to M. piriformis. He found an obturator, instead of a tibial, supply to M. flexor ischiofemoralis. (In some of my specimens I found a tiny blood vessel, appearing much like a nerve, emerging from the obturator foramen and entering M. flexor ischiofemoralis.) Wilcox reports an innervation of M. caudofemoralis pars caudifemoralis from the pudendal plexus, in addition to the usual sciatic one. Wilcox did not distinguish tibial and peroneal components in the thigh. In the shank and foot he misidentified the peroneal nerve as the tibial nerve and therefore gave erroneous innervations for all the muscles supplied by this nerve, except for M. adductor digiti IV, which actually should be supplied by the tibial nerve.

Howell (1938) studied only the hip and thigh musculature of the chicken. He overlooked the femoral nerve supply for M. gluteus profundus.

Romer (1927) studied only the hip and thigh muscles of the chick. He did not distinguish tibial and peroneal components in the thigh. He did not mention any sciatic supply for M. gluteus profundus.

Appleton (1928), studied (in various birds) only those muscles of the hip and thigh that are innervated by the tibial and peroneal nerves. He terms the former "ischiadicus ventralis" and the latter "ischiadicus dorsalis." His findings did not differ from mine.

Many differences in the innervation of specific muscles are reported in the literature, even in the same species (by different workers). Some of these differences may be real; others are probably misinterpretations. Consequently more work needs to be done before a complete understanding can be obtained of the innervation of the leg muscles of birds. Especially needed are studies of the tibial-peroneal nerve relationship, perhaps approached by a method other than gross dissection.

FOOTNOTES: